diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,50386 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999304928060054, + "eval_steps": 1.0, + "global_step": 7193, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "grad_norm": 10.255854128827643, + "learning_rate": 4.6296296296296295e-08, + "loss": 0.7089, + "step": 1 + }, + { + "epoch": 0.0, + "grad_norm": 8.773792360679519, + "learning_rate": 9.259259259259259e-08, + "loss": 0.6414, + "step": 2 + }, + { + "epoch": 0.0, + "grad_norm": 10.454717029458765, + "learning_rate": 1.3888888888888888e-07, + "loss": 0.6782, + "step": 3 + }, + { + "epoch": 0.0, + "grad_norm": 9.402114935002771, + "learning_rate": 1.8518518518518518e-07, + "loss": 0.635, + "step": 4 + }, + { + "epoch": 0.0, + "grad_norm": 9.169183983416053, + "learning_rate": 2.3148148148148148e-07, + "loss": 0.6219, + "step": 5 + }, + { + "epoch": 0.0, + "grad_norm": 10.036191772486438, + "learning_rate": 2.7777777777777776e-07, + "loss": 0.6553, + "step": 6 + }, + { + "epoch": 0.0, + "grad_norm": 11.965713108861875, + "learning_rate": 3.2407407407407406e-07, + "loss": 0.7664, + "step": 7 + }, + { + "epoch": 0.0, + "grad_norm": 11.154265911383941, + "learning_rate": 3.7037037037037036e-07, + "loss": 0.6945, + "step": 8 + }, + { + "epoch": 0.0, + "grad_norm": 10.68626688318975, + "learning_rate": 4.1666666666666667e-07, + "loss": 0.7317, + "step": 9 + }, + { + "epoch": 0.0, + "grad_norm": 13.399415302823204, + "learning_rate": 4.6296296296296297e-07, + "loss": 0.6778, + "step": 10 + }, + { + "epoch": 0.0, + "grad_norm": 8.212753191901033, + "learning_rate": 5.092592592592593e-07, + "loss": 0.63, + "step": 11 + }, + { + "epoch": 0.0, + "grad_norm": 7.796878855920977, + "learning_rate": 5.555555555555555e-07, + "loss": 0.6252, + "step": 12 + }, + { + "epoch": 0.0, + "grad_norm": 11.97829846200218, + "learning_rate": 6.018518518518519e-07, + "loss": 0.6076, + "step": 13 + }, + { + "epoch": 0.0, + "grad_norm": 8.612821227689986, + "learning_rate": 6.481481481481481e-07, + "loss": 0.6745, + "step": 14 + }, + { + "epoch": 0.0, + "grad_norm": 21.00893106426534, + "learning_rate": 6.944444444444446e-07, + "loss": 0.6036, + "step": 15 + }, + { + "epoch": 0.0, + "grad_norm": 9.713683569034302, + "learning_rate": 7.407407407407407e-07, + "loss": 0.5631, + "step": 16 + }, + { + "epoch": 0.0, + "grad_norm": 14.72722347265912, + "learning_rate": 7.870370370370371e-07, + "loss": 0.5615, + "step": 17 + }, + { + "epoch": 0.0, + "grad_norm": 7.544394214922187, + "learning_rate": 8.333333333333333e-07, + "loss": 0.609, + "step": 18 + }, + { + "epoch": 0.0, + "grad_norm": 9.523364730476793, + "learning_rate": 8.796296296296297e-07, + "loss": 0.5802, + "step": 19 + }, + { + "epoch": 0.0, + "grad_norm": 5.419257987223273, + "learning_rate": 9.259259259259259e-07, + "loss": 0.5041, + "step": 20 + }, + { + "epoch": 0.0, + "grad_norm": 9.77850446169523, + "learning_rate": 9.722222222222224e-07, + "loss": 0.5525, + "step": 21 + }, + { + "epoch": 0.0, + "grad_norm": 4.252949025507043, + "learning_rate": 1.0185185185185185e-06, + "loss": 0.4768, + "step": 22 + }, + { + "epoch": 0.0, + "grad_norm": 4.7101311808953685, + "learning_rate": 1.0648148148148149e-06, + "loss": 0.4981, + "step": 23 + }, + { + "epoch": 0.0, + "grad_norm": 4.53366283698979, + "learning_rate": 1.111111111111111e-06, + "loss": 0.5176, + "step": 24 + }, + { + "epoch": 0.0, + "grad_norm": 3.763232944709757, + "learning_rate": 1.1574074074074076e-06, + "loss": 0.5139, + "step": 25 + }, + { + "epoch": 0.0, + "grad_norm": 3.7091692609708717, + "learning_rate": 1.2037037037037037e-06, + "loss": 0.4981, + "step": 26 + }, + { + "epoch": 0.0, + "grad_norm": 15.016693375984712, + "learning_rate": 1.25e-06, + "loss": 0.4577, + "step": 27 + }, + { + "epoch": 0.0, + "grad_norm": 3.6102044253176437, + "learning_rate": 1.2962962962962962e-06, + "loss": 0.4989, + "step": 28 + }, + { + "epoch": 0.0, + "grad_norm": 3.078894895654288, + "learning_rate": 1.3425925925925928e-06, + "loss": 0.4747, + "step": 29 + }, + { + "epoch": 0.0, + "grad_norm": 2.278891574726967, + "learning_rate": 1.3888888888888892e-06, + "loss": 0.4109, + "step": 30 + }, + { + "epoch": 0.0, + "grad_norm": 2.317052004327827, + "learning_rate": 1.4351851851851853e-06, + "loss": 0.447, + "step": 31 + }, + { + "epoch": 0.0, + "grad_norm": 2.4476257193527147, + "learning_rate": 1.4814814814814815e-06, + "loss": 0.4385, + "step": 32 + }, + { + "epoch": 0.0, + "grad_norm": 3.8040634784992315, + "learning_rate": 1.527777777777778e-06, + "loss": 0.435, + "step": 33 + }, + { + "epoch": 0.0, + "grad_norm": 2.610119369335003, + "learning_rate": 1.5740740740740742e-06, + "loss": 0.4321, + "step": 34 + }, + { + "epoch": 0.0, + "grad_norm": 2.6263190952334807, + "learning_rate": 1.6203703703703705e-06, + "loss": 0.4192, + "step": 35 + }, + { + "epoch": 0.01, + "grad_norm": 3.424885660562162, + "learning_rate": 1.6666666666666667e-06, + "loss": 0.4613, + "step": 36 + }, + { + "epoch": 0.01, + "grad_norm": 3.3561962804013494, + "learning_rate": 1.7129629629629632e-06, + "loss": 0.458, + "step": 37 + }, + { + "epoch": 0.01, + "grad_norm": 2.520456771934489, + "learning_rate": 1.7592592592592594e-06, + "loss": 0.4439, + "step": 38 + }, + { + "epoch": 0.01, + "grad_norm": 3.871885426391512, + "learning_rate": 1.8055555555555557e-06, + "loss": 0.4463, + "step": 39 + }, + { + "epoch": 0.01, + "grad_norm": 2.592313829601372, + "learning_rate": 1.8518518518518519e-06, + "loss": 0.4371, + "step": 40 + }, + { + "epoch": 0.01, + "grad_norm": 3.5345587197376505, + "learning_rate": 1.8981481481481484e-06, + "loss": 0.4288, + "step": 41 + }, + { + "epoch": 0.01, + "grad_norm": 3.2766287781088375, + "learning_rate": 1.944444444444445e-06, + "loss": 0.4439, + "step": 42 + }, + { + "epoch": 0.01, + "grad_norm": 2.4974698598721234, + "learning_rate": 1.9907407407407407e-06, + "loss": 0.4335, + "step": 43 + }, + { + "epoch": 0.01, + "grad_norm": 2.446775546130796, + "learning_rate": 2.037037037037037e-06, + "loss": 0.4276, + "step": 44 + }, + { + "epoch": 0.01, + "grad_norm": 3.610162990317946, + "learning_rate": 2.0833333333333334e-06, + "loss": 0.418, + "step": 45 + }, + { + "epoch": 0.01, + "grad_norm": 2.4439951061940675, + "learning_rate": 2.1296296296296298e-06, + "loss": 0.4454, + "step": 46 + }, + { + "epoch": 0.01, + "grad_norm": 3.4040054429684474, + "learning_rate": 2.175925925925926e-06, + "loss": 0.4367, + "step": 47 + }, + { + "epoch": 0.01, + "grad_norm": 2.312735050532508, + "learning_rate": 2.222222222222222e-06, + "loss": 0.4265, + "step": 48 + }, + { + "epoch": 0.01, + "grad_norm": 2.6957192931244576, + "learning_rate": 2.268518518518519e-06, + "loss": 0.4032, + "step": 49 + }, + { + "epoch": 0.01, + "grad_norm": 2.598984536491883, + "learning_rate": 2.314814814814815e-06, + "loss": 0.4118, + "step": 50 + }, + { + "epoch": 0.01, + "grad_norm": 2.5537662533133347, + "learning_rate": 2.361111111111111e-06, + "loss": 0.3956, + "step": 51 + }, + { + "epoch": 0.01, + "grad_norm": 2.6266154138590196, + "learning_rate": 2.4074074074074075e-06, + "loss": 0.4259, + "step": 52 + }, + { + "epoch": 0.01, + "grad_norm": 3.477042477617357, + "learning_rate": 2.453703703703704e-06, + "loss": 0.4311, + "step": 53 + }, + { + "epoch": 0.01, + "grad_norm": 2.985005093220916, + "learning_rate": 2.5e-06, + "loss": 0.468, + "step": 54 + }, + { + "epoch": 0.01, + "grad_norm": 2.8501633741813093, + "learning_rate": 2.5462962962962966e-06, + "loss": 0.4334, + "step": 55 + }, + { + "epoch": 0.01, + "grad_norm": 2.731612824284174, + "learning_rate": 2.5925925925925925e-06, + "loss": 0.4044, + "step": 56 + }, + { + "epoch": 0.01, + "grad_norm": 2.635429910973552, + "learning_rate": 2.6388888888888893e-06, + "loss": 0.4477, + "step": 57 + }, + { + "epoch": 0.01, + "grad_norm": 2.315429583382698, + "learning_rate": 2.6851851851851856e-06, + "loss": 0.4545, + "step": 58 + }, + { + "epoch": 0.01, + "grad_norm": 2.8322617435642976, + "learning_rate": 2.7314814814814816e-06, + "loss": 0.416, + "step": 59 + }, + { + "epoch": 0.01, + "grad_norm": 2.7183081042824786, + "learning_rate": 2.7777777777777783e-06, + "loss": 0.3938, + "step": 60 + }, + { + "epoch": 0.01, + "grad_norm": 2.5562819747238783, + "learning_rate": 2.8240740740740743e-06, + "loss": 0.3987, + "step": 61 + }, + { + "epoch": 0.01, + "grad_norm": 3.6937095292787854, + "learning_rate": 2.8703703703703706e-06, + "loss": 0.4562, + "step": 62 + }, + { + "epoch": 0.01, + "grad_norm": 2.606990917862641, + "learning_rate": 2.916666666666667e-06, + "loss": 0.411, + "step": 63 + }, + { + "epoch": 0.01, + "grad_norm": 2.983323777143174, + "learning_rate": 2.962962962962963e-06, + "loss": 0.4274, + "step": 64 + }, + { + "epoch": 0.01, + "grad_norm": 2.5130825647128834, + "learning_rate": 3.0092592592592597e-06, + "loss": 0.4109, + "step": 65 + }, + { + "epoch": 0.01, + "grad_norm": 4.471999264480983, + "learning_rate": 3.055555555555556e-06, + "loss": 0.4265, + "step": 66 + }, + { + "epoch": 0.01, + "grad_norm": 2.6530405813156763, + "learning_rate": 3.101851851851852e-06, + "loss": 0.4201, + "step": 67 + }, + { + "epoch": 0.01, + "grad_norm": 2.124624576669715, + "learning_rate": 3.1481481481481483e-06, + "loss": 0.4066, + "step": 68 + }, + { + "epoch": 0.01, + "grad_norm": 2.630607085255591, + "learning_rate": 3.1944444444444443e-06, + "loss": 0.3945, + "step": 69 + }, + { + "epoch": 0.01, + "grad_norm": 2.845874367230957, + "learning_rate": 3.240740740740741e-06, + "loss": 0.4325, + "step": 70 + }, + { + "epoch": 0.01, + "grad_norm": 8.069961740943874, + "learning_rate": 3.2870370370370374e-06, + "loss": 0.4038, + "step": 71 + }, + { + "epoch": 0.01, + "grad_norm": 2.488692512085441, + "learning_rate": 3.3333333333333333e-06, + "loss": 0.4638, + "step": 72 + }, + { + "epoch": 0.01, + "grad_norm": 3.0371942256745017, + "learning_rate": 3.37962962962963e-06, + "loss": 0.4097, + "step": 73 + }, + { + "epoch": 0.01, + "grad_norm": 3.220511241197115, + "learning_rate": 3.4259259259259265e-06, + "loss": 0.4003, + "step": 74 + }, + { + "epoch": 0.01, + "grad_norm": 3.633152204811553, + "learning_rate": 3.4722222222222224e-06, + "loss": 0.4464, + "step": 75 + }, + { + "epoch": 0.01, + "grad_norm": 3.4881085145784416, + "learning_rate": 3.5185185185185187e-06, + "loss": 0.4104, + "step": 76 + }, + { + "epoch": 0.01, + "grad_norm": 2.7328395195041746, + "learning_rate": 3.5648148148148147e-06, + "loss": 0.4023, + "step": 77 + }, + { + "epoch": 0.01, + "grad_norm": 2.952179328952469, + "learning_rate": 3.6111111111111115e-06, + "loss": 0.4107, + "step": 78 + }, + { + "epoch": 0.01, + "grad_norm": 3.0145433355840354, + "learning_rate": 3.657407407407408e-06, + "loss": 0.3828, + "step": 79 + }, + { + "epoch": 0.01, + "grad_norm": 3.333631472311848, + "learning_rate": 3.7037037037037037e-06, + "loss": 0.4452, + "step": 80 + }, + { + "epoch": 0.01, + "grad_norm": 2.69753519175676, + "learning_rate": 3.7500000000000005e-06, + "loss": 0.4117, + "step": 81 + }, + { + "epoch": 0.01, + "grad_norm": 3.21176087609264, + "learning_rate": 3.796296296296297e-06, + "loss": 0.3869, + "step": 82 + }, + { + "epoch": 0.01, + "grad_norm": 2.4890191961757817, + "learning_rate": 3.842592592592592e-06, + "loss": 0.4136, + "step": 83 + }, + { + "epoch": 0.01, + "grad_norm": 2.578718140563557, + "learning_rate": 3.88888888888889e-06, + "loss": 0.4368, + "step": 84 + }, + { + "epoch": 0.01, + "grad_norm": 2.466958397571284, + "learning_rate": 3.935185185185186e-06, + "loss": 0.4031, + "step": 85 + }, + { + "epoch": 0.01, + "grad_norm": 2.8900685201894576, + "learning_rate": 3.9814814814814814e-06, + "loss": 0.4191, + "step": 86 + }, + { + "epoch": 0.01, + "grad_norm": 3.3686680325922955, + "learning_rate": 4.027777777777779e-06, + "loss": 0.3907, + "step": 87 + }, + { + "epoch": 0.01, + "grad_norm": 2.7663195782363736, + "learning_rate": 4.074074074074074e-06, + "loss": 0.4172, + "step": 88 + }, + { + "epoch": 0.01, + "grad_norm": 2.3780496870898458, + "learning_rate": 4.1203703703703705e-06, + "loss": 0.4046, + "step": 89 + }, + { + "epoch": 0.01, + "grad_norm": 3.3868434764372934, + "learning_rate": 4.166666666666667e-06, + "loss": 0.4036, + "step": 90 + }, + { + "epoch": 0.01, + "grad_norm": 2.7503898380791716, + "learning_rate": 4.212962962962963e-06, + "loss": 0.4393, + "step": 91 + }, + { + "epoch": 0.01, + "grad_norm": 2.4175323073024435, + "learning_rate": 4.2592592592592596e-06, + "loss": 0.3995, + "step": 92 + }, + { + "epoch": 0.01, + "grad_norm": 3.2598709968883535, + "learning_rate": 4.305555555555556e-06, + "loss": 0.3756, + "step": 93 + }, + { + "epoch": 0.01, + "grad_norm": 2.464987650717689, + "learning_rate": 4.351851851851852e-06, + "loss": 0.3961, + "step": 94 + }, + { + "epoch": 0.01, + "grad_norm": 5.0200344520961195, + "learning_rate": 4.398148148148149e-06, + "loss": 0.4341, + "step": 95 + }, + { + "epoch": 0.01, + "grad_norm": 3.3934409076301946, + "learning_rate": 4.444444444444444e-06, + "loss": 0.3803, + "step": 96 + }, + { + "epoch": 0.01, + "grad_norm": 2.0939684034470716, + "learning_rate": 4.490740740740741e-06, + "loss": 0.4096, + "step": 97 + }, + { + "epoch": 0.01, + "grad_norm": 3.3783022516744, + "learning_rate": 4.537037037037038e-06, + "loss": 0.4205, + "step": 98 + }, + { + "epoch": 0.01, + "grad_norm": 5.413801759638468, + "learning_rate": 4.583333333333333e-06, + "loss": 0.3975, + "step": 99 + }, + { + "epoch": 0.01, + "grad_norm": 2.7398280657726053, + "learning_rate": 4.62962962962963e-06, + "loss": 0.4207, + "step": 100 + }, + { + "epoch": 0.01, + "grad_norm": 2.5402510369008784, + "learning_rate": 4.675925925925927e-06, + "loss": 0.409, + "step": 101 + }, + { + "epoch": 0.01, + "grad_norm": 2.2447264332340486, + "learning_rate": 4.722222222222222e-06, + "loss": 0.396, + "step": 102 + }, + { + "epoch": 0.01, + "grad_norm": 3.55441114778751, + "learning_rate": 4.768518518518519e-06, + "loss": 0.4089, + "step": 103 + }, + { + "epoch": 0.01, + "grad_norm": 4.07838457524117, + "learning_rate": 4.814814814814815e-06, + "loss": 0.3708, + "step": 104 + }, + { + "epoch": 0.01, + "grad_norm": 2.7851693491114786, + "learning_rate": 4.861111111111111e-06, + "loss": 0.4069, + "step": 105 + }, + { + "epoch": 0.01, + "grad_norm": 3.230069779103466, + "learning_rate": 4.907407407407408e-06, + "loss": 0.4133, + "step": 106 + }, + { + "epoch": 0.01, + "grad_norm": 2.6078490758442214, + "learning_rate": 4.953703703703704e-06, + "loss": 0.3846, + "step": 107 + }, + { + "epoch": 0.02, + "grad_norm": 3.145857591672062, + "learning_rate": 5e-06, + "loss": 0.3946, + "step": 108 + }, + { + "epoch": 0.02, + "grad_norm": 2.860373829007912, + "learning_rate": 5.046296296296297e-06, + "loss": 0.4048, + "step": 109 + }, + { + "epoch": 0.02, + "grad_norm": 5.181292163771113, + "learning_rate": 5.092592592592593e-06, + "loss": 0.3778, + "step": 110 + }, + { + "epoch": 0.02, + "grad_norm": 2.2676312338693894, + "learning_rate": 5.138888888888889e-06, + "loss": 0.3934, + "step": 111 + }, + { + "epoch": 0.02, + "grad_norm": 2.5234692901209823, + "learning_rate": 5.185185185185185e-06, + "loss": 0.4269, + "step": 112 + }, + { + "epoch": 0.02, + "grad_norm": 2.5337725988428934, + "learning_rate": 5.231481481481482e-06, + "loss": 0.4008, + "step": 113 + }, + { + "epoch": 0.02, + "grad_norm": 2.608852953085008, + "learning_rate": 5.2777777777777785e-06, + "loss": 0.4367, + "step": 114 + }, + { + "epoch": 0.02, + "grad_norm": 3.2355246857109408, + "learning_rate": 5.324074074074075e-06, + "loss": 0.4185, + "step": 115 + }, + { + "epoch": 0.02, + "grad_norm": 2.837771455408418, + "learning_rate": 5.370370370370371e-06, + "loss": 0.4104, + "step": 116 + }, + { + "epoch": 0.02, + "grad_norm": 3.383503170967786, + "learning_rate": 5.416666666666667e-06, + "loss": 0.4197, + "step": 117 + }, + { + "epoch": 0.02, + "grad_norm": 4.397197479442109, + "learning_rate": 5.462962962962963e-06, + "loss": 0.3684, + "step": 118 + }, + { + "epoch": 0.02, + "grad_norm": 3.1374031884442624, + "learning_rate": 5.5092592592592595e-06, + "loss": 0.4094, + "step": 119 + }, + { + "epoch": 0.02, + "grad_norm": 3.543344757486579, + "learning_rate": 5.555555555555557e-06, + "loss": 0.4067, + "step": 120 + }, + { + "epoch": 0.02, + "grad_norm": 2.1863786999376, + "learning_rate": 5.601851851851853e-06, + "loss": 0.3793, + "step": 121 + }, + { + "epoch": 0.02, + "grad_norm": 2.2217766894286863, + "learning_rate": 5.6481481481481485e-06, + "loss": 0.3809, + "step": 122 + }, + { + "epoch": 0.02, + "grad_norm": 6.544131417153299, + "learning_rate": 5.694444444444445e-06, + "loss": 0.4043, + "step": 123 + }, + { + "epoch": 0.02, + "grad_norm": 3.0111130933671273, + "learning_rate": 5.740740740740741e-06, + "loss": 0.3987, + "step": 124 + }, + { + "epoch": 0.02, + "grad_norm": 3.450861166362308, + "learning_rate": 5.787037037037038e-06, + "loss": 0.402, + "step": 125 + }, + { + "epoch": 0.02, + "grad_norm": 2.90664853284843, + "learning_rate": 5.833333333333334e-06, + "loss": 0.3794, + "step": 126 + }, + { + "epoch": 0.02, + "grad_norm": 3.7115612906279334, + "learning_rate": 5.8796296296296295e-06, + "loss": 0.3958, + "step": 127 + }, + { + "epoch": 0.02, + "grad_norm": 2.8251651795903983, + "learning_rate": 5.925925925925926e-06, + "loss": 0.4155, + "step": 128 + }, + { + "epoch": 0.02, + "grad_norm": 3.536720329943924, + "learning_rate": 5.972222222222222e-06, + "loss": 0.4018, + "step": 129 + }, + { + "epoch": 0.02, + "grad_norm": 3.095105241434553, + "learning_rate": 6.018518518518519e-06, + "loss": 0.4101, + "step": 130 + }, + { + "epoch": 0.02, + "grad_norm": 2.3863804556923283, + "learning_rate": 6.064814814814816e-06, + "loss": 0.3665, + "step": 131 + }, + { + "epoch": 0.02, + "grad_norm": 2.12988678036126, + "learning_rate": 6.111111111111112e-06, + "loss": 0.4129, + "step": 132 + }, + { + "epoch": 0.02, + "grad_norm": 2.6339230413996138, + "learning_rate": 6.157407407407408e-06, + "loss": 0.3602, + "step": 133 + }, + { + "epoch": 0.02, + "grad_norm": 3.086883649805019, + "learning_rate": 6.203703703703704e-06, + "loss": 0.3925, + "step": 134 + }, + { + "epoch": 0.02, + "grad_norm": 2.8630881890034803, + "learning_rate": 6.25e-06, + "loss": 0.3991, + "step": 135 + }, + { + "epoch": 0.02, + "grad_norm": 5.832830470832314, + "learning_rate": 6.296296296296297e-06, + "loss": 0.3822, + "step": 136 + }, + { + "epoch": 0.02, + "grad_norm": 3.155554336249525, + "learning_rate": 6.342592592592594e-06, + "loss": 0.4152, + "step": 137 + }, + { + "epoch": 0.02, + "grad_norm": 2.210110094234751, + "learning_rate": 6.3888888888888885e-06, + "loss": 0.3817, + "step": 138 + }, + { + "epoch": 0.02, + "grad_norm": 2.6684290241753215, + "learning_rate": 6.435185185185186e-06, + "loss": 0.4123, + "step": 139 + }, + { + "epoch": 0.02, + "grad_norm": 3.4651775494160404, + "learning_rate": 6.481481481481482e-06, + "loss": 0.3952, + "step": 140 + }, + { + "epoch": 0.02, + "grad_norm": 2.7795054216788966, + "learning_rate": 6.5277777777777784e-06, + "loss": 0.3871, + "step": 141 + }, + { + "epoch": 0.02, + "grad_norm": 2.4261457167219835, + "learning_rate": 6.574074074074075e-06, + "loss": 0.3693, + "step": 142 + }, + { + "epoch": 0.02, + "grad_norm": 2.7982447932470698, + "learning_rate": 6.620370370370371e-06, + "loss": 0.4014, + "step": 143 + }, + { + "epoch": 0.02, + "grad_norm": 2.6161920186765126, + "learning_rate": 6.666666666666667e-06, + "loss": 0.3935, + "step": 144 + }, + { + "epoch": 0.02, + "grad_norm": 2.8595145804997073, + "learning_rate": 6.712962962962963e-06, + "loss": 0.4079, + "step": 145 + }, + { + "epoch": 0.02, + "grad_norm": 2.638650439556187, + "learning_rate": 6.75925925925926e-06, + "loss": 0.4005, + "step": 146 + }, + { + "epoch": 0.02, + "grad_norm": 2.6351521116339587, + "learning_rate": 6.8055555555555566e-06, + "loss": 0.394, + "step": 147 + }, + { + "epoch": 0.02, + "grad_norm": 2.5240239772862627, + "learning_rate": 6.851851851851853e-06, + "loss": 0.4016, + "step": 148 + }, + { + "epoch": 0.02, + "grad_norm": 2.3911458158822096, + "learning_rate": 6.898148148148148e-06, + "loss": 0.3721, + "step": 149 + }, + { + "epoch": 0.02, + "grad_norm": 2.554819061478118, + "learning_rate": 6.944444444444445e-06, + "loss": 0.3861, + "step": 150 + }, + { + "epoch": 0.02, + "grad_norm": 2.84413328688025, + "learning_rate": 6.990740740740741e-06, + "loss": 0.4175, + "step": 151 + }, + { + "epoch": 0.02, + "grad_norm": 2.247204058545684, + "learning_rate": 7.0370370370370375e-06, + "loss": 0.4141, + "step": 152 + }, + { + "epoch": 0.02, + "grad_norm": 3.794284021960763, + "learning_rate": 7.083333333333335e-06, + "loss": 0.4028, + "step": 153 + }, + { + "epoch": 0.02, + "grad_norm": 3.8029897805515405, + "learning_rate": 7.129629629629629e-06, + "loss": 0.4141, + "step": 154 + }, + { + "epoch": 0.02, + "grad_norm": 3.0642245306154376, + "learning_rate": 7.1759259259259266e-06, + "loss": 0.3733, + "step": 155 + }, + { + "epoch": 0.02, + "grad_norm": 2.5282627153975894, + "learning_rate": 7.222222222222223e-06, + "loss": 0.3635, + "step": 156 + }, + { + "epoch": 0.02, + "grad_norm": 2.5844471230211523, + "learning_rate": 7.268518518518519e-06, + "loss": 0.3887, + "step": 157 + }, + { + "epoch": 0.02, + "grad_norm": 2.8902502031190025, + "learning_rate": 7.314814814814816e-06, + "loss": 0.4025, + "step": 158 + }, + { + "epoch": 0.02, + "grad_norm": 2.4184674724423885, + "learning_rate": 7.361111111111112e-06, + "loss": 0.3886, + "step": 159 + }, + { + "epoch": 0.02, + "grad_norm": 2.350433379430058, + "learning_rate": 7.4074074074074075e-06, + "loss": 0.3926, + "step": 160 + }, + { + "epoch": 0.02, + "grad_norm": 2.410879023537296, + "learning_rate": 7.453703703703704e-06, + "loss": 0.3735, + "step": 161 + }, + { + "epoch": 0.02, + "grad_norm": 2.774671890082111, + "learning_rate": 7.500000000000001e-06, + "loss": 0.3964, + "step": 162 + }, + { + "epoch": 0.02, + "grad_norm": 5.408550595652813, + "learning_rate": 7.546296296296297e-06, + "loss": 0.3886, + "step": 163 + }, + { + "epoch": 0.02, + "grad_norm": 2.5144844458067306, + "learning_rate": 7.592592592592594e-06, + "loss": 0.4011, + "step": 164 + }, + { + "epoch": 0.02, + "grad_norm": 2.2062221267943753, + "learning_rate": 7.638888888888888e-06, + "loss": 0.3934, + "step": 165 + }, + { + "epoch": 0.02, + "grad_norm": 2.4646673477484202, + "learning_rate": 7.685185185185185e-06, + "loss": 0.4079, + "step": 166 + }, + { + "epoch": 0.02, + "grad_norm": 2.211016478246042, + "learning_rate": 7.731481481481483e-06, + "loss": 0.3944, + "step": 167 + }, + { + "epoch": 0.02, + "grad_norm": 2.8306817341982056, + "learning_rate": 7.77777777777778e-06, + "loss": 0.3969, + "step": 168 + }, + { + "epoch": 0.02, + "grad_norm": 2.0189477377183582, + "learning_rate": 7.824074074074076e-06, + "loss": 0.3864, + "step": 169 + }, + { + "epoch": 0.02, + "grad_norm": 3.1359935744283067, + "learning_rate": 7.870370370370372e-06, + "loss": 0.3883, + "step": 170 + }, + { + "epoch": 0.02, + "grad_norm": 2.1675033051506847, + "learning_rate": 7.916666666666667e-06, + "loss": 0.3792, + "step": 171 + }, + { + "epoch": 0.02, + "grad_norm": 2.232837947591229, + "learning_rate": 7.962962962962963e-06, + "loss": 0.386, + "step": 172 + }, + { + "epoch": 0.02, + "grad_norm": 2.397573528752364, + "learning_rate": 8.00925925925926e-06, + "loss": 0.3914, + "step": 173 + }, + { + "epoch": 0.02, + "grad_norm": 2.9173401970279538, + "learning_rate": 8.055555555555557e-06, + "loss": 0.4211, + "step": 174 + }, + { + "epoch": 0.02, + "grad_norm": 2.3163881111155575, + "learning_rate": 8.101851851851854e-06, + "loss": 0.416, + "step": 175 + }, + { + "epoch": 0.02, + "grad_norm": 2.367813538702257, + "learning_rate": 8.148148148148148e-06, + "loss": 0.4094, + "step": 176 + }, + { + "epoch": 0.02, + "grad_norm": 2.292698485039522, + "learning_rate": 8.194444444444445e-06, + "loss": 0.3854, + "step": 177 + }, + { + "epoch": 0.02, + "grad_norm": 3.9651010422341666, + "learning_rate": 8.240740740740741e-06, + "loss": 0.382, + "step": 178 + }, + { + "epoch": 0.02, + "grad_norm": 2.5330404814360112, + "learning_rate": 8.287037037037037e-06, + "loss": 0.3654, + "step": 179 + }, + { + "epoch": 0.03, + "grad_norm": 2.998333546809386, + "learning_rate": 8.333333333333334e-06, + "loss": 0.4114, + "step": 180 + }, + { + "epoch": 0.03, + "grad_norm": 2.6548419145951803, + "learning_rate": 8.37962962962963e-06, + "loss": 0.3994, + "step": 181 + }, + { + "epoch": 0.03, + "grad_norm": 2.5831333500167775, + "learning_rate": 8.425925925925926e-06, + "loss": 0.3886, + "step": 182 + }, + { + "epoch": 0.03, + "grad_norm": 2.27266225837616, + "learning_rate": 8.472222222222223e-06, + "loss": 0.4119, + "step": 183 + }, + { + "epoch": 0.03, + "grad_norm": 3.741513293182775, + "learning_rate": 8.518518518518519e-06, + "loss": 0.4109, + "step": 184 + }, + { + "epoch": 0.03, + "grad_norm": 7.059567772046245, + "learning_rate": 8.564814814814816e-06, + "loss": 0.4142, + "step": 185 + }, + { + "epoch": 0.03, + "grad_norm": 2.7045094193250674, + "learning_rate": 8.611111111111112e-06, + "loss": 0.3939, + "step": 186 + }, + { + "epoch": 0.03, + "grad_norm": 3.2884104885721546, + "learning_rate": 8.657407407407408e-06, + "loss": 0.3893, + "step": 187 + }, + { + "epoch": 0.03, + "grad_norm": 3.628696089287186, + "learning_rate": 8.703703703703705e-06, + "loss": 0.4073, + "step": 188 + }, + { + "epoch": 0.03, + "grad_norm": 2.579362190192803, + "learning_rate": 8.750000000000001e-06, + "loss": 0.3488, + "step": 189 + }, + { + "epoch": 0.03, + "grad_norm": 2.080359881631079, + "learning_rate": 8.796296296296297e-06, + "loss": 0.4009, + "step": 190 + }, + { + "epoch": 0.03, + "grad_norm": 2.226420673550312, + "learning_rate": 8.842592592592594e-06, + "loss": 0.3823, + "step": 191 + }, + { + "epoch": 0.03, + "grad_norm": 2.7427400473972696, + "learning_rate": 8.888888888888888e-06, + "loss": 0.3766, + "step": 192 + }, + { + "epoch": 0.03, + "grad_norm": 2.480725163754228, + "learning_rate": 8.935185185185186e-06, + "loss": 0.3851, + "step": 193 + }, + { + "epoch": 0.03, + "grad_norm": 2.8563757735212523, + "learning_rate": 8.981481481481483e-06, + "loss": 0.4072, + "step": 194 + }, + { + "epoch": 0.03, + "grad_norm": 2.410644972978592, + "learning_rate": 9.027777777777779e-06, + "loss": 0.4012, + "step": 195 + }, + { + "epoch": 0.03, + "grad_norm": 3.659295606710743, + "learning_rate": 9.074074074074075e-06, + "loss": 0.4114, + "step": 196 + }, + { + "epoch": 0.03, + "grad_norm": 2.761358314198226, + "learning_rate": 9.120370370370372e-06, + "loss": 0.3996, + "step": 197 + }, + { + "epoch": 0.03, + "grad_norm": 3.2348973974993718, + "learning_rate": 9.166666666666666e-06, + "loss": 0.3968, + "step": 198 + }, + { + "epoch": 0.03, + "grad_norm": 2.6945154162700216, + "learning_rate": 9.212962962962963e-06, + "loss": 0.405, + "step": 199 + }, + { + "epoch": 0.03, + "grad_norm": 2.4288426622494224, + "learning_rate": 9.25925925925926e-06, + "loss": 0.3837, + "step": 200 + }, + { + "epoch": 0.03, + "grad_norm": 2.8754362562535944, + "learning_rate": 9.305555555555557e-06, + "loss": 0.4191, + "step": 201 + }, + { + "epoch": 0.03, + "grad_norm": 2.190254757680217, + "learning_rate": 9.351851851851854e-06, + "loss": 0.4023, + "step": 202 + }, + { + "epoch": 0.03, + "grad_norm": 2.0237494898523365, + "learning_rate": 9.398148148148148e-06, + "loss": 0.3771, + "step": 203 + }, + { + "epoch": 0.03, + "grad_norm": 2.267735032150335, + "learning_rate": 9.444444444444445e-06, + "loss": 0.3892, + "step": 204 + }, + { + "epoch": 0.03, + "grad_norm": 2.673187739273911, + "learning_rate": 9.490740740740741e-06, + "loss": 0.392, + "step": 205 + }, + { + "epoch": 0.03, + "grad_norm": 2.1912811002528936, + "learning_rate": 9.537037037037037e-06, + "loss": 0.3792, + "step": 206 + }, + { + "epoch": 0.03, + "grad_norm": 3.426379106468188, + "learning_rate": 9.583333333333335e-06, + "loss": 0.4056, + "step": 207 + }, + { + "epoch": 0.03, + "grad_norm": 2.622190742192239, + "learning_rate": 9.62962962962963e-06, + "loss": 0.3826, + "step": 208 + }, + { + "epoch": 0.03, + "grad_norm": 2.6330980830781643, + "learning_rate": 9.675925925925926e-06, + "loss": 0.3802, + "step": 209 + }, + { + "epoch": 0.03, + "grad_norm": 4.004473405526082, + "learning_rate": 9.722222222222223e-06, + "loss": 0.405, + "step": 210 + }, + { + "epoch": 0.03, + "grad_norm": 4.001116884588422, + "learning_rate": 9.768518518518519e-06, + "loss": 0.3793, + "step": 211 + }, + { + "epoch": 0.03, + "grad_norm": 2.314649128390183, + "learning_rate": 9.814814814814815e-06, + "loss": 0.4142, + "step": 212 + }, + { + "epoch": 0.03, + "grad_norm": 2.329180695101604, + "learning_rate": 9.861111111111112e-06, + "loss": 0.3752, + "step": 213 + }, + { + "epoch": 0.03, + "grad_norm": 3.1099185986620514, + "learning_rate": 9.907407407407408e-06, + "loss": 0.4046, + "step": 214 + }, + { + "epoch": 0.03, + "grad_norm": 2.233181099595426, + "learning_rate": 9.953703703703704e-06, + "loss": 0.3845, + "step": 215 + }, + { + "epoch": 0.03, + "grad_norm": 3.105340245366727, + "learning_rate": 1e-05, + "loss": 0.3931, + "step": 216 + }, + { + "epoch": 0.03, + "grad_norm": 2.5297640269171655, + "learning_rate": 9.999999493123334e-06, + "loss": 0.4084, + "step": 217 + }, + { + "epoch": 0.03, + "grad_norm": 3.6222424540260127, + "learning_rate": 9.999997972493432e-06, + "loss": 0.3666, + "step": 218 + }, + { + "epoch": 0.03, + "grad_norm": 3.049819753253031, + "learning_rate": 9.999995438110604e-06, + "loss": 0.4092, + "step": 219 + }, + { + "epoch": 0.03, + "grad_norm": 3.979435315821689, + "learning_rate": 9.999991889975367e-06, + "loss": 0.3918, + "step": 220 + }, + { + "epoch": 0.03, + "grad_norm": 4.333782005775019, + "learning_rate": 9.999987328088438e-06, + "loss": 0.3909, + "step": 221 + }, + { + "epoch": 0.03, + "grad_norm": 2.2302856882470143, + "learning_rate": 9.999981752450742e-06, + "loss": 0.4068, + "step": 222 + }, + { + "epoch": 0.03, + "grad_norm": 3.0862048715286554, + "learning_rate": 9.99997516306341e-06, + "loss": 0.4059, + "step": 223 + }, + { + "epoch": 0.03, + "grad_norm": 1.9401044871350934, + "learning_rate": 9.999967559927774e-06, + "loss": 0.3812, + "step": 224 + }, + { + "epoch": 0.03, + "grad_norm": 2.3490689301477166, + "learning_rate": 9.999958943045383e-06, + "loss": 0.3826, + "step": 225 + }, + { + "epoch": 0.03, + "grad_norm": 4.3580296724564285, + "learning_rate": 9.999949312417978e-06, + "loss": 0.3983, + "step": 226 + }, + { + "epoch": 0.03, + "grad_norm": 3.0017849079679335, + "learning_rate": 9.999938668047516e-06, + "loss": 0.3938, + "step": 227 + }, + { + "epoch": 0.03, + "grad_norm": 2.8971720879447305, + "learning_rate": 9.999927009936152e-06, + "loss": 0.3757, + "step": 228 + }, + { + "epoch": 0.03, + "grad_norm": 3.5281553305953257, + "learning_rate": 9.999914338086249e-06, + "loss": 0.3653, + "step": 229 + }, + { + "epoch": 0.03, + "grad_norm": 3.175527672784214, + "learning_rate": 9.999900652500381e-06, + "loss": 0.3906, + "step": 230 + }, + { + "epoch": 0.03, + "grad_norm": 2.7774732857293856, + "learning_rate": 9.999885953181317e-06, + "loss": 0.3964, + "step": 231 + }, + { + "epoch": 0.03, + "grad_norm": 2.5225225420825703, + "learning_rate": 9.999870240132042e-06, + "loss": 0.398, + "step": 232 + }, + { + "epoch": 0.03, + "grad_norm": 2.8514002981728357, + "learning_rate": 9.999853513355738e-06, + "loss": 0.3815, + "step": 233 + }, + { + "epoch": 0.03, + "grad_norm": 3.1614774719823253, + "learning_rate": 9.9998357728558e-06, + "loss": 0.375, + "step": 234 + }, + { + "epoch": 0.03, + "grad_norm": 2.5484109656345026, + "learning_rate": 9.999817018635821e-06, + "loss": 0.4034, + "step": 235 + }, + { + "epoch": 0.03, + "grad_norm": 2.4150318675061944, + "learning_rate": 9.999797250699606e-06, + "loss": 0.4218, + "step": 236 + }, + { + "epoch": 0.03, + "grad_norm": 2.5519631373100125, + "learning_rate": 9.999776469051164e-06, + "loss": 0.4112, + "step": 237 + }, + { + "epoch": 0.03, + "grad_norm": 2.7270090686318467, + "learning_rate": 9.999754673694705e-06, + "loss": 0.4045, + "step": 238 + }, + { + "epoch": 0.03, + "grad_norm": 3.2057354089106798, + "learning_rate": 9.999731864634651e-06, + "loss": 0.3865, + "step": 239 + }, + { + "epoch": 0.03, + "grad_norm": 2.110521389625803, + "learning_rate": 9.999708041875624e-06, + "loss": 0.4021, + "step": 240 + }, + { + "epoch": 0.03, + "grad_norm": 2.2250370378051207, + "learning_rate": 9.999683205422456e-06, + "loss": 0.3796, + "step": 241 + }, + { + "epoch": 0.03, + "grad_norm": 3.229952953302523, + "learning_rate": 9.999657355280185e-06, + "loss": 0.4, + "step": 242 + }, + { + "epoch": 0.03, + "grad_norm": 2.885164747314696, + "learning_rate": 9.999630491454046e-06, + "loss": 0.3988, + "step": 243 + }, + { + "epoch": 0.03, + "grad_norm": 2.1625233447025947, + "learning_rate": 9.999602613949491e-06, + "loss": 0.379, + "step": 244 + }, + { + "epoch": 0.03, + "grad_norm": 2.3175920836601085, + "learning_rate": 9.99957372277217e-06, + "loss": 0.3999, + "step": 245 + }, + { + "epoch": 0.03, + "grad_norm": 2.917177760291321, + "learning_rate": 9.999543817927939e-06, + "loss": 0.3984, + "step": 246 + }, + { + "epoch": 0.03, + "grad_norm": 2.167640142043269, + "learning_rate": 9.999512899422866e-06, + "loss": 0.388, + "step": 247 + }, + { + "epoch": 0.03, + "grad_norm": 2.5316973084491265, + "learning_rate": 9.999480967263213e-06, + "loss": 0.3681, + "step": 248 + }, + { + "epoch": 0.03, + "grad_norm": 2.609475760653929, + "learning_rate": 9.99944802145546e-06, + "loss": 0.4045, + "step": 249 + }, + { + "epoch": 0.03, + "grad_norm": 3.080115541257489, + "learning_rate": 9.999414062006286e-06, + "loss": 0.3779, + "step": 250 + }, + { + "epoch": 0.03, + "grad_norm": 2.3121438850237666, + "learning_rate": 9.999379088922575e-06, + "loss": 0.3736, + "step": 251 + }, + { + "epoch": 0.04, + "grad_norm": 2.548927172363069, + "learning_rate": 9.999343102211417e-06, + "loss": 0.3754, + "step": 252 + }, + { + "epoch": 0.04, + "grad_norm": 2.5521736908064163, + "learning_rate": 9.999306101880109e-06, + "loss": 0.381, + "step": 253 + }, + { + "epoch": 0.04, + "grad_norm": 5.356583755395139, + "learning_rate": 9.999268087936154e-06, + "loss": 0.3969, + "step": 254 + }, + { + "epoch": 0.04, + "grad_norm": 4.222707242770012, + "learning_rate": 9.999229060387259e-06, + "loss": 0.4022, + "step": 255 + }, + { + "epoch": 0.04, + "grad_norm": 2.944668974608297, + "learning_rate": 9.999189019241335e-06, + "loss": 0.379, + "step": 256 + }, + { + "epoch": 0.04, + "grad_norm": 2.538725419617771, + "learning_rate": 9.999147964506502e-06, + "loss": 0.3779, + "step": 257 + }, + { + "epoch": 0.04, + "grad_norm": 2.558916031031792, + "learning_rate": 9.999105896191085e-06, + "loss": 0.3949, + "step": 258 + }, + { + "epoch": 0.04, + "grad_norm": 2.193622745419291, + "learning_rate": 9.999062814303611e-06, + "loss": 0.3579, + "step": 259 + }, + { + "epoch": 0.04, + "grad_norm": 2.1325665579687567, + "learning_rate": 9.999018718852815e-06, + "loss": 0.3949, + "step": 260 + }, + { + "epoch": 0.04, + "grad_norm": 2.1719565840064083, + "learning_rate": 9.99897360984764e-06, + "loss": 0.3787, + "step": 261 + }, + { + "epoch": 0.04, + "grad_norm": 2.5156973580497337, + "learning_rate": 9.99892748729723e-06, + "loss": 0.3797, + "step": 262 + }, + { + "epoch": 0.04, + "grad_norm": 1.9478086073211056, + "learning_rate": 9.998880351210936e-06, + "loss": 0.3946, + "step": 263 + }, + { + "epoch": 0.04, + "grad_norm": 2.0455359700929376, + "learning_rate": 9.998832201598315e-06, + "loss": 0.3797, + "step": 264 + }, + { + "epoch": 0.04, + "grad_norm": 2.0426932758739853, + "learning_rate": 9.998783038469131e-06, + "loss": 0.3705, + "step": 265 + }, + { + "epoch": 0.04, + "grad_norm": 4.097094564291679, + "learning_rate": 9.99873286183335e-06, + "loss": 0.3365, + "step": 266 + }, + { + "epoch": 0.04, + "grad_norm": 2.92221768736335, + "learning_rate": 9.998681671701145e-06, + "loss": 0.3662, + "step": 267 + }, + { + "epoch": 0.04, + "grad_norm": 2.3594114774981416, + "learning_rate": 9.998629468082896e-06, + "loss": 0.3842, + "step": 268 + }, + { + "epoch": 0.04, + "grad_norm": 2.3985157697699746, + "learning_rate": 9.99857625098919e-06, + "loss": 0.3869, + "step": 269 + }, + { + "epoch": 0.04, + "grad_norm": 2.6714441215778124, + "learning_rate": 9.99852202043081e-06, + "loss": 0.388, + "step": 270 + }, + { + "epoch": 0.04, + "grad_norm": 15.429254596560067, + "learning_rate": 9.998466776418758e-06, + "loss": 0.4012, + "step": 271 + }, + { + "epoch": 0.04, + "grad_norm": 2.337428237412138, + "learning_rate": 9.99841051896423e-06, + "loss": 0.3923, + "step": 272 + }, + { + "epoch": 0.04, + "grad_norm": 3.367293478588156, + "learning_rate": 9.998353248078636e-06, + "loss": 0.4125, + "step": 273 + }, + { + "epoch": 0.04, + "grad_norm": 2.637544665864994, + "learning_rate": 9.998294963773583e-06, + "loss": 0.3789, + "step": 274 + }, + { + "epoch": 0.04, + "grad_norm": 2.638497883770822, + "learning_rate": 9.998235666060893e-06, + "loss": 0.3866, + "step": 275 + }, + { + "epoch": 0.04, + "grad_norm": 2.4198259687675274, + "learning_rate": 9.998175354952589e-06, + "loss": 0.397, + "step": 276 + }, + { + "epoch": 0.04, + "grad_norm": 2.873372497895899, + "learning_rate": 9.998114030460894e-06, + "loss": 0.4007, + "step": 277 + }, + { + "epoch": 0.04, + "grad_norm": 3.0950965291915424, + "learning_rate": 9.998051692598246e-06, + "loss": 0.4188, + "step": 278 + }, + { + "epoch": 0.04, + "grad_norm": 3.318720483702163, + "learning_rate": 9.99798834137728e-06, + "loss": 0.3978, + "step": 279 + }, + { + "epoch": 0.04, + "grad_norm": 22.98179174055231, + "learning_rate": 9.997923976810845e-06, + "loss": 0.3703, + "step": 280 + }, + { + "epoch": 0.04, + "grad_norm": 2.9993748560899403, + "learning_rate": 9.997858598911989e-06, + "loss": 0.3736, + "step": 281 + }, + { + "epoch": 0.04, + "grad_norm": 2.3670250974597145, + "learning_rate": 9.997792207693965e-06, + "loss": 0.4066, + "step": 282 + }, + { + "epoch": 0.04, + "grad_norm": 5.00761835857215, + "learning_rate": 9.99772480317024e-06, + "loss": 0.406, + "step": 283 + }, + { + "epoch": 0.04, + "grad_norm": 2.4894280622845475, + "learning_rate": 9.997656385354475e-06, + "loss": 0.3781, + "step": 284 + }, + { + "epoch": 0.04, + "grad_norm": 2.2509950984317975, + "learning_rate": 9.997586954260543e-06, + "loss": 0.3892, + "step": 285 + }, + { + "epoch": 0.04, + "grad_norm": 3.4046375321591555, + "learning_rate": 9.99751650990252e-06, + "loss": 0.3994, + "step": 286 + }, + { + "epoch": 0.04, + "grad_norm": 2.6763445661421423, + "learning_rate": 9.997445052294693e-06, + "loss": 0.3868, + "step": 287 + }, + { + "epoch": 0.04, + "grad_norm": 2.385114089313256, + "learning_rate": 9.997372581451545e-06, + "loss": 0.3968, + "step": 288 + }, + { + "epoch": 0.04, + "grad_norm": 2.476630787925025, + "learning_rate": 9.997299097387773e-06, + "loss": 0.3845, + "step": 289 + }, + { + "epoch": 0.04, + "grad_norm": 2.8482715742492815, + "learning_rate": 9.997224600118275e-06, + "loss": 0.3823, + "step": 290 + }, + { + "epoch": 0.04, + "grad_norm": 3.1279279380765868, + "learning_rate": 9.997149089658155e-06, + "loss": 0.3662, + "step": 291 + }, + { + "epoch": 0.04, + "grad_norm": 2.1895484940797694, + "learning_rate": 9.997072566022723e-06, + "loss": 0.3693, + "step": 292 + }, + { + "epoch": 0.04, + "grad_norm": 2.9868079913153736, + "learning_rate": 9.996995029227493e-06, + "loss": 0.3787, + "step": 293 + }, + { + "epoch": 0.04, + "grad_norm": 2.52619412516836, + "learning_rate": 9.996916479288185e-06, + "loss": 0.3946, + "step": 294 + }, + { + "epoch": 0.04, + "grad_norm": 2.82595326149051, + "learning_rate": 9.996836916220729e-06, + "loss": 0.3869, + "step": 295 + }, + { + "epoch": 0.04, + "grad_norm": 2.919466565482181, + "learning_rate": 9.996756340041255e-06, + "loss": 0.4068, + "step": 296 + }, + { + "epoch": 0.04, + "grad_norm": 3.0673152913108828, + "learning_rate": 9.996674750766099e-06, + "loss": 0.3769, + "step": 297 + }, + { + "epoch": 0.04, + "grad_norm": 3.6932724381273725, + "learning_rate": 9.9965921484118e-06, + "loss": 0.3863, + "step": 298 + }, + { + "epoch": 0.04, + "grad_norm": 2.5021611942833504, + "learning_rate": 9.996508532995113e-06, + "loss": 0.3915, + "step": 299 + }, + { + "epoch": 0.04, + "grad_norm": 3.526553122277605, + "learning_rate": 9.996423904532985e-06, + "loss": 0.3813, + "step": 300 + }, + { + "epoch": 0.04, + "grad_norm": 2.5451241670390323, + "learning_rate": 9.996338263042577e-06, + "loss": 0.3742, + "step": 301 + }, + { + "epoch": 0.04, + "grad_norm": 2.4935105497044616, + "learning_rate": 9.996251608541253e-06, + "loss": 0.392, + "step": 302 + }, + { + "epoch": 0.04, + "grad_norm": 3.8071742951166567, + "learning_rate": 9.996163941046582e-06, + "loss": 0.3817, + "step": 303 + }, + { + "epoch": 0.04, + "grad_norm": 2.6556964846928466, + "learning_rate": 9.996075260576338e-06, + "loss": 0.3781, + "step": 304 + }, + { + "epoch": 0.04, + "grad_norm": 13.951125739132515, + "learning_rate": 9.9959855671485e-06, + "loss": 0.4062, + "step": 305 + }, + { + "epoch": 0.04, + "grad_norm": 2.689334922157135, + "learning_rate": 9.995894860781256e-06, + "loss": 0.395, + "step": 306 + }, + { + "epoch": 0.04, + "grad_norm": 2.3256406497059023, + "learning_rate": 9.995803141492996e-06, + "loss": 0.4043, + "step": 307 + }, + { + "epoch": 0.04, + "grad_norm": 2.392851205360204, + "learning_rate": 9.995710409302316e-06, + "loss": 0.3987, + "step": 308 + }, + { + "epoch": 0.04, + "grad_norm": 3.047303770801914, + "learning_rate": 9.995616664228016e-06, + "loss": 0.4009, + "step": 309 + }, + { + "epoch": 0.04, + "grad_norm": 2.7726022965440693, + "learning_rate": 9.995521906289106e-06, + "loss": 0.3782, + "step": 310 + }, + { + "epoch": 0.04, + "grad_norm": 2.35500654360542, + "learning_rate": 9.995426135504794e-06, + "loss": 0.3914, + "step": 311 + }, + { + "epoch": 0.04, + "grad_norm": 6.869918670776811, + "learning_rate": 9.995329351894502e-06, + "loss": 0.3921, + "step": 312 + }, + { + "epoch": 0.04, + "grad_norm": 2.4237989041239856, + "learning_rate": 9.99523155547785e-06, + "loss": 0.3741, + "step": 313 + }, + { + "epoch": 0.04, + "grad_norm": 2.100503807914272, + "learning_rate": 9.995132746274668e-06, + "loss": 0.3485, + "step": 314 + }, + { + "epoch": 0.04, + "grad_norm": 2.8373022127990466, + "learning_rate": 9.995032924304986e-06, + "loss": 0.3897, + "step": 315 + }, + { + "epoch": 0.04, + "grad_norm": 6.700975383972722, + "learning_rate": 9.994932089589048e-06, + "loss": 0.3955, + "step": 316 + }, + { + "epoch": 0.04, + "grad_norm": 2.177572945865357, + "learning_rate": 9.994830242147295e-06, + "loss": 0.3943, + "step": 317 + }, + { + "epoch": 0.04, + "grad_norm": 2.466212816814164, + "learning_rate": 9.99472738200038e-06, + "loss": 0.3962, + "step": 318 + }, + { + "epoch": 0.04, + "grad_norm": 5.230872673829239, + "learning_rate": 9.994623509169154e-06, + "loss": 0.3812, + "step": 319 + }, + { + "epoch": 0.04, + "grad_norm": 2.6078488952063537, + "learning_rate": 9.994518623674679e-06, + "loss": 0.3889, + "step": 320 + }, + { + "epoch": 0.04, + "grad_norm": 2.793652451488903, + "learning_rate": 9.99441272553822e-06, + "loss": 0.4132, + "step": 321 + }, + { + "epoch": 0.04, + "grad_norm": 3.168095768278835, + "learning_rate": 9.99430581478125e-06, + "loss": 0.3619, + "step": 322 + }, + { + "epoch": 0.04, + "grad_norm": 2.1446136199978163, + "learning_rate": 9.994197891425443e-06, + "loss": 0.3897, + "step": 323 + }, + { + "epoch": 0.05, + "grad_norm": 2.290075110729867, + "learning_rate": 9.99408895549268e-06, + "loss": 0.3701, + "step": 324 + }, + { + "epoch": 0.05, + "grad_norm": 4.407738424761088, + "learning_rate": 9.993979007005051e-06, + "loss": 0.3862, + "step": 325 + }, + { + "epoch": 0.05, + "grad_norm": 2.5868987033840645, + "learning_rate": 9.993868045984846e-06, + "loss": 0.3916, + "step": 326 + }, + { + "epoch": 0.05, + "grad_norm": 2.0513806840756144, + "learning_rate": 9.993756072454562e-06, + "loss": 0.3936, + "step": 327 + }, + { + "epoch": 0.05, + "grad_norm": 2.8517992886260686, + "learning_rate": 9.993643086436901e-06, + "loss": 0.3644, + "step": 328 + }, + { + "epoch": 0.05, + "grad_norm": 2.2754879588323536, + "learning_rate": 9.993529087954776e-06, + "loss": 0.3815, + "step": 329 + }, + { + "epoch": 0.05, + "grad_norm": 2.3096189879590905, + "learning_rate": 9.993414077031294e-06, + "loss": 0.3822, + "step": 330 + }, + { + "epoch": 0.05, + "grad_norm": 5.029699727230014, + "learning_rate": 9.993298053689775e-06, + "loss": 0.4071, + "step": 331 + }, + { + "epoch": 0.05, + "grad_norm": 2.853240494912146, + "learning_rate": 9.993181017953747e-06, + "loss": 0.3899, + "step": 332 + }, + { + "epoch": 0.05, + "grad_norm": 3.0181243221956273, + "learning_rate": 9.993062969846935e-06, + "loss": 0.4001, + "step": 333 + }, + { + "epoch": 0.05, + "grad_norm": 2.4590142980909664, + "learning_rate": 9.992943909393273e-06, + "loss": 0.3807, + "step": 334 + }, + { + "epoch": 0.05, + "grad_norm": 3.3957440740463354, + "learning_rate": 9.992823836616903e-06, + "loss": 0.3514, + "step": 335 + }, + { + "epoch": 0.05, + "grad_norm": 2.2100828344537815, + "learning_rate": 9.99270275154217e-06, + "loss": 0.371, + "step": 336 + }, + { + "epoch": 0.05, + "grad_norm": 2.410550254608413, + "learning_rate": 9.99258065419362e-06, + "loss": 0.4013, + "step": 337 + }, + { + "epoch": 0.05, + "grad_norm": 2.8521400115890967, + "learning_rate": 9.992457544596012e-06, + "loss": 0.3916, + "step": 338 + }, + { + "epoch": 0.05, + "grad_norm": 3.5959985420131475, + "learning_rate": 9.992333422774308e-06, + "loss": 0.365, + "step": 339 + }, + { + "epoch": 0.05, + "grad_norm": 3.226635678501706, + "learning_rate": 9.99220828875367e-06, + "loss": 0.3826, + "step": 340 + }, + { + "epoch": 0.05, + "grad_norm": 2.6804793707443015, + "learning_rate": 9.99208214255947e-06, + "loss": 0.3855, + "step": 341 + }, + { + "epoch": 0.05, + "grad_norm": 2.6185997998788824, + "learning_rate": 9.991954984217285e-06, + "loss": 0.3726, + "step": 342 + }, + { + "epoch": 0.05, + "grad_norm": 2.613102669844605, + "learning_rate": 9.991826813752897e-06, + "loss": 0.4041, + "step": 343 + }, + { + "epoch": 0.05, + "grad_norm": 2.344373881160252, + "learning_rate": 9.99169763119229e-06, + "loss": 0.3597, + "step": 344 + }, + { + "epoch": 0.05, + "grad_norm": 2.2871342868383793, + "learning_rate": 9.99156743656166e-06, + "loss": 0.3955, + "step": 345 + }, + { + "epoch": 0.05, + "grad_norm": 4.0328687262536675, + "learning_rate": 9.9914362298874e-06, + "loss": 0.4087, + "step": 346 + }, + { + "epoch": 0.05, + "grad_norm": 2.5176935730144954, + "learning_rate": 9.991304011196115e-06, + "loss": 0.3978, + "step": 347 + }, + { + "epoch": 0.05, + "grad_norm": 3.098835328553678, + "learning_rate": 9.991170780514611e-06, + "loss": 0.4193, + "step": 348 + }, + { + "epoch": 0.05, + "grad_norm": 2.770886316635241, + "learning_rate": 9.9910365378699e-06, + "loss": 0.3656, + "step": 349 + }, + { + "epoch": 0.05, + "grad_norm": 3.217356004706249, + "learning_rate": 9.990901283289202e-06, + "loss": 0.3703, + "step": 350 + }, + { + "epoch": 0.05, + "grad_norm": 3.112001265617347, + "learning_rate": 9.99076501679994e-06, + "loss": 0.3571, + "step": 351 + }, + { + "epoch": 0.05, + "grad_norm": 3.7036718164338023, + "learning_rate": 9.99062773842974e-06, + "loss": 0.3674, + "step": 352 + }, + { + "epoch": 0.05, + "grad_norm": 3.391938450897036, + "learning_rate": 9.990489448206436e-06, + "loss": 0.3654, + "step": 353 + }, + { + "epoch": 0.05, + "grad_norm": 2.416733721846635, + "learning_rate": 9.990350146158066e-06, + "loss": 0.3844, + "step": 354 + }, + { + "epoch": 0.05, + "grad_norm": 2.382460052757202, + "learning_rate": 9.990209832312874e-06, + "loss": 0.3776, + "step": 355 + }, + { + "epoch": 0.05, + "grad_norm": 3.2297793967599193, + "learning_rate": 9.990068506699309e-06, + "loss": 0.3879, + "step": 356 + }, + { + "epoch": 0.05, + "grad_norm": 2.3665247815319552, + "learning_rate": 9.989926169346025e-06, + "loss": 0.3641, + "step": 357 + }, + { + "epoch": 0.05, + "grad_norm": 3.226995693412493, + "learning_rate": 9.989782820281881e-06, + "loss": 0.3873, + "step": 358 + }, + { + "epoch": 0.05, + "grad_norm": 2.42977875763257, + "learning_rate": 9.98963845953594e-06, + "loss": 0.3704, + "step": 359 + }, + { + "epoch": 0.05, + "grad_norm": 2.8233407285984753, + "learning_rate": 9.989493087137473e-06, + "loss": 0.3986, + "step": 360 + }, + { + "epoch": 0.05, + "grad_norm": 2.6018614813500416, + "learning_rate": 9.989346703115954e-06, + "loss": 0.3863, + "step": 361 + }, + { + "epoch": 0.05, + "grad_norm": 2.717119135601431, + "learning_rate": 9.98919930750106e-06, + "loss": 0.3839, + "step": 362 + }, + { + "epoch": 0.05, + "grad_norm": 2.3269814132874096, + "learning_rate": 9.98905090032268e-06, + "loss": 0.373, + "step": 363 + }, + { + "epoch": 0.05, + "grad_norm": 2.3179265338941937, + "learning_rate": 9.9889014816109e-06, + "loss": 0.3973, + "step": 364 + }, + { + "epoch": 0.05, + "grad_norm": 2.582258582342007, + "learning_rate": 9.988751051396015e-06, + "loss": 0.3748, + "step": 365 + }, + { + "epoch": 0.05, + "grad_norm": 3.7663563844773003, + "learning_rate": 9.988599609708527e-06, + "loss": 0.3658, + "step": 366 + }, + { + "epoch": 0.05, + "grad_norm": 2.3528035763152064, + "learning_rate": 9.988447156579139e-06, + "loss": 0.395, + "step": 367 + }, + { + "epoch": 0.05, + "grad_norm": 3.12506208801807, + "learning_rate": 9.988293692038762e-06, + "loss": 0.355, + "step": 368 + }, + { + "epoch": 0.05, + "grad_norm": 2.3180047446050924, + "learning_rate": 9.988139216118508e-06, + "loss": 0.3957, + "step": 369 + }, + { + "epoch": 0.05, + "grad_norm": 2.363797145673589, + "learning_rate": 9.987983728849705e-06, + "loss": 0.3651, + "step": 370 + }, + { + "epoch": 0.05, + "grad_norm": 3.567864289600552, + "learning_rate": 9.987827230263867e-06, + "loss": 0.3403, + "step": 371 + }, + { + "epoch": 0.05, + "grad_norm": 2.7579025174238714, + "learning_rate": 9.987669720392734e-06, + "loss": 0.3685, + "step": 372 + }, + { + "epoch": 0.05, + "grad_norm": 2.742819187733749, + "learning_rate": 9.987511199268237e-06, + "loss": 0.3757, + "step": 373 + }, + { + "epoch": 0.05, + "grad_norm": 2.9987928071538463, + "learning_rate": 9.987351666922514e-06, + "loss": 0.3964, + "step": 374 + }, + { + "epoch": 0.05, + "grad_norm": 2.2052866059761302, + "learning_rate": 9.987191123387916e-06, + "loss": 0.4104, + "step": 375 + }, + { + "epoch": 0.05, + "grad_norm": 2.4817814104996967, + "learning_rate": 9.98702956869699e-06, + "loss": 0.4019, + "step": 376 + }, + { + "epoch": 0.05, + "grad_norm": 2.7419154687991534, + "learning_rate": 9.986867002882488e-06, + "loss": 0.3753, + "step": 377 + }, + { + "epoch": 0.05, + "grad_norm": 2.62552754955387, + "learning_rate": 9.986703425977376e-06, + "loss": 0.368, + "step": 378 + }, + { + "epoch": 0.05, + "grad_norm": 2.3254478011613298, + "learning_rate": 9.986538838014817e-06, + "loss": 0.423, + "step": 379 + }, + { + "epoch": 0.05, + "grad_norm": 3.273513647557652, + "learning_rate": 9.986373239028182e-06, + "loss": 0.3976, + "step": 380 + }, + { + "epoch": 0.05, + "grad_norm": 2.895936282967521, + "learning_rate": 9.986206629051045e-06, + "loss": 0.3935, + "step": 381 + }, + { + "epoch": 0.05, + "grad_norm": 3.45481085049066, + "learning_rate": 9.986039008117187e-06, + "loss": 0.385, + "step": 382 + }, + { + "epoch": 0.05, + "grad_norm": 3.523168125924794, + "learning_rate": 9.985870376260594e-06, + "loss": 0.391, + "step": 383 + }, + { + "epoch": 0.05, + "grad_norm": 2.8597559178035494, + "learning_rate": 9.985700733515455e-06, + "loss": 0.3847, + "step": 384 + }, + { + "epoch": 0.05, + "grad_norm": 2.56608520874409, + "learning_rate": 9.985530079916166e-06, + "loss": 0.382, + "step": 385 + }, + { + "epoch": 0.05, + "grad_norm": 3.3062921558356595, + "learning_rate": 9.985358415497326e-06, + "loss": 0.3697, + "step": 386 + }, + { + "epoch": 0.05, + "grad_norm": 2.638409935228937, + "learning_rate": 9.985185740293741e-06, + "loss": 0.3959, + "step": 387 + }, + { + "epoch": 0.05, + "grad_norm": 2.2144344835864587, + "learning_rate": 9.985012054340421e-06, + "loss": 0.3987, + "step": 388 + }, + { + "epoch": 0.05, + "grad_norm": 2.5763999027569073, + "learning_rate": 9.984837357672584e-06, + "loss": 0.4029, + "step": 389 + }, + { + "epoch": 0.05, + "grad_norm": 2.719069785278532, + "learning_rate": 9.984661650325642e-06, + "loss": 0.3586, + "step": 390 + }, + { + "epoch": 0.05, + "grad_norm": 2.8425144698129294, + "learning_rate": 9.984484932335228e-06, + "loss": 0.3602, + "step": 391 + }, + { + "epoch": 0.05, + "grad_norm": 4.652126410114603, + "learning_rate": 9.984307203737166e-06, + "loss": 0.3743, + "step": 392 + }, + { + "epoch": 0.05, + "grad_norm": 2.6024852794041218, + "learning_rate": 9.984128464567493e-06, + "loss": 0.3973, + "step": 393 + }, + { + "epoch": 0.05, + "grad_norm": 2.9025500672455977, + "learning_rate": 9.983948714862449e-06, + "loss": 0.3747, + "step": 394 + }, + { + "epoch": 0.05, + "grad_norm": 3.203628216568939, + "learning_rate": 9.983767954658477e-06, + "loss": 0.3686, + "step": 395 + }, + { + "epoch": 0.06, + "grad_norm": 2.499562446784533, + "learning_rate": 9.983586183992228e-06, + "loss": 0.3723, + "step": 396 + }, + { + "epoch": 0.06, + "grad_norm": 2.298336215112865, + "learning_rate": 9.983403402900554e-06, + "loss": 0.3646, + "step": 397 + }, + { + "epoch": 0.06, + "grad_norm": 2.691867299787411, + "learning_rate": 9.983219611420514e-06, + "loss": 0.3805, + "step": 398 + }, + { + "epoch": 0.06, + "grad_norm": 2.8519468411828703, + "learning_rate": 9.983034809589375e-06, + "loss": 0.3709, + "step": 399 + }, + { + "epoch": 0.06, + "grad_norm": 3.548684517846914, + "learning_rate": 9.982848997444603e-06, + "loss": 0.3676, + "step": 400 + }, + { + "epoch": 0.06, + "grad_norm": 2.415115222890796, + "learning_rate": 9.982662175023871e-06, + "loss": 0.3742, + "step": 401 + }, + { + "epoch": 0.06, + "grad_norm": 7.141037532028862, + "learning_rate": 9.982474342365061e-06, + "loss": 0.3906, + "step": 402 + }, + { + "epoch": 0.06, + "grad_norm": 12.395315619637785, + "learning_rate": 9.98228549950625e-06, + "loss": 0.4095, + "step": 403 + }, + { + "epoch": 0.06, + "grad_norm": 3.429143313310742, + "learning_rate": 9.982095646485732e-06, + "loss": 0.4145, + "step": 404 + }, + { + "epoch": 0.06, + "grad_norm": 2.5543099764666573, + "learning_rate": 9.981904783341998e-06, + "loss": 0.3538, + "step": 405 + }, + { + "epoch": 0.06, + "grad_norm": 2.244399334529943, + "learning_rate": 9.981712910113743e-06, + "loss": 0.3955, + "step": 406 + }, + { + "epoch": 0.06, + "grad_norm": 3.2634200277480176, + "learning_rate": 9.981520026839873e-06, + "loss": 0.3613, + "step": 407 + }, + { + "epoch": 0.06, + "grad_norm": 2.4294166098984493, + "learning_rate": 9.981326133559494e-06, + "loss": 0.3854, + "step": 408 + }, + { + "epoch": 0.06, + "grad_norm": 2.3928530519888964, + "learning_rate": 9.981131230311917e-06, + "loss": 0.3709, + "step": 409 + }, + { + "epoch": 0.06, + "grad_norm": 2.2048772487338426, + "learning_rate": 9.980935317136659e-06, + "loss": 0.383, + "step": 410 + }, + { + "epoch": 0.06, + "grad_norm": 2.252979577012397, + "learning_rate": 9.980738394073444e-06, + "loss": 0.3718, + "step": 411 + }, + { + "epoch": 0.06, + "grad_norm": 3.063149672828551, + "learning_rate": 9.980540461162194e-06, + "loss": 0.371, + "step": 412 + }, + { + "epoch": 0.06, + "grad_norm": 3.4118041729053687, + "learning_rate": 9.980341518443045e-06, + "loss": 0.3626, + "step": 413 + }, + { + "epoch": 0.06, + "grad_norm": 2.6993272104649306, + "learning_rate": 9.980141565956328e-06, + "loss": 0.3874, + "step": 414 + }, + { + "epoch": 0.06, + "grad_norm": 5.14815748802278, + "learning_rate": 9.979940603742584e-06, + "loss": 0.3751, + "step": 415 + }, + { + "epoch": 0.06, + "grad_norm": 2.42964474435871, + "learning_rate": 9.979738631842564e-06, + "loss": 0.3652, + "step": 416 + }, + { + "epoch": 0.06, + "grad_norm": 3.336451848914894, + "learning_rate": 9.979535650297211e-06, + "loss": 0.3861, + "step": 417 + }, + { + "epoch": 0.06, + "grad_norm": 3.0181881078925383, + "learning_rate": 9.979331659147685e-06, + "loss": 0.3683, + "step": 418 + }, + { + "epoch": 0.06, + "grad_norm": 3.9852212148508555, + "learning_rate": 9.97912665843534e-06, + "loss": 0.3821, + "step": 419 + }, + { + "epoch": 0.06, + "grad_norm": 4.154131622652223, + "learning_rate": 9.978920648201743e-06, + "loss": 0.3803, + "step": 420 + }, + { + "epoch": 0.06, + "grad_norm": 2.823803242555759, + "learning_rate": 9.978713628488662e-06, + "loss": 0.3846, + "step": 421 + }, + { + "epoch": 0.06, + "grad_norm": 6.796269941852852, + "learning_rate": 9.978505599338074e-06, + "loss": 0.3832, + "step": 422 + }, + { + "epoch": 0.06, + "grad_norm": 2.4397562760299194, + "learning_rate": 9.97829656079215e-06, + "loss": 0.3962, + "step": 423 + }, + { + "epoch": 0.06, + "grad_norm": 3.2209788400996096, + "learning_rate": 9.97808651289328e-06, + "loss": 0.3823, + "step": 424 + }, + { + "epoch": 0.06, + "grad_norm": 2.8860181735866095, + "learning_rate": 9.977875455684047e-06, + "loss": 0.3745, + "step": 425 + }, + { + "epoch": 0.06, + "grad_norm": 2.9172682645528303, + "learning_rate": 9.977663389207245e-06, + "loss": 0.4054, + "step": 426 + }, + { + "epoch": 0.06, + "grad_norm": 2.394126456967764, + "learning_rate": 9.977450313505868e-06, + "loss": 0.4206, + "step": 427 + }, + { + "epoch": 0.06, + "grad_norm": 2.2819518714189444, + "learning_rate": 9.97723622862312e-06, + "loss": 0.3973, + "step": 428 + }, + { + "epoch": 0.06, + "grad_norm": 4.04257903518687, + "learning_rate": 9.977021134602404e-06, + "loss": 0.3863, + "step": 429 + }, + { + "epoch": 0.06, + "grad_norm": 3.0309197640368937, + "learning_rate": 9.976805031487333e-06, + "loss": 0.396, + "step": 430 + }, + { + "epoch": 0.06, + "grad_norm": 2.5696803158087924, + "learning_rate": 9.97658791932172e-06, + "loss": 0.3705, + "step": 431 + }, + { + "epoch": 0.06, + "grad_norm": 8.551000990450731, + "learning_rate": 9.976369798149588e-06, + "loss": 0.3954, + "step": 432 + }, + { + "epoch": 0.06, + "grad_norm": 4.695338991284906, + "learning_rate": 9.976150668015159e-06, + "loss": 0.3825, + "step": 433 + }, + { + "epoch": 0.06, + "grad_norm": 2.6020117924821182, + "learning_rate": 9.97593052896286e-06, + "loss": 0.3946, + "step": 434 + }, + { + "epoch": 0.06, + "grad_norm": 2.68123131072958, + "learning_rate": 9.975709381037329e-06, + "loss": 0.3526, + "step": 435 + }, + { + "epoch": 0.06, + "grad_norm": 1.984631157513891, + "learning_rate": 9.975487224283398e-06, + "loss": 0.3679, + "step": 436 + }, + { + "epoch": 0.06, + "grad_norm": 2.5574886901725584, + "learning_rate": 9.975264058746115e-06, + "loss": 0.3808, + "step": 437 + }, + { + "epoch": 0.06, + "grad_norm": 2.8223610360733513, + "learning_rate": 9.975039884470723e-06, + "loss": 0.3853, + "step": 438 + }, + { + "epoch": 0.06, + "grad_norm": 2.1964192783751098, + "learning_rate": 9.974814701502675e-06, + "loss": 0.4021, + "step": 439 + }, + { + "epoch": 0.06, + "grad_norm": 2.366419296857818, + "learning_rate": 9.974588509887627e-06, + "loss": 0.3675, + "step": 440 + }, + { + "epoch": 0.06, + "grad_norm": 2.7561308314391524, + "learning_rate": 9.97436130967144e-06, + "loss": 0.3749, + "step": 441 + }, + { + "epoch": 0.06, + "grad_norm": 1.9982266138232563, + "learning_rate": 9.974133100900178e-06, + "loss": 0.3632, + "step": 442 + }, + { + "epoch": 0.06, + "grad_norm": 2.723213837860187, + "learning_rate": 9.973903883620111e-06, + "loss": 0.3825, + "step": 443 + }, + { + "epoch": 0.06, + "grad_norm": 2.5455272537158464, + "learning_rate": 9.973673657877712e-06, + "loss": 0.3717, + "step": 444 + }, + { + "epoch": 0.06, + "grad_norm": 2.0732703164172475, + "learning_rate": 9.973442423719662e-06, + "loss": 0.3683, + "step": 445 + }, + { + "epoch": 0.06, + "grad_norm": 3.379568493941082, + "learning_rate": 9.97321018119284e-06, + "loss": 0.4066, + "step": 446 + }, + { + "epoch": 0.06, + "grad_norm": 2.6177897554915064, + "learning_rate": 9.972976930344338e-06, + "loss": 0.3931, + "step": 447 + }, + { + "epoch": 0.06, + "grad_norm": 2.8784693905227874, + "learning_rate": 9.972742671221444e-06, + "loss": 0.3846, + "step": 448 + }, + { + "epoch": 0.06, + "grad_norm": 2.3197705179751056, + "learning_rate": 9.972507403871656e-06, + "loss": 0.3719, + "step": 449 + }, + { + "epoch": 0.06, + "grad_norm": 3.515545144298594, + "learning_rate": 9.972271128342673e-06, + "loss": 0.3656, + "step": 450 + }, + { + "epoch": 0.06, + "grad_norm": 2.917723040086229, + "learning_rate": 9.972033844682401e-06, + "loss": 0.3991, + "step": 451 + }, + { + "epoch": 0.06, + "grad_norm": 2.5820586003721515, + "learning_rate": 9.97179555293895e-06, + "loss": 0.3895, + "step": 452 + }, + { + "epoch": 0.06, + "grad_norm": 2.387975594657817, + "learning_rate": 9.971556253160634e-06, + "loss": 0.3653, + "step": 453 + }, + { + "epoch": 0.06, + "grad_norm": 2.3537602818987584, + "learning_rate": 9.971315945395971e-06, + "loss": 0.3792, + "step": 454 + }, + { + "epoch": 0.06, + "grad_norm": 3.3134957518374626, + "learning_rate": 9.971074629693682e-06, + "loss": 0.4003, + "step": 455 + }, + { + "epoch": 0.06, + "grad_norm": 3.149841911593156, + "learning_rate": 9.970832306102696e-06, + "loss": 0.3975, + "step": 456 + }, + { + "epoch": 0.06, + "grad_norm": 3.9363643080994386, + "learning_rate": 9.970588974672142e-06, + "loss": 0.3688, + "step": 457 + }, + { + "epoch": 0.06, + "grad_norm": 2.0875421226256297, + "learning_rate": 9.970344635451356e-06, + "loss": 0.3791, + "step": 458 + }, + { + "epoch": 0.06, + "grad_norm": 3.3546889698219324, + "learning_rate": 9.970099288489881e-06, + "loss": 0.3852, + "step": 459 + }, + { + "epoch": 0.06, + "grad_norm": 2.9886379676172314, + "learning_rate": 9.969852933837458e-06, + "loss": 0.3809, + "step": 460 + }, + { + "epoch": 0.06, + "grad_norm": 2.029499050444787, + "learning_rate": 9.969605571544038e-06, + "loss": 0.357, + "step": 461 + }, + { + "epoch": 0.06, + "grad_norm": 2.6990785762216865, + "learning_rate": 9.96935720165977e-06, + "loss": 0.4006, + "step": 462 + }, + { + "epoch": 0.06, + "grad_norm": 2.7950308562082697, + "learning_rate": 9.969107824235016e-06, + "loss": 0.3619, + "step": 463 + }, + { + "epoch": 0.06, + "grad_norm": 2.046040874003261, + "learning_rate": 9.968857439320334e-06, + "loss": 0.3844, + "step": 464 + }, + { + "epoch": 0.06, + "grad_norm": 5.095934463054366, + "learning_rate": 9.96860604696649e-06, + "loss": 0.3847, + "step": 465 + }, + { + "epoch": 0.06, + "grad_norm": 2.702751121284772, + "learning_rate": 9.968353647224457e-06, + "loss": 0.3618, + "step": 466 + }, + { + "epoch": 0.06, + "grad_norm": 2.9082456015200737, + "learning_rate": 9.968100240145406e-06, + "loss": 0.3959, + "step": 467 + }, + { + "epoch": 0.07, + "grad_norm": 2.6903574342218692, + "learning_rate": 9.967845825780717e-06, + "loss": 0.3938, + "step": 468 + }, + { + "epoch": 0.07, + "grad_norm": 2.491454785853338, + "learning_rate": 9.967590404181971e-06, + "loss": 0.3727, + "step": 469 + }, + { + "epoch": 0.07, + "grad_norm": 4.718003618215295, + "learning_rate": 9.967333975400957e-06, + "loss": 0.392, + "step": 470 + }, + { + "epoch": 0.07, + "grad_norm": 2.519863844843847, + "learning_rate": 9.967076539489664e-06, + "loss": 0.3557, + "step": 471 + }, + { + "epoch": 0.07, + "grad_norm": 2.519646585388502, + "learning_rate": 9.966818096500289e-06, + "loss": 0.3558, + "step": 472 + }, + { + "epoch": 0.07, + "grad_norm": 2.2250534353026454, + "learning_rate": 9.966558646485231e-06, + "loss": 0.3544, + "step": 473 + }, + { + "epoch": 0.07, + "grad_norm": 3.404907388716917, + "learning_rate": 9.966298189497094e-06, + "loss": 0.3531, + "step": 474 + }, + { + "epoch": 0.07, + "grad_norm": 2.6659021615419625, + "learning_rate": 9.966036725588686e-06, + "loss": 0.4077, + "step": 475 + }, + { + "epoch": 0.07, + "grad_norm": 3.4159398152346814, + "learning_rate": 9.965774254813017e-06, + "loss": 0.3909, + "step": 476 + }, + { + "epoch": 0.07, + "grad_norm": 2.4428264459998563, + "learning_rate": 9.965510777223304e-06, + "loss": 0.3748, + "step": 477 + }, + { + "epoch": 0.07, + "grad_norm": 3.833694684389677, + "learning_rate": 9.96524629287297e-06, + "loss": 0.38, + "step": 478 + }, + { + "epoch": 0.07, + "grad_norm": 2.849452960791516, + "learning_rate": 9.964980801815636e-06, + "loss": 0.3942, + "step": 479 + }, + { + "epoch": 0.07, + "grad_norm": 3.0356353605518693, + "learning_rate": 9.964714304105133e-06, + "loss": 0.3667, + "step": 480 + }, + { + "epoch": 0.07, + "grad_norm": 3.5794952653751557, + "learning_rate": 9.964446799795488e-06, + "loss": 0.3683, + "step": 481 + }, + { + "epoch": 0.07, + "grad_norm": 2.536534485636776, + "learning_rate": 9.964178288940946e-06, + "loss": 0.3701, + "step": 482 + }, + { + "epoch": 0.07, + "grad_norm": 4.075307372344963, + "learning_rate": 9.963908771595942e-06, + "loss": 0.3754, + "step": 483 + }, + { + "epoch": 0.07, + "grad_norm": 2.891457344326751, + "learning_rate": 9.963638247815125e-06, + "loss": 0.4077, + "step": 484 + }, + { + "epoch": 0.07, + "grad_norm": 2.497289529688001, + "learning_rate": 9.963366717653337e-06, + "loss": 0.3701, + "step": 485 + }, + { + "epoch": 0.07, + "grad_norm": 4.750471081883205, + "learning_rate": 9.96309418116564e-06, + "loss": 0.3959, + "step": 486 + }, + { + "epoch": 0.07, + "grad_norm": 3.363033736919316, + "learning_rate": 9.962820638407284e-06, + "loss": 0.3887, + "step": 487 + }, + { + "epoch": 0.07, + "grad_norm": 3.4507753131389465, + "learning_rate": 9.96254608943373e-06, + "loss": 0.3967, + "step": 488 + }, + { + "epoch": 0.07, + "grad_norm": 3.655774808802044, + "learning_rate": 9.962270534300649e-06, + "loss": 0.3849, + "step": 489 + }, + { + "epoch": 0.07, + "grad_norm": 3.1224988600759898, + "learning_rate": 9.961993973063904e-06, + "loss": 0.361, + "step": 490 + }, + { + "epoch": 0.07, + "grad_norm": 2.6179022170253847, + "learning_rate": 9.961716405779572e-06, + "loss": 0.3649, + "step": 491 + }, + { + "epoch": 0.07, + "grad_norm": 3.0180842898973004, + "learning_rate": 9.961437832503927e-06, + "loss": 0.363, + "step": 492 + }, + { + "epoch": 0.07, + "grad_norm": 2.20392783012301, + "learning_rate": 9.96115825329345e-06, + "loss": 0.3742, + "step": 493 + }, + { + "epoch": 0.07, + "grad_norm": 4.187352196928067, + "learning_rate": 9.96087766820483e-06, + "loss": 0.4186, + "step": 494 + }, + { + "epoch": 0.07, + "grad_norm": 4.782143039519254, + "learning_rate": 9.960596077294951e-06, + "loss": 0.3633, + "step": 495 + }, + { + "epoch": 0.07, + "grad_norm": 2.493715554561522, + "learning_rate": 9.960313480620906e-06, + "loss": 0.3885, + "step": 496 + }, + { + "epoch": 0.07, + "grad_norm": 2.869956104882198, + "learning_rate": 9.960029878239995e-06, + "loss": 0.346, + "step": 497 + }, + { + "epoch": 0.07, + "grad_norm": 2.596048896308056, + "learning_rate": 9.959745270209716e-06, + "loss": 0.3775, + "step": 498 + }, + { + "epoch": 0.07, + "grad_norm": 2.434071879008721, + "learning_rate": 9.959459656587775e-06, + "loss": 0.3914, + "step": 499 + }, + { + "epoch": 0.07, + "grad_norm": 2.9541134877210347, + "learning_rate": 9.959173037432079e-06, + "loss": 0.3583, + "step": 500 + }, + { + "epoch": 0.07, + "grad_norm": 2.768961627380164, + "learning_rate": 9.95888541280074e-06, + "loss": 0.3798, + "step": 501 + }, + { + "epoch": 0.07, + "grad_norm": 2.1196991600931536, + "learning_rate": 9.958596782752075e-06, + "loss": 0.3786, + "step": 502 + }, + { + "epoch": 0.07, + "grad_norm": 2.3554539806139796, + "learning_rate": 9.958307147344605e-06, + "loss": 0.3905, + "step": 503 + }, + { + "epoch": 0.07, + "grad_norm": 2.3168487421184056, + "learning_rate": 9.958016506637052e-06, + "loss": 0.3762, + "step": 504 + }, + { + "epoch": 0.07, + "grad_norm": 3.2951646669476857, + "learning_rate": 9.957724860688343e-06, + "loss": 0.3675, + "step": 505 + }, + { + "epoch": 0.07, + "grad_norm": 2.1282639765637823, + "learning_rate": 9.957432209557613e-06, + "loss": 0.3505, + "step": 506 + }, + { + "epoch": 0.07, + "grad_norm": 2.701106793214317, + "learning_rate": 9.957138553304194e-06, + "loss": 0.3601, + "step": 507 + }, + { + "epoch": 0.07, + "grad_norm": 2.4383854791164437, + "learning_rate": 9.956843891987624e-06, + "loss": 0.3725, + "step": 508 + }, + { + "epoch": 0.07, + "grad_norm": 2.2279895598668196, + "learning_rate": 9.956548225667648e-06, + "loss": 0.3745, + "step": 509 + }, + { + "epoch": 0.07, + "grad_norm": 2.0459387477534277, + "learning_rate": 9.956251554404214e-06, + "loss": 0.3863, + "step": 510 + }, + { + "epoch": 0.07, + "grad_norm": 3.6484727484461272, + "learning_rate": 9.955953878257469e-06, + "loss": 0.3936, + "step": 511 + }, + { + "epoch": 0.07, + "grad_norm": 2.994073686952973, + "learning_rate": 9.955655197287769e-06, + "loss": 0.4, + "step": 512 + }, + { + "epoch": 0.07, + "grad_norm": 2.2005625466027636, + "learning_rate": 9.95535551155567e-06, + "loss": 0.3538, + "step": 513 + }, + { + "epoch": 0.07, + "grad_norm": 2.340142266979258, + "learning_rate": 9.955054821121936e-06, + "loss": 0.3961, + "step": 514 + }, + { + "epoch": 0.07, + "grad_norm": 2.638576865501452, + "learning_rate": 9.95475312604753e-06, + "loss": 0.3732, + "step": 515 + }, + { + "epoch": 0.07, + "grad_norm": 2.1791132920204377, + "learning_rate": 9.954450426393622e-06, + "loss": 0.3761, + "step": 516 + }, + { + "epoch": 0.07, + "grad_norm": 3.2554417073618382, + "learning_rate": 9.954146722221585e-06, + "loss": 0.3852, + "step": 517 + }, + { + "epoch": 0.07, + "grad_norm": 2.169447045038341, + "learning_rate": 9.953842013592994e-06, + "loss": 0.3695, + "step": 518 + }, + { + "epoch": 0.07, + "grad_norm": 3.047937965084294, + "learning_rate": 9.95353630056963e-06, + "loss": 0.357, + "step": 519 + }, + { + "epoch": 0.07, + "grad_norm": 3.1975846187214065, + "learning_rate": 9.953229583213474e-06, + "loss": 0.3812, + "step": 520 + }, + { + "epoch": 0.07, + "grad_norm": 5.891138470106585, + "learning_rate": 9.952921861586718e-06, + "loss": 0.3813, + "step": 521 + }, + { + "epoch": 0.07, + "grad_norm": 3.1383770767366475, + "learning_rate": 9.952613135751749e-06, + "loss": 0.3998, + "step": 522 + }, + { + "epoch": 0.07, + "grad_norm": 5.689422815170085, + "learning_rate": 9.952303405771162e-06, + "loss": 0.354, + "step": 523 + }, + { + "epoch": 0.07, + "grad_norm": 2.6189250042523975, + "learning_rate": 9.951992671707753e-06, + "loss": 0.3868, + "step": 524 + }, + { + "epoch": 0.07, + "grad_norm": 2.4359055257685394, + "learning_rate": 9.951680933624528e-06, + "loss": 0.39, + "step": 525 + }, + { + "epoch": 0.07, + "grad_norm": 3.054610456734872, + "learning_rate": 9.951368191584689e-06, + "loss": 0.3901, + "step": 526 + }, + { + "epoch": 0.07, + "grad_norm": 2.58983371666581, + "learning_rate": 9.951054445651646e-06, + "loss": 0.3695, + "step": 527 + }, + { + "epoch": 0.07, + "grad_norm": 2.286300434821698, + "learning_rate": 9.950739695889011e-06, + "loss": 0.3677, + "step": 528 + }, + { + "epoch": 0.07, + "grad_norm": 6.297848326932995, + "learning_rate": 9.950423942360598e-06, + "loss": 0.3593, + "step": 529 + }, + { + "epoch": 0.07, + "grad_norm": 2.8026639354724447, + "learning_rate": 9.950107185130429e-06, + "loss": 0.3849, + "step": 530 + }, + { + "epoch": 0.07, + "grad_norm": 2.518680239489154, + "learning_rate": 9.949789424262726e-06, + "loss": 0.3876, + "step": 531 + }, + { + "epoch": 0.07, + "grad_norm": 2.664514516226861, + "learning_rate": 9.949470659821913e-06, + "loss": 0.3573, + "step": 532 + }, + { + "epoch": 0.07, + "grad_norm": 2.5086544970824436, + "learning_rate": 9.949150891872621e-06, + "loss": 0.3943, + "step": 533 + }, + { + "epoch": 0.07, + "grad_norm": 5.827944002771721, + "learning_rate": 9.948830120479685e-06, + "loss": 0.3773, + "step": 534 + }, + { + "epoch": 0.07, + "grad_norm": 2.887056832483973, + "learning_rate": 9.948508345708139e-06, + "loss": 0.3741, + "step": 535 + }, + { + "epoch": 0.07, + "grad_norm": 2.1925273484032557, + "learning_rate": 9.948185567623224e-06, + "loss": 0.3861, + "step": 536 + }, + { + "epoch": 0.07, + "grad_norm": 4.294689888666271, + "learning_rate": 9.947861786290385e-06, + "loss": 0.3763, + "step": 537 + }, + { + "epoch": 0.07, + "grad_norm": 7.135729719747838, + "learning_rate": 9.947537001775265e-06, + "loss": 0.354, + "step": 538 + }, + { + "epoch": 0.07, + "grad_norm": 3.442132238987187, + "learning_rate": 9.947211214143719e-06, + "loss": 0.3828, + "step": 539 + }, + { + "epoch": 0.08, + "grad_norm": 2.140669323200884, + "learning_rate": 9.946884423461798e-06, + "loss": 0.3694, + "step": 540 + }, + { + "epoch": 0.08, + "grad_norm": 2.1341162616442815, + "learning_rate": 9.946556629795758e-06, + "loss": 0.3796, + "step": 541 + }, + { + "epoch": 0.08, + "grad_norm": 2.19634759462649, + "learning_rate": 9.946227833212063e-06, + "loss": 0.3815, + "step": 542 + }, + { + "epoch": 0.08, + "grad_norm": 2.4091857097053584, + "learning_rate": 9.945898033777372e-06, + "loss": 0.3973, + "step": 543 + }, + { + "epoch": 0.08, + "grad_norm": 2.89627087407256, + "learning_rate": 9.945567231558557e-06, + "loss": 0.3727, + "step": 544 + }, + { + "epoch": 0.08, + "grad_norm": 5.086414372719643, + "learning_rate": 9.945235426622686e-06, + "loss": 0.3739, + "step": 545 + }, + { + "epoch": 0.08, + "grad_norm": 2.104738501890542, + "learning_rate": 9.944902619037032e-06, + "loss": 0.3638, + "step": 546 + }, + { + "epoch": 0.08, + "grad_norm": 1.8676244179378518, + "learning_rate": 9.944568808869072e-06, + "loss": 0.3689, + "step": 547 + }, + { + "epoch": 0.08, + "grad_norm": 2.6657430068900734, + "learning_rate": 9.944233996186488e-06, + "loss": 0.3557, + "step": 548 + }, + { + "epoch": 0.08, + "grad_norm": 5.015277423006405, + "learning_rate": 9.943898181057161e-06, + "loss": 0.3828, + "step": 549 + }, + { + "epoch": 0.08, + "grad_norm": 3.026725000056818, + "learning_rate": 9.94356136354918e-06, + "loss": 0.3933, + "step": 550 + }, + { + "epoch": 0.08, + "grad_norm": 2.200753853766776, + "learning_rate": 9.943223543730833e-06, + "loss": 0.3837, + "step": 551 + }, + { + "epoch": 0.08, + "grad_norm": 3.285447028379861, + "learning_rate": 9.942884721670616e-06, + "loss": 0.3693, + "step": 552 + }, + { + "epoch": 0.08, + "grad_norm": 2.8113880593051555, + "learning_rate": 9.942544897437222e-06, + "loss": 0.3936, + "step": 553 + }, + { + "epoch": 0.08, + "grad_norm": 2.2139106496266114, + "learning_rate": 9.942204071099553e-06, + "loss": 0.3566, + "step": 554 + }, + { + "epoch": 0.08, + "grad_norm": 2.4768839753568703, + "learning_rate": 9.941862242726712e-06, + "loss": 0.3973, + "step": 555 + }, + { + "epoch": 0.08, + "grad_norm": 3.5462491577892954, + "learning_rate": 9.941519412388002e-06, + "loss": 0.3826, + "step": 556 + }, + { + "epoch": 0.08, + "grad_norm": 2.290747555546339, + "learning_rate": 9.941175580152936e-06, + "loss": 0.355, + "step": 557 + }, + { + "epoch": 0.08, + "grad_norm": 3.1001760767758877, + "learning_rate": 9.940830746091223e-06, + "loss": 0.3946, + "step": 558 + }, + { + "epoch": 0.08, + "grad_norm": 2.4487959343497554, + "learning_rate": 9.94048491027278e-06, + "loss": 0.3696, + "step": 559 + }, + { + "epoch": 0.08, + "grad_norm": 2.374525331533967, + "learning_rate": 9.940138072767724e-06, + "loss": 0.3846, + "step": 560 + }, + { + "epoch": 0.08, + "grad_norm": 3.0877855412453505, + "learning_rate": 9.939790233646378e-06, + "loss": 0.3539, + "step": 561 + }, + { + "epoch": 0.08, + "grad_norm": 3.261117263461534, + "learning_rate": 9.939441392979268e-06, + "loss": 0.3604, + "step": 562 + }, + { + "epoch": 0.08, + "grad_norm": 2.262165757531711, + "learning_rate": 9.939091550837118e-06, + "loss": 0.3894, + "step": 563 + }, + { + "epoch": 0.08, + "grad_norm": 2.7340864914925262, + "learning_rate": 9.93874070729086e-06, + "loss": 0.3631, + "step": 564 + }, + { + "epoch": 0.08, + "grad_norm": 2.1749003545054353, + "learning_rate": 9.93838886241163e-06, + "loss": 0.3747, + "step": 565 + }, + { + "epoch": 0.08, + "grad_norm": 18.3825802922219, + "learning_rate": 9.938036016270763e-06, + "loss": 0.38, + "step": 566 + }, + { + "epoch": 0.08, + "grad_norm": 5.224456467695394, + "learning_rate": 9.937682168939797e-06, + "loss": 0.3812, + "step": 567 + }, + { + "epoch": 0.08, + "grad_norm": 2.5915503429499824, + "learning_rate": 9.937327320490478e-06, + "loss": 0.3816, + "step": 568 + }, + { + "epoch": 0.08, + "grad_norm": 2.3922340577256587, + "learning_rate": 9.936971470994751e-06, + "loss": 0.3536, + "step": 569 + }, + { + "epoch": 0.08, + "grad_norm": 2.5048389644986058, + "learning_rate": 9.936614620524764e-06, + "loss": 0.3882, + "step": 570 + }, + { + "epoch": 0.08, + "grad_norm": 2.201881556037627, + "learning_rate": 9.936256769152869e-06, + "loss": 0.3669, + "step": 571 + }, + { + "epoch": 0.08, + "grad_norm": 2.9036274080885964, + "learning_rate": 9.93589791695162e-06, + "loss": 0.3683, + "step": 572 + }, + { + "epoch": 0.08, + "grad_norm": 2.374856200923602, + "learning_rate": 9.935538063993775e-06, + "loss": 0.4083, + "step": 573 + }, + { + "epoch": 0.08, + "grad_norm": 2.321084443673149, + "learning_rate": 9.935177210352294e-06, + "loss": 0.3812, + "step": 574 + }, + { + "epoch": 0.08, + "grad_norm": 2.440757544368634, + "learning_rate": 9.934815356100342e-06, + "loss": 0.3878, + "step": 575 + }, + { + "epoch": 0.08, + "grad_norm": 2.4211710940719864, + "learning_rate": 9.934452501311283e-06, + "loss": 0.3751, + "step": 576 + }, + { + "epoch": 0.08, + "grad_norm": 2.6143146970721585, + "learning_rate": 9.934088646058686e-06, + "loss": 0.3678, + "step": 577 + }, + { + "epoch": 0.08, + "grad_norm": 2.4462215311215916, + "learning_rate": 9.933723790416326e-06, + "loss": 0.3812, + "step": 578 + }, + { + "epoch": 0.08, + "grad_norm": 3.722956011925295, + "learning_rate": 9.933357934458173e-06, + "loss": 0.3775, + "step": 579 + }, + { + "epoch": 0.08, + "grad_norm": 3.6525910522877076, + "learning_rate": 9.932991078258409e-06, + "loss": 0.3647, + "step": 580 + }, + { + "epoch": 0.08, + "grad_norm": 2.2227516492171056, + "learning_rate": 9.932623221891412e-06, + "loss": 0.374, + "step": 581 + }, + { + "epoch": 0.08, + "grad_norm": 2.9427567372917416, + "learning_rate": 9.932254365431766e-06, + "loss": 0.3899, + "step": 582 + }, + { + "epoch": 0.08, + "grad_norm": 2.069726295744375, + "learning_rate": 9.931884508954254e-06, + "loss": 0.3627, + "step": 583 + }, + { + "epoch": 0.08, + "grad_norm": 2.2442048913080437, + "learning_rate": 9.931513652533871e-06, + "loss": 0.3695, + "step": 584 + }, + { + "epoch": 0.08, + "grad_norm": 2.3469882302327796, + "learning_rate": 9.931141796245803e-06, + "loss": 0.3612, + "step": 585 + }, + { + "epoch": 0.08, + "grad_norm": 2.2590035175450405, + "learning_rate": 9.930768940165445e-06, + "loss": 0.3816, + "step": 586 + }, + { + "epoch": 0.08, + "grad_norm": 2.4171615915832656, + "learning_rate": 9.930395084368395e-06, + "loss": 0.405, + "step": 587 + }, + { + "epoch": 0.08, + "grad_norm": 1.994396739920708, + "learning_rate": 9.930020228930452e-06, + "loss": 0.3649, + "step": 588 + }, + { + "epoch": 0.08, + "grad_norm": 1.9636005082734578, + "learning_rate": 9.929644373927617e-06, + "loss": 0.3932, + "step": 589 + }, + { + "epoch": 0.08, + "grad_norm": 2.3226837090734205, + "learning_rate": 9.929267519436097e-06, + "loss": 0.3941, + "step": 590 + }, + { + "epoch": 0.08, + "grad_norm": 7.158287817530961, + "learning_rate": 9.928889665532298e-06, + "loss": 0.3838, + "step": 591 + }, + { + "epoch": 0.08, + "grad_norm": 2.1546126633562435, + "learning_rate": 9.928510812292832e-06, + "loss": 0.3408, + "step": 592 + }, + { + "epoch": 0.08, + "grad_norm": 2.7436223822094874, + "learning_rate": 9.92813095979451e-06, + "loss": 0.3847, + "step": 593 + }, + { + "epoch": 0.08, + "grad_norm": 2.9378229136907366, + "learning_rate": 9.927750108114347e-06, + "loss": 0.3834, + "step": 594 + }, + { + "epoch": 0.08, + "grad_norm": 2.1075262406563535, + "learning_rate": 9.927368257329561e-06, + "loss": 0.3703, + "step": 595 + }, + { + "epoch": 0.08, + "grad_norm": 2.101225883382935, + "learning_rate": 9.926985407517574e-06, + "loss": 0.3585, + "step": 596 + }, + { + "epoch": 0.08, + "grad_norm": 3.125151315486508, + "learning_rate": 9.926601558756009e-06, + "loss": 0.3979, + "step": 597 + }, + { + "epoch": 0.08, + "grad_norm": 2.5558557274389377, + "learning_rate": 9.92621671112269e-06, + "loss": 0.3526, + "step": 598 + }, + { + "epoch": 0.08, + "grad_norm": 2.1329814354181726, + "learning_rate": 9.925830864695645e-06, + "loss": 0.3802, + "step": 599 + }, + { + "epoch": 0.08, + "grad_norm": 3.299367783954614, + "learning_rate": 9.925444019553107e-06, + "loss": 0.356, + "step": 600 + }, + { + "epoch": 0.08, + "grad_norm": 2.9373482737710073, + "learning_rate": 9.925056175773506e-06, + "loss": 0.3764, + "step": 601 + }, + { + "epoch": 0.08, + "grad_norm": 3.3015765850722585, + "learning_rate": 9.92466733343548e-06, + "loss": 0.3965, + "step": 602 + }, + { + "epoch": 0.08, + "grad_norm": 2.6615059390855618, + "learning_rate": 9.924277492617868e-06, + "loss": 0.3475, + "step": 603 + }, + { + "epoch": 0.08, + "grad_norm": 3.143827175495755, + "learning_rate": 9.923886653399706e-06, + "loss": 0.3971, + "step": 604 + }, + { + "epoch": 0.08, + "grad_norm": 2.3212762567280287, + "learning_rate": 9.923494815860242e-06, + "loss": 0.358, + "step": 605 + }, + { + "epoch": 0.08, + "grad_norm": 3.166420245810776, + "learning_rate": 9.923101980078917e-06, + "loss": 0.393, + "step": 606 + }, + { + "epoch": 0.08, + "grad_norm": 2.9207736323212012, + "learning_rate": 9.922708146135382e-06, + "loss": 0.4124, + "step": 607 + }, + { + "epoch": 0.08, + "grad_norm": 3.34083471199921, + "learning_rate": 9.922313314109485e-06, + "loss": 0.3671, + "step": 608 + }, + { + "epoch": 0.08, + "grad_norm": 2.8466084366380073, + "learning_rate": 9.92191748408128e-06, + "loss": 0.3724, + "step": 609 + }, + { + "epoch": 0.08, + "grad_norm": 2.982758151391327, + "learning_rate": 9.921520656131021e-06, + "loss": 0.3761, + "step": 610 + }, + { + "epoch": 0.08, + "grad_norm": 2.933836562843807, + "learning_rate": 9.921122830339165e-06, + "loss": 0.4002, + "step": 611 + }, + { + "epoch": 0.09, + "grad_norm": 3.3375648322973186, + "learning_rate": 9.920724006786372e-06, + "loss": 0.3665, + "step": 612 + }, + { + "epoch": 0.09, + "grad_norm": 2.858983515874676, + "learning_rate": 9.920324185553504e-06, + "loss": 0.3864, + "step": 613 + }, + { + "epoch": 0.09, + "grad_norm": 3.3403232145122086, + "learning_rate": 9.919923366721623e-06, + "loss": 0.3901, + "step": 614 + }, + { + "epoch": 0.09, + "grad_norm": 2.006163004537021, + "learning_rate": 9.919521550371998e-06, + "loss": 0.3883, + "step": 615 + }, + { + "epoch": 0.09, + "grad_norm": 2.1809715314515605, + "learning_rate": 9.919118736586096e-06, + "loss": 0.391, + "step": 616 + }, + { + "epoch": 0.09, + "grad_norm": 2.0965514912168177, + "learning_rate": 9.918714925445588e-06, + "loss": 0.3588, + "step": 617 + }, + { + "epoch": 0.09, + "grad_norm": 2.5628866922083127, + "learning_rate": 9.918310117032345e-06, + "loss": 0.4007, + "step": 618 + }, + { + "epoch": 0.09, + "grad_norm": 2.0211148749101664, + "learning_rate": 9.917904311428446e-06, + "loss": 0.3791, + "step": 619 + }, + { + "epoch": 0.09, + "grad_norm": 2.7518938909605906, + "learning_rate": 9.917497508716165e-06, + "loss": 0.3765, + "step": 620 + }, + { + "epoch": 0.09, + "grad_norm": 2.2129579933626924, + "learning_rate": 9.917089708977985e-06, + "loss": 0.3783, + "step": 621 + }, + { + "epoch": 0.09, + "grad_norm": 2.4060751587101143, + "learning_rate": 9.916680912296585e-06, + "loss": 0.4076, + "step": 622 + }, + { + "epoch": 0.09, + "grad_norm": 2.0904207204185448, + "learning_rate": 9.916271118754849e-06, + "loss": 0.3577, + "step": 623 + }, + { + "epoch": 0.09, + "grad_norm": 2.455081810991258, + "learning_rate": 9.915860328435863e-06, + "loss": 0.3829, + "step": 624 + }, + { + "epoch": 0.09, + "grad_norm": 9.35658828592991, + "learning_rate": 9.915448541422916e-06, + "loss": 0.3593, + "step": 625 + }, + { + "epoch": 0.09, + "grad_norm": 3.1682687491025314, + "learning_rate": 9.915035757799498e-06, + "loss": 0.3627, + "step": 626 + }, + { + "epoch": 0.09, + "grad_norm": 2.274590836979095, + "learning_rate": 9.914621977649301e-06, + "loss": 0.3968, + "step": 627 + }, + { + "epoch": 0.09, + "grad_norm": 2.181232612141991, + "learning_rate": 9.914207201056218e-06, + "loss": 0.3768, + "step": 628 + }, + { + "epoch": 0.09, + "grad_norm": 2.5263628806102925, + "learning_rate": 9.913791428104347e-06, + "loss": 0.3831, + "step": 629 + }, + { + "epoch": 0.09, + "grad_norm": 2.6044151325383917, + "learning_rate": 9.913374658877983e-06, + "loss": 0.3514, + "step": 630 + }, + { + "epoch": 0.09, + "grad_norm": 2.421800026180831, + "learning_rate": 9.912956893461631e-06, + "loss": 0.3707, + "step": 631 + }, + { + "epoch": 0.09, + "grad_norm": 2.546745221486131, + "learning_rate": 9.91253813193999e-06, + "loss": 0.3605, + "step": 632 + }, + { + "epoch": 0.09, + "grad_norm": 2.066870108629615, + "learning_rate": 9.912118374397965e-06, + "loss": 0.3709, + "step": 633 + }, + { + "epoch": 0.09, + "grad_norm": 3.2199500834910957, + "learning_rate": 9.911697620920662e-06, + "loss": 0.3512, + "step": 634 + }, + { + "epoch": 0.09, + "grad_norm": 2.275658531724209, + "learning_rate": 9.911275871593389e-06, + "loss": 0.3767, + "step": 635 + }, + { + "epoch": 0.09, + "grad_norm": 2.6416408016639052, + "learning_rate": 9.910853126501657e-06, + "loss": 0.406, + "step": 636 + }, + { + "epoch": 0.09, + "grad_norm": 2.4232742771878946, + "learning_rate": 9.910429385731174e-06, + "loss": 0.3619, + "step": 637 + }, + { + "epoch": 0.09, + "grad_norm": 2.9687414068573004, + "learning_rate": 9.91000464936786e-06, + "loss": 0.3654, + "step": 638 + }, + { + "epoch": 0.09, + "grad_norm": 2.680480721642869, + "learning_rate": 9.909578917497827e-06, + "loss": 0.3887, + "step": 639 + }, + { + "epoch": 0.09, + "grad_norm": 2.5366071906708174, + "learning_rate": 9.909152190207391e-06, + "loss": 0.3713, + "step": 640 + }, + { + "epoch": 0.09, + "grad_norm": 2.023742605484717, + "learning_rate": 9.908724467583076e-06, + "loss": 0.3783, + "step": 641 + }, + { + "epoch": 0.09, + "grad_norm": 2.1362566469120225, + "learning_rate": 9.908295749711597e-06, + "loss": 0.3684, + "step": 642 + }, + { + "epoch": 0.09, + "grad_norm": 2.1789166395570967, + "learning_rate": 9.90786603667988e-06, + "loss": 0.3489, + "step": 643 + }, + { + "epoch": 0.09, + "grad_norm": 2.385613697705654, + "learning_rate": 9.90743532857505e-06, + "loss": 0.3842, + "step": 644 + }, + { + "epoch": 0.09, + "grad_norm": 2.6128443238161374, + "learning_rate": 9.907003625484433e-06, + "loss": 0.3548, + "step": 645 + }, + { + "epoch": 0.09, + "grad_norm": 2.172739171265786, + "learning_rate": 9.906570927495555e-06, + "loss": 0.3811, + "step": 646 + }, + { + "epoch": 0.09, + "grad_norm": 2.1323602271470974, + "learning_rate": 9.906137234696151e-06, + "loss": 0.3915, + "step": 647 + }, + { + "epoch": 0.09, + "grad_norm": 2.099216579595944, + "learning_rate": 9.905702547174147e-06, + "loss": 0.3927, + "step": 648 + }, + { + "epoch": 0.09, + "grad_norm": 3.2861034183419022, + "learning_rate": 9.905266865017678e-06, + "loss": 0.3775, + "step": 649 + }, + { + "epoch": 0.09, + "grad_norm": 2.8995870120786673, + "learning_rate": 9.90483018831508e-06, + "loss": 0.3718, + "step": 650 + }, + { + "epoch": 0.09, + "grad_norm": 3.721556857854195, + "learning_rate": 9.90439251715489e-06, + "loss": 0.3821, + "step": 651 + }, + { + "epoch": 0.09, + "grad_norm": 2.4691569396003765, + "learning_rate": 9.903953851625843e-06, + "loss": 0.3817, + "step": 652 + }, + { + "epoch": 0.09, + "grad_norm": 12.877622797632725, + "learning_rate": 9.903514191816882e-06, + "loss": 0.3823, + "step": 653 + }, + { + "epoch": 0.09, + "grad_norm": 2.244482184291508, + "learning_rate": 9.903073537817146e-06, + "loss": 0.3739, + "step": 654 + }, + { + "epoch": 0.09, + "grad_norm": 3.179762895043577, + "learning_rate": 9.902631889715976e-06, + "loss": 0.3793, + "step": 655 + }, + { + "epoch": 0.09, + "grad_norm": 2.7400424399822825, + "learning_rate": 9.902189247602923e-06, + "loss": 0.3862, + "step": 656 + }, + { + "epoch": 0.09, + "grad_norm": 4.297228064695454, + "learning_rate": 9.901745611567727e-06, + "loss": 0.3743, + "step": 657 + }, + { + "epoch": 0.09, + "grad_norm": 2.4865990962142535, + "learning_rate": 9.901300981700338e-06, + "loss": 0.4014, + "step": 658 + }, + { + "epoch": 0.09, + "grad_norm": 3.6105739355744952, + "learning_rate": 9.900855358090906e-06, + "loss": 0.4196, + "step": 659 + }, + { + "epoch": 0.09, + "grad_norm": 2.216883686134313, + "learning_rate": 9.90040874082978e-06, + "loss": 0.3466, + "step": 660 + }, + { + "epoch": 0.09, + "grad_norm": 2.6744791379311446, + "learning_rate": 9.899961130007511e-06, + "loss": 0.3825, + "step": 661 + }, + { + "epoch": 0.09, + "grad_norm": 11.37996689235235, + "learning_rate": 9.899512525714855e-06, + "loss": 0.3778, + "step": 662 + }, + { + "epoch": 0.09, + "grad_norm": 2.4063868648370423, + "learning_rate": 9.899062928042764e-06, + "loss": 0.3911, + "step": 663 + }, + { + "epoch": 0.09, + "grad_norm": 2.7084354047942654, + "learning_rate": 9.898612337082396e-06, + "loss": 0.404, + "step": 664 + }, + { + "epoch": 0.09, + "grad_norm": 2.9146928819938216, + "learning_rate": 9.898160752925109e-06, + "loss": 0.336, + "step": 665 + }, + { + "epoch": 0.09, + "grad_norm": 4.783189156790695, + "learning_rate": 9.89770817566246e-06, + "loss": 0.3904, + "step": 666 + }, + { + "epoch": 0.09, + "grad_norm": 2.5139807075062848, + "learning_rate": 9.897254605386211e-06, + "loss": 0.3724, + "step": 667 + }, + { + "epoch": 0.09, + "grad_norm": 2.1770587424275814, + "learning_rate": 9.896800042188324e-06, + "loss": 0.3696, + "step": 668 + }, + { + "epoch": 0.09, + "grad_norm": 2.1770633108912825, + "learning_rate": 9.89634448616096e-06, + "loss": 0.3766, + "step": 669 + }, + { + "epoch": 0.09, + "grad_norm": 2.7430293796453333, + "learning_rate": 9.895887937396484e-06, + "loss": 0.4038, + "step": 670 + }, + { + "epoch": 0.09, + "grad_norm": 2.5694927277741804, + "learning_rate": 9.895430395987461e-06, + "loss": 0.386, + "step": 671 + }, + { + "epoch": 0.09, + "grad_norm": 2.3270472584622173, + "learning_rate": 9.894971862026662e-06, + "loss": 0.3944, + "step": 672 + }, + { + "epoch": 0.09, + "grad_norm": 2.2572207161114726, + "learning_rate": 9.894512335607052e-06, + "loss": 0.3724, + "step": 673 + }, + { + "epoch": 0.09, + "grad_norm": 2.822295532476373, + "learning_rate": 9.894051816821798e-06, + "loss": 0.3661, + "step": 674 + }, + { + "epoch": 0.09, + "grad_norm": 2.879487501383423, + "learning_rate": 9.893590305764273e-06, + "loss": 0.3905, + "step": 675 + }, + { + "epoch": 0.09, + "grad_norm": 2.4434019970930665, + "learning_rate": 9.893127802528048e-06, + "loss": 0.3949, + "step": 676 + }, + { + "epoch": 0.09, + "grad_norm": 3.401003931365817, + "learning_rate": 9.8926643072069e-06, + "loss": 0.3682, + "step": 677 + }, + { + "epoch": 0.09, + "grad_norm": 5.179765733516757, + "learning_rate": 9.892199819894795e-06, + "loss": 0.3726, + "step": 678 + }, + { + "epoch": 0.09, + "grad_norm": 2.0876124373787484, + "learning_rate": 9.891734340685913e-06, + "loss": 0.3686, + "step": 679 + }, + { + "epoch": 0.09, + "grad_norm": 2.6865881029161893, + "learning_rate": 9.89126786967463e-06, + "loss": 0.3802, + "step": 680 + }, + { + "epoch": 0.09, + "grad_norm": 3.0165587832159737, + "learning_rate": 9.890800406955523e-06, + "loss": 0.3767, + "step": 681 + }, + { + "epoch": 0.09, + "grad_norm": 2.5785071237453043, + "learning_rate": 9.890331952623369e-06, + "loss": 0.4009, + "step": 682 + }, + { + "epoch": 0.09, + "grad_norm": 2.655113113316033, + "learning_rate": 9.88986250677315e-06, + "loss": 0.3832, + "step": 683 + }, + { + "epoch": 0.1, + "grad_norm": 3.093904468149338, + "learning_rate": 9.889392069500046e-06, + "loss": 0.3804, + "step": 684 + }, + { + "epoch": 0.1, + "grad_norm": 2.0103378534872607, + "learning_rate": 9.888920640899435e-06, + "loss": 0.37, + "step": 685 + }, + { + "epoch": 0.1, + "grad_norm": 2.211695341426572, + "learning_rate": 9.888448221066905e-06, + "loss": 0.3655, + "step": 686 + }, + { + "epoch": 0.1, + "grad_norm": 2.278574553567175, + "learning_rate": 9.887974810098233e-06, + "loss": 0.3786, + "step": 687 + }, + { + "epoch": 0.1, + "grad_norm": 2.31303104225155, + "learning_rate": 9.88750040808941e-06, + "loss": 0.3665, + "step": 688 + }, + { + "epoch": 0.1, + "grad_norm": 5.956518299079749, + "learning_rate": 9.887025015136616e-06, + "loss": 0.3803, + "step": 689 + }, + { + "epoch": 0.1, + "grad_norm": 2.3555085893035157, + "learning_rate": 9.886548631336242e-06, + "loss": 0.3622, + "step": 690 + }, + { + "epoch": 0.1, + "grad_norm": 2.298594367771496, + "learning_rate": 9.886071256784871e-06, + "loss": 0.3717, + "step": 691 + }, + { + "epoch": 0.1, + "grad_norm": 3.204867955701134, + "learning_rate": 9.885592891579291e-06, + "loss": 0.3703, + "step": 692 + }, + { + "epoch": 0.1, + "grad_norm": 2.2322856811825615, + "learning_rate": 9.885113535816495e-06, + "loss": 0.3483, + "step": 693 + }, + { + "epoch": 0.1, + "grad_norm": 4.45613116207187, + "learning_rate": 9.884633189593669e-06, + "loss": 0.4048, + "step": 694 + }, + { + "epoch": 0.1, + "grad_norm": 2.6894255289288225, + "learning_rate": 9.884151853008204e-06, + "loss": 0.387, + "step": 695 + }, + { + "epoch": 0.1, + "grad_norm": 2.2236680916770264, + "learning_rate": 9.883669526157694e-06, + "loss": 0.3976, + "step": 696 + }, + { + "epoch": 0.1, + "grad_norm": 2.5353809399932348, + "learning_rate": 9.883186209139928e-06, + "loss": 0.362, + "step": 697 + }, + { + "epoch": 0.1, + "grad_norm": 2.1060959946933764, + "learning_rate": 9.8827019020529e-06, + "loss": 0.3727, + "step": 698 + }, + { + "epoch": 0.1, + "grad_norm": 2.25860297654917, + "learning_rate": 9.882216604994803e-06, + "loss": 0.4248, + "step": 699 + }, + { + "epoch": 0.1, + "grad_norm": 2.0996383339494487, + "learning_rate": 9.88173031806403e-06, + "loss": 0.3498, + "step": 700 + }, + { + "epoch": 0.1, + "grad_norm": 4.219407945227693, + "learning_rate": 9.881243041359181e-06, + "loss": 0.4004, + "step": 701 + }, + { + "epoch": 0.1, + "grad_norm": 4.026599651442923, + "learning_rate": 9.880754774979047e-06, + "loss": 0.3773, + "step": 702 + }, + { + "epoch": 0.1, + "grad_norm": 2.295477541981505, + "learning_rate": 9.880265519022627e-06, + "loss": 0.3584, + "step": 703 + }, + { + "epoch": 0.1, + "grad_norm": 4.3860476807291295, + "learning_rate": 9.879775273589116e-06, + "loss": 0.3868, + "step": 704 + }, + { + "epoch": 0.1, + "grad_norm": 2.44261318742463, + "learning_rate": 9.879284038777912e-06, + "loss": 0.3517, + "step": 705 + }, + { + "epoch": 0.1, + "grad_norm": 2.6377371832748326, + "learning_rate": 9.878791814688613e-06, + "loss": 0.3626, + "step": 706 + }, + { + "epoch": 0.1, + "grad_norm": 3.36093832278382, + "learning_rate": 9.878298601421019e-06, + "loss": 0.3759, + "step": 707 + }, + { + "epoch": 0.1, + "grad_norm": 2.241570663951163, + "learning_rate": 9.87780439907513e-06, + "loss": 0.3702, + "step": 708 + }, + { + "epoch": 0.1, + "grad_norm": 2.2568711928775316, + "learning_rate": 9.877309207751142e-06, + "loss": 0.3885, + "step": 709 + }, + { + "epoch": 0.1, + "grad_norm": 3.3231817927734677, + "learning_rate": 9.87681302754946e-06, + "loss": 0.3709, + "step": 710 + }, + { + "epoch": 0.1, + "grad_norm": 2.323683868919283, + "learning_rate": 9.876315858570681e-06, + "loss": 0.3773, + "step": 711 + }, + { + "epoch": 0.1, + "grad_norm": 2.3989817876346624, + "learning_rate": 9.87581770091561e-06, + "loss": 0.3789, + "step": 712 + }, + { + "epoch": 0.1, + "grad_norm": 3.444852272708474, + "learning_rate": 9.875318554685244e-06, + "loss": 0.3621, + "step": 713 + }, + { + "epoch": 0.1, + "grad_norm": 2.988659419392531, + "learning_rate": 9.874818419980792e-06, + "loss": 0.3726, + "step": 714 + }, + { + "epoch": 0.1, + "grad_norm": 3.175238396742524, + "learning_rate": 9.87431729690365e-06, + "loss": 0.3796, + "step": 715 + }, + { + "epoch": 0.1, + "grad_norm": 2.320051226230734, + "learning_rate": 9.873815185555425e-06, + "loss": 0.3727, + "step": 716 + }, + { + "epoch": 0.1, + "grad_norm": 2.6927585661110154, + "learning_rate": 9.873312086037919e-06, + "loss": 0.3566, + "step": 717 + }, + { + "epoch": 0.1, + "grad_norm": 2.263295513461035, + "learning_rate": 9.872807998453135e-06, + "loss": 0.4014, + "step": 718 + }, + { + "epoch": 0.1, + "grad_norm": 2.266937623642462, + "learning_rate": 9.872302922903278e-06, + "loss": 0.3638, + "step": 719 + }, + { + "epoch": 0.1, + "grad_norm": 3.1099228924804296, + "learning_rate": 9.871796859490752e-06, + "loss": 0.3858, + "step": 720 + }, + { + "epoch": 0.1, + "grad_norm": 2.048430431400738, + "learning_rate": 9.871289808318164e-06, + "loss": 0.3871, + "step": 721 + }, + { + "epoch": 0.1, + "grad_norm": 3.142768382402274, + "learning_rate": 9.870781769488316e-06, + "loss": 0.3756, + "step": 722 + }, + { + "epoch": 0.1, + "grad_norm": 1.952307505198464, + "learning_rate": 9.870272743104215e-06, + "loss": 0.3882, + "step": 723 + }, + { + "epoch": 0.1, + "grad_norm": 3.449290851000038, + "learning_rate": 9.869762729269065e-06, + "loss": 0.3818, + "step": 724 + }, + { + "epoch": 0.1, + "grad_norm": 2.9048955923622963, + "learning_rate": 9.869251728086274e-06, + "loss": 0.3714, + "step": 725 + }, + { + "epoch": 0.1, + "grad_norm": 2.4087701733991356, + "learning_rate": 9.868739739659443e-06, + "loss": 0.3548, + "step": 726 + }, + { + "epoch": 0.1, + "grad_norm": 5.637730971891083, + "learning_rate": 9.868226764092385e-06, + "loss": 0.4093, + "step": 727 + }, + { + "epoch": 0.1, + "grad_norm": 2.1525511720424597, + "learning_rate": 9.8677128014891e-06, + "loss": 0.3839, + "step": 728 + }, + { + "epoch": 0.1, + "grad_norm": 2.0807151380087783, + "learning_rate": 9.867197851953798e-06, + "loss": 0.356, + "step": 729 + }, + { + "epoch": 0.1, + "grad_norm": 2.6874891202459015, + "learning_rate": 9.866681915590884e-06, + "loss": 0.3775, + "step": 730 + }, + { + "epoch": 0.1, + "grad_norm": 2.005795885131233, + "learning_rate": 9.866164992504963e-06, + "loss": 0.3619, + "step": 731 + }, + { + "epoch": 0.1, + "grad_norm": 2.051319427173496, + "learning_rate": 9.865647082800843e-06, + "loss": 0.3828, + "step": 732 + }, + { + "epoch": 0.1, + "grad_norm": 2.262949473915214, + "learning_rate": 9.865128186583531e-06, + "loss": 0.3672, + "step": 733 + }, + { + "epoch": 0.1, + "grad_norm": 2.09374675272666, + "learning_rate": 9.864608303958235e-06, + "loss": 0.3834, + "step": 734 + }, + { + "epoch": 0.1, + "grad_norm": 2.872913251887816, + "learning_rate": 9.864087435030356e-06, + "loss": 0.3519, + "step": 735 + }, + { + "epoch": 0.1, + "grad_norm": 2.340434315535194, + "learning_rate": 9.863565579905506e-06, + "loss": 0.3741, + "step": 736 + }, + { + "epoch": 0.1, + "grad_norm": 2.4326478281381863, + "learning_rate": 9.86304273868949e-06, + "loss": 0.4058, + "step": 737 + }, + { + "epoch": 0.1, + "grad_norm": 2.1708394459794103, + "learning_rate": 9.862518911488312e-06, + "loss": 0.4067, + "step": 738 + }, + { + "epoch": 0.1, + "grad_norm": 2.281463092751463, + "learning_rate": 9.861994098408183e-06, + "loss": 0.3657, + "step": 739 + }, + { + "epoch": 0.1, + "grad_norm": 2.153922768523643, + "learning_rate": 9.861468299555505e-06, + "loss": 0.3374, + "step": 740 + }, + { + "epoch": 0.1, + "grad_norm": 1.9544427587434348, + "learning_rate": 9.860941515036886e-06, + "loss": 0.3628, + "step": 741 + }, + { + "epoch": 0.1, + "grad_norm": 3.3410322609583116, + "learning_rate": 9.86041374495913e-06, + "loss": 0.3793, + "step": 742 + }, + { + "epoch": 0.1, + "grad_norm": 2.2845490918233753, + "learning_rate": 9.859884989429247e-06, + "loss": 0.3606, + "step": 743 + }, + { + "epoch": 0.1, + "grad_norm": 2.072117992673061, + "learning_rate": 9.859355248554437e-06, + "loss": 0.3666, + "step": 744 + }, + { + "epoch": 0.1, + "grad_norm": 2.3234940519785257, + "learning_rate": 9.858824522442109e-06, + "loss": 0.3778, + "step": 745 + }, + { + "epoch": 0.1, + "grad_norm": 2.6425865214232895, + "learning_rate": 9.858292811199867e-06, + "loss": 0.3827, + "step": 746 + }, + { + "epoch": 0.1, + "grad_norm": 2.3690255749599474, + "learning_rate": 9.857760114935518e-06, + "loss": 0.4093, + "step": 747 + }, + { + "epoch": 0.1, + "grad_norm": 2.170730596652672, + "learning_rate": 9.857226433757062e-06, + "loss": 0.3522, + "step": 748 + }, + { + "epoch": 0.1, + "grad_norm": 3.02207235359274, + "learning_rate": 9.856691767772705e-06, + "loss": 0.3573, + "step": 749 + }, + { + "epoch": 0.1, + "grad_norm": 2.186366444228715, + "learning_rate": 9.856156117090855e-06, + "loss": 0.3522, + "step": 750 + }, + { + "epoch": 0.1, + "grad_norm": 2.1326707688722752, + "learning_rate": 9.855619481820109e-06, + "loss": 0.3773, + "step": 751 + }, + { + "epoch": 0.1, + "grad_norm": 4.865752986752456, + "learning_rate": 9.855081862069276e-06, + "loss": 0.3702, + "step": 752 + }, + { + "epoch": 0.1, + "grad_norm": 2.2025732838064216, + "learning_rate": 9.854543257947353e-06, + "loss": 0.3725, + "step": 753 + }, + { + "epoch": 0.1, + "grad_norm": 2.0923671650688034, + "learning_rate": 9.854003669563548e-06, + "loss": 0.3594, + "step": 754 + }, + { + "epoch": 0.1, + "grad_norm": 2.940768655170103, + "learning_rate": 9.853463097027258e-06, + "loss": 0.3766, + "step": 755 + }, + { + "epoch": 0.11, + "grad_norm": 2.3261085088321294, + "learning_rate": 9.852921540448089e-06, + "loss": 0.3628, + "step": 756 + }, + { + "epoch": 0.11, + "grad_norm": 2.64316634338667, + "learning_rate": 9.852378999935839e-06, + "loss": 0.3472, + "step": 757 + }, + { + "epoch": 0.11, + "grad_norm": 2.2688177151931797, + "learning_rate": 9.85183547560051e-06, + "loss": 0.3757, + "step": 758 + }, + { + "epoch": 0.11, + "grad_norm": 2.869387805059771, + "learning_rate": 9.851290967552298e-06, + "loss": 0.3775, + "step": 759 + }, + { + "epoch": 0.11, + "grad_norm": 8.317844942453746, + "learning_rate": 9.850745475901606e-06, + "loss": 0.3577, + "step": 760 + }, + { + "epoch": 0.11, + "grad_norm": 7.34341243379794, + "learning_rate": 9.850199000759033e-06, + "loss": 0.3881, + "step": 761 + }, + { + "epoch": 0.11, + "grad_norm": 2.195982926009558, + "learning_rate": 9.849651542235377e-06, + "loss": 0.3716, + "step": 762 + }, + { + "epoch": 0.11, + "grad_norm": 2.1085665964542892, + "learning_rate": 9.849103100441635e-06, + "loss": 0.408, + "step": 763 + }, + { + "epoch": 0.11, + "grad_norm": 2.060714315261237, + "learning_rate": 9.848553675489005e-06, + "loss": 0.3524, + "step": 764 + }, + { + "epoch": 0.11, + "grad_norm": 2.9135182061753278, + "learning_rate": 9.84800326748888e-06, + "loss": 0.3695, + "step": 765 + }, + { + "epoch": 0.11, + "grad_norm": 2.3107795932902024, + "learning_rate": 9.847451876552858e-06, + "loss": 0.3837, + "step": 766 + }, + { + "epoch": 0.11, + "grad_norm": 1.9960083833330593, + "learning_rate": 9.846899502792735e-06, + "loss": 0.3677, + "step": 767 + }, + { + "epoch": 0.11, + "grad_norm": 2.872556037438994, + "learning_rate": 9.846346146320503e-06, + "loss": 0.3573, + "step": 768 + }, + { + "epoch": 0.11, + "grad_norm": 2.081436436703406, + "learning_rate": 9.845791807248357e-06, + "loss": 0.3638, + "step": 769 + }, + { + "epoch": 0.11, + "grad_norm": 2.645585042386514, + "learning_rate": 9.845236485688689e-06, + "loss": 0.3781, + "step": 770 + }, + { + "epoch": 0.11, + "grad_norm": 2.462836874477033, + "learning_rate": 9.84468018175409e-06, + "loss": 0.341, + "step": 771 + }, + { + "epoch": 0.11, + "grad_norm": 2.7959912815078294, + "learning_rate": 9.84412289555735e-06, + "loss": 0.3845, + "step": 772 + }, + { + "epoch": 0.11, + "grad_norm": 2.2017024692671003, + "learning_rate": 9.843564627211463e-06, + "loss": 0.3629, + "step": 773 + }, + { + "epoch": 0.11, + "grad_norm": 3.0153383431711704, + "learning_rate": 9.843005376829616e-06, + "loss": 0.3813, + "step": 774 + }, + { + "epoch": 0.11, + "grad_norm": 2.49323598113151, + "learning_rate": 9.842445144525197e-06, + "loss": 0.374, + "step": 775 + }, + { + "epoch": 0.11, + "grad_norm": 1.8992291725278971, + "learning_rate": 9.841883930411792e-06, + "loss": 0.381, + "step": 776 + }, + { + "epoch": 0.11, + "grad_norm": 2.155339071516489, + "learning_rate": 9.841321734603194e-06, + "loss": 0.4008, + "step": 777 + }, + { + "epoch": 0.11, + "grad_norm": 2.7870720391026005, + "learning_rate": 9.84075855721338e-06, + "loss": 0.3482, + "step": 778 + }, + { + "epoch": 0.11, + "grad_norm": 2.0025293055639843, + "learning_rate": 9.84019439835654e-06, + "loss": 0.3766, + "step": 779 + }, + { + "epoch": 0.11, + "grad_norm": 2.1379777138672225, + "learning_rate": 9.839629258147056e-06, + "loss": 0.3829, + "step": 780 + }, + { + "epoch": 0.11, + "grad_norm": 2.295059472739771, + "learning_rate": 9.83906313669951e-06, + "loss": 0.3621, + "step": 781 + }, + { + "epoch": 0.11, + "grad_norm": 3.566433024419236, + "learning_rate": 9.838496034128687e-06, + "loss": 0.403, + "step": 782 + }, + { + "epoch": 0.11, + "grad_norm": 3.1062987304772864, + "learning_rate": 9.837927950549564e-06, + "loss": 0.3642, + "step": 783 + }, + { + "epoch": 0.11, + "grad_norm": 2.416249144310666, + "learning_rate": 9.837358886077319e-06, + "loss": 0.3758, + "step": 784 + }, + { + "epoch": 0.11, + "grad_norm": 2.4022166484595573, + "learning_rate": 9.836788840827334e-06, + "loss": 0.3868, + "step": 785 + }, + { + "epoch": 0.11, + "grad_norm": 2.27660842576826, + "learning_rate": 9.836217814915183e-06, + "loss": 0.3869, + "step": 786 + }, + { + "epoch": 0.11, + "grad_norm": 2.6064666528754983, + "learning_rate": 9.835645808456644e-06, + "loss": 0.3802, + "step": 787 + }, + { + "epoch": 0.11, + "grad_norm": 4.616951164061436, + "learning_rate": 9.835072821567691e-06, + "loss": 0.3683, + "step": 788 + }, + { + "epoch": 0.11, + "grad_norm": 2.354256099414065, + "learning_rate": 9.834498854364497e-06, + "loss": 0.3758, + "step": 789 + }, + { + "epoch": 0.11, + "grad_norm": 2.756711234262376, + "learning_rate": 9.833923906963434e-06, + "loss": 0.379, + "step": 790 + }, + { + "epoch": 0.11, + "grad_norm": 2.5002205230131302, + "learning_rate": 9.833347979481075e-06, + "loss": 0.3756, + "step": 791 + }, + { + "epoch": 0.11, + "grad_norm": 2.264436997453065, + "learning_rate": 9.832771072034188e-06, + "loss": 0.3709, + "step": 792 + }, + { + "epoch": 0.11, + "grad_norm": 3.582829379400886, + "learning_rate": 9.83219318473974e-06, + "loss": 0.3839, + "step": 793 + }, + { + "epoch": 0.11, + "grad_norm": 3.3675095578811054, + "learning_rate": 9.831614317714902e-06, + "loss": 0.3724, + "step": 794 + }, + { + "epoch": 0.11, + "grad_norm": 2.8342718899900565, + "learning_rate": 9.831034471077034e-06, + "loss": 0.4038, + "step": 795 + }, + { + "epoch": 0.11, + "grad_norm": 1.9998416034935216, + "learning_rate": 9.830453644943707e-06, + "loss": 0.346, + "step": 796 + }, + { + "epoch": 0.11, + "grad_norm": 2.2616350764472246, + "learning_rate": 9.82987183943268e-06, + "loss": 0.3458, + "step": 797 + }, + { + "epoch": 0.11, + "grad_norm": 3.054611391762719, + "learning_rate": 9.829289054661912e-06, + "loss": 0.3984, + "step": 798 + }, + { + "epoch": 0.11, + "grad_norm": 2.2775825915629166, + "learning_rate": 9.828705290749566e-06, + "loss": 0.3876, + "step": 799 + }, + { + "epoch": 0.11, + "grad_norm": 2.150366910697545, + "learning_rate": 9.828120547814005e-06, + "loss": 0.3761, + "step": 800 + }, + { + "epoch": 0.11, + "grad_norm": 2.393920505067765, + "learning_rate": 9.827534825973779e-06, + "loss": 0.3658, + "step": 801 + }, + { + "epoch": 0.11, + "grad_norm": 6.0407050395244966, + "learning_rate": 9.826948125347644e-06, + "loss": 0.3829, + "step": 802 + }, + { + "epoch": 0.11, + "grad_norm": 2.849267521629154, + "learning_rate": 9.826360446054559e-06, + "loss": 0.3768, + "step": 803 + }, + { + "epoch": 0.11, + "grad_norm": 2.0833388398977473, + "learning_rate": 9.825771788213672e-06, + "loss": 0.3512, + "step": 804 + }, + { + "epoch": 0.11, + "grad_norm": 2.670955319983175, + "learning_rate": 9.825182151944334e-06, + "loss": 0.3781, + "step": 805 + }, + { + "epoch": 0.11, + "grad_norm": 2.6323004291216767, + "learning_rate": 9.824591537366097e-06, + "loss": 0.3221, + "step": 806 + }, + { + "epoch": 0.11, + "grad_norm": 3.032696252744535, + "learning_rate": 9.823999944598705e-06, + "loss": 0.3707, + "step": 807 + }, + { + "epoch": 0.11, + "grad_norm": 3.0449292774213745, + "learning_rate": 9.823407373762108e-06, + "loss": 0.3748, + "step": 808 + }, + { + "epoch": 0.11, + "grad_norm": 2.5100121529299755, + "learning_rate": 9.822813824976446e-06, + "loss": 0.3899, + "step": 809 + }, + { + "epoch": 0.11, + "grad_norm": 2.214858311091561, + "learning_rate": 9.822219298362063e-06, + "loss": 0.3776, + "step": 810 + }, + { + "epoch": 0.11, + "grad_norm": 2.3874149962595124, + "learning_rate": 9.8216237940395e-06, + "loss": 0.3589, + "step": 811 + }, + { + "epoch": 0.11, + "grad_norm": 2.2914144204394162, + "learning_rate": 9.821027312129495e-06, + "loss": 0.3909, + "step": 812 + }, + { + "epoch": 0.11, + "grad_norm": 2.1531441768306627, + "learning_rate": 9.820429852752986e-06, + "loss": 0.3637, + "step": 813 + }, + { + "epoch": 0.11, + "grad_norm": 2.5301662339683064, + "learning_rate": 9.81983141603111e-06, + "loss": 0.3582, + "step": 814 + }, + { + "epoch": 0.11, + "grad_norm": 2.9475559332555354, + "learning_rate": 9.819232002085197e-06, + "loss": 0.3568, + "step": 815 + }, + { + "epoch": 0.11, + "grad_norm": 2.184910219817957, + "learning_rate": 9.818631611036779e-06, + "loss": 0.3687, + "step": 816 + }, + { + "epoch": 0.11, + "grad_norm": 2.4175835848417835, + "learning_rate": 9.818030243007588e-06, + "loss": 0.366, + "step": 817 + }, + { + "epoch": 0.11, + "grad_norm": 2.2283592187333303, + "learning_rate": 9.81742789811955e-06, + "loss": 0.3857, + "step": 818 + }, + { + "epoch": 0.11, + "grad_norm": 2.296206361817545, + "learning_rate": 9.816824576494791e-06, + "loss": 0.3484, + "step": 819 + }, + { + "epoch": 0.11, + "grad_norm": 2.3540589040454254, + "learning_rate": 9.816220278255636e-06, + "loss": 0.3886, + "step": 820 + }, + { + "epoch": 0.11, + "grad_norm": 3.64678394262477, + "learning_rate": 9.815615003524606e-06, + "loss": 0.3461, + "step": 821 + }, + { + "epoch": 0.11, + "grad_norm": 2.4235360683181564, + "learning_rate": 9.81500875242442e-06, + "loss": 0.3815, + "step": 822 + }, + { + "epoch": 0.11, + "grad_norm": 4.415433117131833, + "learning_rate": 9.814401525077999e-06, + "loss": 0.3778, + "step": 823 + }, + { + "epoch": 0.11, + "grad_norm": 2.4335661401114455, + "learning_rate": 9.813793321608454e-06, + "loss": 0.3524, + "step": 824 + }, + { + "epoch": 0.11, + "grad_norm": 1.9453218888292843, + "learning_rate": 9.813184142139102e-06, + "loss": 0.3965, + "step": 825 + }, + { + "epoch": 0.11, + "grad_norm": 2.7632203502608808, + "learning_rate": 9.812573986793454e-06, + "loss": 0.3836, + "step": 826 + }, + { + "epoch": 0.11, + "grad_norm": 3.115680729012198, + "learning_rate": 9.811962855695219e-06, + "loss": 0.3473, + "step": 827 + }, + { + "epoch": 0.12, + "grad_norm": 2.02632596943981, + "learning_rate": 9.811350748968305e-06, + "loss": 0.37, + "step": 828 + }, + { + "epoch": 0.12, + "grad_norm": 1.8646084566856487, + "learning_rate": 9.810737666736816e-06, + "loss": 0.3722, + "step": 829 + }, + { + "epoch": 0.12, + "grad_norm": 2.0560269385123373, + "learning_rate": 9.810123609125054e-06, + "loss": 0.3648, + "step": 830 + }, + { + "epoch": 0.12, + "grad_norm": 2.1549450599008826, + "learning_rate": 9.809508576257522e-06, + "loss": 0.3552, + "step": 831 + }, + { + "epoch": 0.12, + "grad_norm": 1.9048916959057485, + "learning_rate": 9.808892568258917e-06, + "loss": 0.3867, + "step": 832 + }, + { + "epoch": 0.12, + "grad_norm": 1.7544616165506062, + "learning_rate": 9.808275585254134e-06, + "loss": 0.3661, + "step": 833 + }, + { + "epoch": 0.12, + "grad_norm": 2.4811826150559986, + "learning_rate": 9.80765762736827e-06, + "loss": 0.3993, + "step": 834 + }, + { + "epoch": 0.12, + "grad_norm": 1.8405092076895033, + "learning_rate": 9.807038694726613e-06, + "loss": 0.4077, + "step": 835 + }, + { + "epoch": 0.12, + "grad_norm": 4.252178366704906, + "learning_rate": 9.806418787454653e-06, + "loss": 0.3591, + "step": 836 + }, + { + "epoch": 0.12, + "grad_norm": 2.7764156517603604, + "learning_rate": 9.805797905678078e-06, + "loss": 0.3576, + "step": 837 + }, + { + "epoch": 0.12, + "grad_norm": 2.101478835240452, + "learning_rate": 9.80517604952277e-06, + "loss": 0.3683, + "step": 838 + }, + { + "epoch": 0.12, + "grad_norm": 6.217014851323827, + "learning_rate": 9.804553219114812e-06, + "loss": 0.3671, + "step": 839 + }, + { + "epoch": 0.12, + "grad_norm": 2.205436792767768, + "learning_rate": 9.803929414580485e-06, + "loss": 0.3657, + "step": 840 + }, + { + "epoch": 0.12, + "grad_norm": 2.4162038726865656, + "learning_rate": 9.80330463604626e-06, + "loss": 0.3771, + "step": 841 + }, + { + "epoch": 0.12, + "grad_norm": 2.3488872359267896, + "learning_rate": 9.80267888363882e-06, + "loss": 0.3654, + "step": 842 + }, + { + "epoch": 0.12, + "grad_norm": 2.3449760862483173, + "learning_rate": 9.802052157485027e-06, + "loss": 0.3564, + "step": 843 + }, + { + "epoch": 0.12, + "grad_norm": 1.9785809137271635, + "learning_rate": 9.801424457711957e-06, + "loss": 0.3844, + "step": 844 + }, + { + "epoch": 0.12, + "grad_norm": 2.114410747074343, + "learning_rate": 9.800795784446874e-06, + "loss": 0.3475, + "step": 845 + }, + { + "epoch": 0.12, + "grad_norm": 2.58704914735204, + "learning_rate": 9.800166137817245e-06, + "loss": 0.3817, + "step": 846 + }, + { + "epoch": 0.12, + "grad_norm": 1.9873410685247477, + "learning_rate": 9.799535517950727e-06, + "loss": 0.3596, + "step": 847 + }, + { + "epoch": 0.12, + "grad_norm": 1.9850745747687446, + "learning_rate": 9.798903924975181e-06, + "loss": 0.3633, + "step": 848 + }, + { + "epoch": 0.12, + "grad_norm": 2.2975815589464186, + "learning_rate": 9.79827135901866e-06, + "loss": 0.3688, + "step": 849 + }, + { + "epoch": 0.12, + "grad_norm": 2.881698708581663, + "learning_rate": 9.797637820209422e-06, + "loss": 0.3476, + "step": 850 + }, + { + "epoch": 0.12, + "grad_norm": 2.6514701948064765, + "learning_rate": 9.797003308675914e-06, + "loss": 0.3554, + "step": 851 + }, + { + "epoch": 0.12, + "grad_norm": 2.109170077012958, + "learning_rate": 9.796367824546784e-06, + "loss": 0.3651, + "step": 852 + }, + { + "epoch": 0.12, + "grad_norm": 2.034653649508209, + "learning_rate": 9.795731367950878e-06, + "loss": 0.3602, + "step": 853 + }, + { + "epoch": 0.12, + "grad_norm": 1.9361171921180282, + "learning_rate": 9.795093939017237e-06, + "loss": 0.3826, + "step": 854 + }, + { + "epoch": 0.12, + "grad_norm": 2.598621357124829, + "learning_rate": 9.794455537875101e-06, + "loss": 0.3498, + "step": 855 + }, + { + "epoch": 0.12, + "grad_norm": 2.3440528605950703, + "learning_rate": 9.793816164653904e-06, + "loss": 0.3637, + "step": 856 + }, + { + "epoch": 0.12, + "grad_norm": 2.032990379615116, + "learning_rate": 9.793175819483283e-06, + "loss": 0.3878, + "step": 857 + }, + { + "epoch": 0.12, + "grad_norm": 2.3363297938810943, + "learning_rate": 9.792534502493065e-06, + "loss": 0.3534, + "step": 858 + }, + { + "epoch": 0.12, + "grad_norm": 2.1134151203288294, + "learning_rate": 9.791892213813281e-06, + "loss": 0.3549, + "step": 859 + }, + { + "epoch": 0.12, + "grad_norm": 2.040339442425263, + "learning_rate": 9.79124895357415e-06, + "loss": 0.3744, + "step": 860 + }, + { + "epoch": 0.12, + "grad_norm": 1.846331900702117, + "learning_rate": 9.790604721906101e-06, + "loss": 0.3438, + "step": 861 + }, + { + "epoch": 0.12, + "grad_norm": 2.1734578407198546, + "learning_rate": 9.789959518939745e-06, + "loss": 0.3856, + "step": 862 + }, + { + "epoch": 0.12, + "grad_norm": 2.760553604150688, + "learning_rate": 9.789313344805902e-06, + "loss": 0.3819, + "step": 863 + }, + { + "epoch": 0.12, + "grad_norm": 1.9201573535815721, + "learning_rate": 9.788666199635581e-06, + "loss": 0.3797, + "step": 864 + }, + { + "epoch": 0.12, + "grad_norm": 2.3397873194829595, + "learning_rate": 9.788018083559996e-06, + "loss": 0.359, + "step": 865 + }, + { + "epoch": 0.12, + "grad_norm": 3.6955773621828674, + "learning_rate": 9.787368996710548e-06, + "loss": 0.3688, + "step": 866 + }, + { + "epoch": 0.12, + "grad_norm": 2.0449027703071088, + "learning_rate": 9.786718939218843e-06, + "loss": 0.3581, + "step": 867 + }, + { + "epoch": 0.12, + "grad_norm": 3.5537281033985018, + "learning_rate": 9.786067911216676e-06, + "loss": 0.3662, + "step": 868 + }, + { + "epoch": 0.12, + "grad_norm": 1.8942587887824562, + "learning_rate": 9.785415912836049e-06, + "loss": 0.3769, + "step": 869 + }, + { + "epoch": 0.12, + "grad_norm": 1.8255911569673475, + "learning_rate": 9.784762944209152e-06, + "loss": 0.3873, + "step": 870 + }, + { + "epoch": 0.12, + "grad_norm": 2.2316443363133476, + "learning_rate": 9.784109005468377e-06, + "loss": 0.3328, + "step": 871 + }, + { + "epoch": 0.12, + "grad_norm": 2.4530215637732495, + "learning_rate": 9.783454096746307e-06, + "loss": 0.3876, + "step": 872 + }, + { + "epoch": 0.12, + "grad_norm": 2.148452909116464, + "learning_rate": 9.782798218175728e-06, + "loss": 0.3943, + "step": 873 + }, + { + "epoch": 0.12, + "grad_norm": 2.458533455172328, + "learning_rate": 9.78214136988962e-06, + "loss": 0.383, + "step": 874 + }, + { + "epoch": 0.12, + "grad_norm": 3.922186172828907, + "learning_rate": 9.781483552021158e-06, + "loss": 0.3521, + "step": 875 + }, + { + "epoch": 0.12, + "grad_norm": 2.1033771899368467, + "learning_rate": 9.780824764703715e-06, + "loss": 0.3482, + "step": 876 + }, + { + "epoch": 0.12, + "grad_norm": 2.1394086847369254, + "learning_rate": 9.780165008070861e-06, + "loss": 0.3697, + "step": 877 + }, + { + "epoch": 0.12, + "grad_norm": 3.4652206350159327, + "learning_rate": 9.779504282256363e-06, + "loss": 0.3491, + "step": 878 + }, + { + "epoch": 0.12, + "grad_norm": 3.6063956403571478, + "learning_rate": 9.778842587394181e-06, + "loss": 0.3845, + "step": 879 + }, + { + "epoch": 0.12, + "grad_norm": 2.190349099600736, + "learning_rate": 9.778179923618478e-06, + "loss": 0.3529, + "step": 880 + }, + { + "epoch": 0.12, + "grad_norm": 2.2326427417606785, + "learning_rate": 9.777516291063607e-06, + "loss": 0.3778, + "step": 881 + }, + { + "epoch": 0.12, + "grad_norm": 3.9604861851686137, + "learning_rate": 9.77685168986412e-06, + "loss": 0.3969, + "step": 882 + }, + { + "epoch": 0.12, + "grad_norm": 1.950484344654343, + "learning_rate": 9.776186120154764e-06, + "loss": 0.3784, + "step": 883 + }, + { + "epoch": 0.12, + "grad_norm": 3.5768851332092884, + "learning_rate": 9.775519582070487e-06, + "loss": 0.3956, + "step": 884 + }, + { + "epoch": 0.12, + "grad_norm": 2.107818023713643, + "learning_rate": 9.774852075746428e-06, + "loss": 0.412, + "step": 885 + }, + { + "epoch": 0.12, + "grad_norm": 3.028131627381809, + "learning_rate": 9.774183601317926e-06, + "loss": 0.3496, + "step": 886 + }, + { + "epoch": 0.12, + "grad_norm": 2.3187899272526664, + "learning_rate": 9.773514158920512e-06, + "loss": 0.3638, + "step": 887 + }, + { + "epoch": 0.12, + "grad_norm": 1.8517570526945322, + "learning_rate": 9.772843748689918e-06, + "loss": 0.3796, + "step": 888 + }, + { + "epoch": 0.12, + "grad_norm": 1.9065734803558347, + "learning_rate": 9.772172370762067e-06, + "loss": 0.3686, + "step": 889 + }, + { + "epoch": 0.12, + "grad_norm": 2.442358417809032, + "learning_rate": 9.771500025273086e-06, + "loss": 0.3833, + "step": 890 + }, + { + "epoch": 0.12, + "grad_norm": 2.2120236691438246, + "learning_rate": 9.770826712359292e-06, + "loss": 0.379, + "step": 891 + }, + { + "epoch": 0.12, + "grad_norm": 1.999285766973106, + "learning_rate": 9.770152432157198e-06, + "loss": 0.3901, + "step": 892 + }, + { + "epoch": 0.12, + "grad_norm": 1.8817713764454413, + "learning_rate": 9.769477184803515e-06, + "loss": 0.3824, + "step": 893 + }, + { + "epoch": 0.12, + "grad_norm": 2.2660610144236775, + "learning_rate": 9.76880097043515e-06, + "loss": 0.3976, + "step": 894 + }, + { + "epoch": 0.12, + "grad_norm": 2.1088166129554167, + "learning_rate": 9.768123789189208e-06, + "loss": 0.3853, + "step": 895 + }, + { + "epoch": 0.12, + "grad_norm": 2.4157225424583393, + "learning_rate": 9.767445641202984e-06, + "loss": 0.3714, + "step": 896 + }, + { + "epoch": 0.12, + "grad_norm": 1.8756500624676902, + "learning_rate": 9.766766526613978e-06, + "loss": 0.368, + "step": 897 + }, + { + "epoch": 0.12, + "grad_norm": 1.8656585307618054, + "learning_rate": 9.766086445559877e-06, + "loss": 0.3394, + "step": 898 + }, + { + "epoch": 0.12, + "grad_norm": 1.9355633938239492, + "learning_rate": 9.765405398178569e-06, + "loss": 0.3582, + "step": 899 + }, + { + "epoch": 0.13, + "grad_norm": 2.0270053379248876, + "learning_rate": 9.764723384608137e-06, + "loss": 0.388, + "step": 900 + }, + { + "epoch": 0.13, + "grad_norm": 2.684829872542931, + "learning_rate": 9.764040404986858e-06, + "loss": 0.4086, + "step": 901 + }, + { + "epoch": 0.13, + "grad_norm": 2.1601685580535728, + "learning_rate": 9.76335645945321e-06, + "loss": 0.3876, + "step": 902 + }, + { + "epoch": 0.13, + "grad_norm": 1.997291095722403, + "learning_rate": 9.762671548145862e-06, + "loss": 0.3511, + "step": 903 + }, + { + "epoch": 0.13, + "grad_norm": 2.166166875640202, + "learning_rate": 9.76198567120368e-06, + "loss": 0.387, + "step": 904 + }, + { + "epoch": 0.13, + "grad_norm": 1.957855057222124, + "learning_rate": 9.761298828765723e-06, + "loss": 0.3667, + "step": 905 + }, + { + "epoch": 0.13, + "grad_norm": 2.013182593339566, + "learning_rate": 9.760611020971256e-06, + "loss": 0.362, + "step": 906 + }, + { + "epoch": 0.13, + "grad_norm": 2.329160079456012, + "learning_rate": 9.759922247959725e-06, + "loss": 0.3871, + "step": 907 + }, + { + "epoch": 0.13, + "grad_norm": 2.4030405140271722, + "learning_rate": 9.759232509870784e-06, + "loss": 0.393, + "step": 908 + }, + { + "epoch": 0.13, + "grad_norm": 2.2406625059929013, + "learning_rate": 9.758541806844273e-06, + "loss": 0.3811, + "step": 909 + }, + { + "epoch": 0.13, + "grad_norm": 1.787739740506085, + "learning_rate": 9.75785013902024e-06, + "loss": 0.3423, + "step": 910 + }, + { + "epoch": 0.13, + "grad_norm": 2.295934527478593, + "learning_rate": 9.757157506538914e-06, + "loss": 0.3729, + "step": 911 + }, + { + "epoch": 0.13, + "grad_norm": 1.8398358863882494, + "learning_rate": 9.75646390954073e-06, + "loss": 0.3588, + "step": 912 + }, + { + "epoch": 0.13, + "grad_norm": 2.0402925383030857, + "learning_rate": 9.755769348166314e-06, + "loss": 0.3765, + "step": 913 + }, + { + "epoch": 0.13, + "grad_norm": 3.9680015922741316, + "learning_rate": 9.75507382255649e-06, + "loss": 0.3226, + "step": 914 + }, + { + "epoch": 0.13, + "grad_norm": 2.286277524264896, + "learning_rate": 9.754377332852276e-06, + "loss": 0.3749, + "step": 915 + }, + { + "epoch": 0.13, + "grad_norm": 2.011258288417456, + "learning_rate": 9.753679879194887e-06, + "loss": 0.3464, + "step": 916 + }, + { + "epoch": 0.13, + "grad_norm": 1.8312442777820022, + "learning_rate": 9.752981461725728e-06, + "loss": 0.3629, + "step": 917 + }, + { + "epoch": 0.13, + "grad_norm": 3.2441019030173335, + "learning_rate": 9.752282080586409e-06, + "loss": 0.3766, + "step": 918 + }, + { + "epoch": 0.13, + "grad_norm": 4.674820432160112, + "learning_rate": 9.751581735918724e-06, + "loss": 0.3521, + "step": 919 + }, + { + "epoch": 0.13, + "grad_norm": 2.040193602057243, + "learning_rate": 9.750880427864674e-06, + "loss": 0.3824, + "step": 920 + }, + { + "epoch": 0.13, + "grad_norm": 2.4877389547138775, + "learning_rate": 9.750178156566448e-06, + "loss": 0.3874, + "step": 921 + }, + { + "epoch": 0.13, + "grad_norm": 2.650601645365233, + "learning_rate": 9.749474922166429e-06, + "loss": 0.3659, + "step": 922 + }, + { + "epoch": 0.13, + "grad_norm": 2.265466401292786, + "learning_rate": 9.748770724807201e-06, + "loss": 0.407, + "step": 923 + }, + { + "epoch": 0.13, + "grad_norm": 2.8113189887643175, + "learning_rate": 9.748065564631541e-06, + "loss": 0.3636, + "step": 924 + }, + { + "epoch": 0.13, + "grad_norm": 2.4200901306864058, + "learning_rate": 9.747359441782421e-06, + "loss": 0.3862, + "step": 925 + }, + { + "epoch": 0.13, + "grad_norm": 2.348247923898854, + "learning_rate": 9.746652356403004e-06, + "loss": 0.3706, + "step": 926 + }, + { + "epoch": 0.13, + "grad_norm": 2.2293038939046586, + "learning_rate": 9.745944308636657e-06, + "loss": 0.3702, + "step": 927 + }, + { + "epoch": 0.13, + "grad_norm": 3.453908944749896, + "learning_rate": 9.745235298626933e-06, + "loss": 0.3786, + "step": 928 + }, + { + "epoch": 0.13, + "grad_norm": 3.160838931699184, + "learning_rate": 9.744525326517587e-06, + "loss": 0.3991, + "step": 929 + }, + { + "epoch": 0.13, + "grad_norm": 1.9525014056536258, + "learning_rate": 9.743814392452565e-06, + "loss": 0.3871, + "step": 930 + }, + { + "epoch": 0.13, + "grad_norm": 2.6126780064733626, + "learning_rate": 9.743102496576008e-06, + "loss": 0.3706, + "step": 931 + }, + { + "epoch": 0.13, + "grad_norm": 3.08581823832653, + "learning_rate": 9.742389639032257e-06, + "loss": 0.3699, + "step": 932 + }, + { + "epoch": 0.13, + "grad_norm": 2.275521280261364, + "learning_rate": 9.741675819965844e-06, + "loss": 0.3763, + "step": 933 + }, + { + "epoch": 0.13, + "grad_norm": 1.9800373010763195, + "learning_rate": 9.740961039521492e-06, + "loss": 0.3315, + "step": 934 + }, + { + "epoch": 0.13, + "grad_norm": 2.6429904356892315, + "learning_rate": 9.740245297844127e-06, + "loss": 0.3529, + "step": 935 + }, + { + "epoch": 0.13, + "grad_norm": 2.074139296282507, + "learning_rate": 9.739528595078864e-06, + "loss": 0.3582, + "step": 936 + }, + { + "epoch": 0.13, + "grad_norm": 2.420397873825773, + "learning_rate": 9.73881093137102e-06, + "loss": 0.3859, + "step": 937 + }, + { + "epoch": 0.13, + "grad_norm": 3.8215273034643418, + "learning_rate": 9.738092306866092e-06, + "loss": 0.3785, + "step": 938 + }, + { + "epoch": 0.13, + "grad_norm": 1.9696156748313682, + "learning_rate": 9.73737272170979e-06, + "loss": 0.3734, + "step": 939 + }, + { + "epoch": 0.13, + "grad_norm": 2.0194669857701633, + "learning_rate": 9.736652176048009e-06, + "loss": 0.3789, + "step": 940 + }, + { + "epoch": 0.13, + "grad_norm": 3.7699897779918197, + "learning_rate": 9.735930670026839e-06, + "loss": 0.382, + "step": 941 + }, + { + "epoch": 0.13, + "grad_norm": 1.7593994331120393, + "learning_rate": 9.735208203792563e-06, + "loss": 0.3707, + "step": 942 + }, + { + "epoch": 0.13, + "grad_norm": 1.9250112806643354, + "learning_rate": 9.734484777491667e-06, + "loss": 0.3763, + "step": 943 + }, + { + "epoch": 0.13, + "grad_norm": 1.9312347198863704, + "learning_rate": 9.733760391270821e-06, + "loss": 0.3936, + "step": 944 + }, + { + "epoch": 0.13, + "grad_norm": 1.7784419130756404, + "learning_rate": 9.7330350452769e-06, + "loss": 0.3836, + "step": 945 + }, + { + "epoch": 0.13, + "grad_norm": 3.354407450359741, + "learning_rate": 9.732308739656964e-06, + "loss": 0.3858, + "step": 946 + }, + { + "epoch": 0.13, + "grad_norm": 1.8298306493875722, + "learning_rate": 9.731581474558272e-06, + "loss": 0.3637, + "step": 947 + }, + { + "epoch": 0.13, + "grad_norm": 1.7075905657494848, + "learning_rate": 9.73085325012828e-06, + "loss": 0.3775, + "step": 948 + }, + { + "epoch": 0.13, + "grad_norm": 2.627401043694326, + "learning_rate": 9.730124066514637e-06, + "loss": 0.395, + "step": 949 + }, + { + "epoch": 0.13, + "grad_norm": 2.0294535904327566, + "learning_rate": 9.729393923865181e-06, + "loss": 0.3602, + "step": 950 + }, + { + "epoch": 0.13, + "grad_norm": 2.141493657214575, + "learning_rate": 9.728662822327952e-06, + "loss": 0.3547, + "step": 951 + }, + { + "epoch": 0.13, + "grad_norm": 2.7391917764665243, + "learning_rate": 9.727930762051181e-06, + "loss": 0.3713, + "step": 952 + }, + { + "epoch": 0.13, + "grad_norm": 2.0066484948675742, + "learning_rate": 9.727197743183293e-06, + "loss": 0.3871, + "step": 953 + }, + { + "epoch": 0.13, + "grad_norm": 2.7831893182631053, + "learning_rate": 9.72646376587291e-06, + "loss": 0.3614, + "step": 954 + }, + { + "epoch": 0.13, + "grad_norm": 2.1922319885464954, + "learning_rate": 9.725728830268843e-06, + "loss": 0.3709, + "step": 955 + }, + { + "epoch": 0.13, + "grad_norm": 1.7168018570149666, + "learning_rate": 9.724992936520103e-06, + "loss": 0.3455, + "step": 956 + }, + { + "epoch": 0.13, + "grad_norm": 2.268514618461269, + "learning_rate": 9.724256084775895e-06, + "loss": 0.3654, + "step": 957 + }, + { + "epoch": 0.13, + "grad_norm": 2.5493189873865414, + "learning_rate": 9.72351827518561e-06, + "loss": 0.381, + "step": 958 + }, + { + "epoch": 0.13, + "grad_norm": 1.8700310786832726, + "learning_rate": 9.722779507898845e-06, + "loss": 0.3435, + "step": 959 + }, + { + "epoch": 0.13, + "grad_norm": 2.359163548752818, + "learning_rate": 9.722039783065383e-06, + "loss": 0.3706, + "step": 960 + }, + { + "epoch": 0.13, + "grad_norm": 2.546083120759652, + "learning_rate": 9.721299100835203e-06, + "loss": 0.3628, + "step": 961 + }, + { + "epoch": 0.13, + "grad_norm": 2.1336223958891547, + "learning_rate": 9.720557461358482e-06, + "loss": 0.4036, + "step": 962 + }, + { + "epoch": 0.13, + "grad_norm": 2.054300646707425, + "learning_rate": 9.719814864785587e-06, + "loss": 0.3436, + "step": 963 + }, + { + "epoch": 0.13, + "grad_norm": 4.163047691848525, + "learning_rate": 9.719071311267077e-06, + "loss": 0.3766, + "step": 964 + }, + { + "epoch": 0.13, + "grad_norm": 1.908465991552521, + "learning_rate": 9.71832680095371e-06, + "loss": 0.3747, + "step": 965 + }, + { + "epoch": 0.13, + "grad_norm": 2.265977247601528, + "learning_rate": 9.717581333996436e-06, + "loss": 0.3644, + "step": 966 + }, + { + "epoch": 0.13, + "grad_norm": 2.489063685973265, + "learning_rate": 9.7168349105464e-06, + "loss": 0.386, + "step": 967 + }, + { + "epoch": 0.13, + "grad_norm": 2.317831097075742, + "learning_rate": 9.716087530754937e-06, + "loss": 0.3668, + "step": 968 + }, + { + "epoch": 0.13, + "grad_norm": 2.056116402550063, + "learning_rate": 9.715339194773581e-06, + "loss": 0.3675, + "step": 969 + }, + { + "epoch": 0.13, + "grad_norm": 2.4060511156280824, + "learning_rate": 9.714589902754058e-06, + "loss": 0.373, + "step": 970 + }, + { + "epoch": 0.13, + "grad_norm": 2.201858308275542, + "learning_rate": 9.713839654848285e-06, + "loss": 0.3678, + "step": 971 + }, + { + "epoch": 0.14, + "grad_norm": 2.3017621803829114, + "learning_rate": 9.713088451208379e-06, + "loss": 0.3511, + "step": 972 + }, + { + "epoch": 0.14, + "grad_norm": 2.1176995090780273, + "learning_rate": 9.712336291986643e-06, + "loss": 0.388, + "step": 973 + }, + { + "epoch": 0.14, + "grad_norm": 1.8062490233989277, + "learning_rate": 9.711583177335579e-06, + "loss": 0.3753, + "step": 974 + }, + { + "epoch": 0.14, + "grad_norm": 2.268945883295479, + "learning_rate": 9.710829107407884e-06, + "loss": 0.3629, + "step": 975 + }, + { + "epoch": 0.14, + "grad_norm": 1.7718897383957384, + "learning_rate": 9.710074082356444e-06, + "loss": 0.3473, + "step": 976 + }, + { + "epoch": 0.14, + "grad_norm": 1.7613482561033058, + "learning_rate": 9.70931810233434e-06, + "loss": 0.3624, + "step": 977 + }, + { + "epoch": 0.14, + "grad_norm": 2.0160824307120486, + "learning_rate": 9.70856116749485e-06, + "loss": 0.3665, + "step": 978 + }, + { + "epoch": 0.14, + "grad_norm": 2.2492544155196676, + "learning_rate": 9.70780327799144e-06, + "loss": 0.3689, + "step": 979 + }, + { + "epoch": 0.14, + "grad_norm": 2.234583695064284, + "learning_rate": 9.707044433977775e-06, + "loss": 0.3724, + "step": 980 + }, + { + "epoch": 0.14, + "grad_norm": 1.88566371171054, + "learning_rate": 9.70628463560771e-06, + "loss": 0.3498, + "step": 981 + }, + { + "epoch": 0.14, + "grad_norm": 3.533702125220694, + "learning_rate": 9.705523883035297e-06, + "loss": 0.3792, + "step": 982 + }, + { + "epoch": 0.14, + "grad_norm": 2.5954895876904387, + "learning_rate": 9.704762176414774e-06, + "loss": 0.3768, + "step": 983 + }, + { + "epoch": 0.14, + "grad_norm": 3.1786734156923, + "learning_rate": 9.703999515900581e-06, + "loss": 0.3856, + "step": 984 + }, + { + "epoch": 0.14, + "grad_norm": 2.218479089184796, + "learning_rate": 9.70323590164735e-06, + "loss": 0.3882, + "step": 985 + }, + { + "epoch": 0.14, + "grad_norm": 2.9200250054689905, + "learning_rate": 9.702471333809899e-06, + "loss": 0.3763, + "step": 986 + }, + { + "epoch": 0.14, + "grad_norm": 3.2848472563441136, + "learning_rate": 9.701705812543249e-06, + "loss": 0.3668, + "step": 987 + }, + { + "epoch": 0.14, + "grad_norm": 3.2982969870761303, + "learning_rate": 9.700939338002608e-06, + "loss": 0.3471, + "step": 988 + }, + { + "epoch": 0.14, + "grad_norm": 1.882931663585713, + "learning_rate": 9.700171910343381e-06, + "loss": 0.3653, + "step": 989 + }, + { + "epoch": 0.14, + "grad_norm": 2.3151114038701577, + "learning_rate": 9.699403529721162e-06, + "loss": 0.394, + "step": 990 + }, + { + "epoch": 0.14, + "grad_norm": 2.011373849619613, + "learning_rate": 9.698634196291742e-06, + "loss": 0.366, + "step": 991 + }, + { + "epoch": 0.14, + "grad_norm": 2.6951775288940496, + "learning_rate": 9.697863910211103e-06, + "loss": 0.3858, + "step": 992 + }, + { + "epoch": 0.14, + "grad_norm": 2.3689581037418557, + "learning_rate": 9.697092671635421e-06, + "loss": 0.3648, + "step": 993 + }, + { + "epoch": 0.14, + "grad_norm": 2.9066054875538105, + "learning_rate": 9.69632048072107e-06, + "loss": 0.3518, + "step": 994 + }, + { + "epoch": 0.14, + "grad_norm": 2.0601057492448547, + "learning_rate": 9.695547337624603e-06, + "loss": 0.3849, + "step": 995 + }, + { + "epoch": 0.14, + "grad_norm": 2.6902791896690896, + "learning_rate": 9.694773242502782e-06, + "loss": 0.3885, + "step": 996 + }, + { + "epoch": 0.14, + "grad_norm": 2.105032091228643, + "learning_rate": 9.693998195512553e-06, + "loss": 0.35, + "step": 997 + }, + { + "epoch": 0.14, + "grad_norm": 2.5550548525592616, + "learning_rate": 9.693222196811059e-06, + "loss": 0.3661, + "step": 998 + }, + { + "epoch": 0.14, + "grad_norm": 2.2394378207164083, + "learning_rate": 9.692445246555633e-06, + "loss": 0.3639, + "step": 999 + }, + { + "epoch": 0.14, + "grad_norm": 2.0734041622990977, + "learning_rate": 9.691667344903802e-06, + "loss": 0.377, + "step": 1000 + }, + { + "epoch": 0.14, + "grad_norm": 1.9251159833505485, + "learning_rate": 9.690888492013287e-06, + "loss": 0.3544, + "step": 1001 + }, + { + "epoch": 0.14, + "grad_norm": 2.5061408275236214, + "learning_rate": 9.690108688042001e-06, + "loss": 0.3499, + "step": 1002 + }, + { + "epoch": 0.14, + "grad_norm": 2.67034328115048, + "learning_rate": 9.689327933148047e-06, + "loss": 0.3441, + "step": 1003 + }, + { + "epoch": 0.14, + "grad_norm": 2.1072361180225183, + "learning_rate": 9.688546227489727e-06, + "loss": 0.3339, + "step": 1004 + }, + { + "epoch": 0.14, + "grad_norm": 1.8940662428502217, + "learning_rate": 9.687763571225531e-06, + "loss": 0.3834, + "step": 1005 + }, + { + "epoch": 0.14, + "grad_norm": 1.814752516131576, + "learning_rate": 9.686979964514143e-06, + "loss": 0.3603, + "step": 1006 + }, + { + "epoch": 0.14, + "grad_norm": 2.1803369619394206, + "learning_rate": 9.686195407514441e-06, + "loss": 0.3503, + "step": 1007 + }, + { + "epoch": 0.14, + "grad_norm": 4.378769254912763, + "learning_rate": 9.685409900385491e-06, + "loss": 0.3561, + "step": 1008 + }, + { + "epoch": 0.14, + "grad_norm": 4.355345750891033, + "learning_rate": 9.684623443286558e-06, + "loss": 0.3975, + "step": 1009 + }, + { + "epoch": 0.14, + "grad_norm": 2.2141055772949456, + "learning_rate": 9.683836036377098e-06, + "loss": 0.3524, + "step": 1010 + }, + { + "epoch": 0.14, + "grad_norm": 2.5892058116504457, + "learning_rate": 9.683047679816755e-06, + "loss": 0.3746, + "step": 1011 + }, + { + "epoch": 0.14, + "grad_norm": 2.141604985120343, + "learning_rate": 9.68225837376537e-06, + "loss": 0.3741, + "step": 1012 + }, + { + "epoch": 0.14, + "grad_norm": 1.9168559695977674, + "learning_rate": 9.681468118382978e-06, + "loss": 0.3748, + "step": 1013 + }, + { + "epoch": 0.14, + "grad_norm": 2.2369757910166728, + "learning_rate": 9.680676913829799e-06, + "loss": 0.3996, + "step": 1014 + }, + { + "epoch": 0.14, + "grad_norm": 1.9912101717546928, + "learning_rate": 9.679884760266253e-06, + "loss": 0.3736, + "step": 1015 + }, + { + "epoch": 0.14, + "grad_norm": 1.9872113497807558, + "learning_rate": 9.67909165785295e-06, + "loss": 0.3841, + "step": 1016 + }, + { + "epoch": 0.14, + "grad_norm": 1.6708818085276482, + "learning_rate": 9.67829760675069e-06, + "loss": 0.3438, + "step": 1017 + }, + { + "epoch": 0.14, + "grad_norm": 4.228660853631826, + "learning_rate": 9.677502607120471e-06, + "loss": 0.3512, + "step": 1018 + }, + { + "epoch": 0.14, + "grad_norm": 1.7687236820350243, + "learning_rate": 9.676706659123475e-06, + "loss": 0.3786, + "step": 1019 + }, + { + "epoch": 0.14, + "grad_norm": 1.7720784827195377, + "learning_rate": 9.675909762921084e-06, + "loss": 0.358, + "step": 1020 + }, + { + "epoch": 0.14, + "grad_norm": 2.06370131342619, + "learning_rate": 9.67511191867487e-06, + "loss": 0.3452, + "step": 1021 + }, + { + "epoch": 0.14, + "grad_norm": 1.9558394680587798, + "learning_rate": 9.674313126546596e-06, + "loss": 0.3777, + "step": 1022 + }, + { + "epoch": 0.14, + "grad_norm": 1.9798704193607488, + "learning_rate": 9.673513386698215e-06, + "loss": 0.3793, + "step": 1023 + }, + { + "epoch": 0.14, + "grad_norm": 2.1870306385712985, + "learning_rate": 9.672712699291876e-06, + "loss": 0.3848, + "step": 1024 + }, + { + "epoch": 0.14, + "grad_norm": 2.1489024759546647, + "learning_rate": 9.67191106448992e-06, + "loss": 0.3725, + "step": 1025 + }, + { + "epoch": 0.14, + "grad_norm": 2.492659472233847, + "learning_rate": 9.67110848245488e-06, + "loss": 0.3996, + "step": 1026 + }, + { + "epoch": 0.14, + "grad_norm": 2.0239609388597577, + "learning_rate": 9.670304953349476e-06, + "loss": 0.3523, + "step": 1027 + }, + { + "epoch": 0.14, + "grad_norm": 2.1004330437514294, + "learning_rate": 9.669500477336628e-06, + "loss": 0.3789, + "step": 1028 + }, + { + "epoch": 0.14, + "grad_norm": 3.0714171129531427, + "learning_rate": 9.668695054579443e-06, + "loss": 0.3687, + "step": 1029 + }, + { + "epoch": 0.14, + "grad_norm": 2.5615812755408798, + "learning_rate": 9.66788868524122e-06, + "loss": 0.4013, + "step": 1030 + }, + { + "epoch": 0.14, + "grad_norm": 2.695346338788103, + "learning_rate": 9.66708136948545e-06, + "loss": 0.3861, + "step": 1031 + }, + { + "epoch": 0.14, + "grad_norm": 2.0481366583127616, + "learning_rate": 9.666273107475822e-06, + "loss": 0.3561, + "step": 1032 + }, + { + "epoch": 0.14, + "grad_norm": 2.164338700034747, + "learning_rate": 9.665463899376203e-06, + "loss": 0.3751, + "step": 1033 + }, + { + "epoch": 0.14, + "grad_norm": 1.8968790990574609, + "learning_rate": 9.66465374535067e-06, + "loss": 0.3592, + "step": 1034 + }, + { + "epoch": 0.14, + "grad_norm": 2.1484401321598274, + "learning_rate": 9.663842645563474e-06, + "loss": 0.3466, + "step": 1035 + }, + { + "epoch": 0.14, + "grad_norm": 2.0815052632588085, + "learning_rate": 9.663030600179073e-06, + "loss": 0.4, + "step": 1036 + }, + { + "epoch": 0.14, + "grad_norm": 2.900686251675779, + "learning_rate": 9.662217609362106e-06, + "loss": 0.3919, + "step": 1037 + }, + { + "epoch": 0.14, + "grad_norm": 2.3459250970317367, + "learning_rate": 9.661403673277408e-06, + "loss": 0.3815, + "step": 1038 + }, + { + "epoch": 0.14, + "grad_norm": 1.9677376748002777, + "learning_rate": 9.660588792090003e-06, + "loss": 0.391, + "step": 1039 + }, + { + "epoch": 0.14, + "grad_norm": 2.033843326378123, + "learning_rate": 9.659772965965113e-06, + "loss": 0.3509, + "step": 1040 + }, + { + "epoch": 0.14, + "grad_norm": 3.198947590498284, + "learning_rate": 9.658956195068144e-06, + "loss": 0.3697, + "step": 1041 + }, + { + "epoch": 0.14, + "grad_norm": 2.5648524124265615, + "learning_rate": 9.658138479564697e-06, + "loss": 0.3869, + "step": 1042 + }, + { + "epoch": 0.14, + "grad_norm": 2.0583743128028282, + "learning_rate": 9.657319819620566e-06, + "loss": 0.3704, + "step": 1043 + }, + { + "epoch": 0.15, + "grad_norm": 2.3124805056995896, + "learning_rate": 9.656500215401734e-06, + "loss": 0.3681, + "step": 1044 + }, + { + "epoch": 0.15, + "grad_norm": 2.7356708322225938, + "learning_rate": 9.655679667074378e-06, + "loss": 0.3852, + "step": 1045 + }, + { + "epoch": 0.15, + "grad_norm": 2.055117807511721, + "learning_rate": 9.654858174804862e-06, + "loss": 0.3858, + "step": 1046 + }, + { + "epoch": 0.15, + "grad_norm": 3.3554064073578376, + "learning_rate": 9.654035738759745e-06, + "loss": 0.3668, + "step": 1047 + }, + { + "epoch": 0.15, + "grad_norm": 1.9755636275402724, + "learning_rate": 9.653212359105776e-06, + "loss": 0.3721, + "step": 1048 + }, + { + "epoch": 0.15, + "grad_norm": 2.1424161873726133, + "learning_rate": 9.652388036009896e-06, + "loss": 0.3709, + "step": 1049 + }, + { + "epoch": 0.15, + "grad_norm": 1.9438873621574704, + "learning_rate": 9.65156276963924e-06, + "loss": 0.3712, + "step": 1050 + }, + { + "epoch": 0.15, + "grad_norm": 2.2040294912896927, + "learning_rate": 9.650736560161127e-06, + "loss": 0.3729, + "step": 1051 + }, + { + "epoch": 0.15, + "grad_norm": 7.373119136091199, + "learning_rate": 9.649909407743074e-06, + "loss": 0.3572, + "step": 1052 + }, + { + "epoch": 0.15, + "grad_norm": 2.354701556595423, + "learning_rate": 9.649081312552786e-06, + "loss": 0.3793, + "step": 1053 + }, + { + "epoch": 0.15, + "grad_norm": 2.236637350493165, + "learning_rate": 9.648252274758158e-06, + "loss": 0.3507, + "step": 1054 + }, + { + "epoch": 0.15, + "grad_norm": 2.4264456249978243, + "learning_rate": 9.647422294527283e-06, + "loss": 0.3714, + "step": 1055 + }, + { + "epoch": 0.15, + "grad_norm": 2.551745737099854, + "learning_rate": 9.646591372028435e-06, + "loss": 0.366, + "step": 1056 + }, + { + "epoch": 0.15, + "grad_norm": 2.1493138256115247, + "learning_rate": 9.645759507430088e-06, + "loss": 0.378, + "step": 1057 + }, + { + "epoch": 0.15, + "grad_norm": 3.3672673022792057, + "learning_rate": 9.644926700900898e-06, + "loss": 0.3707, + "step": 1058 + }, + { + "epoch": 0.15, + "grad_norm": 1.9623073536629398, + "learning_rate": 9.644092952609724e-06, + "loss": 0.3757, + "step": 1059 + }, + { + "epoch": 0.15, + "grad_norm": 2.191963082186815, + "learning_rate": 9.643258262725602e-06, + "loss": 0.3807, + "step": 1060 + }, + { + "epoch": 0.15, + "grad_norm": 2.2811541011633274, + "learning_rate": 9.64242263141777e-06, + "loss": 0.3597, + "step": 1061 + }, + { + "epoch": 0.15, + "grad_norm": 3.578007071492446, + "learning_rate": 9.641586058855652e-06, + "loss": 0.38, + "step": 1062 + }, + { + "epoch": 0.15, + "grad_norm": 3.019394875239035, + "learning_rate": 9.640748545208862e-06, + "loss": 0.3532, + "step": 1063 + }, + { + "epoch": 0.15, + "grad_norm": 2.26459758799543, + "learning_rate": 9.639910090647211e-06, + "loss": 0.4242, + "step": 1064 + }, + { + "epoch": 0.15, + "grad_norm": 3.4523563111854316, + "learning_rate": 9.639070695340692e-06, + "loss": 0.3549, + "step": 1065 + }, + { + "epoch": 0.15, + "grad_norm": 2.070217936987426, + "learning_rate": 9.638230359459493e-06, + "loss": 0.3516, + "step": 1066 + }, + { + "epoch": 0.15, + "grad_norm": 2.7674439367129198, + "learning_rate": 9.637389083173996e-06, + "loss": 0.3921, + "step": 1067 + }, + { + "epoch": 0.15, + "grad_norm": 2.126836626837346, + "learning_rate": 9.636546866654766e-06, + "loss": 0.4037, + "step": 1068 + }, + { + "epoch": 0.15, + "grad_norm": 2.115847600332281, + "learning_rate": 9.635703710072566e-06, + "loss": 0.3376, + "step": 1069 + }, + { + "epoch": 0.15, + "grad_norm": 6.84717209629175, + "learning_rate": 9.634859613598346e-06, + "loss": 0.3455, + "step": 1070 + }, + { + "epoch": 0.15, + "grad_norm": 4.899221713365408, + "learning_rate": 9.634014577403247e-06, + "loss": 0.3543, + "step": 1071 + }, + { + "epoch": 0.15, + "grad_norm": 1.8914595959349885, + "learning_rate": 9.6331686016586e-06, + "loss": 0.3736, + "step": 1072 + }, + { + "epoch": 0.15, + "grad_norm": 2.117841449340828, + "learning_rate": 9.632321686535928e-06, + "loss": 0.3632, + "step": 1073 + }, + { + "epoch": 0.15, + "grad_norm": 2.343561718651005, + "learning_rate": 9.631473832206944e-06, + "loss": 0.3519, + "step": 1074 + }, + { + "epoch": 0.15, + "grad_norm": 2.13984049372159, + "learning_rate": 9.63062503884355e-06, + "loss": 0.3471, + "step": 1075 + }, + { + "epoch": 0.15, + "grad_norm": 2.540003881058811, + "learning_rate": 9.629775306617838e-06, + "loss": 0.3732, + "step": 1076 + }, + { + "epoch": 0.15, + "grad_norm": 4.340522239511739, + "learning_rate": 9.628924635702095e-06, + "loss": 0.3697, + "step": 1077 + }, + { + "epoch": 0.15, + "grad_norm": 2.0290599215183946, + "learning_rate": 9.628073026268794e-06, + "loss": 0.3408, + "step": 1078 + }, + { + "epoch": 0.15, + "grad_norm": 1.9147571420435043, + "learning_rate": 9.627220478490597e-06, + "loss": 0.3463, + "step": 1079 + }, + { + "epoch": 0.15, + "grad_norm": 3.00124252058213, + "learning_rate": 9.626366992540363e-06, + "loss": 0.3487, + "step": 1080 + }, + { + "epoch": 0.15, + "grad_norm": 2.5728143375376287, + "learning_rate": 9.625512568591133e-06, + "loss": 0.3586, + "step": 1081 + }, + { + "epoch": 0.15, + "grad_norm": 2.0030423061878664, + "learning_rate": 9.624657206816144e-06, + "loss": 0.3676, + "step": 1082 + }, + { + "epoch": 0.15, + "grad_norm": 2.858626394453921, + "learning_rate": 9.623800907388821e-06, + "loss": 0.373, + "step": 1083 + }, + { + "epoch": 0.15, + "grad_norm": 1.7797993700670889, + "learning_rate": 9.622943670482777e-06, + "loss": 0.3649, + "step": 1084 + }, + { + "epoch": 0.15, + "grad_norm": 2.1728557456014945, + "learning_rate": 9.622085496271821e-06, + "loss": 0.3549, + "step": 1085 + }, + { + "epoch": 0.15, + "grad_norm": 3.7261916906585126, + "learning_rate": 9.621226384929945e-06, + "loss": 0.3361, + "step": 1086 + }, + { + "epoch": 0.15, + "grad_norm": 2.1491312452727542, + "learning_rate": 9.620366336631338e-06, + "loss": 0.3794, + "step": 1087 + }, + { + "epoch": 0.15, + "grad_norm": 2.011718804934056, + "learning_rate": 9.619505351550373e-06, + "loss": 0.3668, + "step": 1088 + }, + { + "epoch": 0.15, + "grad_norm": 1.9713387478970226, + "learning_rate": 9.618643429861615e-06, + "loss": 0.3695, + "step": 1089 + }, + { + "epoch": 0.15, + "grad_norm": 2.4723728560519285, + "learning_rate": 9.617780571739821e-06, + "loss": 0.334, + "step": 1090 + }, + { + "epoch": 0.15, + "grad_norm": 2.9634222952696634, + "learning_rate": 9.616916777359933e-06, + "loss": 0.3952, + "step": 1091 + }, + { + "epoch": 0.15, + "grad_norm": 3.063651878331853, + "learning_rate": 9.61605204689709e-06, + "loss": 0.3645, + "step": 1092 + }, + { + "epoch": 0.15, + "grad_norm": 3.4332093036092832, + "learning_rate": 9.615186380526613e-06, + "loss": 0.3572, + "step": 1093 + }, + { + "epoch": 0.15, + "grad_norm": 2.161505029655121, + "learning_rate": 9.61431977842402e-06, + "loss": 0.3771, + "step": 1094 + }, + { + "epoch": 0.15, + "grad_norm": 1.9627393635742698, + "learning_rate": 9.613452240765011e-06, + "loss": 0.3579, + "step": 1095 + }, + { + "epoch": 0.15, + "grad_norm": 1.861070510866028, + "learning_rate": 9.612583767725483e-06, + "loss": 0.3843, + "step": 1096 + }, + { + "epoch": 0.15, + "grad_norm": 1.9734153157995753, + "learning_rate": 9.611714359481518e-06, + "loss": 0.3903, + "step": 1097 + }, + { + "epoch": 0.15, + "grad_norm": 1.8711857390468645, + "learning_rate": 9.610844016209388e-06, + "loss": 0.3211, + "step": 1098 + }, + { + "epoch": 0.15, + "grad_norm": 2.7252810424257756, + "learning_rate": 9.60997273808556e-06, + "loss": 0.3773, + "step": 1099 + }, + { + "epoch": 0.15, + "grad_norm": 2.2148506486198025, + "learning_rate": 9.609100525286684e-06, + "loss": 0.3634, + "step": 1100 + }, + { + "epoch": 0.15, + "grad_norm": 2.5200950241324844, + "learning_rate": 9.608227377989599e-06, + "loss": 0.3679, + "step": 1101 + }, + { + "epoch": 0.15, + "grad_norm": 2.866096971850609, + "learning_rate": 9.607353296371338e-06, + "loss": 0.3584, + "step": 1102 + }, + { + "epoch": 0.15, + "grad_norm": 2.396907259365329, + "learning_rate": 9.606478280609124e-06, + "loss": 0.3515, + "step": 1103 + }, + { + "epoch": 0.15, + "grad_norm": 3.082075851324394, + "learning_rate": 9.605602330880363e-06, + "loss": 0.3585, + "step": 1104 + }, + { + "epoch": 0.15, + "grad_norm": 2.2554893876212887, + "learning_rate": 9.604725447362657e-06, + "loss": 0.3915, + "step": 1105 + }, + { + "epoch": 0.15, + "grad_norm": 3.354083245949654, + "learning_rate": 9.603847630233795e-06, + "loss": 0.3924, + "step": 1106 + }, + { + "epoch": 0.15, + "grad_norm": 1.8998265620620483, + "learning_rate": 9.602968879671754e-06, + "loss": 0.3812, + "step": 1107 + }, + { + "epoch": 0.15, + "grad_norm": 2.8477076040052403, + "learning_rate": 9.602089195854701e-06, + "loss": 0.3997, + "step": 1108 + }, + { + "epoch": 0.15, + "grad_norm": 2.260098232307879, + "learning_rate": 9.601208578960995e-06, + "loss": 0.3777, + "step": 1109 + }, + { + "epoch": 0.15, + "grad_norm": 1.972694786551595, + "learning_rate": 9.600327029169176e-06, + "loss": 0.3833, + "step": 1110 + }, + { + "epoch": 0.15, + "grad_norm": 5.89475099603561, + "learning_rate": 9.599444546657985e-06, + "loss": 0.37, + "step": 1111 + }, + { + "epoch": 0.15, + "grad_norm": 1.7151330647720753, + "learning_rate": 9.598561131606341e-06, + "loss": 0.3632, + "step": 1112 + }, + { + "epoch": 0.15, + "grad_norm": 2.2725330744601933, + "learning_rate": 9.59767678419336e-06, + "loss": 0.3618, + "step": 1113 + }, + { + "epoch": 0.15, + "grad_norm": 2.042966765008003, + "learning_rate": 9.596791504598345e-06, + "loss": 0.3614, + "step": 1114 + }, + { + "epoch": 0.16, + "grad_norm": 2.0911111372083493, + "learning_rate": 9.595905293000785e-06, + "loss": 0.3632, + "step": 1115 + }, + { + "epoch": 0.16, + "grad_norm": 4.718620253403495, + "learning_rate": 9.59501814958036e-06, + "loss": 0.3544, + "step": 1116 + }, + { + "epoch": 0.16, + "grad_norm": 3.163712070931446, + "learning_rate": 9.594130074516938e-06, + "loss": 0.3713, + "step": 1117 + }, + { + "epoch": 0.16, + "grad_norm": 2.1671015442813393, + "learning_rate": 9.59324106799058e-06, + "loss": 0.3659, + "step": 1118 + }, + { + "epoch": 0.16, + "grad_norm": 1.9368198417467493, + "learning_rate": 9.592351130181528e-06, + "loss": 0.3601, + "step": 1119 + }, + { + "epoch": 0.16, + "grad_norm": 2.5425444305644795, + "learning_rate": 9.591460261270223e-06, + "loss": 0.3567, + "step": 1120 + }, + { + "epoch": 0.16, + "grad_norm": 2.095552655854023, + "learning_rate": 9.590568461437285e-06, + "loss": 0.3367, + "step": 1121 + }, + { + "epoch": 0.16, + "grad_norm": 3.6166903909548487, + "learning_rate": 9.58967573086353e-06, + "loss": 0.3682, + "step": 1122 + }, + { + "epoch": 0.16, + "grad_norm": 3.6790021334392398, + "learning_rate": 9.588782069729956e-06, + "loss": 0.3554, + "step": 1123 + }, + { + "epoch": 0.16, + "grad_norm": 3.6881157981279857, + "learning_rate": 9.587887478217756e-06, + "loss": 0.3676, + "step": 1124 + }, + { + "epoch": 0.16, + "grad_norm": 2.2531988870666066, + "learning_rate": 9.58699195650831e-06, + "loss": 0.3836, + "step": 1125 + }, + { + "epoch": 0.16, + "grad_norm": 2.0609002007714667, + "learning_rate": 9.586095504783184e-06, + "loss": 0.3667, + "step": 1126 + }, + { + "epoch": 0.16, + "grad_norm": 2.0985863060324386, + "learning_rate": 9.585198123224134e-06, + "loss": 0.3678, + "step": 1127 + }, + { + "epoch": 0.16, + "grad_norm": 2.007949641894732, + "learning_rate": 9.584299812013105e-06, + "loss": 0.3584, + "step": 1128 + }, + { + "epoch": 0.16, + "grad_norm": 2.22595068117556, + "learning_rate": 9.583400571332231e-06, + "loss": 0.384, + "step": 1129 + }, + { + "epoch": 0.16, + "grad_norm": 1.8990869159543864, + "learning_rate": 9.582500401363833e-06, + "loss": 0.3331, + "step": 1130 + }, + { + "epoch": 0.16, + "grad_norm": 2.200216407371265, + "learning_rate": 9.581599302290421e-06, + "loss": 0.3432, + "step": 1131 + }, + { + "epoch": 0.16, + "grad_norm": 2.9410511627192575, + "learning_rate": 9.580697274294693e-06, + "loss": 0.3638, + "step": 1132 + }, + { + "epoch": 0.16, + "grad_norm": 2.2723708703926473, + "learning_rate": 9.579794317559536e-06, + "loss": 0.3668, + "step": 1133 + }, + { + "epoch": 0.16, + "grad_norm": 2.642164140538499, + "learning_rate": 9.578890432268025e-06, + "loss": 0.3413, + "step": 1134 + }, + { + "epoch": 0.16, + "grad_norm": 2.658766208650648, + "learning_rate": 9.577985618603424e-06, + "loss": 0.3866, + "step": 1135 + }, + { + "epoch": 0.16, + "grad_norm": 2.3572495458346623, + "learning_rate": 9.577079876749185e-06, + "loss": 0.3254, + "step": 1136 + }, + { + "epoch": 0.16, + "grad_norm": 2.583477086514517, + "learning_rate": 9.576173206888949e-06, + "loss": 0.3857, + "step": 1137 + }, + { + "epoch": 0.16, + "grad_norm": 2.065910526590912, + "learning_rate": 9.575265609206539e-06, + "loss": 0.3464, + "step": 1138 + }, + { + "epoch": 0.16, + "grad_norm": 2.4659552640438283, + "learning_rate": 9.574357083885974e-06, + "loss": 0.3448, + "step": 1139 + }, + { + "epoch": 0.16, + "grad_norm": 2.3795684031568785, + "learning_rate": 9.57344763111146e-06, + "loss": 0.3506, + "step": 1140 + }, + { + "epoch": 0.16, + "grad_norm": 2.486966250376723, + "learning_rate": 9.572537251067386e-06, + "loss": 0.3942, + "step": 1141 + }, + { + "epoch": 0.16, + "grad_norm": 2.029949423913536, + "learning_rate": 9.571625943938333e-06, + "loss": 0.349, + "step": 1142 + }, + { + "epoch": 0.16, + "grad_norm": 2.621332679646285, + "learning_rate": 9.570713709909069e-06, + "loss": 0.3559, + "step": 1143 + }, + { + "epoch": 0.16, + "grad_norm": 2.02533904655617, + "learning_rate": 9.569800549164551e-06, + "loss": 0.382, + "step": 1144 + }, + { + "epoch": 0.16, + "grad_norm": 7.1698513142449185, + "learning_rate": 9.568886461889925e-06, + "loss": 0.3592, + "step": 1145 + }, + { + "epoch": 0.16, + "grad_norm": 2.1205883452364422, + "learning_rate": 9.567971448270518e-06, + "loss": 0.3802, + "step": 1146 + }, + { + "epoch": 0.16, + "grad_norm": 3.8474685709405114, + "learning_rate": 9.567055508491854e-06, + "loss": 0.3997, + "step": 1147 + }, + { + "epoch": 0.16, + "grad_norm": 2.168238876722413, + "learning_rate": 9.566138642739636e-06, + "loss": 0.3856, + "step": 1148 + }, + { + "epoch": 0.16, + "grad_norm": 2.1577243793219516, + "learning_rate": 9.565220851199764e-06, + "loss": 0.384, + "step": 1149 + }, + { + "epoch": 0.16, + "grad_norm": 2.1648044255836605, + "learning_rate": 9.564302134058317e-06, + "loss": 0.3467, + "step": 1150 + }, + { + "epoch": 0.16, + "grad_norm": 1.7846710636662575, + "learning_rate": 9.563382491501565e-06, + "loss": 0.3468, + "step": 1151 + }, + { + "epoch": 0.16, + "grad_norm": 9.891095937088908, + "learning_rate": 9.56246192371597e-06, + "loss": 0.355, + "step": 1152 + }, + { + "epoch": 0.16, + "grad_norm": 1.8170539582502374, + "learning_rate": 9.561540430888177e-06, + "loss": 0.3657, + "step": 1153 + }, + { + "epoch": 0.16, + "grad_norm": 3.1927515070531642, + "learning_rate": 9.560618013205017e-06, + "loss": 0.3797, + "step": 1154 + }, + { + "epoch": 0.16, + "grad_norm": 3.014347309286274, + "learning_rate": 9.55969467085351e-06, + "loss": 0.3931, + "step": 1155 + }, + { + "epoch": 0.16, + "grad_norm": 2.1945049049865912, + "learning_rate": 9.558770404020869e-06, + "loss": 0.352, + "step": 1156 + }, + { + "epoch": 0.16, + "grad_norm": 2.83842617601632, + "learning_rate": 9.557845212894486e-06, + "loss": 0.351, + "step": 1157 + }, + { + "epoch": 0.16, + "grad_norm": 2.3649451160096597, + "learning_rate": 9.556919097661944e-06, + "loss": 0.3592, + "step": 1158 + }, + { + "epoch": 0.16, + "grad_norm": 2.463926190589076, + "learning_rate": 9.555992058511014e-06, + "loss": 0.3877, + "step": 1159 + }, + { + "epoch": 0.16, + "grad_norm": 1.6735783309665924, + "learning_rate": 9.555064095629656e-06, + "loss": 0.3628, + "step": 1160 + }, + { + "epoch": 0.16, + "grad_norm": 2.594414614261818, + "learning_rate": 9.554135209206013e-06, + "loss": 0.3441, + "step": 1161 + }, + { + "epoch": 0.16, + "grad_norm": 3.4160373039574194, + "learning_rate": 9.553205399428417e-06, + "loss": 0.3854, + "step": 1162 + }, + { + "epoch": 0.16, + "grad_norm": 2.7052402109287086, + "learning_rate": 9.552274666485388e-06, + "loss": 0.3457, + "step": 1163 + }, + { + "epoch": 0.16, + "grad_norm": 2.4958488800656395, + "learning_rate": 9.551343010565635e-06, + "loss": 0.3609, + "step": 1164 + }, + { + "epoch": 0.16, + "grad_norm": 2.6453572263400846, + "learning_rate": 9.550410431858049e-06, + "loss": 0.3637, + "step": 1165 + }, + { + "epoch": 0.16, + "grad_norm": 2.242192021767899, + "learning_rate": 9.54947693055171e-06, + "loss": 0.3761, + "step": 1166 + }, + { + "epoch": 0.16, + "grad_norm": 2.6005072666232576, + "learning_rate": 9.54854250683589e-06, + "loss": 0.3763, + "step": 1167 + }, + { + "epoch": 0.16, + "grad_norm": 2.2028290586285832, + "learning_rate": 9.547607160900038e-06, + "loss": 0.3346, + "step": 1168 + }, + { + "epoch": 0.16, + "grad_norm": 3.429903468409909, + "learning_rate": 9.546670892933804e-06, + "loss": 0.3696, + "step": 1169 + }, + { + "epoch": 0.16, + "grad_norm": 2.163853503176643, + "learning_rate": 9.54573370312701e-06, + "loss": 0.3761, + "step": 1170 + }, + { + "epoch": 0.16, + "grad_norm": 3.026889859658365, + "learning_rate": 9.544795591669678e-06, + "loss": 0.3844, + "step": 1171 + }, + { + "epoch": 0.16, + "grad_norm": 2.010932073114678, + "learning_rate": 9.543856558752006e-06, + "loss": 0.3827, + "step": 1172 + }, + { + "epoch": 0.16, + "grad_norm": 4.083060808405731, + "learning_rate": 9.542916604564384e-06, + "loss": 0.364, + "step": 1173 + }, + { + "epoch": 0.16, + "grad_norm": 2.3379368359788133, + "learning_rate": 9.54197572929739e-06, + "loss": 0.3661, + "step": 1174 + }, + { + "epoch": 0.16, + "grad_norm": 2.0493544786256064, + "learning_rate": 9.541033933141786e-06, + "loss": 0.3548, + "step": 1175 + }, + { + "epoch": 0.16, + "grad_norm": 1.8725152987670766, + "learning_rate": 9.540091216288522e-06, + "loss": 0.3647, + "step": 1176 + }, + { + "epoch": 0.16, + "grad_norm": 2.9056740690543927, + "learning_rate": 9.539147578928735e-06, + "loss": 0.3612, + "step": 1177 + }, + { + "epoch": 0.16, + "grad_norm": 2.434520121723996, + "learning_rate": 9.538203021253747e-06, + "loss": 0.3695, + "step": 1178 + }, + { + "epoch": 0.16, + "grad_norm": 2.1295602762063384, + "learning_rate": 9.53725754345507e-06, + "loss": 0.3967, + "step": 1179 + }, + { + "epoch": 0.16, + "grad_norm": 2.5668499365260598, + "learning_rate": 9.536311145724398e-06, + "loss": 0.3885, + "step": 1180 + }, + { + "epoch": 0.16, + "grad_norm": 2.1924560243477975, + "learning_rate": 9.535363828253613e-06, + "loss": 0.3768, + "step": 1181 + }, + { + "epoch": 0.16, + "grad_norm": 1.7452252770439616, + "learning_rate": 9.534415591234787e-06, + "loss": 0.3534, + "step": 1182 + }, + { + "epoch": 0.16, + "grad_norm": 2.010932001547967, + "learning_rate": 9.533466434860174e-06, + "loss": 0.3652, + "step": 1183 + }, + { + "epoch": 0.16, + "grad_norm": 2.004381190155336, + "learning_rate": 9.532516359322218e-06, + "loss": 0.3569, + "step": 1184 + }, + { + "epoch": 0.16, + "grad_norm": 2.139431937552467, + "learning_rate": 9.531565364813544e-06, + "loss": 0.3684, + "step": 1185 + }, + { + "epoch": 0.16, + "grad_norm": 3.846213552238962, + "learning_rate": 9.53061345152697e-06, + "loss": 0.4033, + "step": 1186 + }, + { + "epoch": 0.17, + "grad_norm": 1.955783474479744, + "learning_rate": 9.529660619655494e-06, + "loss": 0.3785, + "step": 1187 + }, + { + "epoch": 0.17, + "grad_norm": 1.9245290242199105, + "learning_rate": 9.528706869392307e-06, + "loss": 0.3605, + "step": 1188 + }, + { + "epoch": 0.17, + "grad_norm": 1.8229365412514245, + "learning_rate": 9.52775220093078e-06, + "loss": 0.3443, + "step": 1189 + }, + { + "epoch": 0.17, + "grad_norm": 2.355517394513506, + "learning_rate": 9.526796614464474e-06, + "loss": 0.3717, + "step": 1190 + }, + { + "epoch": 0.17, + "grad_norm": 1.9512489815656289, + "learning_rate": 9.525840110187132e-06, + "loss": 0.3599, + "step": 1191 + }, + { + "epoch": 0.17, + "grad_norm": 2.2937900515127687, + "learning_rate": 9.524882688292688e-06, + "loss": 0.3667, + "step": 1192 + }, + { + "epoch": 0.17, + "grad_norm": 1.8626708495831679, + "learning_rate": 9.523924348975261e-06, + "loss": 0.3663, + "step": 1193 + }, + { + "epoch": 0.17, + "grad_norm": 2.4294067642071493, + "learning_rate": 9.522965092429154e-06, + "loss": 0.3646, + "step": 1194 + }, + { + "epoch": 0.17, + "grad_norm": 2.240874813516777, + "learning_rate": 9.522004918848857e-06, + "loss": 0.3714, + "step": 1195 + }, + { + "epoch": 0.17, + "grad_norm": 2.1513814833710088, + "learning_rate": 9.521043828429044e-06, + "loss": 0.3555, + "step": 1196 + }, + { + "epoch": 0.17, + "grad_norm": 2.2388290144926266, + "learning_rate": 9.520081821364578e-06, + "loss": 0.3633, + "step": 1197 + }, + { + "epoch": 0.17, + "grad_norm": 2.34341333187234, + "learning_rate": 9.519118897850509e-06, + "loss": 0.3616, + "step": 1198 + }, + { + "epoch": 0.17, + "grad_norm": 2.7229989538849666, + "learning_rate": 9.518155058082066e-06, + "loss": 0.3749, + "step": 1199 + }, + { + "epoch": 0.17, + "grad_norm": 2.04459009305544, + "learning_rate": 9.517190302254671e-06, + "loss": 0.3726, + "step": 1200 + }, + { + "epoch": 0.17, + "grad_norm": 1.8148512178816867, + "learning_rate": 9.516224630563928e-06, + "loss": 0.3805, + "step": 1201 + }, + { + "epoch": 0.17, + "grad_norm": 1.9411568800764762, + "learning_rate": 9.515258043205628e-06, + "loss": 0.3861, + "step": 1202 + }, + { + "epoch": 0.17, + "grad_norm": 1.7686124768842317, + "learning_rate": 9.514290540375749e-06, + "loss": 0.3675, + "step": 1203 + }, + { + "epoch": 0.17, + "grad_norm": 2.0070145543317395, + "learning_rate": 9.513322122270449e-06, + "loss": 0.3818, + "step": 1204 + }, + { + "epoch": 0.17, + "grad_norm": 1.7016668144149885, + "learning_rate": 9.512352789086078e-06, + "loss": 0.3463, + "step": 1205 + }, + { + "epoch": 0.17, + "grad_norm": 1.9470660802857211, + "learning_rate": 9.511382541019167e-06, + "loss": 0.3783, + "step": 1206 + }, + { + "epoch": 0.17, + "grad_norm": 2.5011887661703014, + "learning_rate": 9.510411378266437e-06, + "loss": 0.3438, + "step": 1207 + }, + { + "epoch": 0.17, + "grad_norm": 2.960034009699605, + "learning_rate": 9.50943930102479e-06, + "loss": 0.3649, + "step": 1208 + }, + { + "epoch": 0.17, + "grad_norm": 4.468435032951581, + "learning_rate": 9.508466309491315e-06, + "loss": 0.3588, + "step": 1209 + }, + { + "epoch": 0.17, + "grad_norm": 5.932763517290316, + "learning_rate": 9.50749240386329e-06, + "loss": 0.3804, + "step": 1210 + }, + { + "epoch": 0.17, + "grad_norm": 2.0387090213197876, + "learning_rate": 9.50651758433817e-06, + "loss": 0.3799, + "step": 1211 + }, + { + "epoch": 0.17, + "grad_norm": 1.9968043594731677, + "learning_rate": 9.505541851113604e-06, + "loss": 0.3762, + "step": 1212 + }, + { + "epoch": 0.17, + "grad_norm": 1.9146468060023623, + "learning_rate": 9.504565204387422e-06, + "loss": 0.3713, + "step": 1213 + }, + { + "epoch": 0.17, + "grad_norm": 2.696209622948401, + "learning_rate": 9.503587644357639e-06, + "loss": 0.3559, + "step": 1214 + }, + { + "epoch": 0.17, + "grad_norm": 2.1987575517373346, + "learning_rate": 9.502609171222456e-06, + "loss": 0.3691, + "step": 1215 + }, + { + "epoch": 0.17, + "grad_norm": 2.2667762681097954, + "learning_rate": 9.50162978518026e-06, + "loss": 0.3721, + "step": 1216 + }, + { + "epoch": 0.17, + "grad_norm": 2.6767691516061025, + "learning_rate": 9.500649486429621e-06, + "loss": 0.3842, + "step": 1217 + }, + { + "epoch": 0.17, + "grad_norm": 2.1811181805500612, + "learning_rate": 9.499668275169294e-06, + "loss": 0.3561, + "step": 1218 + }, + { + "epoch": 0.17, + "grad_norm": 1.9423034398650803, + "learning_rate": 9.498686151598224e-06, + "loss": 0.3624, + "step": 1219 + }, + { + "epoch": 0.17, + "grad_norm": 2.0865843397455515, + "learning_rate": 9.497703115915533e-06, + "loss": 0.3682, + "step": 1220 + }, + { + "epoch": 0.17, + "grad_norm": 2.050852854777589, + "learning_rate": 9.496719168320536e-06, + "loss": 0.3682, + "step": 1221 + }, + { + "epoch": 0.17, + "grad_norm": 4.766207289594002, + "learning_rate": 9.495734309012728e-06, + "loss": 0.3804, + "step": 1222 + }, + { + "epoch": 0.17, + "grad_norm": 2.343269789461014, + "learning_rate": 9.494748538191787e-06, + "loss": 0.3396, + "step": 1223 + }, + { + "epoch": 0.17, + "grad_norm": 2.0276744706957026, + "learning_rate": 9.493761856057582e-06, + "loss": 0.3657, + "step": 1224 + }, + { + "epoch": 0.17, + "grad_norm": 2.447313427390787, + "learning_rate": 9.492774262810163e-06, + "loss": 0.3709, + "step": 1225 + }, + { + "epoch": 0.17, + "grad_norm": 2.2557328356496438, + "learning_rate": 9.491785758649764e-06, + "loss": 0.3463, + "step": 1226 + }, + { + "epoch": 0.17, + "grad_norm": 2.2622872702951478, + "learning_rate": 9.490796343776805e-06, + "loss": 0.3436, + "step": 1227 + }, + { + "epoch": 0.17, + "grad_norm": 1.934398007315207, + "learning_rate": 9.489806018391892e-06, + "loss": 0.38, + "step": 1228 + }, + { + "epoch": 0.17, + "grad_norm": 2.402041398508939, + "learning_rate": 9.488814782695812e-06, + "loss": 0.3647, + "step": 1229 + }, + { + "epoch": 0.17, + "grad_norm": 2.069224345334658, + "learning_rate": 9.48782263688954e-06, + "loss": 0.3609, + "step": 1230 + }, + { + "epoch": 0.17, + "grad_norm": 2.5657816471744432, + "learning_rate": 9.486829581174235e-06, + "loss": 0.3862, + "step": 1231 + }, + { + "epoch": 0.17, + "grad_norm": 2.2691310450791464, + "learning_rate": 9.485835615751237e-06, + "loss": 0.3817, + "step": 1232 + }, + { + "epoch": 0.17, + "grad_norm": 2.3247365393240718, + "learning_rate": 9.484840740822074e-06, + "loss": 0.3574, + "step": 1233 + }, + { + "epoch": 0.17, + "grad_norm": 2.1499931493778948, + "learning_rate": 9.483844956588462e-06, + "loss": 0.3611, + "step": 1234 + }, + { + "epoch": 0.17, + "grad_norm": 9.294834826322472, + "learning_rate": 9.48284826325229e-06, + "loss": 0.3789, + "step": 1235 + }, + { + "epoch": 0.17, + "grad_norm": 2.550569193076414, + "learning_rate": 9.481850661015643e-06, + "loss": 0.3667, + "step": 1236 + }, + { + "epoch": 0.17, + "grad_norm": 2.367862997274986, + "learning_rate": 9.480852150080784e-06, + "loss": 0.3747, + "step": 1237 + }, + { + "epoch": 0.17, + "grad_norm": 2.7501819334558704, + "learning_rate": 9.47985273065016e-06, + "loss": 0.3686, + "step": 1238 + }, + { + "epoch": 0.17, + "grad_norm": 1.6471392668204439, + "learning_rate": 9.47885240292641e-06, + "loss": 0.3501, + "step": 1239 + }, + { + "epoch": 0.17, + "grad_norm": 2.151239011286587, + "learning_rate": 9.477851167112344e-06, + "loss": 0.3561, + "step": 1240 + }, + { + "epoch": 0.17, + "grad_norm": 2.2865754933225686, + "learning_rate": 9.476849023410965e-06, + "loss": 0.3487, + "step": 1241 + }, + { + "epoch": 0.17, + "grad_norm": 2.6300985397121317, + "learning_rate": 9.475845972025462e-06, + "loss": 0.3685, + "step": 1242 + }, + { + "epoch": 0.17, + "grad_norm": 1.9745594218627454, + "learning_rate": 9.4748420131592e-06, + "loss": 0.3692, + "step": 1243 + }, + { + "epoch": 0.17, + "grad_norm": 4.28758963313779, + "learning_rate": 9.473837147015734e-06, + "loss": 0.3511, + "step": 1244 + }, + { + "epoch": 0.17, + "grad_norm": 1.9214585897105314, + "learning_rate": 9.472831373798802e-06, + "loss": 0.3548, + "step": 1245 + }, + { + "epoch": 0.17, + "grad_norm": 2.4956974319650094, + "learning_rate": 9.471824693712325e-06, + "loss": 0.3362, + "step": 1246 + }, + { + "epoch": 0.17, + "grad_norm": 2.703113151082221, + "learning_rate": 9.470817106960407e-06, + "loss": 0.3918, + "step": 1247 + }, + { + "epoch": 0.17, + "grad_norm": 2.49360691902255, + "learning_rate": 9.469808613747338e-06, + "loss": 0.3455, + "step": 1248 + }, + { + "epoch": 0.17, + "grad_norm": 2.6585335577262548, + "learning_rate": 9.468799214277591e-06, + "loss": 0.3682, + "step": 1249 + }, + { + "epoch": 0.17, + "grad_norm": 2.1592289052713514, + "learning_rate": 9.467788908755821e-06, + "loss": 0.3888, + "step": 1250 + }, + { + "epoch": 0.17, + "grad_norm": 1.9331694572451836, + "learning_rate": 9.466777697386868e-06, + "loss": 0.3605, + "step": 1251 + }, + { + "epoch": 0.17, + "grad_norm": 2.224225936606841, + "learning_rate": 9.465765580375757e-06, + "loss": 0.3509, + "step": 1252 + }, + { + "epoch": 0.17, + "grad_norm": 3.285637317323941, + "learning_rate": 9.464752557927695e-06, + "loss": 0.3681, + "step": 1253 + }, + { + "epoch": 0.17, + "grad_norm": 2.0070201718337906, + "learning_rate": 9.463738630248073e-06, + "loss": 0.3574, + "step": 1254 + }, + { + "epoch": 0.17, + "grad_norm": 1.7638792246367772, + "learning_rate": 9.462723797542465e-06, + "loss": 0.3801, + "step": 1255 + }, + { + "epoch": 0.17, + "grad_norm": 2.5700230802063526, + "learning_rate": 9.46170806001663e-06, + "loss": 0.3755, + "step": 1256 + }, + { + "epoch": 0.17, + "grad_norm": 2.7482422693717155, + "learning_rate": 9.46069141787651e-06, + "loss": 0.3832, + "step": 1257 + }, + { + "epoch": 0.17, + "grad_norm": 2.153825377057438, + "learning_rate": 9.459673871328225e-06, + "loss": 0.3637, + "step": 1258 + }, + { + "epoch": 0.18, + "grad_norm": 1.8020038888337957, + "learning_rate": 9.45865542057809e-06, + "loss": 0.3611, + "step": 1259 + }, + { + "epoch": 0.18, + "grad_norm": 2.248033455801858, + "learning_rate": 9.457636065832592e-06, + "loss": 0.3865, + "step": 1260 + }, + { + "epoch": 0.18, + "grad_norm": 2.0590674767673143, + "learning_rate": 9.456615807298408e-06, + "loss": 0.361, + "step": 1261 + }, + { + "epoch": 0.18, + "grad_norm": 2.7589014151333764, + "learning_rate": 9.455594645182396e-06, + "loss": 0.3821, + "step": 1262 + }, + { + "epoch": 0.18, + "grad_norm": 1.9817697398659642, + "learning_rate": 9.454572579691596e-06, + "loss": 0.3538, + "step": 1263 + }, + { + "epoch": 0.18, + "grad_norm": 3.0577407126649594, + "learning_rate": 9.453549611033233e-06, + "loss": 0.322, + "step": 1264 + }, + { + "epoch": 0.18, + "grad_norm": 1.8078460489215782, + "learning_rate": 9.452525739414716e-06, + "loss": 0.3516, + "step": 1265 + }, + { + "epoch": 0.18, + "grad_norm": 3.3927508708375513, + "learning_rate": 9.451500965043633e-06, + "loss": 0.3648, + "step": 1266 + }, + { + "epoch": 0.18, + "grad_norm": 1.917741430471497, + "learning_rate": 9.450475288127761e-06, + "loss": 0.3628, + "step": 1267 + }, + { + "epoch": 0.18, + "grad_norm": 2.204247493273934, + "learning_rate": 9.449448708875053e-06, + "loss": 0.3381, + "step": 1268 + }, + { + "epoch": 0.18, + "grad_norm": 1.8783304788284458, + "learning_rate": 9.448421227493651e-06, + "loss": 0.3746, + "step": 1269 + }, + { + "epoch": 0.18, + "grad_norm": 2.014208421370514, + "learning_rate": 9.447392844191879e-06, + "loss": 0.3802, + "step": 1270 + }, + { + "epoch": 0.18, + "grad_norm": 3.4942835129938974, + "learning_rate": 9.446363559178238e-06, + "loss": 0.3552, + "step": 1271 + }, + { + "epoch": 0.18, + "grad_norm": 1.6473723950137598, + "learning_rate": 9.445333372661421e-06, + "loss": 0.3637, + "step": 1272 + }, + { + "epoch": 0.18, + "grad_norm": 2.6089064653844294, + "learning_rate": 9.444302284850295e-06, + "loss": 0.3635, + "step": 1273 + }, + { + "epoch": 0.18, + "grad_norm": 2.0832589953667395, + "learning_rate": 9.443270295953914e-06, + "loss": 0.352, + "step": 1274 + }, + { + "epoch": 0.18, + "grad_norm": 2.0727846691341774, + "learning_rate": 9.442237406181518e-06, + "loss": 0.3523, + "step": 1275 + }, + { + "epoch": 0.18, + "grad_norm": 2.3852868516971393, + "learning_rate": 9.441203615742524e-06, + "loss": 0.3613, + "step": 1276 + }, + { + "epoch": 0.18, + "grad_norm": 2.157436551290936, + "learning_rate": 9.440168924846533e-06, + "loss": 0.3515, + "step": 1277 + }, + { + "epoch": 0.18, + "grad_norm": 2.039466583761825, + "learning_rate": 9.43913333370333e-06, + "loss": 0.3558, + "step": 1278 + }, + { + "epoch": 0.18, + "grad_norm": 2.240747358784542, + "learning_rate": 9.43809684252288e-06, + "loss": 0.3717, + "step": 1279 + }, + { + "epoch": 0.18, + "grad_norm": 1.929913763700084, + "learning_rate": 9.437059451515335e-06, + "loss": 0.3344, + "step": 1280 + }, + { + "epoch": 0.18, + "grad_norm": 3.2236352108743875, + "learning_rate": 9.436021160891027e-06, + "loss": 0.3842, + "step": 1281 + }, + { + "epoch": 0.18, + "grad_norm": 1.8668660208370609, + "learning_rate": 9.434981970860468e-06, + "loss": 0.3692, + "step": 1282 + }, + { + "epoch": 0.18, + "grad_norm": 2.0739620220940473, + "learning_rate": 9.433941881634357e-06, + "loss": 0.3855, + "step": 1283 + }, + { + "epoch": 0.18, + "grad_norm": 2.163740195941227, + "learning_rate": 9.432900893423567e-06, + "loss": 0.3535, + "step": 1284 + }, + { + "epoch": 0.18, + "grad_norm": 1.915836327701064, + "learning_rate": 9.431859006439166e-06, + "loss": 0.3701, + "step": 1285 + }, + { + "epoch": 0.18, + "grad_norm": 2.2173870390756645, + "learning_rate": 9.430816220892394e-06, + "loss": 0.3814, + "step": 1286 + }, + { + "epoch": 0.18, + "grad_norm": 2.587853955584811, + "learning_rate": 9.429772536994676e-06, + "loss": 0.3592, + "step": 1287 + }, + { + "epoch": 0.18, + "grad_norm": 2.4086344815947642, + "learning_rate": 9.428727954957622e-06, + "loss": 0.3635, + "step": 1288 + }, + { + "epoch": 0.18, + "grad_norm": 2.2298222703209807, + "learning_rate": 9.42768247499302e-06, + "loss": 0.3588, + "step": 1289 + }, + { + "epoch": 0.18, + "grad_norm": 2.2812259573546827, + "learning_rate": 9.42663609731284e-06, + "loss": 0.3386, + "step": 1290 + }, + { + "epoch": 0.18, + "grad_norm": 2.8263509001886638, + "learning_rate": 9.425588822129239e-06, + "loss": 0.3815, + "step": 1291 + }, + { + "epoch": 0.18, + "grad_norm": 2.239586178010626, + "learning_rate": 9.424540649654552e-06, + "loss": 0.3403, + "step": 1292 + }, + { + "epoch": 0.18, + "grad_norm": 1.7913654997493131, + "learning_rate": 9.423491580101296e-06, + "loss": 0.3481, + "step": 1293 + }, + { + "epoch": 0.18, + "grad_norm": 3.901204278173805, + "learning_rate": 9.422441613682169e-06, + "loss": 0.3789, + "step": 1294 + }, + { + "epoch": 0.18, + "grad_norm": 2.1425565149425205, + "learning_rate": 9.421390750610056e-06, + "loss": 0.3526, + "step": 1295 + }, + { + "epoch": 0.18, + "grad_norm": 2.3762258622545556, + "learning_rate": 9.420338991098018e-06, + "loss": 0.3771, + "step": 1296 + }, + { + "epoch": 0.18, + "grad_norm": 1.8636272629883026, + "learning_rate": 9.419286335359299e-06, + "loss": 0.3535, + "step": 1297 + }, + { + "epoch": 0.18, + "grad_norm": 1.691100124927992, + "learning_rate": 9.418232783607327e-06, + "loss": 0.3635, + "step": 1298 + }, + { + "epoch": 0.18, + "grad_norm": 6.752678409913943, + "learning_rate": 9.41717833605571e-06, + "loss": 0.351, + "step": 1299 + }, + { + "epoch": 0.18, + "grad_norm": 2.1359568628185746, + "learning_rate": 9.41612299291824e-06, + "loss": 0.3866, + "step": 1300 + }, + { + "epoch": 0.18, + "grad_norm": 3.167285062923421, + "learning_rate": 9.415066754408886e-06, + "loss": 0.3741, + "step": 1301 + }, + { + "epoch": 0.18, + "grad_norm": 2.258560302423304, + "learning_rate": 9.4140096207418e-06, + "loss": 0.3387, + "step": 1302 + }, + { + "epoch": 0.18, + "grad_norm": 3.013157549382052, + "learning_rate": 9.41295159213132e-06, + "loss": 0.3728, + "step": 1303 + }, + { + "epoch": 0.18, + "grad_norm": 2.1197635656708833, + "learning_rate": 9.41189266879196e-06, + "loss": 0.4041, + "step": 1304 + }, + { + "epoch": 0.18, + "grad_norm": 2.114813695637405, + "learning_rate": 9.410832850938417e-06, + "loss": 0.3703, + "step": 1305 + }, + { + "epoch": 0.18, + "grad_norm": 3.043400857840909, + "learning_rate": 9.40977213878557e-06, + "loss": 0.362, + "step": 1306 + }, + { + "epoch": 0.18, + "grad_norm": 1.9739805498861134, + "learning_rate": 9.40871053254848e-06, + "loss": 0.3567, + "step": 1307 + }, + { + "epoch": 0.18, + "grad_norm": 2.4196988652208726, + "learning_rate": 9.40764803244239e-06, + "loss": 0.3784, + "step": 1308 + }, + { + "epoch": 0.18, + "grad_norm": 1.9509093910882307, + "learning_rate": 9.406584638682717e-06, + "loss": 0.3583, + "step": 1309 + }, + { + "epoch": 0.18, + "grad_norm": 1.863528346126843, + "learning_rate": 9.405520351485071e-06, + "loss": 0.3602, + "step": 1310 + }, + { + "epoch": 0.18, + "grad_norm": 2.9061160183535963, + "learning_rate": 9.404455171065232e-06, + "loss": 0.3815, + "step": 1311 + }, + { + "epoch": 0.18, + "grad_norm": 2.6577708430493154, + "learning_rate": 9.40338909763917e-06, + "loss": 0.3705, + "step": 1312 + }, + { + "epoch": 0.18, + "grad_norm": 2.5072052761646586, + "learning_rate": 9.40232213142303e-06, + "loss": 0.3657, + "step": 1313 + }, + { + "epoch": 0.18, + "grad_norm": 1.980547565758528, + "learning_rate": 9.401254272633141e-06, + "loss": 0.3727, + "step": 1314 + }, + { + "epoch": 0.18, + "grad_norm": 1.9074029746474537, + "learning_rate": 9.400185521486011e-06, + "loss": 0.3573, + "step": 1315 + }, + { + "epoch": 0.18, + "grad_norm": 2.1007518362617352, + "learning_rate": 9.39911587819833e-06, + "loss": 0.3643, + "step": 1316 + }, + { + "epoch": 0.18, + "grad_norm": 2.5545863724481928, + "learning_rate": 9.39804534298697e-06, + "loss": 0.3585, + "step": 1317 + }, + { + "epoch": 0.18, + "grad_norm": 2.3006965297234183, + "learning_rate": 9.396973916068983e-06, + "loss": 0.3514, + "step": 1318 + }, + { + "epoch": 0.18, + "grad_norm": 2.076370595053014, + "learning_rate": 9.395901597661601e-06, + "loss": 0.3559, + "step": 1319 + }, + { + "epoch": 0.18, + "grad_norm": 4.128410865280985, + "learning_rate": 9.394828387982235e-06, + "loss": 0.3786, + "step": 1320 + }, + { + "epoch": 0.18, + "grad_norm": 2.191669013487254, + "learning_rate": 9.393754287248482e-06, + "loss": 0.377, + "step": 1321 + }, + { + "epoch": 0.18, + "grad_norm": 2.608103339628542, + "learning_rate": 9.392679295678116e-06, + "loss": 0.3412, + "step": 1322 + }, + { + "epoch": 0.18, + "grad_norm": 3.6566965854093603, + "learning_rate": 9.391603413489092e-06, + "loss": 0.3652, + "step": 1323 + }, + { + "epoch": 0.18, + "grad_norm": 2.1791157623114454, + "learning_rate": 9.390526640899545e-06, + "loss": 0.3604, + "step": 1324 + }, + { + "epoch": 0.18, + "grad_norm": 6.521521552150946, + "learning_rate": 9.389448978127793e-06, + "loss": 0.3626, + "step": 1325 + }, + { + "epoch": 0.18, + "grad_norm": 3.842930841526663, + "learning_rate": 9.38837042539233e-06, + "loss": 0.3712, + "step": 1326 + }, + { + "epoch": 0.18, + "grad_norm": 1.970341809453208, + "learning_rate": 9.387290982911838e-06, + "loss": 0.3405, + "step": 1327 + }, + { + "epoch": 0.18, + "grad_norm": 15.879703947457518, + "learning_rate": 9.38621065090517e-06, + "loss": 0.3755, + "step": 1328 + }, + { + "epoch": 0.18, + "grad_norm": 2.2103512066816804, + "learning_rate": 9.385129429591367e-06, + "loss": 0.359, + "step": 1329 + }, + { + "epoch": 0.18, + "grad_norm": 3.1019637156919044, + "learning_rate": 9.384047319189647e-06, + "loss": 0.3893, + "step": 1330 + }, + { + "epoch": 0.19, + "grad_norm": 2.7688083001914428, + "learning_rate": 9.382964319919406e-06, + "loss": 0.3745, + "step": 1331 + }, + { + "epoch": 0.19, + "grad_norm": 2.4870248647892508, + "learning_rate": 9.381880432000226e-06, + "loss": 0.3739, + "step": 1332 + }, + { + "epoch": 0.19, + "grad_norm": 1.874837851227523, + "learning_rate": 9.380795655651863e-06, + "loss": 0.364, + "step": 1333 + }, + { + "epoch": 0.19, + "grad_norm": 2.6060058446782897, + "learning_rate": 9.379709991094258e-06, + "loss": 0.3837, + "step": 1334 + }, + { + "epoch": 0.19, + "grad_norm": 2.500554423271266, + "learning_rate": 9.378623438547532e-06, + "loss": 0.3782, + "step": 1335 + }, + { + "epoch": 0.19, + "grad_norm": 2.6411857491973554, + "learning_rate": 9.377535998231981e-06, + "loss": 0.3752, + "step": 1336 + }, + { + "epoch": 0.19, + "grad_norm": 1.9230205248164371, + "learning_rate": 9.376447670368086e-06, + "loss": 0.3269, + "step": 1337 + }, + { + "epoch": 0.19, + "grad_norm": 2.228035340856152, + "learning_rate": 9.375358455176505e-06, + "loss": 0.3905, + "step": 1338 + }, + { + "epoch": 0.19, + "grad_norm": 2.157321534052014, + "learning_rate": 9.374268352878078e-06, + "loss": 0.3641, + "step": 1339 + }, + { + "epoch": 0.19, + "grad_norm": 2.4254385226155795, + "learning_rate": 9.373177363693825e-06, + "loss": 0.3535, + "step": 1340 + }, + { + "epoch": 0.19, + "grad_norm": 3.351048631490815, + "learning_rate": 9.372085487844943e-06, + "loss": 0.3586, + "step": 1341 + }, + { + "epoch": 0.19, + "grad_norm": 1.8807465735014992, + "learning_rate": 9.37099272555281e-06, + "loss": 0.3755, + "step": 1342 + }, + { + "epoch": 0.19, + "grad_norm": 1.9169068372247988, + "learning_rate": 9.369899077038987e-06, + "loss": 0.3747, + "step": 1343 + }, + { + "epoch": 0.19, + "grad_norm": 2.2066961935048233, + "learning_rate": 9.368804542525211e-06, + "loss": 0.3658, + "step": 1344 + }, + { + "epoch": 0.19, + "grad_norm": 2.1972579541330632, + "learning_rate": 9.367709122233396e-06, + "loss": 0.3695, + "step": 1345 + }, + { + "epoch": 0.19, + "grad_norm": 2.790740500241218, + "learning_rate": 9.366612816385646e-06, + "loss": 0.3613, + "step": 1346 + }, + { + "epoch": 0.19, + "grad_norm": 1.8318475389793103, + "learning_rate": 9.36551562520423e-06, + "loss": 0.3633, + "step": 1347 + }, + { + "epoch": 0.19, + "grad_norm": 2.3477306681150014, + "learning_rate": 9.364417548911612e-06, + "loss": 0.3598, + "step": 1348 + }, + { + "epoch": 0.19, + "grad_norm": 2.347233637832629, + "learning_rate": 9.36331858773042e-06, + "loss": 0.3515, + "step": 1349 + }, + { + "epoch": 0.19, + "grad_norm": 1.7894114914278767, + "learning_rate": 9.362218741883477e-06, + "loss": 0.3746, + "step": 1350 + }, + { + "epoch": 0.19, + "grad_norm": 2.1278032000075453, + "learning_rate": 9.361118011593772e-06, + "loss": 0.3743, + "step": 1351 + }, + { + "epoch": 0.19, + "grad_norm": 2.4486524660469517, + "learning_rate": 9.360016397084479e-06, + "loss": 0.3547, + "step": 1352 + }, + { + "epoch": 0.19, + "grad_norm": 2.794011592703434, + "learning_rate": 9.358913898578954e-06, + "loss": 0.3679, + "step": 1353 + }, + { + "epoch": 0.19, + "grad_norm": 2.1464316062969346, + "learning_rate": 9.357810516300726e-06, + "loss": 0.3663, + "step": 1354 + }, + { + "epoch": 0.19, + "grad_norm": 2.065417325619045, + "learning_rate": 9.35670625047351e-06, + "loss": 0.3799, + "step": 1355 + }, + { + "epoch": 0.19, + "grad_norm": 2.72105623322669, + "learning_rate": 9.355601101321195e-06, + "loss": 0.3863, + "step": 1356 + }, + { + "epoch": 0.19, + "grad_norm": 1.8377065057298574, + "learning_rate": 9.354495069067849e-06, + "loss": 0.3737, + "step": 1357 + }, + { + "epoch": 0.19, + "grad_norm": 2.2932623472351152, + "learning_rate": 9.353388153937723e-06, + "loss": 0.3755, + "step": 1358 + }, + { + "epoch": 0.19, + "grad_norm": 1.902609496600716, + "learning_rate": 9.352280356155244e-06, + "loss": 0.3595, + "step": 1359 + }, + { + "epoch": 0.19, + "grad_norm": 2.02516679166649, + "learning_rate": 9.35117167594502e-06, + "loss": 0.3703, + "step": 1360 + }, + { + "epoch": 0.19, + "grad_norm": 2.2870405373170133, + "learning_rate": 9.350062113531832e-06, + "loss": 0.3667, + "step": 1361 + }, + { + "epoch": 0.19, + "grad_norm": 2.287446379214703, + "learning_rate": 9.348951669140652e-06, + "loss": 0.3571, + "step": 1362 + }, + { + "epoch": 0.19, + "grad_norm": 3.9848324932461794, + "learning_rate": 9.347840342996617e-06, + "loss": 0.358, + "step": 1363 + }, + { + "epoch": 0.19, + "grad_norm": 1.7648871490908027, + "learning_rate": 9.346728135325051e-06, + "loss": 0.3549, + "step": 1364 + }, + { + "epoch": 0.19, + "grad_norm": 2.024651766865241, + "learning_rate": 9.345615046351457e-06, + "loss": 0.3508, + "step": 1365 + }, + { + "epoch": 0.19, + "grad_norm": 2.5032156872894635, + "learning_rate": 9.344501076301513e-06, + "loss": 0.3693, + "step": 1366 + }, + { + "epoch": 0.19, + "grad_norm": 2.527474161411308, + "learning_rate": 9.343386225401076e-06, + "loss": 0.3862, + "step": 1367 + }, + { + "epoch": 0.19, + "grad_norm": 1.9642586309011276, + "learning_rate": 9.342270493876185e-06, + "loss": 0.3696, + "step": 1368 + }, + { + "epoch": 0.19, + "grad_norm": 3.3414850675268535, + "learning_rate": 9.341153881953052e-06, + "loss": 0.3432, + "step": 1369 + }, + { + "epoch": 0.19, + "grad_norm": 2.0132998202395442, + "learning_rate": 9.340036389858075e-06, + "loss": 0.3619, + "step": 1370 + }, + { + "epoch": 0.19, + "grad_norm": 1.9478107103330953, + "learning_rate": 9.338918017817821e-06, + "loss": 0.3631, + "step": 1371 + }, + { + "epoch": 0.19, + "grad_norm": 1.8162425202567807, + "learning_rate": 9.337798766059047e-06, + "loss": 0.3651, + "step": 1372 + }, + { + "epoch": 0.19, + "grad_norm": 2.838728734468256, + "learning_rate": 9.336678634808678e-06, + "loss": 0.3771, + "step": 1373 + }, + { + "epoch": 0.19, + "grad_norm": 4.259857295257755, + "learning_rate": 9.335557624293823e-06, + "loss": 0.3535, + "step": 1374 + }, + { + "epoch": 0.19, + "grad_norm": 2.5034100177744736, + "learning_rate": 9.334435734741766e-06, + "loss": 0.3791, + "step": 1375 + }, + { + "epoch": 0.19, + "grad_norm": 2.609629257179483, + "learning_rate": 9.333312966379973e-06, + "loss": 0.3513, + "step": 1376 + }, + { + "epoch": 0.19, + "grad_norm": 2.0304686038559567, + "learning_rate": 9.332189319436082e-06, + "loss": 0.3597, + "step": 1377 + }, + { + "epoch": 0.19, + "grad_norm": 2.299479778895593, + "learning_rate": 9.331064794137918e-06, + "loss": 0.3691, + "step": 1378 + }, + { + "epoch": 0.19, + "grad_norm": 2.3398387925218187, + "learning_rate": 9.329939390713476e-06, + "loss": 0.3492, + "step": 1379 + }, + { + "epoch": 0.19, + "grad_norm": 2.031849899185623, + "learning_rate": 9.328813109390936e-06, + "loss": 0.3763, + "step": 1380 + }, + { + "epoch": 0.19, + "grad_norm": 2.2418914827929433, + "learning_rate": 9.327685950398647e-06, + "loss": 0.352, + "step": 1381 + }, + { + "epoch": 0.19, + "grad_norm": 3.767488989470614, + "learning_rate": 9.326557913965146e-06, + "loss": 0.3444, + "step": 1382 + }, + { + "epoch": 0.19, + "grad_norm": 2.5477393425379122, + "learning_rate": 9.32542900031914e-06, + "loss": 0.3668, + "step": 1383 + }, + { + "epoch": 0.19, + "grad_norm": 2.561730983945127, + "learning_rate": 9.324299209689519e-06, + "loss": 0.353, + "step": 1384 + }, + { + "epoch": 0.19, + "grad_norm": 8.819817665134451, + "learning_rate": 9.323168542305345e-06, + "loss": 0.3892, + "step": 1385 + }, + { + "epoch": 0.19, + "grad_norm": 2.5433801723147007, + "learning_rate": 9.322036998395866e-06, + "loss": 0.3432, + "step": 1386 + }, + { + "epoch": 0.19, + "grad_norm": 1.8821621704550504, + "learning_rate": 9.320904578190504e-06, + "loss": 0.3414, + "step": 1387 + }, + { + "epoch": 0.19, + "grad_norm": 2.7919476435783785, + "learning_rate": 9.319771281918854e-06, + "loss": 0.3608, + "step": 1388 + }, + { + "epoch": 0.19, + "grad_norm": 2.2610687523988715, + "learning_rate": 9.318637109810695e-06, + "loss": 0.3521, + "step": 1389 + }, + { + "epoch": 0.19, + "grad_norm": 2.169647344268509, + "learning_rate": 9.31750206209598e-06, + "loss": 0.3672, + "step": 1390 + }, + { + "epoch": 0.19, + "grad_norm": 2.148024413204392, + "learning_rate": 9.316366139004842e-06, + "loss": 0.3517, + "step": 1391 + }, + { + "epoch": 0.19, + "grad_norm": 2.816028016558708, + "learning_rate": 9.315229340767588e-06, + "loss": 0.3511, + "step": 1392 + }, + { + "epoch": 0.19, + "grad_norm": 2.103562788778611, + "learning_rate": 9.314091667614706e-06, + "loss": 0.3391, + "step": 1393 + }, + { + "epoch": 0.19, + "grad_norm": 4.263939740129281, + "learning_rate": 9.312953119776863e-06, + "loss": 0.3798, + "step": 1394 + }, + { + "epoch": 0.19, + "grad_norm": 2.0006212127030345, + "learning_rate": 9.311813697484893e-06, + "loss": 0.3627, + "step": 1395 + }, + { + "epoch": 0.19, + "grad_norm": 2.2058158696501615, + "learning_rate": 9.31067340096982e-06, + "loss": 0.3626, + "step": 1396 + }, + { + "epoch": 0.19, + "grad_norm": 2.677056497334697, + "learning_rate": 9.309532230462839e-06, + "loss": 0.3309, + "step": 1397 + }, + { + "epoch": 0.19, + "grad_norm": 2.726495378498076, + "learning_rate": 9.308390186195324e-06, + "loss": 0.3466, + "step": 1398 + }, + { + "epoch": 0.19, + "grad_norm": 2.959795507750859, + "learning_rate": 9.307247268398822e-06, + "loss": 0.3439, + "step": 1399 + }, + { + "epoch": 0.19, + "grad_norm": 2.732535635676669, + "learning_rate": 9.306103477305065e-06, + "loss": 0.3436, + "step": 1400 + }, + { + "epoch": 0.19, + "grad_norm": 2.1888053233352305, + "learning_rate": 9.304958813145954e-06, + "loss": 0.3453, + "step": 1401 + }, + { + "epoch": 0.19, + "grad_norm": 4.097493697661495, + "learning_rate": 9.30381327615357e-06, + "loss": 0.3647, + "step": 1402 + }, + { + "epoch": 0.2, + "grad_norm": 3.2431094065151007, + "learning_rate": 9.302666866560172e-06, + "loss": 0.3635, + "step": 1403 + }, + { + "epoch": 0.2, + "grad_norm": 2.3215358128120136, + "learning_rate": 9.301519584598199e-06, + "loss": 0.3793, + "step": 1404 + }, + { + "epoch": 0.2, + "grad_norm": 2.1064970995431125, + "learning_rate": 9.30037143050026e-06, + "loss": 0.3428, + "step": 1405 + }, + { + "epoch": 0.2, + "grad_norm": 2.408994605837863, + "learning_rate": 9.299222404499141e-06, + "loss": 0.3692, + "step": 1406 + }, + { + "epoch": 0.2, + "grad_norm": 1.8375576771514672, + "learning_rate": 9.298072506827814e-06, + "loss": 0.3514, + "step": 1407 + }, + { + "epoch": 0.2, + "grad_norm": 2.4828826643736366, + "learning_rate": 9.296921737719415e-06, + "loss": 0.3588, + "step": 1408 + }, + { + "epoch": 0.2, + "grad_norm": 2.2285524155249754, + "learning_rate": 9.295770097407269e-06, + "loss": 0.3694, + "step": 1409 + }, + { + "epoch": 0.2, + "grad_norm": 2.8658482923368505, + "learning_rate": 9.29461758612487e-06, + "loss": 0.3678, + "step": 1410 + }, + { + "epoch": 0.2, + "grad_norm": 2.756787814043346, + "learning_rate": 9.293464204105889e-06, + "loss": 0.3921, + "step": 1411 + }, + { + "epoch": 0.2, + "grad_norm": 2.879136365161853, + "learning_rate": 9.292309951584176e-06, + "loss": 0.3399, + "step": 1412 + }, + { + "epoch": 0.2, + "grad_norm": 2.1273522937735425, + "learning_rate": 9.291154828793756e-06, + "loss": 0.3511, + "step": 1413 + }, + { + "epoch": 0.2, + "grad_norm": 2.15181712188809, + "learning_rate": 9.289998835968832e-06, + "loss": 0.3606, + "step": 1414 + }, + { + "epoch": 0.2, + "grad_norm": 2.3775633519701316, + "learning_rate": 9.288841973343781e-06, + "loss": 0.3899, + "step": 1415 + }, + { + "epoch": 0.2, + "grad_norm": 2.554209089041417, + "learning_rate": 9.287684241153159e-06, + "loss": 0.3573, + "step": 1416 + }, + { + "epoch": 0.2, + "grad_norm": 2.01934039905452, + "learning_rate": 9.286525639631695e-06, + "loss": 0.3708, + "step": 1417 + }, + { + "epoch": 0.2, + "grad_norm": 3.253250945678716, + "learning_rate": 9.2853661690143e-06, + "loss": 0.3707, + "step": 1418 + }, + { + "epoch": 0.2, + "grad_norm": 2.4619861774512004, + "learning_rate": 9.284205829536052e-06, + "loss": 0.3693, + "step": 1419 + }, + { + "epoch": 0.2, + "grad_norm": 2.9060145744381094, + "learning_rate": 9.283044621432216e-06, + "loss": 0.3698, + "step": 1420 + }, + { + "epoch": 0.2, + "grad_norm": 3.6931040618806867, + "learning_rate": 9.281882544938224e-06, + "loss": 0.3809, + "step": 1421 + }, + { + "epoch": 0.2, + "grad_norm": 2.2911874323054997, + "learning_rate": 9.280719600289688e-06, + "loss": 0.3719, + "step": 1422 + }, + { + "epoch": 0.2, + "grad_norm": 2.1124657116760783, + "learning_rate": 9.279555787722398e-06, + "loss": 0.3674, + "step": 1423 + }, + { + "epoch": 0.2, + "grad_norm": 2.3388236382607483, + "learning_rate": 9.278391107472316e-06, + "loss": 0.3565, + "step": 1424 + }, + { + "epoch": 0.2, + "grad_norm": 4.1213153171151005, + "learning_rate": 9.277225559775584e-06, + "loss": 0.3789, + "step": 1425 + }, + { + "epoch": 0.2, + "grad_norm": 2.4384838843563585, + "learning_rate": 9.276059144868515e-06, + "loss": 0.3842, + "step": 1426 + }, + { + "epoch": 0.2, + "grad_norm": 2.4396101307022504, + "learning_rate": 9.274891862987601e-06, + "loss": 0.37, + "step": 1427 + }, + { + "epoch": 0.2, + "grad_norm": 2.300967785568793, + "learning_rate": 9.273723714369509e-06, + "loss": 0.3509, + "step": 1428 + }, + { + "epoch": 0.2, + "grad_norm": 2.9024278773237318, + "learning_rate": 9.272554699251081e-06, + "loss": 0.3432, + "step": 1429 + }, + { + "epoch": 0.2, + "grad_norm": 2.2912172580773764, + "learning_rate": 9.27138481786934e-06, + "loss": 0.3662, + "step": 1430 + }, + { + "epoch": 0.2, + "grad_norm": 2.4783570988679626, + "learning_rate": 9.270214070461476e-06, + "loss": 0.3503, + "step": 1431 + }, + { + "epoch": 0.2, + "grad_norm": 2.8214108129776823, + "learning_rate": 9.26904245726486e-06, + "loss": 0.3592, + "step": 1432 + }, + { + "epoch": 0.2, + "grad_norm": 2.3874367859170573, + "learning_rate": 9.267869978517036e-06, + "loss": 0.3451, + "step": 1433 + }, + { + "epoch": 0.2, + "grad_norm": 2.1403806584742617, + "learning_rate": 9.266696634455729e-06, + "loss": 0.3693, + "step": 1434 + }, + { + "epoch": 0.2, + "grad_norm": 3.1238872861445275, + "learning_rate": 9.26552242531883e-06, + "loss": 0.3397, + "step": 1435 + }, + { + "epoch": 0.2, + "grad_norm": 2.5212173078940183, + "learning_rate": 9.264347351344415e-06, + "loss": 0.3702, + "step": 1436 + }, + { + "epoch": 0.2, + "grad_norm": 2.234574470867367, + "learning_rate": 9.263171412770728e-06, + "loss": 0.364, + "step": 1437 + }, + { + "epoch": 0.2, + "grad_norm": 1.8941233548645044, + "learning_rate": 9.261994609836194e-06, + "loss": 0.3632, + "step": 1438 + }, + { + "epoch": 0.2, + "grad_norm": 2.3651975729590755, + "learning_rate": 9.260816942779409e-06, + "loss": 0.3543, + "step": 1439 + }, + { + "epoch": 0.2, + "grad_norm": 3.0192767314337967, + "learning_rate": 9.259638411839144e-06, + "loss": 0.351, + "step": 1440 + }, + { + "epoch": 0.2, + "grad_norm": 2.600772629086889, + "learning_rate": 9.258459017254352e-06, + "loss": 0.3505, + "step": 1441 + }, + { + "epoch": 0.2, + "grad_norm": 2.4164290629793816, + "learning_rate": 9.257278759264149e-06, + "loss": 0.3523, + "step": 1442 + }, + { + "epoch": 0.2, + "grad_norm": 13.501998360509315, + "learning_rate": 9.256097638107839e-06, + "loss": 0.3477, + "step": 1443 + }, + { + "epoch": 0.2, + "grad_norm": 2.1228534038858142, + "learning_rate": 9.254915654024894e-06, + "loss": 0.3717, + "step": 1444 + }, + { + "epoch": 0.2, + "grad_norm": 2.9242156118484064, + "learning_rate": 9.253732807254959e-06, + "loss": 0.3572, + "step": 1445 + }, + { + "epoch": 0.2, + "grad_norm": 3.176216110688619, + "learning_rate": 9.252549098037859e-06, + "loss": 0.3703, + "step": 1446 + }, + { + "epoch": 0.2, + "grad_norm": 2.818644021081555, + "learning_rate": 9.251364526613594e-06, + "loss": 0.3756, + "step": 1447 + }, + { + "epoch": 0.2, + "grad_norm": 2.144221327095277, + "learning_rate": 9.250179093222335e-06, + "loss": 0.3623, + "step": 1448 + }, + { + "epoch": 0.2, + "grad_norm": 2.2210998654987324, + "learning_rate": 9.248992798104427e-06, + "loss": 0.3873, + "step": 1449 + }, + { + "epoch": 0.2, + "grad_norm": 2.4005954340463873, + "learning_rate": 9.247805641500396e-06, + "loss": 0.3446, + "step": 1450 + }, + { + "epoch": 0.2, + "grad_norm": 2.5316683210491076, + "learning_rate": 9.246617623650935e-06, + "loss": 0.3513, + "step": 1451 + }, + { + "epoch": 0.2, + "grad_norm": 2.8550501638398904, + "learning_rate": 9.245428744796919e-06, + "loss": 0.3592, + "step": 1452 + }, + { + "epoch": 0.2, + "grad_norm": 2.2003132157569523, + "learning_rate": 9.24423900517939e-06, + "loss": 0.3392, + "step": 1453 + }, + { + "epoch": 0.2, + "grad_norm": 1.9116414775767272, + "learning_rate": 9.243048405039574e-06, + "loss": 0.3462, + "step": 1454 + }, + { + "epoch": 0.2, + "grad_norm": 1.9483283067904889, + "learning_rate": 9.241856944618861e-06, + "loss": 0.3636, + "step": 1455 + }, + { + "epoch": 0.2, + "grad_norm": 3.6003445036916446, + "learning_rate": 9.240664624158822e-06, + "loss": 0.3527, + "step": 1456 + }, + { + "epoch": 0.2, + "grad_norm": 2.488594023843255, + "learning_rate": 9.239471443901202e-06, + "loss": 0.3526, + "step": 1457 + }, + { + "epoch": 0.2, + "grad_norm": 3.3227529975411407, + "learning_rate": 9.238277404087918e-06, + "loss": 0.3672, + "step": 1458 + }, + { + "epoch": 0.2, + "grad_norm": 2.2387302221693375, + "learning_rate": 9.23708250496106e-06, + "loss": 0.3401, + "step": 1459 + }, + { + "epoch": 0.2, + "grad_norm": 2.9857561982966305, + "learning_rate": 9.2358867467629e-06, + "loss": 0.3254, + "step": 1460 + }, + { + "epoch": 0.2, + "grad_norm": 2.713729707016033, + "learning_rate": 9.234690129735875e-06, + "loss": 0.3406, + "step": 1461 + }, + { + "epoch": 0.2, + "grad_norm": 2.4594258446814337, + "learning_rate": 9.2334926541226e-06, + "loss": 0.3448, + "step": 1462 + }, + { + "epoch": 0.2, + "grad_norm": 2.195776449868004, + "learning_rate": 9.232294320165865e-06, + "loss": 0.3355, + "step": 1463 + }, + { + "epoch": 0.2, + "grad_norm": 2.073996159353538, + "learning_rate": 9.231095128108634e-06, + "loss": 0.3376, + "step": 1464 + }, + { + "epoch": 0.2, + "grad_norm": 2.062626949793734, + "learning_rate": 9.22989507819404e-06, + "loss": 0.3605, + "step": 1465 + }, + { + "epoch": 0.2, + "grad_norm": 2.918929626824144, + "learning_rate": 9.2286941706654e-06, + "loss": 0.3653, + "step": 1466 + }, + { + "epoch": 0.2, + "grad_norm": 3.3260556541823485, + "learning_rate": 9.227492405766193e-06, + "loss": 0.3689, + "step": 1467 + }, + { + "epoch": 0.2, + "grad_norm": 1.9058524466174238, + "learning_rate": 9.22628978374008e-06, + "loss": 0.334, + "step": 1468 + }, + { + "epoch": 0.2, + "grad_norm": 3.264124123313891, + "learning_rate": 9.225086304830896e-06, + "loss": 0.3641, + "step": 1469 + }, + { + "epoch": 0.2, + "grad_norm": 2.100213322609, + "learning_rate": 9.22388196928264e-06, + "loss": 0.3439, + "step": 1470 + }, + { + "epoch": 0.2, + "grad_norm": 3.260186078606789, + "learning_rate": 9.222676777339498e-06, + "loss": 0.3734, + "step": 1471 + }, + { + "epoch": 0.2, + "grad_norm": 2.1580926379201752, + "learning_rate": 9.221470729245826e-06, + "loss": 0.345, + "step": 1472 + }, + { + "epoch": 0.2, + "grad_norm": 2.208330393813212, + "learning_rate": 9.220263825246143e-06, + "loss": 0.37, + "step": 1473 + }, + { + "epoch": 0.2, + "grad_norm": 2.0425365152676656, + "learning_rate": 9.219056065585154e-06, + "loss": 0.3806, + "step": 1474 + }, + { + "epoch": 0.21, + "grad_norm": 2.480627669608923, + "learning_rate": 9.217847450507733e-06, + "loss": 0.347, + "step": 1475 + }, + { + "epoch": 0.21, + "grad_norm": 2.7181710754804063, + "learning_rate": 9.216637980258926e-06, + "loss": 0.3694, + "step": 1476 + }, + { + "epoch": 0.21, + "grad_norm": 2.2542736141588584, + "learning_rate": 9.215427655083957e-06, + "loss": 0.3602, + "step": 1477 + }, + { + "epoch": 0.21, + "grad_norm": 3.239672549913999, + "learning_rate": 9.214216475228217e-06, + "loss": 0.3676, + "step": 1478 + }, + { + "epoch": 0.21, + "grad_norm": 3.858716667084395, + "learning_rate": 9.213004440937277e-06, + "loss": 0.354, + "step": 1479 + }, + { + "epoch": 0.21, + "grad_norm": 2.2942713725182537, + "learning_rate": 9.211791552456873e-06, + "loss": 0.3726, + "step": 1480 + }, + { + "epoch": 0.21, + "grad_norm": 3.425518151235285, + "learning_rate": 9.210577810032923e-06, + "loss": 0.3679, + "step": 1481 + }, + { + "epoch": 0.21, + "grad_norm": 2.5551607423596514, + "learning_rate": 9.209363213911512e-06, + "loss": 0.3657, + "step": 1482 + }, + { + "epoch": 0.21, + "grad_norm": 2.200458843140561, + "learning_rate": 9.208147764338902e-06, + "loss": 0.3714, + "step": 1483 + }, + { + "epoch": 0.21, + "grad_norm": 2.482215344953573, + "learning_rate": 9.206931461561524e-06, + "loss": 0.3563, + "step": 1484 + }, + { + "epoch": 0.21, + "grad_norm": 1.9983371002136014, + "learning_rate": 9.205714305825983e-06, + "loss": 0.3624, + "step": 1485 + }, + { + "epoch": 0.21, + "grad_norm": 3.046868529490069, + "learning_rate": 9.204496297379063e-06, + "loss": 0.371, + "step": 1486 + }, + { + "epoch": 0.21, + "grad_norm": 2.0159379447928827, + "learning_rate": 9.203277436467713e-06, + "loss": 0.349, + "step": 1487 + }, + { + "epoch": 0.21, + "grad_norm": 2.7165861782148286, + "learning_rate": 9.20205772333906e-06, + "loss": 0.3531, + "step": 1488 + }, + { + "epoch": 0.21, + "grad_norm": 1.9602064393693868, + "learning_rate": 9.200837158240396e-06, + "loss": 0.3525, + "step": 1489 + }, + { + "epoch": 0.21, + "grad_norm": 2.8473966742300525, + "learning_rate": 9.199615741419196e-06, + "loss": 0.3538, + "step": 1490 + }, + { + "epoch": 0.21, + "grad_norm": 2.15065385417656, + "learning_rate": 9.198393473123105e-06, + "loss": 0.3453, + "step": 1491 + }, + { + "epoch": 0.21, + "grad_norm": 2.335233494584752, + "learning_rate": 9.197170353599935e-06, + "loss": 0.3605, + "step": 1492 + }, + { + "epoch": 0.21, + "grad_norm": 2.687435712397397, + "learning_rate": 9.195946383097674e-06, + "loss": 0.3274, + "step": 1493 + }, + { + "epoch": 0.21, + "grad_norm": 1.9124098213747043, + "learning_rate": 9.194721561864485e-06, + "loss": 0.3779, + "step": 1494 + }, + { + "epoch": 0.21, + "grad_norm": 2.4256936736337966, + "learning_rate": 9.193495890148702e-06, + "loss": 0.3622, + "step": 1495 + }, + { + "epoch": 0.21, + "grad_norm": 2.555916921441627, + "learning_rate": 9.192269368198827e-06, + "loss": 0.379, + "step": 1496 + }, + { + "epoch": 0.21, + "grad_norm": 2.1259090639702367, + "learning_rate": 9.191041996263542e-06, + "loss": 0.3512, + "step": 1497 + }, + { + "epoch": 0.21, + "grad_norm": 2.1877510804524043, + "learning_rate": 9.189813774591693e-06, + "loss": 0.3764, + "step": 1498 + }, + { + "epoch": 0.21, + "grad_norm": 2.6768782283666197, + "learning_rate": 9.18858470343231e-06, + "loss": 0.3639, + "step": 1499 + }, + { + "epoch": 0.21, + "grad_norm": 2.211496487835044, + "learning_rate": 9.18735478303458e-06, + "loss": 0.3389, + "step": 1500 + }, + { + "epoch": 0.21, + "grad_norm": 2.1463860324110478, + "learning_rate": 9.186124013647875e-06, + "loss": 0.3234, + "step": 1501 + }, + { + "epoch": 0.21, + "grad_norm": 2.3922185058951326, + "learning_rate": 9.184892395521734e-06, + "loss": 0.3749, + "step": 1502 + }, + { + "epoch": 0.21, + "grad_norm": 2.0560253679203253, + "learning_rate": 9.183659928905867e-06, + "loss": 0.3358, + "step": 1503 + }, + { + "epoch": 0.21, + "grad_norm": 2.7724877126353116, + "learning_rate": 9.182426614050157e-06, + "loss": 0.3428, + "step": 1504 + }, + { + "epoch": 0.21, + "grad_norm": 2.415728028148105, + "learning_rate": 9.181192451204662e-06, + "loss": 0.3771, + "step": 1505 + }, + { + "epoch": 0.21, + "grad_norm": 2.033828361348063, + "learning_rate": 9.179957440619606e-06, + "loss": 0.3672, + "step": 1506 + }, + { + "epoch": 0.21, + "grad_norm": 2.0723381103278458, + "learning_rate": 9.178721582545391e-06, + "loss": 0.3529, + "step": 1507 + }, + { + "epoch": 0.21, + "grad_norm": 2.8159165820525325, + "learning_rate": 9.177484877232588e-06, + "loss": 0.3538, + "step": 1508 + }, + { + "epoch": 0.21, + "grad_norm": 2.69900380869936, + "learning_rate": 9.176247324931937e-06, + "loss": 0.3544, + "step": 1509 + }, + { + "epoch": 0.21, + "grad_norm": 59.128666341124124, + "learning_rate": 9.175008925894355e-06, + "loss": 0.3583, + "step": 1510 + }, + { + "epoch": 0.21, + "grad_norm": 3.0464332953090927, + "learning_rate": 9.173769680370928e-06, + "loss": 0.3659, + "step": 1511 + }, + { + "epoch": 0.21, + "grad_norm": 2.3567004107030685, + "learning_rate": 9.172529588612912e-06, + "loss": 0.3369, + "step": 1512 + }, + { + "epoch": 0.21, + "grad_norm": 2.0067184926445307, + "learning_rate": 9.171288650871742e-06, + "loss": 0.3545, + "step": 1513 + }, + { + "epoch": 0.21, + "grad_norm": 3.1015659518092433, + "learning_rate": 9.17004686739901e-06, + "loss": 0.3843, + "step": 1514 + }, + { + "epoch": 0.21, + "grad_norm": 2.502623575028134, + "learning_rate": 9.168804238446495e-06, + "loss": 0.3574, + "step": 1515 + }, + { + "epoch": 0.21, + "grad_norm": 3.096982241527454, + "learning_rate": 9.167560764266138e-06, + "loss": 0.3735, + "step": 1516 + }, + { + "epoch": 0.21, + "grad_norm": 4.648378525769068, + "learning_rate": 9.166316445110058e-06, + "loss": 0.3646, + "step": 1517 + }, + { + "epoch": 0.21, + "grad_norm": 4.563026561179445, + "learning_rate": 9.165071281230538e-06, + "loss": 0.347, + "step": 1518 + }, + { + "epoch": 0.21, + "grad_norm": 1.9796477659121714, + "learning_rate": 9.163825272880036e-06, + "loss": 0.3641, + "step": 1519 + }, + { + "epoch": 0.21, + "grad_norm": 2.007906393742986, + "learning_rate": 9.162578420311182e-06, + "loss": 0.3464, + "step": 1520 + }, + { + "epoch": 0.21, + "grad_norm": 2.1287333954424765, + "learning_rate": 9.161330723776775e-06, + "loss": 0.3825, + "step": 1521 + }, + { + "epoch": 0.21, + "grad_norm": 2.2798140967688325, + "learning_rate": 9.160082183529788e-06, + "loss": 0.3834, + "step": 1522 + }, + { + "epoch": 0.21, + "grad_norm": 2.167438593866005, + "learning_rate": 9.158832799823363e-06, + "loss": 0.3683, + "step": 1523 + }, + { + "epoch": 0.21, + "grad_norm": 2.257399491330866, + "learning_rate": 9.157582572910813e-06, + "loss": 0.3687, + "step": 1524 + }, + { + "epoch": 0.21, + "grad_norm": 2.045697550959512, + "learning_rate": 9.15633150304562e-06, + "loss": 0.3524, + "step": 1525 + }, + { + "epoch": 0.21, + "grad_norm": 2.4450197403127585, + "learning_rate": 9.155079590481446e-06, + "loss": 0.352, + "step": 1526 + }, + { + "epoch": 0.21, + "grad_norm": 2.12648150360745, + "learning_rate": 9.153826835472108e-06, + "loss": 0.3749, + "step": 1527 + }, + { + "epoch": 0.21, + "grad_norm": 2.0015893493794663, + "learning_rate": 9.15257323827161e-06, + "loss": 0.3638, + "step": 1528 + }, + { + "epoch": 0.21, + "grad_norm": 2.342170319144583, + "learning_rate": 9.151318799134118e-06, + "loss": 0.3375, + "step": 1529 + }, + { + "epoch": 0.21, + "grad_norm": 2.104095866065225, + "learning_rate": 9.150063518313968e-06, + "loss": 0.3472, + "step": 1530 + }, + { + "epoch": 0.21, + "grad_norm": 3.066877052756292, + "learning_rate": 9.148807396065671e-06, + "loss": 0.3442, + "step": 1531 + }, + { + "epoch": 0.21, + "grad_norm": 5.019982956759269, + "learning_rate": 9.147550432643906e-06, + "loss": 0.368, + "step": 1532 + }, + { + "epoch": 0.21, + "grad_norm": 2.139658736411206, + "learning_rate": 9.146292628303523e-06, + "loss": 0.3796, + "step": 1533 + }, + { + "epoch": 0.21, + "grad_norm": 2.3321444136006986, + "learning_rate": 9.145033983299545e-06, + "loss": 0.3621, + "step": 1534 + }, + { + "epoch": 0.21, + "grad_norm": 2.653465997561299, + "learning_rate": 9.143774497887158e-06, + "loss": 0.3472, + "step": 1535 + }, + { + "epoch": 0.21, + "grad_norm": 2.3798708801400714, + "learning_rate": 9.14251417232173e-06, + "loss": 0.3906, + "step": 1536 + }, + { + "epoch": 0.21, + "grad_norm": 2.82717582525801, + "learning_rate": 9.14125300685879e-06, + "loss": 0.3562, + "step": 1537 + }, + { + "epoch": 0.21, + "grad_norm": 3.739965314798417, + "learning_rate": 9.139991001754037e-06, + "loss": 0.3443, + "step": 1538 + }, + { + "epoch": 0.21, + "grad_norm": 2.1720585141544317, + "learning_rate": 9.138728157263347e-06, + "loss": 0.3503, + "step": 1539 + }, + { + "epoch": 0.21, + "grad_norm": 2.436978925643727, + "learning_rate": 9.137464473642762e-06, + "loss": 0.3522, + "step": 1540 + }, + { + "epoch": 0.21, + "grad_norm": 3.0937915873589628, + "learning_rate": 9.136199951148495e-06, + "loss": 0.3569, + "step": 1541 + }, + { + "epoch": 0.21, + "grad_norm": 2.8074787864095265, + "learning_rate": 9.134934590036926e-06, + "loss": 0.3506, + "step": 1542 + }, + { + "epoch": 0.21, + "grad_norm": 2.322815779138555, + "learning_rate": 9.133668390564611e-06, + "loss": 0.3894, + "step": 1543 + }, + { + "epoch": 0.21, + "grad_norm": 2.000653583201319, + "learning_rate": 9.132401352988272e-06, + "loss": 0.3663, + "step": 1544 + }, + { + "epoch": 0.21, + "grad_norm": 3.334429259022657, + "learning_rate": 9.131133477564802e-06, + "loss": 0.3536, + "step": 1545 + }, + { + "epoch": 0.21, + "grad_norm": 1.9573628136832681, + "learning_rate": 9.129864764551261e-06, + "loss": 0.3245, + "step": 1546 + }, + { + "epoch": 0.22, + "grad_norm": 3.554722917624787, + "learning_rate": 9.128595214204885e-06, + "loss": 0.3673, + "step": 1547 + }, + { + "epoch": 0.22, + "grad_norm": 2.944834390347597, + "learning_rate": 9.127324826783075e-06, + "loss": 0.3647, + "step": 1548 + }, + { + "epoch": 0.22, + "grad_norm": 6.620170130708857, + "learning_rate": 9.126053602543403e-06, + "loss": 0.3549, + "step": 1549 + }, + { + "epoch": 0.22, + "grad_norm": 3.2751991435834666, + "learning_rate": 9.124781541743609e-06, + "loss": 0.3766, + "step": 1550 + }, + { + "epoch": 0.22, + "grad_norm": 2.4859361941373495, + "learning_rate": 9.123508644641604e-06, + "loss": 0.3262, + "step": 1551 + }, + { + "epoch": 0.22, + "grad_norm": 2.1338983019899134, + "learning_rate": 9.122234911495472e-06, + "loss": 0.3612, + "step": 1552 + }, + { + "epoch": 0.22, + "grad_norm": 2.503789321784492, + "learning_rate": 9.12096034256346e-06, + "loss": 0.342, + "step": 1553 + }, + { + "epoch": 0.22, + "grad_norm": 3.8980352693701024, + "learning_rate": 9.11968493810399e-06, + "loss": 0.3543, + "step": 1554 + }, + { + "epoch": 0.22, + "grad_norm": 3.248394790421149, + "learning_rate": 9.118408698375649e-06, + "loss": 0.3835, + "step": 1555 + }, + { + "epoch": 0.22, + "grad_norm": 2.30852031684993, + "learning_rate": 9.117131623637197e-06, + "loss": 0.3446, + "step": 1556 + }, + { + "epoch": 0.22, + "grad_norm": 3.047973850912468, + "learning_rate": 9.11585371414756e-06, + "loss": 0.342, + "step": 1557 + }, + { + "epoch": 0.22, + "grad_norm": 1.940993072539323, + "learning_rate": 9.114574970165838e-06, + "loss": 0.3659, + "step": 1558 + }, + { + "epoch": 0.22, + "grad_norm": 2.3611226209577256, + "learning_rate": 9.113295391951295e-06, + "loss": 0.3609, + "step": 1559 + }, + { + "epoch": 0.22, + "grad_norm": 2.6052810172562997, + "learning_rate": 9.112014979763366e-06, + "loss": 0.3366, + "step": 1560 + }, + { + "epoch": 0.22, + "grad_norm": 2.357883090640589, + "learning_rate": 9.110733733861657e-06, + "loss": 0.3387, + "step": 1561 + }, + { + "epoch": 0.22, + "grad_norm": 2.8023665160617988, + "learning_rate": 9.10945165450594e-06, + "loss": 0.3426, + "step": 1562 + }, + { + "epoch": 0.22, + "grad_norm": 2.0316967976035936, + "learning_rate": 9.10816874195616e-06, + "loss": 0.3633, + "step": 1563 + }, + { + "epoch": 0.22, + "grad_norm": 2.2816233530920593, + "learning_rate": 9.106884996472424e-06, + "loss": 0.3811, + "step": 1564 + }, + { + "epoch": 0.22, + "grad_norm": 2.3066945765709614, + "learning_rate": 9.105600418315017e-06, + "loss": 0.3679, + "step": 1565 + }, + { + "epoch": 0.22, + "grad_norm": 2.1685115855462165, + "learning_rate": 9.104315007744385e-06, + "loss": 0.34, + "step": 1566 + }, + { + "epoch": 0.22, + "grad_norm": 2.228569156963965, + "learning_rate": 9.103028765021146e-06, + "loss": 0.3531, + "step": 1567 + }, + { + "epoch": 0.22, + "grad_norm": 3.568372462987173, + "learning_rate": 9.101741690406088e-06, + "loss": 0.3826, + "step": 1568 + }, + { + "epoch": 0.22, + "grad_norm": 1.7590963714347128, + "learning_rate": 9.100453784160165e-06, + "loss": 0.3616, + "step": 1569 + }, + { + "epoch": 0.22, + "grad_norm": 1.8462094092886068, + "learning_rate": 9.099165046544501e-06, + "loss": 0.3476, + "step": 1570 + }, + { + "epoch": 0.22, + "grad_norm": 2.3103973852560706, + "learning_rate": 9.097875477820389e-06, + "loss": 0.3434, + "step": 1571 + }, + { + "epoch": 0.22, + "grad_norm": 2.414780769134637, + "learning_rate": 9.096585078249289e-06, + "loss": 0.3319, + "step": 1572 + }, + { + "epoch": 0.22, + "grad_norm": 5.416523629255541, + "learning_rate": 9.095293848092833e-06, + "loss": 0.3415, + "step": 1573 + }, + { + "epoch": 0.22, + "grad_norm": 2.1406002656493173, + "learning_rate": 9.094001787612816e-06, + "loss": 0.3523, + "step": 1574 + }, + { + "epoch": 0.22, + "grad_norm": 1.9997698191476172, + "learning_rate": 9.092708897071206e-06, + "loss": 0.3433, + "step": 1575 + }, + { + "epoch": 0.22, + "grad_norm": 2.11477279849082, + "learning_rate": 9.091415176730135e-06, + "loss": 0.3816, + "step": 1576 + }, + { + "epoch": 0.22, + "grad_norm": 2.2825122974720373, + "learning_rate": 9.090120626851906e-06, + "loss": 0.3747, + "step": 1577 + }, + { + "epoch": 0.22, + "grad_norm": 2.404267581169105, + "learning_rate": 9.088825247698992e-06, + "loss": 0.3141, + "step": 1578 + }, + { + "epoch": 0.22, + "grad_norm": 2.3295210918034805, + "learning_rate": 9.08752903953403e-06, + "loss": 0.3402, + "step": 1579 + }, + { + "epoch": 0.22, + "grad_norm": 2.0350801607009785, + "learning_rate": 9.086232002619828e-06, + "loss": 0.3781, + "step": 1580 + }, + { + "epoch": 0.22, + "grad_norm": 3.0201927038430445, + "learning_rate": 9.084934137219363e-06, + "loss": 0.3639, + "step": 1581 + }, + { + "epoch": 0.22, + "grad_norm": 6.20993387709582, + "learning_rate": 9.083635443595774e-06, + "loss": 0.401, + "step": 1582 + }, + { + "epoch": 0.22, + "grad_norm": 2.505088362015913, + "learning_rate": 9.082335922012375e-06, + "loss": 0.3559, + "step": 1583 + }, + { + "epoch": 0.22, + "grad_norm": 2.236354517148501, + "learning_rate": 9.081035572732644e-06, + "loss": 0.3392, + "step": 1584 + }, + { + "epoch": 0.22, + "grad_norm": 3.8699592029274283, + "learning_rate": 9.079734396020227e-06, + "loss": 0.3789, + "step": 1585 + }, + { + "epoch": 0.22, + "grad_norm": 2.524741644150287, + "learning_rate": 9.078432392138941e-06, + "loss": 0.3795, + "step": 1586 + }, + { + "epoch": 0.22, + "grad_norm": 2.2949933532708005, + "learning_rate": 9.077129561352764e-06, + "loss": 0.3479, + "step": 1587 + }, + { + "epoch": 0.22, + "grad_norm": 2.8659971815939005, + "learning_rate": 9.075825903925849e-06, + "loss": 0.3587, + "step": 1588 + }, + { + "epoch": 0.22, + "grad_norm": 3.4387586945250455, + "learning_rate": 9.074521420122513e-06, + "loss": 0.3573, + "step": 1589 + }, + { + "epoch": 0.22, + "grad_norm": 4.529772931823923, + "learning_rate": 9.073216110207238e-06, + "loss": 0.3765, + "step": 1590 + }, + { + "epoch": 0.22, + "grad_norm": 2.0526846313404223, + "learning_rate": 9.071909974444682e-06, + "loss": 0.358, + "step": 1591 + }, + { + "epoch": 0.22, + "grad_norm": 1.8005517610694504, + "learning_rate": 9.07060301309966e-06, + "loss": 0.3413, + "step": 1592 + }, + { + "epoch": 0.22, + "grad_norm": 2.1104189260517723, + "learning_rate": 9.069295226437163e-06, + "loss": 0.3477, + "step": 1593 + }, + { + "epoch": 0.22, + "grad_norm": 2.227052782870714, + "learning_rate": 9.067986614722342e-06, + "loss": 0.3319, + "step": 1594 + }, + { + "epoch": 0.22, + "grad_norm": 2.0132598796320433, + "learning_rate": 9.066677178220519e-06, + "loss": 0.3605, + "step": 1595 + }, + { + "epoch": 0.22, + "grad_norm": 2.593738535422066, + "learning_rate": 9.065366917197187e-06, + "loss": 0.3627, + "step": 1596 + }, + { + "epoch": 0.22, + "grad_norm": 2.4622842233936444, + "learning_rate": 9.064055831918e-06, + "loss": 0.3737, + "step": 1597 + }, + { + "epoch": 0.22, + "grad_norm": 2.4726208851778133, + "learning_rate": 9.062743922648783e-06, + "loss": 0.3532, + "step": 1598 + }, + { + "epoch": 0.22, + "grad_norm": 2.608782582363567, + "learning_rate": 9.061431189655523e-06, + "loss": 0.3626, + "step": 1599 + }, + { + "epoch": 0.22, + "grad_norm": 2.8888451754895765, + "learning_rate": 9.060117633204379e-06, + "loss": 0.3698, + "step": 1600 + }, + { + "epoch": 0.22, + "grad_norm": 2.6437017642913947, + "learning_rate": 9.058803253561676e-06, + "loss": 0.3669, + "step": 1601 + }, + { + "epoch": 0.22, + "grad_norm": 2.0092828755376115, + "learning_rate": 9.057488050993906e-06, + "loss": 0.3312, + "step": 1602 + }, + { + "epoch": 0.22, + "grad_norm": 2.4476340244564807, + "learning_rate": 9.056172025767725e-06, + "loss": 0.3732, + "step": 1603 + }, + { + "epoch": 0.22, + "grad_norm": 2.0570079756414468, + "learning_rate": 9.05485517814996e-06, + "loss": 0.339, + "step": 1604 + }, + { + "epoch": 0.22, + "grad_norm": 3.330737893610654, + "learning_rate": 9.053537508407603e-06, + "loss": 0.338, + "step": 1605 + }, + { + "epoch": 0.22, + "grad_norm": 2.082824813150249, + "learning_rate": 9.05221901680781e-06, + "loss": 0.3566, + "step": 1606 + }, + { + "epoch": 0.22, + "grad_norm": 2.0106439343773017, + "learning_rate": 9.05089970361791e-06, + "loss": 0.3683, + "step": 1607 + }, + { + "epoch": 0.22, + "grad_norm": 4.024612376507351, + "learning_rate": 9.04957956910539e-06, + "loss": 0.3764, + "step": 1608 + }, + { + "epoch": 0.22, + "grad_norm": 2.589313582972412, + "learning_rate": 9.048258613537912e-06, + "loss": 0.372, + "step": 1609 + }, + { + "epoch": 0.22, + "grad_norm": 2.0737954280566746, + "learning_rate": 9.046936837183297e-06, + "loss": 0.3817, + "step": 1610 + }, + { + "epoch": 0.22, + "grad_norm": 2.0605978351745646, + "learning_rate": 9.045614240309539e-06, + "loss": 0.3453, + "step": 1611 + }, + { + "epoch": 0.22, + "grad_norm": 2.137110813337936, + "learning_rate": 9.044290823184796e-06, + "loss": 0.3497, + "step": 1612 + }, + { + "epoch": 0.22, + "grad_norm": 2.2771018810561703, + "learning_rate": 9.042966586077389e-06, + "loss": 0.3501, + "step": 1613 + }, + { + "epoch": 0.22, + "grad_norm": 2.086640973941434, + "learning_rate": 9.041641529255808e-06, + "loss": 0.355, + "step": 1614 + }, + { + "epoch": 0.22, + "grad_norm": 2.4503898500601125, + "learning_rate": 9.040315652988712e-06, + "loss": 0.3683, + "step": 1615 + }, + { + "epoch": 0.22, + "grad_norm": 2.8687926667726082, + "learning_rate": 9.038988957544919e-06, + "loss": 0.3862, + "step": 1616 + }, + { + "epoch": 0.22, + "grad_norm": 2.9516486100974735, + "learning_rate": 9.037661443193422e-06, + "loss": 0.3647, + "step": 1617 + }, + { + "epoch": 0.22, + "grad_norm": 2.888272881667731, + "learning_rate": 9.036333110203373e-06, + "loss": 0.3508, + "step": 1618 + }, + { + "epoch": 0.23, + "grad_norm": 2.079328640081481, + "learning_rate": 9.035003958844094e-06, + "loss": 0.3699, + "step": 1619 + }, + { + "epoch": 0.23, + "grad_norm": 2.2599222900799263, + "learning_rate": 9.033673989385068e-06, + "loss": 0.368, + "step": 1620 + }, + { + "epoch": 0.23, + "grad_norm": 2.265777345747243, + "learning_rate": 9.03234320209595e-06, + "loss": 0.3752, + "step": 1621 + }, + { + "epoch": 0.23, + "grad_norm": 2.929269278444532, + "learning_rate": 9.031011597246557e-06, + "loss": 0.3666, + "step": 1622 + }, + { + "epoch": 0.23, + "grad_norm": 2.5682391159916342, + "learning_rate": 9.029679175106874e-06, + "loss": 0.3608, + "step": 1623 + }, + { + "epoch": 0.23, + "grad_norm": 2.6038317606513606, + "learning_rate": 9.028345935947049e-06, + "loss": 0.3531, + "step": 1624 + }, + { + "epoch": 0.23, + "grad_norm": 2.1854894090907315, + "learning_rate": 9.027011880037397e-06, + "loss": 0.3583, + "step": 1625 + }, + { + "epoch": 0.23, + "grad_norm": 2.42499250366774, + "learning_rate": 9.025677007648402e-06, + "loss": 0.3553, + "step": 1626 + }, + { + "epoch": 0.23, + "grad_norm": 1.8430651264560063, + "learning_rate": 9.024341319050704e-06, + "loss": 0.3337, + "step": 1627 + }, + { + "epoch": 0.23, + "grad_norm": 1.9853700526133808, + "learning_rate": 9.023004814515121e-06, + "loss": 0.3307, + "step": 1628 + }, + { + "epoch": 0.23, + "grad_norm": 2.1359943342237706, + "learning_rate": 9.021667494312628e-06, + "loss": 0.3764, + "step": 1629 + }, + { + "epoch": 0.23, + "grad_norm": 2.256211138497533, + "learning_rate": 9.020329358714364e-06, + "loss": 0.3492, + "step": 1630 + }, + { + "epoch": 0.23, + "grad_norm": 2.731986580216264, + "learning_rate": 9.018990407991643e-06, + "loss": 0.3461, + "step": 1631 + }, + { + "epoch": 0.23, + "grad_norm": 2.9401242557298612, + "learning_rate": 9.017650642415933e-06, + "loss": 0.3829, + "step": 1632 + }, + { + "epoch": 0.23, + "grad_norm": 3.1324385214053643, + "learning_rate": 9.016310062258876e-06, + "loss": 0.3784, + "step": 1633 + }, + { + "epoch": 0.23, + "grad_norm": 1.9172400567480088, + "learning_rate": 9.014968667792275e-06, + "loss": 0.3611, + "step": 1634 + }, + { + "epoch": 0.23, + "grad_norm": 1.9849259104969792, + "learning_rate": 9.013626459288097e-06, + "loss": 0.3066, + "step": 1635 + }, + { + "epoch": 0.23, + "grad_norm": 2.3282580388832415, + "learning_rate": 9.012283437018474e-06, + "loss": 0.3798, + "step": 1636 + }, + { + "epoch": 0.23, + "grad_norm": 2.109710839310121, + "learning_rate": 9.01093960125571e-06, + "loss": 0.3521, + "step": 1637 + }, + { + "epoch": 0.23, + "grad_norm": 2.548442255520924, + "learning_rate": 9.009594952272266e-06, + "loss": 0.3656, + "step": 1638 + }, + { + "epoch": 0.23, + "grad_norm": 2.0597615877936732, + "learning_rate": 9.008249490340767e-06, + "loss": 0.3584, + "step": 1639 + }, + { + "epoch": 0.23, + "grad_norm": 3.9329451230283903, + "learning_rate": 9.006903215734011e-06, + "loss": 0.359, + "step": 1640 + }, + { + "epoch": 0.23, + "grad_norm": 2.3787442052991112, + "learning_rate": 9.005556128724954e-06, + "loss": 0.3649, + "step": 1641 + }, + { + "epoch": 0.23, + "grad_norm": 1.984881881143664, + "learning_rate": 9.004208229586721e-06, + "loss": 0.3363, + "step": 1642 + }, + { + "epoch": 0.23, + "grad_norm": 2.291608894860086, + "learning_rate": 9.002859518592598e-06, + "loss": 0.3387, + "step": 1643 + }, + { + "epoch": 0.23, + "grad_norm": 1.940365516994036, + "learning_rate": 9.001509996016034e-06, + "loss": 0.3572, + "step": 1644 + }, + { + "epoch": 0.23, + "grad_norm": 2.217213496645248, + "learning_rate": 9.000159662130649e-06, + "loss": 0.3597, + "step": 1645 + }, + { + "epoch": 0.23, + "grad_norm": 1.8175896893191694, + "learning_rate": 8.998808517210225e-06, + "loss": 0.3396, + "step": 1646 + }, + { + "epoch": 0.23, + "grad_norm": 2.8606830329687902, + "learning_rate": 8.997456561528706e-06, + "loss": 0.3534, + "step": 1647 + }, + { + "epoch": 0.23, + "grad_norm": 1.8355762053993576, + "learning_rate": 8.9961037953602e-06, + "loss": 0.374, + "step": 1648 + }, + { + "epoch": 0.23, + "grad_norm": 2.6472223494588585, + "learning_rate": 8.994750218978985e-06, + "loss": 0.3348, + "step": 1649 + }, + { + "epoch": 0.23, + "grad_norm": 2.2749974354154205, + "learning_rate": 8.993395832659497e-06, + "loss": 0.3505, + "step": 1650 + }, + { + "epoch": 0.23, + "grad_norm": 3.8362336243039907, + "learning_rate": 8.99204063667634e-06, + "loss": 0.3563, + "step": 1651 + }, + { + "epoch": 0.23, + "grad_norm": 2.0013150190305633, + "learning_rate": 8.99068463130428e-06, + "loss": 0.3634, + "step": 1652 + }, + { + "epoch": 0.23, + "grad_norm": 10.47941583893039, + "learning_rate": 8.989327816818248e-06, + "loss": 0.3748, + "step": 1653 + }, + { + "epoch": 0.23, + "grad_norm": 1.8992902131528766, + "learning_rate": 8.98797019349334e-06, + "loss": 0.3327, + "step": 1654 + }, + { + "epoch": 0.23, + "grad_norm": 2.4475531216374793, + "learning_rate": 8.986611761604814e-06, + "loss": 0.3537, + "step": 1655 + }, + { + "epoch": 0.23, + "grad_norm": 1.995842235657491, + "learning_rate": 8.985252521428094e-06, + "loss": 0.3532, + "step": 1656 + }, + { + "epoch": 0.23, + "grad_norm": 4.360070538831659, + "learning_rate": 8.983892473238765e-06, + "loss": 0.3256, + "step": 1657 + }, + { + "epoch": 0.23, + "grad_norm": 3.1121949528770694, + "learning_rate": 8.98253161731258e-06, + "loss": 0.3796, + "step": 1658 + }, + { + "epoch": 0.23, + "grad_norm": 4.335209647473415, + "learning_rate": 8.981169953925452e-06, + "loss": 0.3532, + "step": 1659 + }, + { + "epoch": 0.23, + "grad_norm": 2.186387808563297, + "learning_rate": 8.979807483353459e-06, + "loss": 0.3553, + "step": 1660 + }, + { + "epoch": 0.23, + "grad_norm": 2.182327997110226, + "learning_rate": 8.978444205872844e-06, + "loss": 0.3485, + "step": 1661 + }, + { + "epoch": 0.23, + "grad_norm": 2.3066615162897337, + "learning_rate": 8.977080121760012e-06, + "loss": 0.3334, + "step": 1662 + }, + { + "epoch": 0.23, + "grad_norm": 2.4624722956872094, + "learning_rate": 8.97571523129153e-06, + "loss": 0.3745, + "step": 1663 + }, + { + "epoch": 0.23, + "grad_norm": 2.1248819238542684, + "learning_rate": 8.974349534744133e-06, + "loss": 0.3757, + "step": 1664 + }, + { + "epoch": 0.23, + "grad_norm": 2.3414036094105204, + "learning_rate": 8.972983032394716e-06, + "loss": 0.3374, + "step": 1665 + }, + { + "epoch": 0.23, + "grad_norm": 2.2831865155713387, + "learning_rate": 8.97161572452034e-06, + "loss": 0.3485, + "step": 1666 + }, + { + "epoch": 0.23, + "grad_norm": 2.3459323544353774, + "learning_rate": 8.970247611398223e-06, + "loss": 0.3529, + "step": 1667 + }, + { + "epoch": 0.23, + "grad_norm": 3.4351474225219873, + "learning_rate": 8.968878693305756e-06, + "loss": 0.3967, + "step": 1668 + }, + { + "epoch": 0.23, + "grad_norm": 2.2480657394216923, + "learning_rate": 8.967508970520485e-06, + "loss": 0.3526, + "step": 1669 + }, + { + "epoch": 0.23, + "grad_norm": 2.4220156159292308, + "learning_rate": 8.966138443320124e-06, + "loss": 0.3537, + "step": 1670 + }, + { + "epoch": 0.23, + "grad_norm": 1.9806226280473958, + "learning_rate": 8.964767111982547e-06, + "loss": 0.372, + "step": 1671 + }, + { + "epoch": 0.23, + "grad_norm": 1.7415805325285079, + "learning_rate": 8.96339497678579e-06, + "loss": 0.3482, + "step": 1672 + }, + { + "epoch": 0.23, + "grad_norm": 2.3649476471791075, + "learning_rate": 8.962022038008059e-06, + "loss": 0.3539, + "step": 1673 + }, + { + "epoch": 0.23, + "grad_norm": 1.8695884374111074, + "learning_rate": 8.960648295927716e-06, + "loss": 0.3397, + "step": 1674 + }, + { + "epoch": 0.23, + "grad_norm": 2.0605597897972685, + "learning_rate": 8.95927375082329e-06, + "loss": 0.343, + "step": 1675 + }, + { + "epoch": 0.23, + "grad_norm": 2.5338013584669135, + "learning_rate": 8.957898402973467e-06, + "loss": 0.3273, + "step": 1676 + }, + { + "epoch": 0.23, + "grad_norm": 3.0291610394593276, + "learning_rate": 8.956522252657102e-06, + "loss": 0.3747, + "step": 1677 + }, + { + "epoch": 0.23, + "grad_norm": 2.208256105949683, + "learning_rate": 8.955145300153213e-06, + "loss": 0.3439, + "step": 1678 + }, + { + "epoch": 0.23, + "grad_norm": 2.2234115132653027, + "learning_rate": 8.953767545740972e-06, + "loss": 0.3711, + "step": 1679 + }, + { + "epoch": 0.23, + "grad_norm": 1.7504626949776267, + "learning_rate": 8.952388989699724e-06, + "loss": 0.3515, + "step": 1680 + }, + { + "epoch": 0.23, + "grad_norm": 2.2348233775574258, + "learning_rate": 8.951009632308973e-06, + "loss": 0.3555, + "step": 1681 + }, + { + "epoch": 0.23, + "grad_norm": 2.45994661476635, + "learning_rate": 8.949629473848382e-06, + "loss": 0.3459, + "step": 1682 + }, + { + "epoch": 0.23, + "grad_norm": 2.0179569841870353, + "learning_rate": 8.948248514597779e-06, + "loss": 0.3186, + "step": 1683 + }, + { + "epoch": 0.23, + "grad_norm": 4.2969272835703505, + "learning_rate": 8.946866754837156e-06, + "loss": 0.3627, + "step": 1684 + }, + { + "epoch": 0.23, + "grad_norm": 2.5705529068896102, + "learning_rate": 8.945484194846666e-06, + "loss": 0.3772, + "step": 1685 + }, + { + "epoch": 0.23, + "grad_norm": 1.8973580892960562, + "learning_rate": 8.944100834906622e-06, + "loss": 0.3604, + "step": 1686 + }, + { + "epoch": 0.23, + "grad_norm": 3.4936583705917115, + "learning_rate": 8.942716675297502e-06, + "loss": 0.3519, + "step": 1687 + }, + { + "epoch": 0.23, + "grad_norm": 1.93718688589324, + "learning_rate": 8.941331716299944e-06, + "loss": 0.3556, + "step": 1688 + }, + { + "epoch": 0.23, + "grad_norm": 3.050480831261928, + "learning_rate": 8.939945958194753e-06, + "loss": 0.3536, + "step": 1689 + }, + { + "epoch": 0.23, + "grad_norm": 1.9479809092907991, + "learning_rate": 8.93855940126289e-06, + "loss": 0.3695, + "step": 1690 + }, + { + "epoch": 0.24, + "grad_norm": 2.4920884363695457, + "learning_rate": 8.937172045785478e-06, + "loss": 0.347, + "step": 1691 + }, + { + "epoch": 0.24, + "grad_norm": 1.8475449574120308, + "learning_rate": 8.935783892043809e-06, + "loss": 0.3534, + "step": 1692 + }, + { + "epoch": 0.24, + "grad_norm": 1.987637947619648, + "learning_rate": 8.934394940319329e-06, + "loss": 0.4001, + "step": 1693 + }, + { + "epoch": 0.24, + "grad_norm": 2.355124208380363, + "learning_rate": 8.933005190893651e-06, + "loss": 0.3442, + "step": 1694 + }, + { + "epoch": 0.24, + "grad_norm": 2.1841456138782394, + "learning_rate": 8.931614644048545e-06, + "loss": 0.3567, + "step": 1695 + }, + { + "epoch": 0.24, + "grad_norm": 2.4100175111432485, + "learning_rate": 8.930223300065947e-06, + "loss": 0.3701, + "step": 1696 + }, + { + "epoch": 0.24, + "grad_norm": 4.330520565797649, + "learning_rate": 8.928831159227952e-06, + "loss": 0.3267, + "step": 1697 + }, + { + "epoch": 0.24, + "grad_norm": 2.4713658108340795, + "learning_rate": 8.92743822181682e-06, + "loss": 0.3686, + "step": 1698 + }, + { + "epoch": 0.24, + "grad_norm": 2.240464735053102, + "learning_rate": 8.926044488114966e-06, + "loss": 0.3623, + "step": 1699 + }, + { + "epoch": 0.24, + "grad_norm": 2.5921658701698247, + "learning_rate": 8.924649958404973e-06, + "loss": 0.3257, + "step": 1700 + }, + { + "epoch": 0.24, + "grad_norm": 2.0580504172660037, + "learning_rate": 8.923254632969582e-06, + "loss": 0.3334, + "step": 1701 + }, + { + "epoch": 0.24, + "grad_norm": 2.2650347240133217, + "learning_rate": 8.921858512091697e-06, + "loss": 0.3382, + "step": 1702 + }, + { + "epoch": 0.24, + "grad_norm": 2.1593915634075715, + "learning_rate": 8.920461596054382e-06, + "loss": 0.369, + "step": 1703 + }, + { + "epoch": 0.24, + "grad_norm": 3.1050916080832773, + "learning_rate": 8.919063885140862e-06, + "loss": 0.3507, + "step": 1704 + }, + { + "epoch": 0.24, + "grad_norm": 2.354241375218865, + "learning_rate": 8.917665379634524e-06, + "loss": 0.3443, + "step": 1705 + }, + { + "epoch": 0.24, + "grad_norm": 2.2542965741696563, + "learning_rate": 8.916266079818918e-06, + "loss": 0.3821, + "step": 1706 + }, + { + "epoch": 0.24, + "grad_norm": 2.1735665439928167, + "learning_rate": 8.91486598597775e-06, + "loss": 0.35, + "step": 1707 + }, + { + "epoch": 0.24, + "grad_norm": 2.4424039088911553, + "learning_rate": 8.913465098394892e-06, + "loss": 0.3558, + "step": 1708 + }, + { + "epoch": 0.24, + "grad_norm": 2.5611447135061036, + "learning_rate": 8.912063417354374e-06, + "loss": 0.335, + "step": 1709 + }, + { + "epoch": 0.24, + "grad_norm": 3.4460736991928687, + "learning_rate": 8.910660943140388e-06, + "loss": 0.3708, + "step": 1710 + }, + { + "epoch": 0.24, + "grad_norm": 1.932847664552096, + "learning_rate": 8.909257676037287e-06, + "loss": 0.3628, + "step": 1711 + }, + { + "epoch": 0.24, + "grad_norm": 2.011736510926362, + "learning_rate": 8.907853616329582e-06, + "loss": 0.3758, + "step": 1712 + }, + { + "epoch": 0.24, + "grad_norm": 1.9655214133225463, + "learning_rate": 8.90644876430195e-06, + "loss": 0.3317, + "step": 1713 + }, + { + "epoch": 0.24, + "grad_norm": 1.9674070246153224, + "learning_rate": 8.905043120239223e-06, + "loss": 0.3526, + "step": 1714 + }, + { + "epoch": 0.24, + "grad_norm": 4.0885409751768576, + "learning_rate": 8.9036366844264e-06, + "loss": 0.3669, + "step": 1715 + }, + { + "epoch": 0.24, + "grad_norm": 5.4136536393617956, + "learning_rate": 8.902229457148634e-06, + "loss": 0.3272, + "step": 1716 + }, + { + "epoch": 0.24, + "grad_norm": 2.6793476192865135, + "learning_rate": 8.900821438691242e-06, + "loss": 0.3436, + "step": 1717 + }, + { + "epoch": 0.24, + "grad_norm": 1.9530617501960494, + "learning_rate": 8.899412629339699e-06, + "loss": 0.353, + "step": 1718 + }, + { + "epoch": 0.24, + "grad_norm": 2.4113671482173937, + "learning_rate": 8.898003029379646e-06, + "loss": 0.3498, + "step": 1719 + }, + { + "epoch": 0.24, + "grad_norm": 2.8847589830822353, + "learning_rate": 8.896592639096877e-06, + "loss": 0.3699, + "step": 1720 + }, + { + "epoch": 0.24, + "grad_norm": 2.1768110371839478, + "learning_rate": 8.89518145877735e-06, + "loss": 0.3535, + "step": 1721 + }, + { + "epoch": 0.24, + "grad_norm": 1.7984383758825988, + "learning_rate": 8.893769488707183e-06, + "loss": 0.3416, + "step": 1722 + }, + { + "epoch": 0.24, + "grad_norm": 2.161430165683251, + "learning_rate": 8.892356729172656e-06, + "loss": 0.3439, + "step": 1723 + }, + { + "epoch": 0.24, + "grad_norm": 2.255432683463923, + "learning_rate": 8.890943180460203e-06, + "loss": 0.3747, + "step": 1724 + }, + { + "epoch": 0.24, + "grad_norm": 2.13192311286334, + "learning_rate": 8.889528842856424e-06, + "loss": 0.3519, + "step": 1725 + }, + { + "epoch": 0.24, + "grad_norm": 2.018547761011207, + "learning_rate": 8.88811371664808e-06, + "loss": 0.3606, + "step": 1726 + }, + { + "epoch": 0.24, + "grad_norm": 2.2764811559806737, + "learning_rate": 8.886697802122082e-06, + "loss": 0.3815, + "step": 1727 + }, + { + "epoch": 0.24, + "grad_norm": 2.2395252260237553, + "learning_rate": 8.885281099565512e-06, + "loss": 0.3552, + "step": 1728 + }, + { + "epoch": 0.24, + "grad_norm": 2.371757416122058, + "learning_rate": 8.883863609265609e-06, + "loss": 0.3437, + "step": 1729 + }, + { + "epoch": 0.24, + "grad_norm": 1.769129531764842, + "learning_rate": 8.882445331509766e-06, + "loss": 0.3663, + "step": 1730 + }, + { + "epoch": 0.24, + "grad_norm": 2.0508348238295335, + "learning_rate": 8.881026266585542e-06, + "loss": 0.3425, + "step": 1731 + }, + { + "epoch": 0.24, + "grad_norm": 6.5110772174217955, + "learning_rate": 8.879606414780652e-06, + "loss": 0.3352, + "step": 1732 + }, + { + "epoch": 0.24, + "grad_norm": 3.238606497498135, + "learning_rate": 8.878185776382974e-06, + "loss": 0.3749, + "step": 1733 + }, + { + "epoch": 0.24, + "grad_norm": 2.6653341259864614, + "learning_rate": 8.876764351680542e-06, + "loss": 0.3638, + "step": 1734 + }, + { + "epoch": 0.24, + "grad_norm": 2.386769808813547, + "learning_rate": 8.87534214096155e-06, + "loss": 0.3384, + "step": 1735 + }, + { + "epoch": 0.24, + "grad_norm": 2.0445421338902743, + "learning_rate": 8.873919144514355e-06, + "loss": 0.3741, + "step": 1736 + }, + { + "epoch": 0.24, + "grad_norm": 2.640727585790127, + "learning_rate": 8.872495362627468e-06, + "loss": 0.3399, + "step": 1737 + }, + { + "epoch": 0.24, + "grad_norm": 2.020904201887481, + "learning_rate": 8.871070795589562e-06, + "loss": 0.3367, + "step": 1738 + }, + { + "epoch": 0.24, + "grad_norm": 3.0624540423923325, + "learning_rate": 8.86964544368947e-06, + "loss": 0.3519, + "step": 1739 + }, + { + "epoch": 0.24, + "grad_norm": 2.1630607262355728, + "learning_rate": 8.86821930721618e-06, + "loss": 0.3733, + "step": 1740 + }, + { + "epoch": 0.24, + "grad_norm": 2.6443056658324364, + "learning_rate": 8.866792386458848e-06, + "loss": 0.3547, + "step": 1741 + }, + { + "epoch": 0.24, + "grad_norm": 2.0408250324302064, + "learning_rate": 8.865364681706776e-06, + "loss": 0.3256, + "step": 1742 + }, + { + "epoch": 0.24, + "grad_norm": 2.4762436427866805, + "learning_rate": 8.863936193249439e-06, + "loss": 0.3745, + "step": 1743 + }, + { + "epoch": 0.24, + "grad_norm": 2.4865320520108467, + "learning_rate": 8.86250692137646e-06, + "loss": 0.3487, + "step": 1744 + }, + { + "epoch": 0.24, + "grad_norm": 2.83926076262984, + "learning_rate": 8.861076866377625e-06, + "loss": 0.3526, + "step": 1745 + }, + { + "epoch": 0.24, + "grad_norm": 1.9986585945604518, + "learning_rate": 8.859646028542877e-06, + "loss": 0.3659, + "step": 1746 + }, + { + "epoch": 0.24, + "grad_norm": 1.9621870157848862, + "learning_rate": 8.858214408162325e-06, + "loss": 0.3749, + "step": 1747 + }, + { + "epoch": 0.24, + "grad_norm": 2.392757691530449, + "learning_rate": 8.856782005526226e-06, + "loss": 0.3497, + "step": 1748 + }, + { + "epoch": 0.24, + "grad_norm": 3.914220449993792, + "learning_rate": 8.855348820925001e-06, + "loss": 0.3408, + "step": 1749 + }, + { + "epoch": 0.24, + "grad_norm": 3.207371438186789, + "learning_rate": 8.853914854649232e-06, + "loss": 0.3599, + "step": 1750 + }, + { + "epoch": 0.24, + "grad_norm": 2.902756913302656, + "learning_rate": 8.852480106989654e-06, + "loss": 0.3639, + "step": 1751 + }, + { + "epoch": 0.24, + "grad_norm": 3.1960916589754116, + "learning_rate": 8.851044578237163e-06, + "loss": 0.3883, + "step": 1752 + }, + { + "epoch": 0.24, + "grad_norm": 1.9829440028476422, + "learning_rate": 8.849608268682814e-06, + "loss": 0.3684, + "step": 1753 + }, + { + "epoch": 0.24, + "grad_norm": 2.2336473309390454, + "learning_rate": 8.848171178617822e-06, + "loss": 0.3562, + "step": 1754 + }, + { + "epoch": 0.24, + "grad_norm": 2.401658702436871, + "learning_rate": 8.846733308333555e-06, + "loss": 0.3405, + "step": 1755 + }, + { + "epoch": 0.24, + "grad_norm": 2.4258591005716297, + "learning_rate": 8.845294658121542e-06, + "loss": 0.3492, + "step": 1756 + }, + { + "epoch": 0.24, + "grad_norm": 1.8212949477424634, + "learning_rate": 8.843855228273472e-06, + "loss": 0.3261, + "step": 1757 + }, + { + "epoch": 0.24, + "grad_norm": 1.9445689221655482, + "learning_rate": 8.84241501908119e-06, + "loss": 0.3591, + "step": 1758 + }, + { + "epoch": 0.24, + "grad_norm": 2.0513827563610376, + "learning_rate": 8.8409740308367e-06, + "loss": 0.3473, + "step": 1759 + }, + { + "epoch": 0.24, + "grad_norm": 2.036373128799953, + "learning_rate": 8.83953226383216e-06, + "loss": 0.3586, + "step": 1760 + }, + { + "epoch": 0.24, + "grad_norm": 2.057072641701227, + "learning_rate": 8.838089718359895e-06, + "loss": 0.3426, + "step": 1761 + }, + { + "epoch": 0.24, + "grad_norm": 2.68446446509157, + "learning_rate": 8.836646394712376e-06, + "loss": 0.3383, + "step": 1762 + }, + { + "epoch": 0.25, + "grad_norm": 1.8155021795811577, + "learning_rate": 8.835202293182242e-06, + "loss": 0.3483, + "step": 1763 + }, + { + "epoch": 0.25, + "grad_norm": 2.152299769066046, + "learning_rate": 8.833757414062284e-06, + "loss": 0.4017, + "step": 1764 + }, + { + "epoch": 0.25, + "grad_norm": 2.194582768280641, + "learning_rate": 8.832311757645452e-06, + "loss": 0.3815, + "step": 1765 + }, + { + "epoch": 0.25, + "grad_norm": 2.4785363592752736, + "learning_rate": 8.830865324224854e-06, + "loss": 0.3618, + "step": 1766 + }, + { + "epoch": 0.25, + "grad_norm": 2.345426973841663, + "learning_rate": 8.829418114093756e-06, + "loss": 0.3412, + "step": 1767 + }, + { + "epoch": 0.25, + "grad_norm": 2.351294097560187, + "learning_rate": 8.82797012754558e-06, + "loss": 0.3447, + "step": 1768 + }, + { + "epoch": 0.25, + "grad_norm": 2.315006033326096, + "learning_rate": 8.826521364873906e-06, + "loss": 0.3543, + "step": 1769 + }, + { + "epoch": 0.25, + "grad_norm": 2.0415156514056965, + "learning_rate": 8.825071826372474e-06, + "loss": 0.338, + "step": 1770 + }, + { + "epoch": 0.25, + "grad_norm": 2.423664151718502, + "learning_rate": 8.823621512335175e-06, + "loss": 0.3645, + "step": 1771 + }, + { + "epoch": 0.25, + "grad_norm": 2.970152902161558, + "learning_rate": 8.822170423056062e-06, + "loss": 0.3482, + "step": 1772 + }, + { + "epoch": 0.25, + "grad_norm": 2.355292528625311, + "learning_rate": 8.82071855882935e-06, + "loss": 0.3477, + "step": 1773 + }, + { + "epoch": 0.25, + "grad_norm": 2.4409726404085537, + "learning_rate": 8.819265919949395e-06, + "loss": 0.3354, + "step": 1774 + }, + { + "epoch": 0.25, + "grad_norm": 15.236647496024847, + "learning_rate": 8.81781250671073e-06, + "loss": 0.3345, + "step": 1775 + }, + { + "epoch": 0.25, + "grad_norm": 2.6117565393945017, + "learning_rate": 8.816358319408031e-06, + "loss": 0.3718, + "step": 1776 + }, + { + "epoch": 0.25, + "grad_norm": 7.51316453121629, + "learning_rate": 8.814903358336137e-06, + "loss": 0.3911, + "step": 1777 + }, + { + "epoch": 0.25, + "grad_norm": 2.3652324891032053, + "learning_rate": 8.813447623790043e-06, + "loss": 0.3595, + "step": 1778 + }, + { + "epoch": 0.25, + "grad_norm": 7.400316775811969, + "learning_rate": 8.811991116064896e-06, + "loss": 0.3551, + "step": 1779 + }, + { + "epoch": 0.25, + "grad_norm": 3.5347103597864313, + "learning_rate": 8.81053383545601e-06, + "loss": 0.341, + "step": 1780 + }, + { + "epoch": 0.25, + "grad_norm": 5.586870687719069, + "learning_rate": 8.809075782258842e-06, + "loss": 0.3714, + "step": 1781 + }, + { + "epoch": 0.25, + "grad_norm": 3.2891125620737958, + "learning_rate": 8.80761695676902e-06, + "loss": 0.3429, + "step": 1782 + }, + { + "epoch": 0.25, + "grad_norm": 1.9112824512154871, + "learning_rate": 8.80615735928232e-06, + "loss": 0.3684, + "step": 1783 + }, + { + "epoch": 0.25, + "grad_norm": 2.876467983635815, + "learning_rate": 8.804696990094677e-06, + "loss": 0.356, + "step": 1784 + }, + { + "epoch": 0.25, + "grad_norm": 2.1565845121746476, + "learning_rate": 8.803235849502178e-06, + "loss": 0.3501, + "step": 1785 + }, + { + "epoch": 0.25, + "grad_norm": 2.0871232776566955, + "learning_rate": 8.801773937801075e-06, + "loss": 0.3772, + "step": 1786 + }, + { + "epoch": 0.25, + "grad_norm": 4.124928998937251, + "learning_rate": 8.800311255287768e-06, + "loss": 0.3846, + "step": 1787 + }, + { + "epoch": 0.25, + "grad_norm": 3.319673428625679, + "learning_rate": 8.79884780225882e-06, + "loss": 0.3598, + "step": 1788 + }, + { + "epoch": 0.25, + "grad_norm": 3.480116993045024, + "learning_rate": 8.797383579010943e-06, + "loss": 0.3252, + "step": 1789 + }, + { + "epoch": 0.25, + "grad_norm": 2.9368570946206574, + "learning_rate": 8.795918585841013e-06, + "loss": 0.3804, + "step": 1790 + }, + { + "epoch": 0.25, + "grad_norm": 2.4376411358460244, + "learning_rate": 8.794452823046056e-06, + "loss": 0.364, + "step": 1791 + }, + { + "epoch": 0.25, + "grad_norm": 2.5312285220778517, + "learning_rate": 8.792986290923258e-06, + "loss": 0.3293, + "step": 1792 + }, + { + "epoch": 0.25, + "grad_norm": 2.083241447949004, + "learning_rate": 8.791518989769959e-06, + "loss": 0.3619, + "step": 1793 + }, + { + "epoch": 0.25, + "grad_norm": 2.813478207133536, + "learning_rate": 8.790050919883655e-06, + "loss": 0.3807, + "step": 1794 + }, + { + "epoch": 0.25, + "grad_norm": 2.699827409886475, + "learning_rate": 8.788582081561998e-06, + "loss": 0.3283, + "step": 1795 + }, + { + "epoch": 0.25, + "grad_norm": 1.935191646710915, + "learning_rate": 8.787112475102796e-06, + "loss": 0.3502, + "step": 1796 + }, + { + "epoch": 0.25, + "grad_norm": 2.314486897483611, + "learning_rate": 8.785642100804013e-06, + "loss": 0.3757, + "step": 1797 + }, + { + "epoch": 0.25, + "grad_norm": 2.057350782977668, + "learning_rate": 8.784170958963768e-06, + "loss": 0.3564, + "step": 1798 + }, + { + "epoch": 0.25, + "grad_norm": 1.851692635461214, + "learning_rate": 8.782699049880336e-06, + "loss": 0.3922, + "step": 1799 + }, + { + "epoch": 0.25, + "grad_norm": 2.774307959216126, + "learning_rate": 8.781226373852148e-06, + "loss": 0.3451, + "step": 1800 + }, + { + "epoch": 0.25, + "grad_norm": 1.8565077548134263, + "learning_rate": 8.77975293117779e-06, + "loss": 0.3436, + "step": 1801 + }, + { + "epoch": 0.25, + "grad_norm": 1.9471214768583494, + "learning_rate": 8.778278722156003e-06, + "loss": 0.3583, + "step": 1802 + }, + { + "epoch": 0.25, + "grad_norm": 2.1197176642512194, + "learning_rate": 8.776803747085686e-06, + "loss": 0.3464, + "step": 1803 + }, + { + "epoch": 0.25, + "grad_norm": 1.926411009775773, + "learning_rate": 8.775328006265886e-06, + "loss": 0.3619, + "step": 1804 + }, + { + "epoch": 0.25, + "grad_norm": 2.0974790892968826, + "learning_rate": 8.773851499995815e-06, + "loss": 0.3211, + "step": 1805 + }, + { + "epoch": 0.25, + "grad_norm": 3.9933511786974987, + "learning_rate": 8.772374228574836e-06, + "loss": 0.3626, + "step": 1806 + }, + { + "epoch": 0.25, + "grad_norm": 1.875951630443157, + "learning_rate": 8.770896192302463e-06, + "loss": 0.3501, + "step": 1807 + }, + { + "epoch": 0.25, + "grad_norm": 2.305971536893168, + "learning_rate": 8.769417391478372e-06, + "loss": 0.3423, + "step": 1808 + }, + { + "epoch": 0.25, + "grad_norm": 2.3679808226166146, + "learning_rate": 8.767937826402388e-06, + "loss": 0.3677, + "step": 1809 + }, + { + "epoch": 0.25, + "grad_norm": 1.7988006384424764, + "learning_rate": 8.766457497374498e-06, + "loss": 0.346, + "step": 1810 + }, + { + "epoch": 0.25, + "grad_norm": 1.8272281431868809, + "learning_rate": 8.764976404694837e-06, + "loss": 0.3534, + "step": 1811 + }, + { + "epoch": 0.25, + "grad_norm": 3.3281378539293396, + "learning_rate": 8.763494548663698e-06, + "loss": 0.3387, + "step": 1812 + }, + { + "epoch": 0.25, + "grad_norm": 2.312249440617897, + "learning_rate": 8.762011929581526e-06, + "loss": 0.3444, + "step": 1813 + }, + { + "epoch": 0.25, + "grad_norm": 1.9318777582271007, + "learning_rate": 8.760528547748928e-06, + "loss": 0.3481, + "step": 1814 + }, + { + "epoch": 0.25, + "grad_norm": 2.170484925812013, + "learning_rate": 8.759044403466657e-06, + "loss": 0.353, + "step": 1815 + }, + { + "epoch": 0.25, + "grad_norm": 1.8355326845855704, + "learning_rate": 8.757559497035623e-06, + "loss": 0.3434, + "step": 1816 + }, + { + "epoch": 0.25, + "grad_norm": 2.895938118026698, + "learning_rate": 8.756073828756895e-06, + "loss": 0.3588, + "step": 1817 + }, + { + "epoch": 0.25, + "grad_norm": 1.9778068618584557, + "learning_rate": 8.754587398931693e-06, + "loss": 0.3652, + "step": 1818 + }, + { + "epoch": 0.25, + "grad_norm": 2.2353832509728746, + "learning_rate": 8.753100207861392e-06, + "loss": 0.3549, + "step": 1819 + }, + { + "epoch": 0.25, + "grad_norm": 2.0767783001456372, + "learning_rate": 8.751612255847517e-06, + "loss": 0.3574, + "step": 1820 + }, + { + "epoch": 0.25, + "grad_norm": 6.265565687151176, + "learning_rate": 8.750123543191754e-06, + "loss": 0.3487, + "step": 1821 + }, + { + "epoch": 0.25, + "grad_norm": 2.572513299748265, + "learning_rate": 8.748634070195942e-06, + "loss": 0.3814, + "step": 1822 + }, + { + "epoch": 0.25, + "grad_norm": 2.002293455215471, + "learning_rate": 8.747143837162071e-06, + "loss": 0.3382, + "step": 1823 + }, + { + "epoch": 0.25, + "grad_norm": 1.993941145703842, + "learning_rate": 8.745652844392285e-06, + "loss": 0.3515, + "step": 1824 + }, + { + "epoch": 0.25, + "grad_norm": 2.159093817497519, + "learning_rate": 8.744161092188887e-06, + "loss": 0.3608, + "step": 1825 + }, + { + "epoch": 0.25, + "grad_norm": 2.351685024578397, + "learning_rate": 8.74266858085433e-06, + "loss": 0.3592, + "step": 1826 + }, + { + "epoch": 0.25, + "grad_norm": 2.565325063136761, + "learning_rate": 8.74117531069122e-06, + "loss": 0.3556, + "step": 1827 + }, + { + "epoch": 0.25, + "grad_norm": 1.8832021381566844, + "learning_rate": 8.739681282002318e-06, + "loss": 0.3422, + "step": 1828 + }, + { + "epoch": 0.25, + "grad_norm": 1.9691405743646753, + "learning_rate": 8.738186495090543e-06, + "loss": 0.3429, + "step": 1829 + }, + { + "epoch": 0.25, + "grad_norm": 1.7569805815628126, + "learning_rate": 8.736690950258962e-06, + "loss": 0.3561, + "step": 1830 + }, + { + "epoch": 0.25, + "grad_norm": 2.6126723861068815, + "learning_rate": 8.735194647810797e-06, + "loss": 0.3505, + "step": 1831 + }, + { + "epoch": 0.25, + "grad_norm": 2.1992044988266746, + "learning_rate": 8.733697588049426e-06, + "loss": 0.335, + "step": 1832 + }, + { + "epoch": 0.25, + "grad_norm": 2.12076995348315, + "learning_rate": 8.732199771278376e-06, + "loss": 0.3363, + "step": 1833 + }, + { + "epoch": 0.25, + "grad_norm": 4.073994657316734, + "learning_rate": 8.730701197801333e-06, + "loss": 0.3508, + "step": 1834 + }, + { + "epoch": 0.26, + "grad_norm": 6.428334878002039, + "learning_rate": 8.729201867922132e-06, + "loss": 0.3561, + "step": 1835 + }, + { + "epoch": 0.26, + "grad_norm": 2.5569970516180387, + "learning_rate": 8.727701781944766e-06, + "loss": 0.3486, + "step": 1836 + }, + { + "epoch": 0.26, + "grad_norm": 3.8389325606651483, + "learning_rate": 8.726200940173374e-06, + "loss": 0.36, + "step": 1837 + }, + { + "epoch": 0.26, + "grad_norm": 2.0580492568092166, + "learning_rate": 8.724699342912258e-06, + "loss": 0.3716, + "step": 1838 + }, + { + "epoch": 0.26, + "grad_norm": 1.7861801475895251, + "learning_rate": 8.723196990465863e-06, + "loss": 0.3698, + "step": 1839 + }, + { + "epoch": 0.26, + "grad_norm": 1.9160812928065178, + "learning_rate": 8.721693883138795e-06, + "loss": 0.362, + "step": 1840 + }, + { + "epoch": 0.26, + "grad_norm": 2.8731273283020395, + "learning_rate": 8.720190021235808e-06, + "loss": 0.3231, + "step": 1841 + }, + { + "epoch": 0.26, + "grad_norm": 2.0066622169782704, + "learning_rate": 8.718685405061813e-06, + "loss": 0.3682, + "step": 1842 + }, + { + "epoch": 0.26, + "grad_norm": 1.7516486992721054, + "learning_rate": 8.717180034921871e-06, + "loss": 0.3454, + "step": 1843 + }, + { + "epoch": 0.26, + "grad_norm": 2.3627264764192, + "learning_rate": 8.715673911121196e-06, + "loss": 0.3518, + "step": 1844 + }, + { + "epoch": 0.26, + "grad_norm": 2.0436740948037517, + "learning_rate": 8.714167033965156e-06, + "loss": 0.375, + "step": 1845 + }, + { + "epoch": 0.26, + "grad_norm": 3.3046547283461263, + "learning_rate": 8.712659403759273e-06, + "loss": 0.3265, + "step": 1846 + }, + { + "epoch": 0.26, + "grad_norm": 2.255527111544609, + "learning_rate": 8.711151020809217e-06, + "loss": 0.3666, + "step": 1847 + }, + { + "epoch": 0.26, + "grad_norm": 3.7822573733935245, + "learning_rate": 8.709641885420817e-06, + "loss": 0.3601, + "step": 1848 + }, + { + "epoch": 0.26, + "grad_norm": 2.4425547885717047, + "learning_rate": 8.70813199790005e-06, + "loss": 0.3726, + "step": 1849 + }, + { + "epoch": 0.26, + "grad_norm": 2.0180134930646774, + "learning_rate": 8.706621358553044e-06, + "loss": 0.365, + "step": 1850 + }, + { + "epoch": 0.26, + "grad_norm": 2.009968836417301, + "learning_rate": 8.705109967686086e-06, + "loss": 0.3567, + "step": 1851 + }, + { + "epoch": 0.26, + "grad_norm": 2.3037414247408377, + "learning_rate": 8.703597825605608e-06, + "loss": 0.3758, + "step": 1852 + }, + { + "epoch": 0.26, + "grad_norm": 2.1317118560659734, + "learning_rate": 8.702084932618202e-06, + "loss": 0.3396, + "step": 1853 + }, + { + "epoch": 0.26, + "grad_norm": 1.6762351366793806, + "learning_rate": 8.700571289030603e-06, + "loss": 0.3532, + "step": 1854 + }, + { + "epoch": 0.26, + "grad_norm": 1.7752382105287197, + "learning_rate": 8.699056895149708e-06, + "loss": 0.3401, + "step": 1855 + }, + { + "epoch": 0.26, + "grad_norm": 2.4407903176060546, + "learning_rate": 8.69754175128256e-06, + "loss": 0.3441, + "step": 1856 + }, + { + "epoch": 0.26, + "grad_norm": 4.051975742600371, + "learning_rate": 8.696025857736355e-06, + "loss": 0.369, + "step": 1857 + }, + { + "epoch": 0.26, + "grad_norm": 3.5090378156518187, + "learning_rate": 8.69450921481844e-06, + "loss": 0.371, + "step": 1858 + }, + { + "epoch": 0.26, + "grad_norm": 5.740013860256348, + "learning_rate": 8.692991822836316e-06, + "loss": 0.3458, + "step": 1859 + }, + { + "epoch": 0.26, + "grad_norm": 2.832829113485366, + "learning_rate": 8.691473682097637e-06, + "loss": 0.3395, + "step": 1860 + }, + { + "epoch": 0.26, + "grad_norm": 2.099841383829608, + "learning_rate": 8.689954792910207e-06, + "loss": 0.352, + "step": 1861 + }, + { + "epoch": 0.26, + "grad_norm": 3.3572442000601974, + "learning_rate": 8.68843515558198e-06, + "loss": 0.3314, + "step": 1862 + }, + { + "epoch": 0.26, + "grad_norm": 2.1728352273443314, + "learning_rate": 8.686914770421063e-06, + "loss": 0.3464, + "step": 1863 + }, + { + "epoch": 0.26, + "grad_norm": 2.397557719747904, + "learning_rate": 8.685393637735718e-06, + "loss": 0.3517, + "step": 1864 + }, + { + "epoch": 0.26, + "grad_norm": 2.3835913626631315, + "learning_rate": 8.683871757834353e-06, + "loss": 0.3469, + "step": 1865 + }, + { + "epoch": 0.26, + "grad_norm": 2.171898835271978, + "learning_rate": 8.68234913102553e-06, + "loss": 0.3537, + "step": 1866 + }, + { + "epoch": 0.26, + "grad_norm": 4.088967999286961, + "learning_rate": 8.680825757617966e-06, + "loss": 0.3525, + "step": 1867 + }, + { + "epoch": 0.26, + "grad_norm": 2.5536271131515256, + "learning_rate": 8.67930163792052e-06, + "loss": 0.393, + "step": 1868 + }, + { + "epoch": 0.26, + "grad_norm": 7.413753006358147, + "learning_rate": 8.677776772242216e-06, + "loss": 0.3419, + "step": 1869 + }, + { + "epoch": 0.26, + "grad_norm": 2.418715208522257, + "learning_rate": 8.676251160892216e-06, + "loss": 0.3717, + "step": 1870 + }, + { + "epoch": 0.26, + "grad_norm": 1.8728972351843087, + "learning_rate": 8.67472480417984e-06, + "loss": 0.3563, + "step": 1871 + }, + { + "epoch": 0.26, + "grad_norm": 2.2814649740853783, + "learning_rate": 8.673197702414559e-06, + "loss": 0.3721, + "step": 1872 + }, + { + "epoch": 0.26, + "grad_norm": 2.852407451489275, + "learning_rate": 8.67166985590599e-06, + "loss": 0.3297, + "step": 1873 + }, + { + "epoch": 0.26, + "grad_norm": 2.2064236182599175, + "learning_rate": 8.670141264963912e-06, + "loss": 0.3521, + "step": 1874 + }, + { + "epoch": 0.26, + "grad_norm": 2.057788340148867, + "learning_rate": 8.668611929898242e-06, + "loss": 0.3616, + "step": 1875 + }, + { + "epoch": 0.26, + "grad_norm": 1.993556415709249, + "learning_rate": 8.667081851019054e-06, + "loss": 0.3439, + "step": 1876 + }, + { + "epoch": 0.26, + "grad_norm": 1.7134930382594291, + "learning_rate": 8.665551028636576e-06, + "loss": 0.3343, + "step": 1877 + }, + { + "epoch": 0.26, + "grad_norm": 2.842545537416324, + "learning_rate": 8.66401946306118e-06, + "loss": 0.3554, + "step": 1878 + }, + { + "epoch": 0.26, + "grad_norm": 2.1486806541190138, + "learning_rate": 8.662487154603394e-06, + "loss": 0.3698, + "step": 1879 + }, + { + "epoch": 0.26, + "grad_norm": 1.768816848570785, + "learning_rate": 8.660954103573893e-06, + "loss": 0.363, + "step": 1880 + }, + { + "epoch": 0.26, + "grad_norm": 2.0211085044800727, + "learning_rate": 8.659420310283507e-06, + "loss": 0.3626, + "step": 1881 + }, + { + "epoch": 0.26, + "grad_norm": 2.2787299239160927, + "learning_rate": 8.657885775043208e-06, + "loss": 0.3471, + "step": 1882 + }, + { + "epoch": 0.26, + "grad_norm": 2.270209351869361, + "learning_rate": 8.65635049816413e-06, + "loss": 0.3504, + "step": 1883 + }, + { + "epoch": 0.26, + "grad_norm": 2.221862132018184, + "learning_rate": 8.654814479957547e-06, + "loss": 0.3859, + "step": 1884 + }, + { + "epoch": 0.26, + "grad_norm": 2.367789788874681, + "learning_rate": 8.653277720734891e-06, + "loss": 0.3579, + "step": 1885 + }, + { + "epoch": 0.26, + "grad_norm": 2.094871624333058, + "learning_rate": 8.651740220807739e-06, + "loss": 0.3391, + "step": 1886 + }, + { + "epoch": 0.26, + "grad_norm": 1.541128534914522, + "learning_rate": 8.65020198048782e-06, + "loss": 0.3181, + "step": 1887 + }, + { + "epoch": 0.26, + "grad_norm": 2.427827837459815, + "learning_rate": 8.648663000087015e-06, + "loss": 0.347, + "step": 1888 + }, + { + "epoch": 0.26, + "grad_norm": 2.858557286946919, + "learning_rate": 8.647123279917352e-06, + "loss": 0.3563, + "step": 1889 + }, + { + "epoch": 0.26, + "grad_norm": 1.9285924151512779, + "learning_rate": 8.645582820291009e-06, + "loss": 0.3763, + "step": 1890 + }, + { + "epoch": 0.26, + "grad_norm": 15.029563238046116, + "learning_rate": 8.64404162152032e-06, + "loss": 0.3292, + "step": 1891 + }, + { + "epoch": 0.26, + "grad_norm": 1.9430571340733247, + "learning_rate": 8.642499683917759e-06, + "loss": 0.3597, + "step": 1892 + }, + { + "epoch": 0.26, + "grad_norm": 1.890268020738639, + "learning_rate": 8.640957007795956e-06, + "loss": 0.3896, + "step": 1893 + }, + { + "epoch": 0.26, + "grad_norm": 1.9639537228870239, + "learning_rate": 8.639413593467691e-06, + "loss": 0.3767, + "step": 1894 + }, + { + "epoch": 0.26, + "grad_norm": 3.2785644501424587, + "learning_rate": 8.637869441245892e-06, + "loss": 0.3468, + "step": 1895 + }, + { + "epoch": 0.26, + "grad_norm": 2.1064555556113786, + "learning_rate": 8.636324551443636e-06, + "loss": 0.3576, + "step": 1896 + }, + { + "epoch": 0.26, + "grad_norm": 2.0401403029840846, + "learning_rate": 8.634778924374152e-06, + "loss": 0.3334, + "step": 1897 + }, + { + "epoch": 0.26, + "grad_norm": 2.139996251162376, + "learning_rate": 8.633232560350814e-06, + "loss": 0.367, + "step": 1898 + }, + { + "epoch": 0.26, + "grad_norm": 3.56546417390585, + "learning_rate": 8.631685459687152e-06, + "loss": 0.337, + "step": 1899 + }, + { + "epoch": 0.26, + "grad_norm": 2.139197914288517, + "learning_rate": 8.630137622696839e-06, + "loss": 0.3406, + "step": 1900 + }, + { + "epoch": 0.26, + "grad_norm": 1.989481007328386, + "learning_rate": 8.6285890496937e-06, + "loss": 0.3394, + "step": 1901 + }, + { + "epoch": 0.26, + "grad_norm": 2.117771329156402, + "learning_rate": 8.627039740991712e-06, + "loss": 0.3412, + "step": 1902 + }, + { + "epoch": 0.26, + "grad_norm": 1.9975265848280135, + "learning_rate": 8.625489696904996e-06, + "loss": 0.3463, + "step": 1903 + }, + { + "epoch": 0.26, + "grad_norm": 2.3482161251605524, + "learning_rate": 8.623938917747824e-06, + "loss": 0.3316, + "step": 1904 + }, + { + "epoch": 0.26, + "grad_norm": 1.9819834283062594, + "learning_rate": 8.62238740383462e-06, + "loss": 0.3489, + "step": 1905 + }, + { + "epoch": 0.26, + "grad_norm": 2.42249311480117, + "learning_rate": 8.620835155479951e-06, + "loss": 0.3606, + "step": 1906 + }, + { + "epoch": 0.27, + "grad_norm": 2.512308258129878, + "learning_rate": 8.619282172998539e-06, + "loss": 0.3228, + "step": 1907 + }, + { + "epoch": 0.27, + "grad_norm": 2.510124497244114, + "learning_rate": 8.617728456705253e-06, + "loss": 0.3671, + "step": 1908 + }, + { + "epoch": 0.27, + "grad_norm": 2.3908542012379477, + "learning_rate": 8.616174006915105e-06, + "loss": 0.3573, + "step": 1909 + }, + { + "epoch": 0.27, + "grad_norm": 109.70484963032625, + "learning_rate": 8.614618823943268e-06, + "loss": 0.3485, + "step": 1910 + }, + { + "epoch": 0.27, + "grad_norm": 2.3118953504720476, + "learning_rate": 8.613062908105051e-06, + "loss": 0.3592, + "step": 1911 + }, + { + "epoch": 0.27, + "grad_norm": 2.0922997274050243, + "learning_rate": 8.611506259715918e-06, + "loss": 0.357, + "step": 1912 + }, + { + "epoch": 0.27, + "grad_norm": 2.267873695136359, + "learning_rate": 8.609948879091482e-06, + "loss": 0.3483, + "step": 1913 + }, + { + "epoch": 0.27, + "grad_norm": 2.2073089265359442, + "learning_rate": 8.608390766547504e-06, + "loss": 0.3174, + "step": 1914 + }, + { + "epoch": 0.27, + "grad_norm": 3.50202121692801, + "learning_rate": 8.606831922399888e-06, + "loss": 0.3423, + "step": 1915 + }, + { + "epoch": 0.27, + "grad_norm": 2.021153872963197, + "learning_rate": 8.605272346964692e-06, + "loss": 0.3713, + "step": 1916 + }, + { + "epoch": 0.27, + "grad_norm": 2.0062743701398964, + "learning_rate": 8.603712040558126e-06, + "loss": 0.3384, + "step": 1917 + }, + { + "epoch": 0.27, + "grad_norm": 1.7904444366957033, + "learning_rate": 8.602151003496536e-06, + "loss": 0.365, + "step": 1918 + }, + { + "epoch": 0.27, + "grad_norm": 1.8401471318356102, + "learning_rate": 8.600589236096428e-06, + "loss": 0.3555, + "step": 1919 + }, + { + "epoch": 0.27, + "grad_norm": 2.1087640101070777, + "learning_rate": 8.59902673867445e-06, + "loss": 0.3773, + "step": 1920 + }, + { + "epoch": 0.27, + "grad_norm": 2.988876836528279, + "learning_rate": 8.597463511547399e-06, + "loss": 0.3311, + "step": 1921 + }, + { + "epoch": 0.27, + "grad_norm": 3.104988687360701, + "learning_rate": 8.59589955503222e-06, + "loss": 0.3206, + "step": 1922 + }, + { + "epoch": 0.27, + "grad_norm": 5.402969739661756, + "learning_rate": 8.594334869446008e-06, + "loss": 0.3649, + "step": 1923 + }, + { + "epoch": 0.27, + "grad_norm": 2.2515111617925307, + "learning_rate": 8.592769455106002e-06, + "loss": 0.3401, + "step": 1924 + }, + { + "epoch": 0.27, + "grad_norm": 2.5060779304924456, + "learning_rate": 8.591203312329594e-06, + "loss": 0.3862, + "step": 1925 + }, + { + "epoch": 0.27, + "grad_norm": 2.469468462855278, + "learning_rate": 8.589636441434317e-06, + "loss": 0.3594, + "step": 1926 + }, + { + "epoch": 0.27, + "grad_norm": 2.189886568147109, + "learning_rate": 8.588068842737856e-06, + "loss": 0.3688, + "step": 1927 + }, + { + "epoch": 0.27, + "grad_norm": 1.9449214051153723, + "learning_rate": 8.586500516558044e-06, + "loss": 0.3523, + "step": 1928 + }, + { + "epoch": 0.27, + "grad_norm": 2.0922636078767147, + "learning_rate": 8.584931463212859e-06, + "loss": 0.3692, + "step": 1929 + }, + { + "epoch": 0.27, + "grad_norm": 2.105351756918438, + "learning_rate": 8.583361683020426e-06, + "loss": 0.3366, + "step": 1930 + }, + { + "epoch": 0.27, + "grad_norm": 1.8655624710131207, + "learning_rate": 8.581791176299022e-06, + "loss": 0.3519, + "step": 1931 + }, + { + "epoch": 0.27, + "grad_norm": 1.7692395814753044, + "learning_rate": 8.580219943367068e-06, + "loss": 0.3405, + "step": 1932 + }, + { + "epoch": 0.27, + "grad_norm": 2.1841387021312224, + "learning_rate": 8.578647984543132e-06, + "loss": 0.3453, + "step": 1933 + }, + { + "epoch": 0.27, + "grad_norm": 2.959690198534538, + "learning_rate": 8.577075300145928e-06, + "loss": 0.3295, + "step": 1934 + }, + { + "epoch": 0.27, + "grad_norm": 1.7797613416065627, + "learning_rate": 8.57550189049432e-06, + "loss": 0.373, + "step": 1935 + }, + { + "epoch": 0.27, + "grad_norm": 2.1908695255870305, + "learning_rate": 8.573927755907318e-06, + "loss": 0.3371, + "step": 1936 + }, + { + "epoch": 0.27, + "grad_norm": 2.8620895969190334, + "learning_rate": 8.57235289670408e-06, + "loss": 0.3623, + "step": 1937 + }, + { + "epoch": 0.27, + "grad_norm": 3.4688967504236907, + "learning_rate": 8.570777313203908e-06, + "loss": 0.3587, + "step": 1938 + }, + { + "epoch": 0.27, + "grad_norm": 2.0251766326455014, + "learning_rate": 8.569201005726255e-06, + "loss": 0.3651, + "step": 1939 + }, + { + "epoch": 0.27, + "grad_norm": 9.875886019658815, + "learning_rate": 8.567623974590714e-06, + "loss": 0.3262, + "step": 1940 + }, + { + "epoch": 0.27, + "grad_norm": 2.118450320605872, + "learning_rate": 8.566046220117032e-06, + "loss": 0.3471, + "step": 1941 + }, + { + "epoch": 0.27, + "grad_norm": 2.4746780852132946, + "learning_rate": 8.5644677426251e-06, + "loss": 0.3448, + "step": 1942 + }, + { + "epoch": 0.27, + "grad_norm": 4.052983702119374, + "learning_rate": 8.562888542434956e-06, + "loss": 0.3929, + "step": 1943 + }, + { + "epoch": 0.27, + "grad_norm": 2.116056636336954, + "learning_rate": 8.561308619866782e-06, + "loss": 0.3747, + "step": 1944 + }, + { + "epoch": 0.27, + "grad_norm": 2.1318641204283506, + "learning_rate": 8.559727975240909e-06, + "loss": 0.3597, + "step": 1945 + }, + { + "epoch": 0.27, + "grad_norm": 2.8291170536971353, + "learning_rate": 8.558146608877814e-06, + "loss": 0.3751, + "step": 1946 + }, + { + "epoch": 0.27, + "grad_norm": 8.15025039566333, + "learning_rate": 8.55656452109812e-06, + "loss": 0.3494, + "step": 1947 + }, + { + "epoch": 0.27, + "grad_norm": 2.3320562478363147, + "learning_rate": 8.554981712222597e-06, + "loss": 0.3302, + "step": 1948 + }, + { + "epoch": 0.27, + "grad_norm": 2.310063966128424, + "learning_rate": 8.553398182572158e-06, + "loss": 0.3455, + "step": 1949 + }, + { + "epoch": 0.27, + "grad_norm": 2.307193012118905, + "learning_rate": 8.551813932467867e-06, + "loss": 0.3451, + "step": 1950 + }, + { + "epoch": 0.27, + "grad_norm": 2.0570998957276068, + "learning_rate": 8.550228962230934e-06, + "loss": 0.3339, + "step": 1951 + }, + { + "epoch": 0.27, + "grad_norm": 2.714521971147656, + "learning_rate": 8.548643272182707e-06, + "loss": 0.3579, + "step": 1952 + }, + { + "epoch": 0.27, + "grad_norm": 2.862525036658205, + "learning_rate": 8.547056862644689e-06, + "loss": 0.3601, + "step": 1953 + }, + { + "epoch": 0.27, + "grad_norm": 2.2486016114554443, + "learning_rate": 8.545469733938528e-06, + "loss": 0.3645, + "step": 1954 + }, + { + "epoch": 0.27, + "grad_norm": 2.742345211493815, + "learning_rate": 8.54388188638601e-06, + "loss": 0.36, + "step": 1955 + }, + { + "epoch": 0.27, + "grad_norm": 2.733461767279722, + "learning_rate": 8.542293320309077e-06, + "loss": 0.3573, + "step": 1956 + }, + { + "epoch": 0.27, + "grad_norm": 2.1331249189178703, + "learning_rate": 8.540704036029807e-06, + "loss": 0.36, + "step": 1957 + }, + { + "epoch": 0.27, + "grad_norm": 2.103724423528566, + "learning_rate": 8.539114033870434e-06, + "loss": 0.3501, + "step": 1958 + }, + { + "epoch": 0.27, + "grad_norm": 2.2832420844127013, + "learning_rate": 8.537523314153328e-06, + "loss": 0.3526, + "step": 1959 + }, + { + "epoch": 0.27, + "grad_norm": 2.3038014880943787, + "learning_rate": 8.535931877201012e-06, + "loss": 0.355, + "step": 1960 + }, + { + "epoch": 0.27, + "grad_norm": 2.0891694458875887, + "learning_rate": 8.534339723336144e-06, + "loss": 0.359, + "step": 1961 + }, + { + "epoch": 0.27, + "grad_norm": 1.9008333784074556, + "learning_rate": 8.532746852881543e-06, + "loss": 0.3532, + "step": 1962 + }, + { + "epoch": 0.27, + "grad_norm": 3.2320127704818034, + "learning_rate": 8.53115326616016e-06, + "loss": 0.3642, + "step": 1963 + }, + { + "epoch": 0.27, + "grad_norm": 2.3019821406259795, + "learning_rate": 8.529558963495094e-06, + "loss": 0.3427, + "step": 1964 + }, + { + "epoch": 0.27, + "grad_norm": 2.0121752203321064, + "learning_rate": 8.527963945209594e-06, + "loss": 0.3557, + "step": 1965 + }, + { + "epoch": 0.27, + "grad_norm": 2.0306856862089138, + "learning_rate": 8.526368211627051e-06, + "loss": 0.3421, + "step": 1966 + }, + { + "epoch": 0.27, + "grad_norm": 2.368657704882052, + "learning_rate": 8.524771763071e-06, + "loss": 0.3552, + "step": 1967 + }, + { + "epoch": 0.27, + "grad_norm": 2.0899993110404056, + "learning_rate": 8.523174599865122e-06, + "loss": 0.3437, + "step": 1968 + }, + { + "epoch": 0.27, + "grad_norm": 2.1259004713193983, + "learning_rate": 8.521576722333242e-06, + "loss": 0.3487, + "step": 1969 + }, + { + "epoch": 0.27, + "grad_norm": 3.446181609861232, + "learning_rate": 8.519978130799334e-06, + "loss": 0.3524, + "step": 1970 + }, + { + "epoch": 0.27, + "grad_norm": 2.2172328123695095, + "learning_rate": 8.51837882558751e-06, + "loss": 0.3393, + "step": 1971 + }, + { + "epoch": 0.27, + "grad_norm": 2.8859593406463753, + "learning_rate": 8.516778807022032e-06, + "loss": 0.3107, + "step": 1972 + }, + { + "epoch": 0.27, + "grad_norm": 2.4147267252370512, + "learning_rate": 8.515178075427306e-06, + "loss": 0.3563, + "step": 1973 + }, + { + "epoch": 0.27, + "grad_norm": 2.1303568994542252, + "learning_rate": 8.513576631127876e-06, + "loss": 0.358, + "step": 1974 + }, + { + "epoch": 0.27, + "grad_norm": 2.2996406963384075, + "learning_rate": 8.511974474448442e-06, + "loss": 0.3579, + "step": 1975 + }, + { + "epoch": 0.27, + "grad_norm": 2.4099237828624696, + "learning_rate": 8.510371605713841e-06, + "loss": 0.358, + "step": 1976 + }, + { + "epoch": 0.27, + "grad_norm": 2.268229607768355, + "learning_rate": 8.508768025249052e-06, + "loss": 0.3088, + "step": 1977 + }, + { + "epoch": 0.27, + "grad_norm": 2.184244988227538, + "learning_rate": 8.507163733379207e-06, + "loss": 0.3622, + "step": 1978 + }, + { + "epoch": 0.28, + "grad_norm": 2.043039252439401, + "learning_rate": 8.505558730429574e-06, + "loss": 0.3656, + "step": 1979 + }, + { + "epoch": 0.28, + "grad_norm": 1.9316061202539203, + "learning_rate": 8.50395301672557e-06, + "loss": 0.3424, + "step": 1980 + }, + { + "epoch": 0.28, + "grad_norm": 2.045448232070744, + "learning_rate": 8.502346592592753e-06, + "loss": 0.3625, + "step": 1981 + }, + { + "epoch": 0.28, + "grad_norm": 1.9331287923834783, + "learning_rate": 8.500739458356827e-06, + "loss": 0.3486, + "step": 1982 + }, + { + "epoch": 0.28, + "grad_norm": 2.080600184108194, + "learning_rate": 8.499131614343643e-06, + "loss": 0.3141, + "step": 1983 + }, + { + "epoch": 0.28, + "grad_norm": 2.456095233245677, + "learning_rate": 8.497523060879188e-06, + "loss": 0.3459, + "step": 1984 + }, + { + "epoch": 0.28, + "grad_norm": 1.7387588135652168, + "learning_rate": 8.495913798289596e-06, + "loss": 0.354, + "step": 1985 + }, + { + "epoch": 0.28, + "grad_norm": 2.676129421082605, + "learning_rate": 8.494303826901154e-06, + "loss": 0.3469, + "step": 1986 + }, + { + "epoch": 0.28, + "grad_norm": 2.7634690527976042, + "learning_rate": 8.492693147040275e-06, + "loss": 0.3211, + "step": 1987 + }, + { + "epoch": 0.28, + "grad_norm": 1.8036692516809365, + "learning_rate": 8.491081759033533e-06, + "loss": 0.346, + "step": 1988 + }, + { + "epoch": 0.28, + "grad_norm": 2.075798402119156, + "learning_rate": 8.489469663207634e-06, + "loss": 0.336, + "step": 1989 + }, + { + "epoch": 0.28, + "grad_norm": 2.1260581951763697, + "learning_rate": 8.487856859889432e-06, + "loss": 0.3744, + "step": 1990 + }, + { + "epoch": 0.28, + "grad_norm": 5.69146499478951, + "learning_rate": 8.486243349405923e-06, + "loss": 0.3437, + "step": 1991 + }, + { + "epoch": 0.28, + "grad_norm": 2.3920566149115596, + "learning_rate": 8.484629132084252e-06, + "loss": 0.3549, + "step": 1992 + }, + { + "epoch": 0.28, + "grad_norm": 2.4979060423923674, + "learning_rate": 8.483014208251698e-06, + "loss": 0.3526, + "step": 1993 + }, + { + "epoch": 0.28, + "grad_norm": 2.2966300133258364, + "learning_rate": 8.481398578235689e-06, + "loss": 0.355, + "step": 1994 + }, + { + "epoch": 0.28, + "grad_norm": 2.1629685870956705, + "learning_rate": 8.479782242363796e-06, + "loss": 0.3455, + "step": 1995 + }, + { + "epoch": 0.28, + "grad_norm": 2.01839054375142, + "learning_rate": 8.47816520096373e-06, + "loss": 0.3501, + "step": 1996 + }, + { + "epoch": 0.28, + "grad_norm": 2.2178587339054214, + "learning_rate": 8.476547454363351e-06, + "loss": 0.3309, + "step": 1997 + }, + { + "epoch": 0.28, + "grad_norm": 2.462992120001083, + "learning_rate": 8.474929002890654e-06, + "loss": 0.3618, + "step": 1998 + }, + { + "epoch": 0.28, + "grad_norm": 2.183349502051506, + "learning_rate": 8.473309846873785e-06, + "loss": 0.3459, + "step": 1999 + }, + { + "epoch": 0.28, + "grad_norm": 1.9006891802220436, + "learning_rate": 8.471689986641028e-06, + "loss": 0.3132, + "step": 2000 + }, + { + "epoch": 0.28, + "grad_norm": 1.8704340788252853, + "learning_rate": 8.470069422520807e-06, + "loss": 0.352, + "step": 2001 + }, + { + "epoch": 0.28, + "grad_norm": 2.050688623318934, + "learning_rate": 8.468448154841695e-06, + "loss": 0.3298, + "step": 2002 + }, + { + "epoch": 0.28, + "grad_norm": 2.3046838077425305, + "learning_rate": 8.466826183932408e-06, + "loss": 0.3578, + "step": 2003 + }, + { + "epoch": 0.28, + "grad_norm": 2.6266352686071177, + "learning_rate": 8.465203510121798e-06, + "loss": 0.3554, + "step": 2004 + }, + { + "epoch": 0.28, + "grad_norm": 2.023197922902269, + "learning_rate": 8.463580133738864e-06, + "loss": 0.3415, + "step": 2005 + }, + { + "epoch": 0.28, + "grad_norm": 2.5553111487522893, + "learning_rate": 8.461956055112747e-06, + "loss": 0.3663, + "step": 2006 + }, + { + "epoch": 0.28, + "grad_norm": 3.560828382663605, + "learning_rate": 8.460331274572729e-06, + "loss": 0.3519, + "step": 2007 + }, + { + "epoch": 0.28, + "grad_norm": 1.831814830848626, + "learning_rate": 8.458705792448236e-06, + "loss": 0.3586, + "step": 2008 + }, + { + "epoch": 0.28, + "grad_norm": 1.940802597896443, + "learning_rate": 8.457079609068837e-06, + "loss": 0.3301, + "step": 2009 + }, + { + "epoch": 0.28, + "grad_norm": 2.118982025208206, + "learning_rate": 8.455452724764242e-06, + "loss": 0.3733, + "step": 2010 + }, + { + "epoch": 0.28, + "grad_norm": 2.1598504215449505, + "learning_rate": 8.4538251398643e-06, + "loss": 0.3299, + "step": 2011 + }, + { + "epoch": 0.28, + "grad_norm": 2.00875458156595, + "learning_rate": 8.452196854699005e-06, + "loss": 0.3322, + "step": 2012 + }, + { + "epoch": 0.28, + "grad_norm": 2.426666493945255, + "learning_rate": 8.450567869598496e-06, + "loss": 0.3584, + "step": 2013 + }, + { + "epoch": 0.28, + "grad_norm": 2.0697421855308695, + "learning_rate": 8.448938184893049e-06, + "loss": 0.3384, + "step": 2014 + }, + { + "epoch": 0.28, + "grad_norm": 2.760969176247034, + "learning_rate": 8.447307800913084e-06, + "loss": 0.3493, + "step": 2015 + }, + { + "epoch": 0.28, + "grad_norm": 6.959596191765999, + "learning_rate": 8.445676717989162e-06, + "loss": 0.3707, + "step": 2016 + }, + { + "epoch": 0.28, + "grad_norm": 1.9212015104061921, + "learning_rate": 8.444044936451986e-06, + "loss": 0.3587, + "step": 2017 + }, + { + "epoch": 0.28, + "grad_norm": 2.771184342503705, + "learning_rate": 8.442412456632401e-06, + "loss": 0.3534, + "step": 2018 + }, + { + "epoch": 0.28, + "grad_norm": 2.279437347275153, + "learning_rate": 8.440779278861395e-06, + "loss": 0.3315, + "step": 2019 + }, + { + "epoch": 0.28, + "grad_norm": 2.228582643564673, + "learning_rate": 8.439145403470092e-06, + "loss": 0.3609, + "step": 2020 + }, + { + "epoch": 0.28, + "grad_norm": 2.3124990341613754, + "learning_rate": 8.437510830789767e-06, + "loss": 0.3803, + "step": 2021 + }, + { + "epoch": 0.28, + "grad_norm": 1.7543084897413876, + "learning_rate": 8.435875561151824e-06, + "loss": 0.3334, + "step": 2022 + }, + { + "epoch": 0.28, + "grad_norm": 3.1585904089299577, + "learning_rate": 8.43423959488782e-06, + "loss": 0.3393, + "step": 2023 + }, + { + "epoch": 0.28, + "grad_norm": 4.016998645420543, + "learning_rate": 8.432602932329448e-06, + "loss": 0.3465, + "step": 2024 + }, + { + "epoch": 0.28, + "grad_norm": 2.276254109257021, + "learning_rate": 8.430965573808538e-06, + "loss": 0.3372, + "step": 2025 + }, + { + "epoch": 0.28, + "grad_norm": 1.99557913802582, + "learning_rate": 8.429327519657071e-06, + "loss": 0.3346, + "step": 2026 + }, + { + "epoch": 0.28, + "grad_norm": 2.126402088340201, + "learning_rate": 8.427688770207162e-06, + "loss": 0.3252, + "step": 2027 + }, + { + "epoch": 0.28, + "grad_norm": 2.1046838420046887, + "learning_rate": 8.426049325791066e-06, + "loss": 0.3332, + "step": 2028 + }, + { + "epoch": 0.28, + "grad_norm": 2.171488956393148, + "learning_rate": 8.424409186741185e-06, + "loss": 0.3368, + "step": 2029 + }, + { + "epoch": 0.28, + "grad_norm": 2.2005135517327483, + "learning_rate": 8.422768353390054e-06, + "loss": 0.3617, + "step": 2030 + }, + { + "epoch": 0.28, + "grad_norm": 4.648275371065415, + "learning_rate": 8.421126826070358e-06, + "loss": 0.375, + "step": 2031 + }, + { + "epoch": 0.28, + "grad_norm": 2.428400129823582, + "learning_rate": 8.419484605114914e-06, + "loss": 0.3094, + "step": 2032 + }, + { + "epoch": 0.28, + "grad_norm": 2.4001293663979375, + "learning_rate": 8.417841690856685e-06, + "loss": 0.341, + "step": 2033 + }, + { + "epoch": 0.28, + "grad_norm": 1.9882718856727122, + "learning_rate": 8.416198083628773e-06, + "loss": 0.3402, + "step": 2034 + }, + { + "epoch": 0.28, + "grad_norm": 5.241873044520965, + "learning_rate": 8.414553783764421e-06, + "loss": 0.3495, + "step": 2035 + }, + { + "epoch": 0.28, + "grad_norm": 2.3356680771619547, + "learning_rate": 8.41290879159701e-06, + "loss": 0.3529, + "step": 2036 + }, + { + "epoch": 0.28, + "grad_norm": 1.8128147794970257, + "learning_rate": 8.411263107460064e-06, + "loss": 0.3691, + "step": 2037 + }, + { + "epoch": 0.28, + "grad_norm": 2.2515039635953746, + "learning_rate": 8.409616731687247e-06, + "loss": 0.3348, + "step": 2038 + }, + { + "epoch": 0.28, + "grad_norm": 2.9265268430167257, + "learning_rate": 8.407969664612364e-06, + "loss": 0.3377, + "step": 2039 + }, + { + "epoch": 0.28, + "grad_norm": 2.106994160771417, + "learning_rate": 8.406321906569355e-06, + "loss": 0.3168, + "step": 2040 + }, + { + "epoch": 0.28, + "grad_norm": 1.9785847927786608, + "learning_rate": 8.40467345789231e-06, + "loss": 0.3363, + "step": 2041 + }, + { + "epoch": 0.28, + "grad_norm": 5.915964670881892, + "learning_rate": 8.403024318915446e-06, + "loss": 0.3645, + "step": 2042 + }, + { + "epoch": 0.28, + "grad_norm": 2.150163836954914, + "learning_rate": 8.401374489973133e-06, + "loss": 0.3651, + "step": 2043 + }, + { + "epoch": 0.28, + "grad_norm": 2.0261331378530882, + "learning_rate": 8.399723971399874e-06, + "loss": 0.37, + "step": 2044 + }, + { + "epoch": 0.28, + "grad_norm": 2.101149776062917, + "learning_rate": 8.39807276353031e-06, + "loss": 0.3504, + "step": 2045 + }, + { + "epoch": 0.28, + "grad_norm": 1.8276108027655515, + "learning_rate": 8.396420866699224e-06, + "loss": 0.3631, + "step": 2046 + }, + { + "epoch": 0.28, + "grad_norm": 2.0042582586697826, + "learning_rate": 8.394768281241543e-06, + "loss": 0.3486, + "step": 2047 + }, + { + "epoch": 0.28, + "grad_norm": 1.882494611830747, + "learning_rate": 8.393115007492327e-06, + "loss": 0.3662, + "step": 2048 + }, + { + "epoch": 0.28, + "grad_norm": 2.581832476379239, + "learning_rate": 8.39146104578678e-06, + "loss": 0.3472, + "step": 2049 + }, + { + "epoch": 0.28, + "grad_norm": 2.1632230331654863, + "learning_rate": 8.389806396460244e-06, + "loss": 0.3338, + "step": 2050 + }, + { + "epoch": 0.29, + "grad_norm": 2.3147843126305934, + "learning_rate": 8.388151059848197e-06, + "loss": 0.3554, + "step": 2051 + }, + { + "epoch": 0.29, + "grad_norm": 1.9217766164672232, + "learning_rate": 8.386495036286264e-06, + "loss": 0.3137, + "step": 2052 + }, + { + "epoch": 0.29, + "grad_norm": 4.280877949630869, + "learning_rate": 8.384838326110202e-06, + "loss": 0.3529, + "step": 2053 + }, + { + "epoch": 0.29, + "grad_norm": 2.5135190122607622, + "learning_rate": 8.383180929655912e-06, + "loss": 0.3497, + "step": 2054 + }, + { + "epoch": 0.29, + "grad_norm": 2.9613968027394364, + "learning_rate": 8.38152284725943e-06, + "loss": 0.3814, + "step": 2055 + }, + { + "epoch": 0.29, + "grad_norm": 2.1300659690281862, + "learning_rate": 8.379864079256935e-06, + "loss": 0.3555, + "step": 2056 + }, + { + "epoch": 0.29, + "grad_norm": 2.040891912730706, + "learning_rate": 8.378204625984743e-06, + "loss": 0.3316, + "step": 2057 + }, + { + "epoch": 0.29, + "grad_norm": 2.221585907228375, + "learning_rate": 8.376544487779307e-06, + "loss": 0.3385, + "step": 2058 + }, + { + "epoch": 0.29, + "grad_norm": 1.8219037666653028, + "learning_rate": 8.374883664977225e-06, + "loss": 0.3618, + "step": 2059 + }, + { + "epoch": 0.29, + "grad_norm": 1.9548762741613157, + "learning_rate": 8.373222157915229e-06, + "loss": 0.3572, + "step": 2060 + }, + { + "epoch": 0.29, + "grad_norm": 1.81138294479201, + "learning_rate": 8.371559966930189e-06, + "loss": 0.3531, + "step": 2061 + }, + { + "epoch": 0.29, + "grad_norm": 2.373712736913635, + "learning_rate": 8.369897092359115e-06, + "loss": 0.3718, + "step": 2062 + }, + { + "epoch": 0.29, + "grad_norm": 2.067226576811413, + "learning_rate": 8.368233534539158e-06, + "loss": 0.3132, + "step": 2063 + }, + { + "epoch": 0.29, + "grad_norm": 2.2075117461399687, + "learning_rate": 8.366569293807605e-06, + "loss": 0.3542, + "step": 2064 + }, + { + "epoch": 0.29, + "grad_norm": 3.835946318772492, + "learning_rate": 8.36490437050188e-06, + "loss": 0.3551, + "step": 2065 + }, + { + "epoch": 0.29, + "grad_norm": 2.0863625624746467, + "learning_rate": 8.36323876495955e-06, + "loss": 0.3777, + "step": 2066 + }, + { + "epoch": 0.29, + "grad_norm": 2.1960080075546444, + "learning_rate": 8.361572477518315e-06, + "loss": 0.3684, + "step": 2067 + }, + { + "epoch": 0.29, + "grad_norm": 2.600840011024807, + "learning_rate": 8.359905508516018e-06, + "loss": 0.3499, + "step": 2068 + }, + { + "epoch": 0.29, + "grad_norm": 2.472187288922723, + "learning_rate": 8.358237858290637e-06, + "loss": 0.3561, + "step": 2069 + }, + { + "epoch": 0.29, + "grad_norm": 2.868781390185572, + "learning_rate": 8.356569527180292e-06, + "loss": 0.3341, + "step": 2070 + }, + { + "epoch": 0.29, + "grad_norm": 1.951407791661315, + "learning_rate": 8.354900515523232e-06, + "loss": 0.3192, + "step": 2071 + }, + { + "epoch": 0.29, + "grad_norm": 2.3081684161953877, + "learning_rate": 8.353230823657857e-06, + "loss": 0.328, + "step": 2072 + }, + { + "epoch": 0.29, + "grad_norm": 2.068907357620654, + "learning_rate": 8.351560451922693e-06, + "loss": 0.3029, + "step": 2073 + }, + { + "epoch": 0.29, + "grad_norm": 2.1797857168138055, + "learning_rate": 8.349889400656412e-06, + "loss": 0.345, + "step": 2074 + }, + { + "epoch": 0.29, + "grad_norm": 2.0366931054328954, + "learning_rate": 8.34821767019782e-06, + "loss": 0.3742, + "step": 2075 + }, + { + "epoch": 0.29, + "grad_norm": 2.048044825716263, + "learning_rate": 8.346545260885862e-06, + "loss": 0.3648, + "step": 2076 + }, + { + "epoch": 0.29, + "grad_norm": 2.347693918713975, + "learning_rate": 8.344872173059618e-06, + "loss": 0.3678, + "step": 2077 + }, + { + "epoch": 0.29, + "grad_norm": 2.5710154374277066, + "learning_rate": 8.343198407058311e-06, + "loss": 0.3496, + "step": 2078 + }, + { + "epoch": 0.29, + "grad_norm": 2.3072230017951822, + "learning_rate": 8.341523963221294e-06, + "loss": 0.3183, + "step": 2079 + }, + { + "epoch": 0.29, + "grad_norm": 1.998045737776981, + "learning_rate": 8.339848841888065e-06, + "loss": 0.3382, + "step": 2080 + }, + { + "epoch": 0.29, + "grad_norm": 2.6140879865109805, + "learning_rate": 8.338173043398255e-06, + "loss": 0.3665, + "step": 2081 + }, + { + "epoch": 0.29, + "grad_norm": 2.5847183323942216, + "learning_rate": 8.336496568091633e-06, + "loss": 0.3524, + "step": 2082 + }, + { + "epoch": 0.29, + "grad_norm": 2.3637655615166606, + "learning_rate": 8.334819416308105e-06, + "loss": 0.3263, + "step": 2083 + }, + { + "epoch": 0.29, + "grad_norm": 1.75806062303984, + "learning_rate": 8.333141588387717e-06, + "loss": 0.3581, + "step": 2084 + }, + { + "epoch": 0.29, + "grad_norm": 1.922161863370118, + "learning_rate": 8.331463084670645e-06, + "loss": 0.3221, + "step": 2085 + }, + { + "epoch": 0.29, + "grad_norm": 3.5809066469757562, + "learning_rate": 8.329783905497212e-06, + "loss": 0.3357, + "step": 2086 + }, + { + "epoch": 0.29, + "grad_norm": 4.487731493924078, + "learning_rate": 8.32810405120787e-06, + "loss": 0.3561, + "step": 2087 + }, + { + "epoch": 0.29, + "grad_norm": 3.7571496921281207, + "learning_rate": 8.326423522143213e-06, + "loss": 0.3461, + "step": 2088 + }, + { + "epoch": 0.29, + "grad_norm": 2.4102912815354447, + "learning_rate": 8.324742318643966e-06, + "loss": 0.3786, + "step": 2089 + }, + { + "epoch": 0.29, + "grad_norm": 2.8748139468259497, + "learning_rate": 8.323060441050995e-06, + "loss": 0.3701, + "step": 2090 + }, + { + "epoch": 0.29, + "grad_norm": 1.9703854656457362, + "learning_rate": 8.321377889705302e-06, + "loss": 0.376, + "step": 2091 + }, + { + "epoch": 0.29, + "grad_norm": 2.126364959214484, + "learning_rate": 8.319694664948026e-06, + "loss": 0.389, + "step": 2092 + }, + { + "epoch": 0.29, + "grad_norm": 3.0891910825131066, + "learning_rate": 8.318010767120442e-06, + "loss": 0.3244, + "step": 2093 + }, + { + "epoch": 0.29, + "grad_norm": 2.381103796587183, + "learning_rate": 8.31632619656396e-06, + "loss": 0.345, + "step": 2094 + }, + { + "epoch": 0.29, + "grad_norm": 1.9999829610687239, + "learning_rate": 8.31464095362013e-06, + "loss": 0.3575, + "step": 2095 + }, + { + "epoch": 0.29, + "grad_norm": 6.257391843677199, + "learning_rate": 8.312955038630635e-06, + "loss": 0.3514, + "step": 2096 + }, + { + "epoch": 0.29, + "grad_norm": 2.229809895530916, + "learning_rate": 8.311268451937296e-06, + "loss": 0.3531, + "step": 2097 + }, + { + "epoch": 0.29, + "grad_norm": 2.1458467095235227, + "learning_rate": 8.309581193882068e-06, + "loss": 0.3652, + "step": 2098 + }, + { + "epoch": 0.29, + "grad_norm": 1.6949507323388746, + "learning_rate": 8.307893264807044e-06, + "loss": 0.3591, + "step": 2099 + }, + { + "epoch": 0.29, + "grad_norm": 1.915511640817953, + "learning_rate": 8.306204665054455e-06, + "loss": 0.3312, + "step": 2100 + }, + { + "epoch": 0.29, + "grad_norm": 7.113439281478692, + "learning_rate": 8.304515394966663e-06, + "loss": 0.373, + "step": 2101 + }, + { + "epoch": 0.29, + "grad_norm": 2.6758386902598996, + "learning_rate": 8.302825454886172e-06, + "loss": 0.341, + "step": 2102 + }, + { + "epoch": 0.29, + "grad_norm": 3.6614432628979383, + "learning_rate": 8.301134845155614e-06, + "loss": 0.3498, + "step": 2103 + }, + { + "epoch": 0.29, + "grad_norm": 2.5449871400731885, + "learning_rate": 8.299443566117763e-06, + "loss": 0.3405, + "step": 2104 + }, + { + "epoch": 0.29, + "grad_norm": 2.7989526031957794, + "learning_rate": 8.29775161811553e-06, + "loss": 0.3629, + "step": 2105 + }, + { + "epoch": 0.29, + "grad_norm": 2.4608184664314865, + "learning_rate": 8.296059001491955e-06, + "loss": 0.3309, + "step": 2106 + }, + { + "epoch": 0.29, + "grad_norm": 4.384869045890157, + "learning_rate": 8.294365716590218e-06, + "loss": 0.3595, + "step": 2107 + }, + { + "epoch": 0.29, + "grad_norm": 1.9623190591051387, + "learning_rate": 8.292671763753635e-06, + "loss": 0.3486, + "step": 2108 + }, + { + "epoch": 0.29, + "grad_norm": 2.3222678745264638, + "learning_rate": 8.290977143325653e-06, + "loss": 0.3458, + "step": 2109 + }, + { + "epoch": 0.29, + "grad_norm": 2.0207655026202236, + "learning_rate": 8.28928185564986e-06, + "loss": 0.351, + "step": 2110 + }, + { + "epoch": 0.29, + "grad_norm": 2.277750210162525, + "learning_rate": 8.287585901069977e-06, + "loss": 0.3467, + "step": 2111 + }, + { + "epoch": 0.29, + "grad_norm": 4.059552399499174, + "learning_rate": 8.285889279929857e-06, + "loss": 0.3469, + "step": 2112 + }, + { + "epoch": 0.29, + "grad_norm": 2.404266557181124, + "learning_rate": 8.284191992573496e-06, + "loss": 0.3565, + "step": 2113 + }, + { + "epoch": 0.29, + "grad_norm": 2.96365701425356, + "learning_rate": 8.282494039345015e-06, + "loss": 0.3552, + "step": 2114 + }, + { + "epoch": 0.29, + "grad_norm": 1.8121291526463947, + "learning_rate": 8.280795420588679e-06, + "loss": 0.3347, + "step": 2115 + }, + { + "epoch": 0.29, + "grad_norm": 2.0568210654970214, + "learning_rate": 8.27909613664888e-06, + "loss": 0.361, + "step": 2116 + }, + { + "epoch": 0.29, + "grad_norm": 2.953712663791127, + "learning_rate": 8.277396187870153e-06, + "loss": 0.3658, + "step": 2117 + }, + { + "epoch": 0.29, + "grad_norm": 2.351016885462782, + "learning_rate": 8.275695574597162e-06, + "loss": 0.3472, + "step": 2118 + }, + { + "epoch": 0.29, + "grad_norm": 1.958941619743219, + "learning_rate": 8.273994297174707e-06, + "loss": 0.3631, + "step": 2119 + }, + { + "epoch": 0.29, + "grad_norm": 1.9174107154479392, + "learning_rate": 8.272292355947724e-06, + "loss": 0.3613, + "step": 2120 + }, + { + "epoch": 0.29, + "grad_norm": 2.158060878714958, + "learning_rate": 8.270589751261285e-06, + "loss": 0.3301, + "step": 2121 + }, + { + "epoch": 0.29, + "grad_norm": 2.292699604796072, + "learning_rate": 8.268886483460587e-06, + "loss": 0.374, + "step": 2122 + }, + { + "epoch": 0.3, + "grad_norm": 2.442657329660222, + "learning_rate": 8.267182552890978e-06, + "loss": 0.337, + "step": 2123 + }, + { + "epoch": 0.3, + "grad_norm": 2.058263787642913, + "learning_rate": 8.265477959897924e-06, + "loss": 0.343, + "step": 2124 + }, + { + "epoch": 0.3, + "grad_norm": 2.182277188780185, + "learning_rate": 8.263772704827036e-06, + "loss": 0.3547, + "step": 2125 + }, + { + "epoch": 0.3, + "grad_norm": 2.524550955901232, + "learning_rate": 8.262066788024052e-06, + "loss": 0.3469, + "step": 2126 + }, + { + "epoch": 0.3, + "grad_norm": 2.3144520501109143, + "learning_rate": 8.260360209834852e-06, + "loss": 0.3432, + "step": 2127 + }, + { + "epoch": 0.3, + "grad_norm": 1.9653661422105464, + "learning_rate": 8.258652970605444e-06, + "loss": 0.3437, + "step": 2128 + }, + { + "epoch": 0.3, + "grad_norm": 2.3807696187793765, + "learning_rate": 8.256945070681972e-06, + "loss": 0.3462, + "step": 2129 + }, + { + "epoch": 0.3, + "grad_norm": 2.2544293640585926, + "learning_rate": 8.255236510410714e-06, + "loss": 0.342, + "step": 2130 + }, + { + "epoch": 0.3, + "grad_norm": 2.0911971689006714, + "learning_rate": 8.253527290138082e-06, + "loss": 0.3287, + "step": 2131 + }, + { + "epoch": 0.3, + "grad_norm": 2.504619324239555, + "learning_rate": 8.25181741021062e-06, + "loss": 0.3627, + "step": 2132 + }, + { + "epoch": 0.3, + "grad_norm": 2.7053167182711917, + "learning_rate": 8.250106870975008e-06, + "loss": 0.3511, + "step": 2133 + }, + { + "epoch": 0.3, + "grad_norm": 3.6804800339517794, + "learning_rate": 8.24839567277806e-06, + "loss": 0.3624, + "step": 2134 + }, + { + "epoch": 0.3, + "grad_norm": 2.0076142448125442, + "learning_rate": 8.24668381596672e-06, + "loss": 0.3465, + "step": 2135 + }, + { + "epoch": 0.3, + "grad_norm": 2.1834546352502997, + "learning_rate": 8.244971300888074e-06, + "loss": 0.3801, + "step": 2136 + }, + { + "epoch": 0.3, + "grad_norm": 2.233691359436469, + "learning_rate": 8.243258127889328e-06, + "loss": 0.3518, + "step": 2137 + }, + { + "epoch": 0.3, + "grad_norm": 2.68857086095495, + "learning_rate": 8.241544297317835e-06, + "loss": 0.3542, + "step": 2138 + }, + { + "epoch": 0.3, + "grad_norm": 1.9485148178094975, + "learning_rate": 8.239829809521071e-06, + "loss": 0.327, + "step": 2139 + }, + { + "epoch": 0.3, + "grad_norm": 2.072439895376774, + "learning_rate": 8.238114664846653e-06, + "loss": 0.3379, + "step": 2140 + }, + { + "epoch": 0.3, + "grad_norm": 2.8031131753246763, + "learning_rate": 8.236398863642326e-06, + "loss": 0.3526, + "step": 2141 + }, + { + "epoch": 0.3, + "grad_norm": 2.0308081345824878, + "learning_rate": 8.23468240625597e-06, + "loss": 0.3607, + "step": 2142 + }, + { + "epoch": 0.3, + "grad_norm": 2.545416066740476, + "learning_rate": 8.232965293035598e-06, + "loss": 0.3388, + "step": 2143 + }, + { + "epoch": 0.3, + "grad_norm": 2.5670364539992354, + "learning_rate": 8.231247524329354e-06, + "loss": 0.3617, + "step": 2144 + }, + { + "epoch": 0.3, + "grad_norm": 2.0525114359615313, + "learning_rate": 8.22952910048552e-06, + "loss": 0.3522, + "step": 2145 + }, + { + "epoch": 0.3, + "grad_norm": 2.0399536858841474, + "learning_rate": 8.227810021852505e-06, + "loss": 0.3439, + "step": 2146 + }, + { + "epoch": 0.3, + "grad_norm": 2.4331968069329473, + "learning_rate": 8.226090288778855e-06, + "loss": 0.3462, + "step": 2147 + }, + { + "epoch": 0.3, + "grad_norm": 2.335680881887103, + "learning_rate": 8.224369901613247e-06, + "loss": 0.3349, + "step": 2148 + }, + { + "epoch": 0.3, + "grad_norm": 2.639860192093094, + "learning_rate": 8.22264886070449e-06, + "loss": 0.3442, + "step": 2149 + }, + { + "epoch": 0.3, + "grad_norm": 2.4357971802377447, + "learning_rate": 8.220927166401525e-06, + "loss": 0.3401, + "step": 2150 + }, + { + "epoch": 0.3, + "grad_norm": 2.1182650831273144, + "learning_rate": 8.21920481905343e-06, + "loss": 0.3432, + "step": 2151 + }, + { + "epoch": 0.3, + "grad_norm": 2.261310700531172, + "learning_rate": 8.217481819009408e-06, + "loss": 0.3784, + "step": 2152 + }, + { + "epoch": 0.3, + "grad_norm": 2.4878213881320526, + "learning_rate": 8.215758166618801e-06, + "loss": 0.4014, + "step": 2153 + }, + { + "epoch": 0.3, + "grad_norm": 3.064612237912592, + "learning_rate": 8.214033862231081e-06, + "loss": 0.3257, + "step": 2154 + }, + { + "epoch": 0.3, + "grad_norm": 2.260181040051796, + "learning_rate": 8.21230890619585e-06, + "loss": 0.3507, + "step": 2155 + }, + { + "epoch": 0.3, + "grad_norm": 2.737086836009125, + "learning_rate": 8.210583298862844e-06, + "loss": 0.3512, + "step": 2156 + }, + { + "epoch": 0.3, + "grad_norm": 2.580648478721182, + "learning_rate": 8.208857040581934e-06, + "loss": 0.3371, + "step": 2157 + }, + { + "epoch": 0.3, + "grad_norm": 2.135286494139655, + "learning_rate": 8.207130131703117e-06, + "loss": 0.36, + "step": 2158 + }, + { + "epoch": 0.3, + "grad_norm": 2.0928838904617186, + "learning_rate": 8.205402572576527e-06, + "loss": 0.3332, + "step": 2159 + }, + { + "epoch": 0.3, + "grad_norm": 2.030874919708704, + "learning_rate": 8.203674363552424e-06, + "loss": 0.3541, + "step": 2160 + }, + { + "epoch": 0.3, + "grad_norm": 2.099081377719468, + "learning_rate": 8.201945504981207e-06, + "loss": 0.3367, + "step": 2161 + }, + { + "epoch": 0.3, + "grad_norm": 2.737630788319806, + "learning_rate": 8.200215997213406e-06, + "loss": 0.3435, + "step": 2162 + }, + { + "epoch": 0.3, + "grad_norm": 2.4007390964104145, + "learning_rate": 8.198485840599673e-06, + "loss": 0.343, + "step": 2163 + }, + { + "epoch": 0.3, + "grad_norm": 2.4331582977595945, + "learning_rate": 8.196755035490802e-06, + "loss": 0.3598, + "step": 2164 + }, + { + "epoch": 0.3, + "grad_norm": 2.4774855256415274, + "learning_rate": 8.195023582237715e-06, + "loss": 0.3953, + "step": 2165 + }, + { + "epoch": 0.3, + "grad_norm": 3.9700150625772945, + "learning_rate": 8.193291481191464e-06, + "loss": 0.3427, + "step": 2166 + }, + { + "epoch": 0.3, + "grad_norm": 2.0406317050581837, + "learning_rate": 8.191558732703235e-06, + "loss": 0.3343, + "step": 2167 + }, + { + "epoch": 0.3, + "grad_norm": 2.263653109912056, + "learning_rate": 8.189825337124343e-06, + "loss": 0.3619, + "step": 2168 + }, + { + "epoch": 0.3, + "grad_norm": 2.2648185969580856, + "learning_rate": 8.188091294806237e-06, + "loss": 0.3492, + "step": 2169 + }, + { + "epoch": 0.3, + "grad_norm": 1.9335177796452938, + "learning_rate": 8.186356606100493e-06, + "loss": 0.3287, + "step": 2170 + }, + { + "epoch": 0.3, + "grad_norm": 1.7118232785364882, + "learning_rate": 8.184621271358823e-06, + "loss": 0.3464, + "step": 2171 + }, + { + "epoch": 0.3, + "grad_norm": 1.8227518689306037, + "learning_rate": 8.182885290933063e-06, + "loss": 0.3549, + "step": 2172 + }, + { + "epoch": 0.3, + "grad_norm": 2.6053637842534116, + "learning_rate": 8.181148665175187e-06, + "loss": 0.3755, + "step": 2173 + }, + { + "epoch": 0.3, + "grad_norm": 2.634832292775382, + "learning_rate": 8.179411394437297e-06, + "loss": 0.3403, + "step": 2174 + }, + { + "epoch": 0.3, + "grad_norm": 2.111944826685948, + "learning_rate": 8.177673479071626e-06, + "loss": 0.3638, + "step": 2175 + }, + { + "epoch": 0.3, + "grad_norm": 2.414587900936821, + "learning_rate": 8.175934919430534e-06, + "loss": 0.348, + "step": 2176 + }, + { + "epoch": 0.3, + "grad_norm": 2.503066823533205, + "learning_rate": 8.174195715866523e-06, + "loss": 0.3576, + "step": 2177 + }, + { + "epoch": 0.3, + "grad_norm": 3.2398535009781244, + "learning_rate": 8.172455868732207e-06, + "loss": 0.363, + "step": 2178 + }, + { + "epoch": 0.3, + "grad_norm": 1.8424136873780415, + "learning_rate": 8.17071537838035e-06, + "loss": 0.3631, + "step": 2179 + }, + { + "epoch": 0.3, + "grad_norm": 1.9382578162741146, + "learning_rate": 8.168974245163835e-06, + "loss": 0.3613, + "step": 2180 + }, + { + "epoch": 0.3, + "grad_norm": 2.6523386406131335, + "learning_rate": 8.167232469435676e-06, + "loss": 0.3579, + "step": 2181 + }, + { + "epoch": 0.3, + "grad_norm": 2.18833551957683, + "learning_rate": 8.165490051549022e-06, + "loss": 0.3483, + "step": 2182 + }, + { + "epoch": 0.3, + "grad_norm": 2.299754423416497, + "learning_rate": 8.163746991857147e-06, + "loss": 0.3411, + "step": 2183 + }, + { + "epoch": 0.3, + "grad_norm": 3.719518936110764, + "learning_rate": 8.162003290713458e-06, + "loss": 0.3362, + "step": 2184 + }, + { + "epoch": 0.3, + "grad_norm": 2.211307205961508, + "learning_rate": 8.160258948471491e-06, + "loss": 0.3663, + "step": 2185 + }, + { + "epoch": 0.3, + "grad_norm": 2.0675535415251343, + "learning_rate": 8.158513965484915e-06, + "loss": 0.3631, + "step": 2186 + }, + { + "epoch": 0.3, + "grad_norm": 2.0452303517522608, + "learning_rate": 8.156768342107526e-06, + "loss": 0.3424, + "step": 2187 + }, + { + "epoch": 0.3, + "grad_norm": 2.8014430443642806, + "learning_rate": 8.15502207869325e-06, + "loss": 0.3454, + "step": 2188 + }, + { + "epoch": 0.3, + "grad_norm": 2.5044929432722567, + "learning_rate": 8.15327517559614e-06, + "loss": 0.3362, + "step": 2189 + }, + { + "epoch": 0.3, + "grad_norm": 2.294484554183171, + "learning_rate": 8.151527633170384e-06, + "loss": 0.3234, + "step": 2190 + }, + { + "epoch": 0.3, + "grad_norm": 2.331631336361069, + "learning_rate": 8.149779451770298e-06, + "loss": 0.3523, + "step": 2191 + }, + { + "epoch": 0.3, + "grad_norm": 2.6176443566245147, + "learning_rate": 8.14803063175033e-06, + "loss": 0.3444, + "step": 2192 + }, + { + "epoch": 0.3, + "grad_norm": 2.3000290158531898, + "learning_rate": 8.146281173465048e-06, + "loss": 0.3562, + "step": 2193 + }, + { + "epoch": 0.3, + "grad_norm": 2.427038315014441, + "learning_rate": 8.144531077269157e-06, + "loss": 0.3236, + "step": 2194 + }, + { + "epoch": 0.31, + "grad_norm": 2.273120189521116, + "learning_rate": 8.142780343517496e-06, + "loss": 0.3442, + "step": 2195 + }, + { + "epoch": 0.31, + "grad_norm": 2.006627548531321, + "learning_rate": 8.141028972565021e-06, + "loss": 0.3622, + "step": 2196 + }, + { + "epoch": 0.31, + "grad_norm": 2.4237509845686334, + "learning_rate": 8.139276964766827e-06, + "loss": 0.349, + "step": 2197 + }, + { + "epoch": 0.31, + "grad_norm": 2.345492071237475, + "learning_rate": 8.137524320478131e-06, + "loss": 0.3383, + "step": 2198 + }, + { + "epoch": 0.31, + "grad_norm": 2.526751180607056, + "learning_rate": 8.135771040054289e-06, + "loss": 0.373, + "step": 2199 + }, + { + "epoch": 0.31, + "grad_norm": 2.8387869519506435, + "learning_rate": 8.134017123850774e-06, + "loss": 0.3562, + "step": 2200 + }, + { + "epoch": 0.31, + "grad_norm": 2.2615109886700044, + "learning_rate": 8.1322625722232e-06, + "loss": 0.3345, + "step": 2201 + }, + { + "epoch": 0.31, + "grad_norm": 2.4931009216763567, + "learning_rate": 8.130507385527295e-06, + "loss": 0.3674, + "step": 2202 + }, + { + "epoch": 0.31, + "grad_norm": 2.4838323284683006, + "learning_rate": 8.12875156411893e-06, + "loss": 0.3576, + "step": 2203 + }, + { + "epoch": 0.31, + "grad_norm": 2.706591039793965, + "learning_rate": 8.1269951083541e-06, + "loss": 0.3409, + "step": 2204 + }, + { + "epoch": 0.31, + "grad_norm": 2.6906331734726616, + "learning_rate": 8.125238018588923e-06, + "loss": 0.3347, + "step": 2205 + }, + { + "epoch": 0.31, + "grad_norm": 2.4943641129922125, + "learning_rate": 8.123480295179654e-06, + "loss": 0.3317, + "step": 2206 + }, + { + "epoch": 0.31, + "grad_norm": 1.9936468945735297, + "learning_rate": 8.12172193848267e-06, + "loss": 0.3574, + "step": 2207 + }, + { + "epoch": 0.31, + "grad_norm": 2.392284567375314, + "learning_rate": 8.119962948854481e-06, + "loss": 0.3493, + "step": 2208 + }, + { + "epoch": 0.31, + "grad_norm": 2.2750263162842734, + "learning_rate": 8.118203326651722e-06, + "loss": 0.372, + "step": 2209 + }, + { + "epoch": 0.31, + "grad_norm": 2.2573485130105255, + "learning_rate": 8.116443072231156e-06, + "loss": 0.3445, + "step": 2210 + }, + { + "epoch": 0.31, + "grad_norm": 2.8732979431421137, + "learning_rate": 8.114682185949681e-06, + "loss": 0.3608, + "step": 2211 + }, + { + "epoch": 0.31, + "grad_norm": 2.415795052023318, + "learning_rate": 8.112920668164313e-06, + "loss": 0.3253, + "step": 2212 + }, + { + "epoch": 0.31, + "grad_norm": 5.165334285566745, + "learning_rate": 8.111158519232203e-06, + "loss": 0.337, + "step": 2213 + }, + { + "epoch": 0.31, + "grad_norm": 1.9622831349602827, + "learning_rate": 8.109395739510626e-06, + "loss": 0.3278, + "step": 2214 + }, + { + "epoch": 0.31, + "grad_norm": 2.9807142184327255, + "learning_rate": 8.107632329356989e-06, + "loss": 0.3343, + "step": 2215 + }, + { + "epoch": 0.31, + "grad_norm": 1.900140286784658, + "learning_rate": 8.105868289128824e-06, + "loss": 0.3452, + "step": 2216 + }, + { + "epoch": 0.31, + "grad_norm": 2.364385281915764, + "learning_rate": 8.10410361918379e-06, + "loss": 0.3355, + "step": 2217 + }, + { + "epoch": 0.31, + "grad_norm": 2.0100269359778933, + "learning_rate": 8.102338319879678e-06, + "loss": 0.3354, + "step": 2218 + }, + { + "epoch": 0.31, + "grad_norm": 3.0704063453445047, + "learning_rate": 8.100572391574399e-06, + "loss": 0.3557, + "step": 2219 + }, + { + "epoch": 0.31, + "grad_norm": 2.7131445309652764, + "learning_rate": 8.098805834626003e-06, + "loss": 0.3436, + "step": 2220 + }, + { + "epoch": 0.31, + "grad_norm": 2.197744102427651, + "learning_rate": 8.097038649392654e-06, + "loss": 0.3527, + "step": 2221 + }, + { + "epoch": 0.31, + "grad_norm": 2.6620543761458273, + "learning_rate": 8.095270836232652e-06, + "loss": 0.3434, + "step": 2222 + }, + { + "epoch": 0.31, + "grad_norm": 2.1548975531520425, + "learning_rate": 8.093502395504422e-06, + "loss": 0.3419, + "step": 2223 + }, + { + "epoch": 0.31, + "grad_norm": 2.3297331007348934, + "learning_rate": 8.091733327566517e-06, + "loss": 0.3382, + "step": 2224 + }, + { + "epoch": 0.31, + "grad_norm": 1.8468031463802559, + "learning_rate": 8.08996363277762e-06, + "loss": 0.3294, + "step": 2225 + }, + { + "epoch": 0.31, + "grad_norm": 2.7401927669008312, + "learning_rate": 8.088193311496533e-06, + "loss": 0.3243, + "step": 2226 + }, + { + "epoch": 0.31, + "grad_norm": 2.45186712136449, + "learning_rate": 8.086422364082192e-06, + "loss": 0.3511, + "step": 2227 + }, + { + "epoch": 0.31, + "grad_norm": 1.882910923819779, + "learning_rate": 8.084650790893657e-06, + "loss": 0.3789, + "step": 2228 + }, + { + "epoch": 0.31, + "grad_norm": 2.544989254920362, + "learning_rate": 8.082878592290116e-06, + "loss": 0.3454, + "step": 2229 + }, + { + "epoch": 0.31, + "grad_norm": 2.139772843001241, + "learning_rate": 8.081105768630883e-06, + "loss": 0.3368, + "step": 2230 + }, + { + "epoch": 0.31, + "grad_norm": 2.516727310941703, + "learning_rate": 8.079332320275399e-06, + "loss": 0.3343, + "step": 2231 + }, + { + "epoch": 0.31, + "grad_norm": 2.7192890445995688, + "learning_rate": 8.077558247583233e-06, + "loss": 0.367, + "step": 2232 + }, + { + "epoch": 0.31, + "grad_norm": 1.8778310312008508, + "learning_rate": 8.07578355091408e-06, + "loss": 0.3338, + "step": 2233 + }, + { + "epoch": 0.31, + "grad_norm": 4.146686592249456, + "learning_rate": 8.07400823062776e-06, + "loss": 0.3631, + "step": 2234 + }, + { + "epoch": 0.31, + "grad_norm": 2.2299384856755453, + "learning_rate": 8.072232287084219e-06, + "loss": 0.3307, + "step": 2235 + }, + { + "epoch": 0.31, + "grad_norm": 3.0353012277314044, + "learning_rate": 8.070455720643533e-06, + "loss": 0.3611, + "step": 2236 + }, + { + "epoch": 0.31, + "grad_norm": 2.9757277731698193, + "learning_rate": 8.068678531665899e-06, + "loss": 0.3453, + "step": 2237 + }, + { + "epoch": 0.31, + "grad_norm": 2.9135390950971822, + "learning_rate": 8.066900720511647e-06, + "loss": 0.3456, + "step": 2238 + }, + { + "epoch": 0.31, + "grad_norm": 2.1506392292655856, + "learning_rate": 8.065122287541228e-06, + "loss": 0.3735, + "step": 2239 + }, + { + "epoch": 0.31, + "grad_norm": 2.944123433127856, + "learning_rate": 8.063343233115218e-06, + "loss": 0.3512, + "step": 2240 + }, + { + "epoch": 0.31, + "grad_norm": 1.9656833237591618, + "learning_rate": 8.061563557594325e-06, + "loss": 0.3264, + "step": 2241 + }, + { + "epoch": 0.31, + "grad_norm": 2.3915420370018046, + "learning_rate": 8.059783261339377e-06, + "loss": 0.334, + "step": 2242 + }, + { + "epoch": 0.31, + "grad_norm": 1.9902163213854827, + "learning_rate": 8.05800234471133e-06, + "loss": 0.3481, + "step": 2243 + }, + { + "epoch": 0.31, + "grad_norm": 2.2317293545610037, + "learning_rate": 8.05622080807127e-06, + "loss": 0.336, + "step": 2244 + }, + { + "epoch": 0.31, + "grad_norm": 2.305489194185094, + "learning_rate": 8.0544386517804e-06, + "loss": 0.3312, + "step": 2245 + }, + { + "epoch": 0.31, + "grad_norm": 2.238915171715029, + "learning_rate": 8.052655876200058e-06, + "loss": 0.3349, + "step": 2246 + }, + { + "epoch": 0.31, + "grad_norm": 3.3042331380192738, + "learning_rate": 8.050872481691699e-06, + "loss": 0.3473, + "step": 2247 + }, + { + "epoch": 0.31, + "grad_norm": 2.5463843886925743, + "learning_rate": 8.049088468616908e-06, + "loss": 0.3373, + "step": 2248 + }, + { + "epoch": 0.31, + "grad_norm": 2.469439987231937, + "learning_rate": 8.047303837337396e-06, + "loss": 0.3512, + "step": 2249 + }, + { + "epoch": 0.31, + "grad_norm": 1.930306651350828, + "learning_rate": 8.045518588214997e-06, + "loss": 0.363, + "step": 2250 + }, + { + "epoch": 0.31, + "grad_norm": 2.000690769099861, + "learning_rate": 8.043732721611672e-06, + "loss": 0.343, + "step": 2251 + }, + { + "epoch": 0.31, + "grad_norm": 1.903066774137539, + "learning_rate": 8.041946237889509e-06, + "loss": 0.3258, + "step": 2252 + }, + { + "epoch": 0.31, + "grad_norm": 2.2535985211632705, + "learning_rate": 8.040159137410715e-06, + "loss": 0.3439, + "step": 2253 + }, + { + "epoch": 0.31, + "grad_norm": 1.8997462447525182, + "learning_rate": 8.038371420537626e-06, + "loss": 0.3427, + "step": 2254 + }, + { + "epoch": 0.31, + "grad_norm": 2.6083422167247474, + "learning_rate": 8.036583087632706e-06, + "loss": 0.3669, + "step": 2255 + }, + { + "epoch": 0.31, + "grad_norm": 2.256190577466268, + "learning_rate": 8.034794139058539e-06, + "loss": 0.3665, + "step": 2256 + }, + { + "epoch": 0.31, + "grad_norm": 3.464393454087205, + "learning_rate": 8.033004575177836e-06, + "loss": 0.3313, + "step": 2257 + }, + { + "epoch": 0.31, + "grad_norm": 2.304714419378602, + "learning_rate": 8.03121439635343e-06, + "loss": 0.3309, + "step": 2258 + }, + { + "epoch": 0.31, + "grad_norm": 2.1741898133242232, + "learning_rate": 8.029423602948283e-06, + "loss": 0.3533, + "step": 2259 + }, + { + "epoch": 0.31, + "grad_norm": 2.4956218696870973, + "learning_rate": 8.027632195325481e-06, + "loss": 0.3536, + "step": 2260 + }, + { + "epoch": 0.31, + "grad_norm": 2.726367216827653, + "learning_rate": 8.02584017384823e-06, + "loss": 0.3576, + "step": 2261 + }, + { + "epoch": 0.31, + "grad_norm": 2.579530809006685, + "learning_rate": 8.024047538879864e-06, + "loss": 0.3289, + "step": 2262 + }, + { + "epoch": 0.31, + "grad_norm": 7.8578830124025485, + "learning_rate": 8.022254290783841e-06, + "loss": 0.3468, + "step": 2263 + }, + { + "epoch": 0.31, + "grad_norm": 2.619632406310583, + "learning_rate": 8.020460429923747e-06, + "loss": 0.3478, + "step": 2264 + }, + { + "epoch": 0.31, + "grad_norm": 1.9934498485763608, + "learning_rate": 8.018665956663284e-06, + "loss": 0.3289, + "step": 2265 + }, + { + "epoch": 0.32, + "grad_norm": 2.104083107717779, + "learning_rate": 8.016870871366285e-06, + "loss": 0.362, + "step": 2266 + }, + { + "epoch": 0.32, + "grad_norm": 2.7106177607253943, + "learning_rate": 8.015075174396703e-06, + "loss": 0.3475, + "step": 2267 + }, + { + "epoch": 0.32, + "grad_norm": 2.0614769461804583, + "learning_rate": 8.013278866118617e-06, + "loss": 0.3337, + "step": 2268 + }, + { + "epoch": 0.32, + "grad_norm": 2.0762505989803293, + "learning_rate": 8.011481946896232e-06, + "loss": 0.3452, + "step": 2269 + }, + { + "epoch": 0.32, + "grad_norm": 2.1524780747553858, + "learning_rate": 8.009684417093872e-06, + "loss": 0.3342, + "step": 2270 + }, + { + "epoch": 0.32, + "grad_norm": 2.4447601260962477, + "learning_rate": 8.007886277075987e-06, + "loss": 0.3274, + "step": 2271 + }, + { + "epoch": 0.32, + "grad_norm": 2.339854738422332, + "learning_rate": 8.006087527207154e-06, + "loss": 0.3611, + "step": 2272 + }, + { + "epoch": 0.32, + "grad_norm": 3.1014708203385597, + "learning_rate": 8.004288167852068e-06, + "loss": 0.3623, + "step": 2273 + }, + { + "epoch": 0.32, + "grad_norm": 2.067816527324774, + "learning_rate": 8.002488199375552e-06, + "loss": 0.3446, + "step": 2274 + }, + { + "epoch": 0.32, + "grad_norm": 9.5418499600885, + "learning_rate": 8.000687622142553e-06, + "loss": 0.336, + "step": 2275 + }, + { + "epoch": 0.32, + "grad_norm": 2.3966419712883695, + "learning_rate": 7.998886436518134e-06, + "loss": 0.3312, + "step": 2276 + }, + { + "epoch": 0.32, + "grad_norm": 3.2828278420971895, + "learning_rate": 7.997084642867488e-06, + "loss": 0.3195, + "step": 2277 + }, + { + "epoch": 0.32, + "grad_norm": 2.2329623718053777, + "learning_rate": 7.995282241555931e-06, + "loss": 0.3484, + "step": 2278 + }, + { + "epoch": 0.32, + "grad_norm": 2.026859863018875, + "learning_rate": 7.993479232948904e-06, + "loss": 0.3388, + "step": 2279 + }, + { + "epoch": 0.32, + "grad_norm": 2.51358859872986, + "learning_rate": 7.991675617411961e-06, + "loss": 0.3389, + "step": 2280 + }, + { + "epoch": 0.32, + "grad_norm": 2.5194193033015457, + "learning_rate": 7.989871395310793e-06, + "loss": 0.3448, + "step": 2281 + }, + { + "epoch": 0.32, + "grad_norm": 2.164650901196226, + "learning_rate": 7.988066567011206e-06, + "loss": 0.3696, + "step": 2282 + }, + { + "epoch": 0.32, + "grad_norm": 2.4967937447590627, + "learning_rate": 7.986261132879125e-06, + "loss": 0.3573, + "step": 2283 + }, + { + "epoch": 0.32, + "grad_norm": 2.386657239390454, + "learning_rate": 7.98445509328061e-06, + "loss": 0.331, + "step": 2284 + }, + { + "epoch": 0.32, + "grad_norm": 2.9669688757013484, + "learning_rate": 7.982648448581832e-06, + "loss": 0.3491, + "step": 2285 + }, + { + "epoch": 0.32, + "grad_norm": 2.601284632365897, + "learning_rate": 7.980841199149091e-06, + "loss": 0.3366, + "step": 2286 + }, + { + "epoch": 0.32, + "grad_norm": 2.102410332848691, + "learning_rate": 7.979033345348808e-06, + "loss": 0.3511, + "step": 2287 + }, + { + "epoch": 0.32, + "grad_norm": 3.3848610912751917, + "learning_rate": 7.977224887547527e-06, + "loss": 0.3644, + "step": 2288 + }, + { + "epoch": 0.32, + "grad_norm": 3.213190163165974, + "learning_rate": 7.975415826111914e-06, + "loss": 0.3269, + "step": 2289 + }, + { + "epoch": 0.32, + "grad_norm": 2.290521998675272, + "learning_rate": 7.973606161408756e-06, + "loss": 0.3655, + "step": 2290 + }, + { + "epoch": 0.32, + "grad_norm": 3.401647368517366, + "learning_rate": 7.971795893804965e-06, + "loss": 0.3404, + "step": 2291 + }, + { + "epoch": 0.32, + "grad_norm": 2.446290278982956, + "learning_rate": 7.969985023667575e-06, + "loss": 0.3516, + "step": 2292 + }, + { + "epoch": 0.32, + "grad_norm": 2.4612194670262095, + "learning_rate": 7.968173551363738e-06, + "loss": 0.362, + "step": 2293 + }, + { + "epoch": 0.32, + "grad_norm": 2.051197342698838, + "learning_rate": 7.966361477260733e-06, + "loss": 0.3541, + "step": 2294 + }, + { + "epoch": 0.32, + "grad_norm": 2.3760042113694273, + "learning_rate": 7.96454880172596e-06, + "loss": 0.3474, + "step": 2295 + }, + { + "epoch": 0.32, + "grad_norm": 2.892667712679432, + "learning_rate": 7.96273552512694e-06, + "loss": 0.3326, + "step": 2296 + }, + { + "epoch": 0.32, + "grad_norm": 2.1493890376768583, + "learning_rate": 7.960921647831314e-06, + "loss": 0.3271, + "step": 2297 + }, + { + "epoch": 0.32, + "grad_norm": 4.716202639803364, + "learning_rate": 7.95910717020685e-06, + "loss": 0.3496, + "step": 2298 + }, + { + "epoch": 0.32, + "grad_norm": 1.8398212900283903, + "learning_rate": 7.957292092621432e-06, + "loss": 0.3427, + "step": 2299 + }, + { + "epoch": 0.32, + "grad_norm": 2.143540659586766, + "learning_rate": 7.955476415443069e-06, + "loss": 0.3435, + "step": 2300 + }, + { + "epoch": 0.32, + "grad_norm": 2.9384007911356096, + "learning_rate": 7.95366013903989e-06, + "loss": 0.3414, + "step": 2301 + }, + { + "epoch": 0.32, + "grad_norm": 2.620167153358771, + "learning_rate": 7.951843263780148e-06, + "loss": 0.316, + "step": 2302 + }, + { + "epoch": 0.32, + "grad_norm": 2.467522179143929, + "learning_rate": 7.950025790032214e-06, + "loss": 0.3153, + "step": 2303 + }, + { + "epoch": 0.32, + "grad_norm": 2.292353171654257, + "learning_rate": 7.948207718164584e-06, + "loss": 0.3574, + "step": 2304 + }, + { + "epoch": 0.32, + "grad_norm": 2.1108421012660123, + "learning_rate": 7.946389048545871e-06, + "loss": 0.3493, + "step": 2305 + }, + { + "epoch": 0.32, + "grad_norm": 1.9725081019195208, + "learning_rate": 7.944569781544814e-06, + "loss": 0.3136, + "step": 2306 + }, + { + "epoch": 0.32, + "grad_norm": 2.555812836407514, + "learning_rate": 7.942749917530266e-06, + "loss": 0.3616, + "step": 2307 + }, + { + "epoch": 0.32, + "grad_norm": 2.3651043435120265, + "learning_rate": 7.940929456871211e-06, + "loss": 0.3859, + "step": 2308 + }, + { + "epoch": 0.32, + "grad_norm": 2.7977771977756745, + "learning_rate": 7.939108399936747e-06, + "loss": 0.3159, + "step": 2309 + }, + { + "epoch": 0.32, + "grad_norm": 2.194657796878919, + "learning_rate": 7.937286747096093e-06, + "loss": 0.371, + "step": 2310 + }, + { + "epoch": 0.32, + "grad_norm": 1.9854556373812002, + "learning_rate": 7.93546449871859e-06, + "loss": 0.3237, + "step": 2311 + }, + { + "epoch": 0.32, + "grad_norm": 2.370632351656442, + "learning_rate": 7.933641655173703e-06, + "loss": 0.336, + "step": 2312 + }, + { + "epoch": 0.32, + "grad_norm": 2.774332671665691, + "learning_rate": 7.931818216831014e-06, + "loss": 0.3602, + "step": 2313 + }, + { + "epoch": 0.32, + "grad_norm": 2.9121906195678524, + "learning_rate": 7.929994184060221e-06, + "loss": 0.3448, + "step": 2314 + }, + { + "epoch": 0.32, + "grad_norm": 3.2702134130371547, + "learning_rate": 7.928169557231156e-06, + "loss": 0.3455, + "step": 2315 + }, + { + "epoch": 0.32, + "grad_norm": 2.8783180618806474, + "learning_rate": 7.926344336713758e-06, + "loss": 0.3438, + "step": 2316 + }, + { + "epoch": 0.32, + "grad_norm": 2.0896262661100256, + "learning_rate": 7.924518522878093e-06, + "loss": 0.3378, + "step": 2317 + }, + { + "epoch": 0.32, + "grad_norm": 3.20228381913423, + "learning_rate": 7.922692116094349e-06, + "loss": 0.3728, + "step": 2318 + }, + { + "epoch": 0.32, + "grad_norm": 2.461165295720281, + "learning_rate": 7.920865116732824e-06, + "loss": 0.3529, + "step": 2319 + }, + { + "epoch": 0.32, + "grad_norm": 1.84590277234616, + "learning_rate": 7.91903752516395e-06, + "loss": 0.3448, + "step": 2320 + }, + { + "epoch": 0.32, + "grad_norm": 2.6095911086002195, + "learning_rate": 7.91720934175827e-06, + "loss": 0.3194, + "step": 2321 + }, + { + "epoch": 0.32, + "grad_norm": 2.0699341492944985, + "learning_rate": 7.91538056688645e-06, + "loss": 0.3258, + "step": 2322 + }, + { + "epoch": 0.32, + "grad_norm": 2.401286398006276, + "learning_rate": 7.913551200919274e-06, + "loss": 0.3721, + "step": 2323 + }, + { + "epoch": 0.32, + "grad_norm": 2.5835541085906506, + "learning_rate": 7.911721244227647e-06, + "loss": 0.33, + "step": 2324 + }, + { + "epoch": 0.32, + "grad_norm": 1.7542871721423148, + "learning_rate": 7.909890697182598e-06, + "loss": 0.3188, + "step": 2325 + }, + { + "epoch": 0.32, + "grad_norm": 2.1615711030525526, + "learning_rate": 7.908059560155267e-06, + "loss": 0.344, + "step": 2326 + }, + { + "epoch": 0.32, + "grad_norm": 2.225972559857873, + "learning_rate": 7.90622783351692e-06, + "loss": 0.3559, + "step": 2327 + }, + { + "epoch": 0.32, + "grad_norm": 8.028769273632316, + "learning_rate": 7.90439551763894e-06, + "loss": 0.3321, + "step": 2328 + }, + { + "epoch": 0.32, + "grad_norm": 2.6053119613193427, + "learning_rate": 7.90256261289283e-06, + "loss": 0.349, + "step": 2329 + }, + { + "epoch": 0.32, + "grad_norm": 2.5418070167441575, + "learning_rate": 7.900729119650215e-06, + "loss": 0.3691, + "step": 2330 + }, + { + "epoch": 0.32, + "grad_norm": 2.2206790079555603, + "learning_rate": 7.898895038282836e-06, + "loss": 0.3522, + "step": 2331 + }, + { + "epoch": 0.32, + "grad_norm": 2.643070694877896, + "learning_rate": 7.897060369162552e-06, + "loss": 0.347, + "step": 2332 + }, + { + "epoch": 0.32, + "grad_norm": 2.5244611832103576, + "learning_rate": 7.895225112661347e-06, + "loss": 0.3561, + "step": 2333 + }, + { + "epoch": 0.32, + "grad_norm": 3.2498164555799844, + "learning_rate": 7.893389269151319e-06, + "loss": 0.349, + "step": 2334 + }, + { + "epoch": 0.32, + "grad_norm": 2.3211558342515644, + "learning_rate": 7.891552839004685e-06, + "loss": 0.341, + "step": 2335 + }, + { + "epoch": 0.32, + "grad_norm": 4.6635427883473435, + "learning_rate": 7.889715822593783e-06, + "loss": 0.3423, + "step": 2336 + }, + { + "epoch": 0.32, + "grad_norm": 2.2571443469288806, + "learning_rate": 7.887878220291072e-06, + "loss": 0.339, + "step": 2337 + }, + { + "epoch": 0.33, + "grad_norm": 2.199868347332378, + "learning_rate": 7.886040032469122e-06, + "loss": 0.3419, + "step": 2338 + }, + { + "epoch": 0.33, + "grad_norm": 2.543594754249072, + "learning_rate": 7.884201259500631e-06, + "loss": 0.3438, + "step": 2339 + }, + { + "epoch": 0.33, + "grad_norm": 2.6876309633138167, + "learning_rate": 7.88236190175841e-06, + "loss": 0.3373, + "step": 2340 + }, + { + "epoch": 0.33, + "grad_norm": 2.274218933978384, + "learning_rate": 7.880521959615392e-06, + "loss": 0.3231, + "step": 2341 + }, + { + "epoch": 0.33, + "grad_norm": 2.1389816106929325, + "learning_rate": 7.878681433444624e-06, + "loss": 0.3209, + "step": 2342 + }, + { + "epoch": 0.33, + "grad_norm": 2.370270999081007, + "learning_rate": 7.876840323619273e-06, + "loss": 0.3686, + "step": 2343 + }, + { + "epoch": 0.33, + "grad_norm": 2.4288559111958006, + "learning_rate": 7.87499863051263e-06, + "loss": 0.3436, + "step": 2344 + }, + { + "epoch": 0.33, + "grad_norm": 2.3969726000777203, + "learning_rate": 7.873156354498092e-06, + "loss": 0.3481, + "step": 2345 + }, + { + "epoch": 0.33, + "grad_norm": 2.2031635007616237, + "learning_rate": 7.87131349594919e-06, + "loss": 0.3256, + "step": 2346 + }, + { + "epoch": 0.33, + "grad_norm": 2.1102587765547227, + "learning_rate": 7.869470055239558e-06, + "loss": 0.3365, + "step": 2347 + }, + { + "epoch": 0.33, + "grad_norm": 1.8380076082098187, + "learning_rate": 7.86762603274296e-06, + "loss": 0.3489, + "step": 2348 + }, + { + "epoch": 0.33, + "grad_norm": 3.1539229476193427, + "learning_rate": 7.865781428833269e-06, + "loss": 0.3437, + "step": 2349 + }, + { + "epoch": 0.33, + "grad_norm": 1.9813552481878716, + "learning_rate": 7.863936243884482e-06, + "loss": 0.3386, + "step": 2350 + }, + { + "epoch": 0.33, + "grad_norm": 3.0006380054803974, + "learning_rate": 7.862090478270711e-06, + "loss": 0.3332, + "step": 2351 + }, + { + "epoch": 0.33, + "grad_norm": 2.6690034339542317, + "learning_rate": 7.860244132366184e-06, + "loss": 0.3436, + "step": 2352 + }, + { + "epoch": 0.33, + "grad_norm": 2.241062974589399, + "learning_rate": 7.858397206545251e-06, + "loss": 0.3367, + "step": 2353 + }, + { + "epoch": 0.33, + "grad_norm": 2.2916131469351466, + "learning_rate": 7.856549701182379e-06, + "loss": 0.3223, + "step": 2354 + }, + { + "epoch": 0.33, + "grad_norm": 2.7164191257218353, + "learning_rate": 7.85470161665215e-06, + "loss": 0.3587, + "step": 2355 + }, + { + "epoch": 0.33, + "grad_norm": 3.217514808375243, + "learning_rate": 7.852852953329263e-06, + "loss": 0.3624, + "step": 2356 + }, + { + "epoch": 0.33, + "grad_norm": 2.5448761150708754, + "learning_rate": 7.851003711588532e-06, + "loss": 0.3404, + "step": 2357 + }, + { + "epoch": 0.33, + "grad_norm": 1.9707843896868267, + "learning_rate": 7.8491538918049e-06, + "loss": 0.3469, + "step": 2358 + }, + { + "epoch": 0.33, + "grad_norm": 2.2605678183616016, + "learning_rate": 7.847303494353416e-06, + "loss": 0.3321, + "step": 2359 + }, + { + "epoch": 0.33, + "grad_norm": 2.6214319380328166, + "learning_rate": 7.845452519609246e-06, + "loss": 0.3315, + "step": 2360 + }, + { + "epoch": 0.33, + "grad_norm": 2.823049192716383, + "learning_rate": 7.84360096794768e-06, + "loss": 0.3333, + "step": 2361 + }, + { + "epoch": 0.33, + "grad_norm": 2.584609200731347, + "learning_rate": 7.841748839744121e-06, + "loss": 0.3327, + "step": 2362 + }, + { + "epoch": 0.33, + "grad_norm": 2.569865126059081, + "learning_rate": 7.839896135374088e-06, + "loss": 0.3358, + "step": 2363 + }, + { + "epoch": 0.33, + "grad_norm": 2.2682205955487253, + "learning_rate": 7.83804285521322e-06, + "loss": 0.3733, + "step": 2364 + }, + { + "epoch": 0.33, + "grad_norm": 2.1876520605963203, + "learning_rate": 7.836188999637268e-06, + "loss": 0.3502, + "step": 2365 + }, + { + "epoch": 0.33, + "grad_norm": 1.8687820052238782, + "learning_rate": 7.834334569022106e-06, + "loss": 0.3421, + "step": 2366 + }, + { + "epoch": 0.33, + "grad_norm": 2.6022301612861622, + "learning_rate": 7.832479563743717e-06, + "loss": 0.3297, + "step": 2367 + }, + { + "epoch": 0.33, + "grad_norm": 6.1233576230333515, + "learning_rate": 7.830623984178207e-06, + "loss": 0.3433, + "step": 2368 + }, + { + "epoch": 0.33, + "grad_norm": 2.63977469754912, + "learning_rate": 7.828767830701796e-06, + "loss": 0.3275, + "step": 2369 + }, + { + "epoch": 0.33, + "grad_norm": 2.053697002632728, + "learning_rate": 7.82691110369082e-06, + "loss": 0.3343, + "step": 2370 + }, + { + "epoch": 0.33, + "grad_norm": 2.890408748000691, + "learning_rate": 7.825053803521731e-06, + "loss": 0.3408, + "step": 2371 + }, + { + "epoch": 0.33, + "grad_norm": 2.4860178415492697, + "learning_rate": 7.823195930571099e-06, + "loss": 0.3356, + "step": 2372 + }, + { + "epoch": 0.33, + "grad_norm": 2.495470613862421, + "learning_rate": 7.821337485215607e-06, + "loss": 0.3425, + "step": 2373 + }, + { + "epoch": 0.33, + "grad_norm": 3.8775026202735825, + "learning_rate": 7.819478467832058e-06, + "loss": 0.3645, + "step": 2374 + }, + { + "epoch": 0.33, + "grad_norm": 3.0860768841514656, + "learning_rate": 7.81761887879737e-06, + "loss": 0.3483, + "step": 2375 + }, + { + "epoch": 0.33, + "grad_norm": 3.066825156500302, + "learning_rate": 7.815758718488573e-06, + "loss": 0.3548, + "step": 2376 + }, + { + "epoch": 0.33, + "grad_norm": 1.9536790015426528, + "learning_rate": 7.813897987282816e-06, + "loss": 0.3297, + "step": 2377 + }, + { + "epoch": 0.33, + "grad_norm": 2.519955168788722, + "learning_rate": 7.812036685557367e-06, + "loss": 0.3509, + "step": 2378 + }, + { + "epoch": 0.33, + "grad_norm": 3.038148540459448, + "learning_rate": 7.810174813689601e-06, + "loss": 0.35, + "step": 2379 + }, + { + "epoch": 0.33, + "grad_norm": 2.305476938991848, + "learning_rate": 7.808312372057019e-06, + "loss": 0.342, + "step": 2380 + }, + { + "epoch": 0.33, + "grad_norm": 1.9198011394979686, + "learning_rate": 7.806449361037229e-06, + "loss": 0.332, + "step": 2381 + }, + { + "epoch": 0.33, + "grad_norm": 5.380152256865416, + "learning_rate": 7.804585781007958e-06, + "loss": 0.321, + "step": 2382 + }, + { + "epoch": 0.33, + "grad_norm": 3.2610007890399397, + "learning_rate": 7.802721632347048e-06, + "loss": 0.3389, + "step": 2383 + }, + { + "epoch": 0.33, + "grad_norm": 2.4725679373674487, + "learning_rate": 7.800856915432458e-06, + "loss": 0.3302, + "step": 2384 + }, + { + "epoch": 0.33, + "grad_norm": 2.68005311591632, + "learning_rate": 7.79899163064226e-06, + "loss": 0.3382, + "step": 2385 + }, + { + "epoch": 0.33, + "grad_norm": 3.139425658529219, + "learning_rate": 7.797125778354642e-06, + "loss": 0.3546, + "step": 2386 + }, + { + "epoch": 0.33, + "grad_norm": 2.1679392674067657, + "learning_rate": 7.795259358947903e-06, + "loss": 0.3459, + "step": 2387 + }, + { + "epoch": 0.33, + "grad_norm": 3.07354863091647, + "learning_rate": 7.793392372800466e-06, + "loss": 0.323, + "step": 2388 + }, + { + "epoch": 0.33, + "grad_norm": 3.58442108913341, + "learning_rate": 7.791524820290862e-06, + "loss": 0.3635, + "step": 2389 + }, + { + "epoch": 0.33, + "grad_norm": 1.891552294347941, + "learning_rate": 7.78965670179774e-06, + "loss": 0.3469, + "step": 2390 + }, + { + "epoch": 0.33, + "grad_norm": 2.775653768649609, + "learning_rate": 7.787788017699855e-06, + "loss": 0.3567, + "step": 2391 + }, + { + "epoch": 0.33, + "grad_norm": 3.172604233472699, + "learning_rate": 7.785918768376093e-06, + "loss": 0.3453, + "step": 2392 + }, + { + "epoch": 0.33, + "grad_norm": 3.0329534692416313, + "learning_rate": 7.784048954205441e-06, + "loss": 0.352, + "step": 2393 + }, + { + "epoch": 0.33, + "grad_norm": 2.996574524409091, + "learning_rate": 7.782178575567005e-06, + "loss": 0.3626, + "step": 2394 + }, + { + "epoch": 0.33, + "grad_norm": 2.7058570081208355, + "learning_rate": 7.780307632840009e-06, + "loss": 0.3233, + "step": 2395 + }, + { + "epoch": 0.33, + "grad_norm": 2.5215282696411365, + "learning_rate": 7.778436126403784e-06, + "loss": 0.3425, + "step": 2396 + }, + { + "epoch": 0.33, + "grad_norm": 4.362887002669442, + "learning_rate": 7.77656405663778e-06, + "loss": 0.3555, + "step": 2397 + }, + { + "epoch": 0.33, + "grad_norm": 2.372508839712804, + "learning_rate": 7.774691423921561e-06, + "loss": 0.3465, + "step": 2398 + }, + { + "epoch": 0.33, + "grad_norm": 2.148457387291239, + "learning_rate": 7.772818228634805e-06, + "loss": 0.3451, + "step": 2399 + }, + { + "epoch": 0.33, + "grad_norm": 1.9800575579637316, + "learning_rate": 7.770944471157302e-06, + "loss": 0.3278, + "step": 2400 + }, + { + "epoch": 0.33, + "grad_norm": 2.4539320671876212, + "learning_rate": 7.76907015186896e-06, + "loss": 0.3269, + "step": 2401 + }, + { + "epoch": 0.33, + "grad_norm": 2.125222417595608, + "learning_rate": 7.767195271149797e-06, + "loss": 0.3632, + "step": 2402 + }, + { + "epoch": 0.33, + "grad_norm": 2.541830243680224, + "learning_rate": 7.765319829379944e-06, + "loss": 0.3681, + "step": 2403 + }, + { + "epoch": 0.33, + "grad_norm": 2.345727314681814, + "learning_rate": 7.763443826939654e-06, + "loss": 0.3372, + "step": 2404 + }, + { + "epoch": 0.33, + "grad_norm": 2.4090819293619155, + "learning_rate": 7.761567264209282e-06, + "loss": 0.3436, + "step": 2405 + }, + { + "epoch": 0.33, + "grad_norm": 2.713592041569188, + "learning_rate": 7.759690141569305e-06, + "loss": 0.3293, + "step": 2406 + }, + { + "epoch": 0.33, + "grad_norm": 2.2818131704076716, + "learning_rate": 7.757812459400311e-06, + "loss": 0.3495, + "step": 2407 + }, + { + "epoch": 0.33, + "grad_norm": 2.9453540009931163, + "learning_rate": 7.755934218083e-06, + "loss": 0.3518, + "step": 2408 + }, + { + "epoch": 0.33, + "grad_norm": 2.0628245119343616, + "learning_rate": 7.754055417998189e-06, + "loss": 0.3575, + "step": 2409 + }, + { + "epoch": 0.34, + "grad_norm": 2.1162214238387533, + "learning_rate": 7.752176059526803e-06, + "loss": 0.3314, + "step": 2410 + }, + { + "epoch": 0.34, + "grad_norm": 2.758327294876577, + "learning_rate": 7.750296143049885e-06, + "loss": 0.3541, + "step": 2411 + }, + { + "epoch": 0.34, + "grad_norm": 1.9025429514628587, + "learning_rate": 7.74841566894859e-06, + "loss": 0.3254, + "step": 2412 + }, + { + "epoch": 0.34, + "grad_norm": 2.376290572925721, + "learning_rate": 7.746534637604184e-06, + "loss": 0.3577, + "step": 2413 + }, + { + "epoch": 0.34, + "grad_norm": 2.201943245533939, + "learning_rate": 7.744653049398045e-06, + "loss": 0.342, + "step": 2414 + }, + { + "epoch": 0.34, + "grad_norm": 2.4402869227592365, + "learning_rate": 7.742770904711672e-06, + "loss": 0.3196, + "step": 2415 + }, + { + "epoch": 0.34, + "grad_norm": 2.4159975655780093, + "learning_rate": 7.740888203926666e-06, + "loss": 0.3315, + "step": 2416 + }, + { + "epoch": 0.34, + "grad_norm": 2.1401036314111885, + "learning_rate": 7.739004947424748e-06, + "loss": 0.3209, + "step": 2417 + }, + { + "epoch": 0.34, + "grad_norm": 2.1529411836820556, + "learning_rate": 7.73712113558775e-06, + "loss": 0.3671, + "step": 2418 + }, + { + "epoch": 0.34, + "grad_norm": 2.6068837304779384, + "learning_rate": 7.735236768797617e-06, + "loss": 0.3214, + "step": 2419 + }, + { + "epoch": 0.34, + "grad_norm": 2.2569619192928334, + "learning_rate": 7.733351847436403e-06, + "loss": 0.3475, + "step": 2420 + }, + { + "epoch": 0.34, + "grad_norm": 3.0010252550472254, + "learning_rate": 7.731466371886277e-06, + "loss": 0.3487, + "step": 2421 + }, + { + "epoch": 0.34, + "grad_norm": 2.525198013185836, + "learning_rate": 7.729580342529522e-06, + "loss": 0.3744, + "step": 2422 + }, + { + "epoch": 0.34, + "grad_norm": 3.7102059367192757, + "learning_rate": 7.727693759748531e-06, + "loss": 0.3371, + "step": 2423 + }, + { + "epoch": 0.34, + "grad_norm": 2.480239980846062, + "learning_rate": 7.725806623925812e-06, + "loss": 0.3381, + "step": 2424 + }, + { + "epoch": 0.34, + "grad_norm": 2.1799877842963014, + "learning_rate": 7.723918935443981e-06, + "loss": 0.3371, + "step": 2425 + }, + { + "epoch": 0.34, + "grad_norm": 2.2804069441328636, + "learning_rate": 7.722030694685768e-06, + "loss": 0.3585, + "step": 2426 + }, + { + "epoch": 0.34, + "grad_norm": 2.423765075659841, + "learning_rate": 7.720141902034013e-06, + "loss": 0.3463, + "step": 2427 + }, + { + "epoch": 0.34, + "grad_norm": 3.1448730706358368, + "learning_rate": 7.718252557871677e-06, + "loss": 0.3484, + "step": 2428 + }, + { + "epoch": 0.34, + "grad_norm": 2.8197390720897775, + "learning_rate": 7.716362662581818e-06, + "loss": 0.354, + "step": 2429 + }, + { + "epoch": 0.34, + "grad_norm": 2.270796970909256, + "learning_rate": 7.714472216547617e-06, + "loss": 0.3247, + "step": 2430 + }, + { + "epoch": 0.34, + "grad_norm": 2.2595869785171603, + "learning_rate": 7.712581220152363e-06, + "loss": 0.3216, + "step": 2431 + }, + { + "epoch": 0.34, + "grad_norm": 2.7351243848747258, + "learning_rate": 7.710689673779456e-06, + "loss": 0.3468, + "step": 2432 + }, + { + "epoch": 0.34, + "grad_norm": 2.2943792483835757, + "learning_rate": 7.708797577812412e-06, + "loss": 0.3387, + "step": 2433 + }, + { + "epoch": 0.34, + "grad_norm": 2.306457333314307, + "learning_rate": 7.706904932634849e-06, + "loss": 0.3436, + "step": 2434 + }, + { + "epoch": 0.34, + "grad_norm": 2.004536200098642, + "learning_rate": 7.705011738630506e-06, + "loss": 0.3473, + "step": 2435 + }, + { + "epoch": 0.34, + "grad_norm": 4.724782648897764, + "learning_rate": 7.703117996183227e-06, + "loss": 0.3297, + "step": 2436 + }, + { + "epoch": 0.34, + "grad_norm": 3.3400517775077003, + "learning_rate": 7.701223705676972e-06, + "loss": 0.3174, + "step": 2437 + }, + { + "epoch": 0.34, + "grad_norm": 2.865936769533719, + "learning_rate": 7.699328867495807e-06, + "loss": 0.3503, + "step": 2438 + }, + { + "epoch": 0.34, + "grad_norm": 2.729172460604864, + "learning_rate": 7.697433482023915e-06, + "loss": 0.3403, + "step": 2439 + }, + { + "epoch": 0.34, + "grad_norm": 3.0154473694130117, + "learning_rate": 7.695537549645582e-06, + "loss": 0.3275, + "step": 2440 + }, + { + "epoch": 0.34, + "grad_norm": 3.6905893923441546, + "learning_rate": 7.693641070745214e-06, + "loss": 0.3547, + "step": 2441 + }, + { + "epoch": 0.34, + "grad_norm": 2.4581782587973344, + "learning_rate": 7.69174404570732e-06, + "loss": 0.3261, + "step": 2442 + }, + { + "epoch": 0.34, + "grad_norm": 2.5568597099293884, + "learning_rate": 7.689846474916528e-06, + "loss": 0.325, + "step": 2443 + }, + { + "epoch": 0.34, + "grad_norm": 2.403814472305729, + "learning_rate": 7.687948358757567e-06, + "loss": 0.3387, + "step": 2444 + }, + { + "epoch": 0.34, + "grad_norm": 4.307360305334001, + "learning_rate": 7.686049697615281e-06, + "loss": 0.3502, + "step": 2445 + }, + { + "epoch": 0.34, + "grad_norm": 2.337837460586784, + "learning_rate": 7.68415049187463e-06, + "loss": 0.3189, + "step": 2446 + }, + { + "epoch": 0.34, + "grad_norm": 2.1034867460619227, + "learning_rate": 7.682250741920674e-06, + "loss": 0.3703, + "step": 2447 + }, + { + "epoch": 0.34, + "grad_norm": 2.4820384212608144, + "learning_rate": 7.68035044813859e-06, + "loss": 0.3149, + "step": 2448 + }, + { + "epoch": 0.34, + "grad_norm": 2.512304498038403, + "learning_rate": 7.678449610913663e-06, + "loss": 0.3448, + "step": 2449 + }, + { + "epoch": 0.34, + "grad_norm": 3.382109225252942, + "learning_rate": 7.676548230631293e-06, + "loss": 0.3673, + "step": 2450 + }, + { + "epoch": 0.34, + "grad_norm": 1.8685922914430604, + "learning_rate": 7.674646307676981e-06, + "loss": 0.3143, + "step": 2451 + }, + { + "epoch": 0.34, + "grad_norm": 2.430041815305568, + "learning_rate": 7.672743842436346e-06, + "loss": 0.3493, + "step": 2452 + }, + { + "epoch": 0.34, + "grad_norm": 2.3680825973368167, + "learning_rate": 7.670840835295116e-06, + "loss": 0.3337, + "step": 2453 + }, + { + "epoch": 0.34, + "grad_norm": 3.0662738986339306, + "learning_rate": 7.668937286639122e-06, + "loss": 0.3357, + "step": 2454 + }, + { + "epoch": 0.34, + "grad_norm": 2.3299901478758462, + "learning_rate": 7.667033196854314e-06, + "loss": 0.3197, + "step": 2455 + }, + { + "epoch": 0.34, + "grad_norm": 2.660433175501756, + "learning_rate": 7.665128566326746e-06, + "loss": 0.3704, + "step": 2456 + }, + { + "epoch": 0.34, + "grad_norm": 2.516637592277573, + "learning_rate": 7.66322339544258e-06, + "loss": 0.3388, + "step": 2457 + }, + { + "epoch": 0.34, + "grad_norm": 2.3582292291929403, + "learning_rate": 7.661317684588097e-06, + "loss": 0.3408, + "step": 2458 + }, + { + "epoch": 0.34, + "grad_norm": 2.5933440573456887, + "learning_rate": 7.659411434149675e-06, + "loss": 0.348, + "step": 2459 + }, + { + "epoch": 0.34, + "grad_norm": 2.829027985448064, + "learning_rate": 7.657504644513814e-06, + "loss": 0.3648, + "step": 2460 + }, + { + "epoch": 0.34, + "grad_norm": 2.167417934786794, + "learning_rate": 7.655597316067109e-06, + "loss": 0.3445, + "step": 2461 + }, + { + "epoch": 0.34, + "grad_norm": 2.4940617861408687, + "learning_rate": 7.653689449196277e-06, + "loss": 0.3465, + "step": 2462 + }, + { + "epoch": 0.34, + "grad_norm": 5.660332967189399, + "learning_rate": 7.65178104428814e-06, + "loss": 0.3253, + "step": 2463 + }, + { + "epoch": 0.34, + "grad_norm": 1.8601698435986758, + "learning_rate": 7.649872101729627e-06, + "loss": 0.3451, + "step": 2464 + }, + { + "epoch": 0.34, + "grad_norm": 2.3711936021976188, + "learning_rate": 7.647962621907776e-06, + "loss": 0.3116, + "step": 2465 + }, + { + "epoch": 0.34, + "grad_norm": 4.211262434225153, + "learning_rate": 7.646052605209738e-06, + "loss": 0.3602, + "step": 2466 + }, + { + "epoch": 0.34, + "grad_norm": 1.8945394838732919, + "learning_rate": 7.64414205202277e-06, + "loss": 0.3294, + "step": 2467 + }, + { + "epoch": 0.34, + "grad_norm": 1.9241215261934468, + "learning_rate": 7.642230962734234e-06, + "loss": 0.341, + "step": 2468 + }, + { + "epoch": 0.34, + "grad_norm": 2.842939256766711, + "learning_rate": 7.640319337731608e-06, + "loss": 0.3518, + "step": 2469 + }, + { + "epoch": 0.34, + "grad_norm": 2.3610176522009834, + "learning_rate": 7.638407177402474e-06, + "loss": 0.3623, + "step": 2470 + }, + { + "epoch": 0.34, + "grad_norm": 1.8529989414378023, + "learning_rate": 7.636494482134527e-06, + "loss": 0.3296, + "step": 2471 + }, + { + "epoch": 0.34, + "grad_norm": 2.2341466969060946, + "learning_rate": 7.634581252315563e-06, + "loss": 0.3536, + "step": 2472 + }, + { + "epoch": 0.34, + "grad_norm": 2.181531036346788, + "learning_rate": 7.632667488333492e-06, + "loss": 0.3488, + "step": 2473 + }, + { + "epoch": 0.34, + "grad_norm": 2.497800902314139, + "learning_rate": 7.630753190576331e-06, + "loss": 0.3835, + "step": 2474 + }, + { + "epoch": 0.34, + "grad_norm": 2.6764775410877113, + "learning_rate": 7.628838359432205e-06, + "loss": 0.3478, + "step": 2475 + }, + { + "epoch": 0.34, + "grad_norm": 2.1780328928177113, + "learning_rate": 7.626922995289348e-06, + "loss": 0.3295, + "step": 2476 + }, + { + "epoch": 0.34, + "grad_norm": 2.8554632588694684, + "learning_rate": 7.625007098536102e-06, + "loss": 0.3405, + "step": 2477 + }, + { + "epoch": 0.34, + "grad_norm": 2.1042372616292084, + "learning_rate": 7.623090669560913e-06, + "loss": 0.3377, + "step": 2478 + }, + { + "epoch": 0.34, + "grad_norm": 2.446908493190573, + "learning_rate": 7.621173708752343e-06, + "loss": 0.344, + "step": 2479 + }, + { + "epoch": 0.34, + "grad_norm": 2.679866973165594, + "learning_rate": 7.619256216499054e-06, + "loss": 0.348, + "step": 2480 + }, + { + "epoch": 0.34, + "grad_norm": 2.2809472146495833, + "learning_rate": 7.617338193189818e-06, + "loss": 0.3335, + "step": 2481 + }, + { + "epoch": 0.35, + "grad_norm": 2.3937246928802245, + "learning_rate": 7.615419639213519e-06, + "loss": 0.3194, + "step": 2482 + }, + { + "epoch": 0.35, + "grad_norm": 2.2946326910830033, + "learning_rate": 7.613500554959141e-06, + "loss": 0.3251, + "step": 2483 + }, + { + "epoch": 0.35, + "grad_norm": 2.1885665318785694, + "learning_rate": 7.611580940815783e-06, + "loss": 0.3121, + "step": 2484 + }, + { + "epoch": 0.35, + "grad_norm": 2.392918617014723, + "learning_rate": 7.6096607971726475e-06, + "loss": 0.3459, + "step": 2485 + }, + { + "epoch": 0.35, + "grad_norm": 2.5869259929827995, + "learning_rate": 7.6077401244190444e-06, + "loss": 0.3267, + "step": 2486 + }, + { + "epoch": 0.35, + "grad_norm": 2.2175621682838633, + "learning_rate": 7.605818922944388e-06, + "loss": 0.3399, + "step": 2487 + }, + { + "epoch": 0.35, + "grad_norm": 1.9997121495143675, + "learning_rate": 7.6038971931382074e-06, + "loss": 0.3432, + "step": 2488 + }, + { + "epoch": 0.35, + "grad_norm": 2.2371568264406525, + "learning_rate": 7.601974935390136e-06, + "loss": 0.3535, + "step": 2489 + }, + { + "epoch": 0.35, + "grad_norm": 2.6898955591179354, + "learning_rate": 7.60005215008991e-06, + "loss": 0.3336, + "step": 2490 + }, + { + "epoch": 0.35, + "grad_norm": 2.5605176341284777, + "learning_rate": 7.598128837627372e-06, + "loss": 0.3556, + "step": 2491 + }, + { + "epoch": 0.35, + "grad_norm": 2.6400300973703326, + "learning_rate": 7.596204998392482e-06, + "loss": 0.3345, + "step": 2492 + }, + { + "epoch": 0.35, + "grad_norm": 2.1429681136452405, + "learning_rate": 7.594280632775296e-06, + "loss": 0.3376, + "step": 2493 + }, + { + "epoch": 0.35, + "grad_norm": 2.3057068290180744, + "learning_rate": 7.59235574116598e-06, + "loss": 0.3408, + "step": 2494 + }, + { + "epoch": 0.35, + "grad_norm": 2.2584963272612364, + "learning_rate": 7.590430323954807e-06, + "loss": 0.3339, + "step": 2495 + }, + { + "epoch": 0.35, + "grad_norm": 3.3487149850860543, + "learning_rate": 7.588504381532158e-06, + "loss": 0.3317, + "step": 2496 + }, + { + "epoch": 0.35, + "grad_norm": 1.9888282817317733, + "learning_rate": 7.58657791428852e-06, + "loss": 0.3251, + "step": 2497 + }, + { + "epoch": 0.35, + "grad_norm": 1.8913267762747263, + "learning_rate": 7.584650922614483e-06, + "loss": 0.3196, + "step": 2498 + }, + { + "epoch": 0.35, + "grad_norm": 2.6639036916987613, + "learning_rate": 7.582723406900747e-06, + "loss": 0.3138, + "step": 2499 + }, + { + "epoch": 0.35, + "grad_norm": 1.934943416424287, + "learning_rate": 7.580795367538116e-06, + "loss": 0.3503, + "step": 2500 + }, + { + "epoch": 0.35, + "grad_norm": 3.0209840547789746, + "learning_rate": 7.5788668049175025e-06, + "loss": 0.3431, + "step": 2501 + }, + { + "epoch": 0.35, + "grad_norm": 2.03889462156148, + "learning_rate": 7.576937719429925e-06, + "loss": 0.3373, + "step": 2502 + }, + { + "epoch": 0.35, + "grad_norm": 2.1188911394972125, + "learning_rate": 7.575008111466504e-06, + "loss": 0.3537, + "step": 2503 + }, + { + "epoch": 0.35, + "grad_norm": 2.421245964803, + "learning_rate": 7.573077981418471e-06, + "loss": 0.352, + "step": 2504 + }, + { + "epoch": 0.35, + "grad_norm": 2.192053285484204, + "learning_rate": 7.571147329677161e-06, + "loss": 0.3515, + "step": 2505 + }, + { + "epoch": 0.35, + "grad_norm": 3.062550485576645, + "learning_rate": 7.5692161566340135e-06, + "loss": 0.3623, + "step": 2506 + }, + { + "epoch": 0.35, + "grad_norm": 1.906095195325121, + "learning_rate": 7.567284462680575e-06, + "loss": 0.3206, + "step": 2507 + }, + { + "epoch": 0.35, + "grad_norm": 3.2000237024674028, + "learning_rate": 7.5653522482085006e-06, + "loss": 0.3633, + "step": 2508 + }, + { + "epoch": 0.35, + "grad_norm": 2.9812804340718815, + "learning_rate": 7.5634195136095455e-06, + "loss": 0.3464, + "step": 2509 + }, + { + "epoch": 0.35, + "grad_norm": 2.393951273533495, + "learning_rate": 7.561486259275574e-06, + "loss": 0.3362, + "step": 2510 + }, + { + "epoch": 0.35, + "grad_norm": 2.2225403753911985, + "learning_rate": 7.559552485598553e-06, + "loss": 0.3412, + "step": 2511 + }, + { + "epoch": 0.35, + "grad_norm": 2.212882458062359, + "learning_rate": 7.557618192970559e-06, + "loss": 0.3719, + "step": 2512 + }, + { + "epoch": 0.35, + "grad_norm": 2.291260220009652, + "learning_rate": 7.555683381783771e-06, + "loss": 0.3285, + "step": 2513 + }, + { + "epoch": 0.35, + "grad_norm": 6.037370607992909, + "learning_rate": 7.553748052430469e-06, + "loss": 0.3345, + "step": 2514 + }, + { + "epoch": 0.35, + "grad_norm": 1.7866009329318444, + "learning_rate": 7.5518122053030485e-06, + "loss": 0.3217, + "step": 2515 + }, + { + "epoch": 0.35, + "grad_norm": 2.6511941616896446, + "learning_rate": 7.549875840794e-06, + "loss": 0.3314, + "step": 2516 + }, + { + "epoch": 0.35, + "grad_norm": 2.0900139547510905, + "learning_rate": 7.547938959295922e-06, + "loss": 0.311, + "step": 2517 + }, + { + "epoch": 0.35, + "grad_norm": 2.299623295287365, + "learning_rate": 7.546001561201521e-06, + "loss": 0.3412, + "step": 2518 + }, + { + "epoch": 0.35, + "grad_norm": 2.3398010417296575, + "learning_rate": 7.5440636469036045e-06, + "loss": 0.3536, + "step": 2519 + }, + { + "epoch": 0.35, + "grad_norm": 2.095623926256206, + "learning_rate": 7.542125216795086e-06, + "loss": 0.3402, + "step": 2520 + }, + { + "epoch": 0.35, + "grad_norm": 2.1162044165701546, + "learning_rate": 7.540186271268983e-06, + "loss": 0.3414, + "step": 2521 + }, + { + "epoch": 0.35, + "grad_norm": 4.415663329089738, + "learning_rate": 7.5382468107184214e-06, + "loss": 0.3486, + "step": 2522 + }, + { + "epoch": 0.35, + "grad_norm": 2.6117192479056937, + "learning_rate": 7.536306835536622e-06, + "loss": 0.3565, + "step": 2523 + }, + { + "epoch": 0.35, + "grad_norm": 1.9103960865591938, + "learning_rate": 7.534366346116922e-06, + "loss": 0.3446, + "step": 2524 + }, + { + "epoch": 0.35, + "grad_norm": 2.3000290755725943, + "learning_rate": 7.532425342852753e-06, + "loss": 0.3472, + "step": 2525 + }, + { + "epoch": 0.35, + "grad_norm": 2.888522516088044, + "learning_rate": 7.530483826137655e-06, + "loss": 0.3652, + "step": 2526 + }, + { + "epoch": 0.35, + "grad_norm": 2.788116086559966, + "learning_rate": 7.528541796365274e-06, + "loss": 0.3492, + "step": 2527 + }, + { + "epoch": 0.35, + "grad_norm": 1.6628171459824737, + "learning_rate": 7.526599253929357e-06, + "loss": 0.3381, + "step": 2528 + }, + { + "epoch": 0.35, + "grad_norm": 2.2891382270928653, + "learning_rate": 7.524656199223757e-06, + "loss": 0.3411, + "step": 2529 + }, + { + "epoch": 0.35, + "grad_norm": 2.3768354386531185, + "learning_rate": 7.522712632642424e-06, + "loss": 0.3542, + "step": 2530 + }, + { + "epoch": 0.35, + "grad_norm": 2.765238304396076, + "learning_rate": 7.520768554579425e-06, + "loss": 0.3408, + "step": 2531 + }, + { + "epoch": 0.35, + "grad_norm": 2.024167559112422, + "learning_rate": 7.518823965428918e-06, + "loss": 0.3278, + "step": 2532 + }, + { + "epoch": 0.35, + "grad_norm": 2.4702800619046514, + "learning_rate": 7.516878865585171e-06, + "loss": 0.3116, + "step": 2533 + }, + { + "epoch": 0.35, + "grad_norm": 2.130199197549435, + "learning_rate": 7.514933255442556e-06, + "loss": 0.3239, + "step": 2534 + }, + { + "epoch": 0.35, + "grad_norm": 2.4481156581819232, + "learning_rate": 7.512987135395544e-06, + "loss": 0.3378, + "step": 2535 + }, + { + "epoch": 0.35, + "grad_norm": 2.1053748159445718, + "learning_rate": 7.5110405058387155e-06, + "loss": 0.3573, + "step": 2536 + }, + { + "epoch": 0.35, + "grad_norm": 2.0363473536435475, + "learning_rate": 7.509093367166748e-06, + "loss": 0.339, + "step": 2537 + }, + { + "epoch": 0.35, + "grad_norm": 1.9997202137863326, + "learning_rate": 7.507145719774426e-06, + "loss": 0.3445, + "step": 2538 + }, + { + "epoch": 0.35, + "grad_norm": 3.941661964392704, + "learning_rate": 7.505197564056637e-06, + "loss": 0.3309, + "step": 2539 + }, + { + "epoch": 0.35, + "grad_norm": 5.351950075833039, + "learning_rate": 7.5032489004083685e-06, + "loss": 0.3495, + "step": 2540 + }, + { + "epoch": 0.35, + "grad_norm": 2.3210695568815085, + "learning_rate": 7.501299729224718e-06, + "loss": 0.3696, + "step": 2541 + }, + { + "epoch": 0.35, + "grad_norm": 2.833862005000848, + "learning_rate": 7.499350050900877e-06, + "loss": 0.3156, + "step": 2542 + }, + { + "epoch": 0.35, + "grad_norm": 1.7687901150479477, + "learning_rate": 7.497399865832145e-06, + "loss": 0.3267, + "step": 2543 + }, + { + "epoch": 0.35, + "grad_norm": 2.2524398696592796, + "learning_rate": 7.495449174413925e-06, + "loss": 0.3621, + "step": 2544 + }, + { + "epoch": 0.35, + "grad_norm": 2.2984159690983836, + "learning_rate": 7.493497977041718e-06, + "loss": 0.3405, + "step": 2545 + }, + { + "epoch": 0.35, + "grad_norm": 3.927712960853225, + "learning_rate": 7.491546274111136e-06, + "loss": 0.3729, + "step": 2546 + }, + { + "epoch": 0.35, + "grad_norm": 3.074201270164846, + "learning_rate": 7.48959406601788e-06, + "loss": 0.3674, + "step": 2547 + }, + { + "epoch": 0.35, + "grad_norm": 2.102010268092819, + "learning_rate": 7.48764135315777e-06, + "loss": 0.3337, + "step": 2548 + }, + { + "epoch": 0.35, + "grad_norm": 3.221216030504068, + "learning_rate": 7.485688135926713e-06, + "loss": 0.342, + "step": 2549 + }, + { + "epoch": 0.35, + "grad_norm": 3.11174161437411, + "learning_rate": 7.483734414720729e-06, + "loss": 0.335, + "step": 2550 + }, + { + "epoch": 0.35, + "grad_norm": 1.9617469156801393, + "learning_rate": 7.481780189935937e-06, + "loss": 0.3336, + "step": 2551 + }, + { + "epoch": 0.35, + "grad_norm": 2.2176932798971247, + "learning_rate": 7.4798254619685516e-06, + "loss": 0.3545, + "step": 2552 + }, + { + "epoch": 0.35, + "grad_norm": 2.6588280183063073, + "learning_rate": 7.477870231214903e-06, + "loss": 0.3246, + "step": 2553 + }, + { + "epoch": 0.36, + "grad_norm": 2.2995199812534453, + "learning_rate": 7.47591449807141e-06, + "loss": 0.3298, + "step": 2554 + }, + { + "epoch": 0.36, + "grad_norm": 2.1030678220375627, + "learning_rate": 7.4739582629346e-06, + "loss": 0.351, + "step": 2555 + }, + { + "epoch": 0.36, + "grad_norm": 2.6645485075292177, + "learning_rate": 7.472001526201102e-06, + "loss": 0.3466, + "step": 2556 + }, + { + "epoch": 0.36, + "grad_norm": 2.008746693288422, + "learning_rate": 7.470044288267645e-06, + "loss": 0.3421, + "step": 2557 + }, + { + "epoch": 0.36, + "grad_norm": 2.117765615315883, + "learning_rate": 7.468086549531062e-06, + "loss": 0.3492, + "step": 2558 + }, + { + "epoch": 0.36, + "grad_norm": 2.597162724909585, + "learning_rate": 7.466128310388283e-06, + "loss": 0.344, + "step": 2559 + }, + { + "epoch": 0.36, + "grad_norm": 1.929809500730167, + "learning_rate": 7.464169571236344e-06, + "loss": 0.3297, + "step": 2560 + }, + { + "epoch": 0.36, + "grad_norm": 2.231040837117052, + "learning_rate": 7.462210332472379e-06, + "loss": 0.3609, + "step": 2561 + }, + { + "epoch": 0.36, + "grad_norm": 2.0209281344442287, + "learning_rate": 7.460250594493628e-06, + "loss": 0.3435, + "step": 2562 + }, + { + "epoch": 0.36, + "grad_norm": 2.189003387231218, + "learning_rate": 7.4582903576974265e-06, + "loss": 0.306, + "step": 2563 + }, + { + "epoch": 0.36, + "grad_norm": 2.568049058437364, + "learning_rate": 7.456329622481214e-06, + "loss": 0.3456, + "step": 2564 + }, + { + "epoch": 0.36, + "grad_norm": 3.6989052370984927, + "learning_rate": 7.454368389242533e-06, + "loss": 0.3463, + "step": 2565 + }, + { + "epoch": 0.36, + "grad_norm": 2.742895615248565, + "learning_rate": 7.452406658379024e-06, + "loss": 0.3433, + "step": 2566 + }, + { + "epoch": 0.36, + "grad_norm": 2.4160981345692814, + "learning_rate": 7.450444430288427e-06, + "loss": 0.3411, + "step": 2567 + }, + { + "epoch": 0.36, + "grad_norm": 2.6251805779237114, + "learning_rate": 7.448481705368588e-06, + "loss": 0.3454, + "step": 2568 + }, + { + "epoch": 0.36, + "grad_norm": 1.742375392880048, + "learning_rate": 7.446518484017448e-06, + "loss": 0.3442, + "step": 2569 + }, + { + "epoch": 0.36, + "grad_norm": 1.8597156108124988, + "learning_rate": 7.444554766633054e-06, + "loss": 0.3403, + "step": 2570 + }, + { + "epoch": 0.36, + "grad_norm": 2.6775095553664086, + "learning_rate": 7.4425905536135505e-06, + "loss": 0.3251, + "step": 2571 + }, + { + "epoch": 0.36, + "grad_norm": 2.2619418913706753, + "learning_rate": 7.440625845357182e-06, + "loss": 0.32, + "step": 2572 + }, + { + "epoch": 0.36, + "grad_norm": 2.3753027867828678, + "learning_rate": 7.4386606422622945e-06, + "loss": 0.3499, + "step": 2573 + }, + { + "epoch": 0.36, + "grad_norm": 2.9584432852246274, + "learning_rate": 7.436694944727335e-06, + "loss": 0.3557, + "step": 2574 + }, + { + "epoch": 0.36, + "grad_norm": 27.663332788086908, + "learning_rate": 7.43472875315085e-06, + "loss": 0.3391, + "step": 2575 + }, + { + "epoch": 0.36, + "grad_norm": 2.1180695656502153, + "learning_rate": 7.432762067931485e-06, + "loss": 0.3503, + "step": 2576 + }, + { + "epoch": 0.36, + "grad_norm": 1.705168965495483, + "learning_rate": 7.430794889467988e-06, + "loss": 0.3145, + "step": 2577 + }, + { + "epoch": 0.36, + "grad_norm": 1.8213252122990466, + "learning_rate": 7.428827218159205e-06, + "loss": 0.3377, + "step": 2578 + }, + { + "epoch": 0.36, + "grad_norm": 3.187272795376978, + "learning_rate": 7.426859054404084e-06, + "loss": 0.3343, + "step": 2579 + }, + { + "epoch": 0.36, + "grad_norm": 2.5494779498245834, + "learning_rate": 7.4248903986016695e-06, + "loss": 0.3425, + "step": 2580 + }, + { + "epoch": 0.36, + "grad_norm": 2.1877419551040465, + "learning_rate": 7.422921251151109e-06, + "loss": 0.3529, + "step": 2581 + }, + { + "epoch": 0.36, + "grad_norm": 2.3013563500225067, + "learning_rate": 7.420951612451648e-06, + "loss": 0.3651, + "step": 2582 + }, + { + "epoch": 0.36, + "grad_norm": 1.9474773461894186, + "learning_rate": 7.418981482902632e-06, + "loss": 0.3529, + "step": 2583 + }, + { + "epoch": 0.36, + "grad_norm": 2.4114807281118082, + "learning_rate": 7.4170108629035064e-06, + "loss": 0.3572, + "step": 2584 + }, + { + "epoch": 0.36, + "grad_norm": 3.490948987207873, + "learning_rate": 7.415039752853816e-06, + "loss": 0.3297, + "step": 2585 + }, + { + "epoch": 0.36, + "grad_norm": 3.8569932570340413, + "learning_rate": 7.413068153153205e-06, + "loss": 0.3484, + "step": 2586 + }, + { + "epoch": 0.36, + "grad_norm": 2.2059543309242873, + "learning_rate": 7.411096064201414e-06, + "loss": 0.3229, + "step": 2587 + }, + { + "epoch": 0.36, + "grad_norm": 2.0179358480862075, + "learning_rate": 7.409123486398288e-06, + "loss": 0.3123, + "step": 2588 + }, + { + "epoch": 0.36, + "grad_norm": 3.0000541471456637, + "learning_rate": 7.40715042014377e-06, + "loss": 0.3503, + "step": 2589 + }, + { + "epoch": 0.36, + "grad_norm": 1.9898283004482202, + "learning_rate": 7.405176865837894e-06, + "loss": 0.348, + "step": 2590 + }, + { + "epoch": 0.36, + "grad_norm": 2.176566774984802, + "learning_rate": 7.403202823880807e-06, + "loss": 0.3155, + "step": 2591 + }, + { + "epoch": 0.36, + "grad_norm": 2.94715148243927, + "learning_rate": 7.401228294672743e-06, + "loss": 0.3426, + "step": 2592 + }, + { + "epoch": 0.36, + "grad_norm": 2.0646333259024963, + "learning_rate": 7.39925327861404e-06, + "loss": 0.3546, + "step": 2593 + }, + { + "epoch": 0.36, + "grad_norm": 3.6607023949608415, + "learning_rate": 7.397277776105134e-06, + "loss": 0.3214, + "step": 2594 + }, + { + "epoch": 0.36, + "grad_norm": 2.2254981562752363, + "learning_rate": 7.395301787546558e-06, + "loss": 0.3404, + "step": 2595 + }, + { + "epoch": 0.36, + "grad_norm": 1.8896761074661754, + "learning_rate": 7.393325313338949e-06, + "loss": 0.3365, + "step": 2596 + }, + { + "epoch": 0.36, + "grad_norm": 2.219711749802034, + "learning_rate": 7.391348353883035e-06, + "loss": 0.3349, + "step": 2597 + }, + { + "epoch": 0.36, + "grad_norm": 2.3472628313741994, + "learning_rate": 7.389370909579646e-06, + "loss": 0.3713, + "step": 2598 + }, + { + "epoch": 0.36, + "grad_norm": 2.6335094516879867, + "learning_rate": 7.3873929808297096e-06, + "loss": 0.3367, + "step": 2599 + }, + { + "epoch": 0.36, + "grad_norm": 2.370210565285639, + "learning_rate": 7.385414568034256e-06, + "loss": 0.3362, + "step": 2600 + }, + { + "epoch": 0.36, + "grad_norm": 2.1776043849149604, + "learning_rate": 7.383435671594405e-06, + "loss": 0.3242, + "step": 2601 + }, + { + "epoch": 0.36, + "grad_norm": 2.3244342735075576, + "learning_rate": 7.381456291911383e-06, + "loss": 0.3278, + "step": 2602 + }, + { + "epoch": 0.36, + "grad_norm": 1.933053913188878, + "learning_rate": 7.379476429386507e-06, + "loss": 0.3347, + "step": 2603 + }, + { + "epoch": 0.36, + "grad_norm": 1.8753460535022637, + "learning_rate": 7.377496084421199e-06, + "loss": 0.3357, + "step": 2604 + }, + { + "epoch": 0.36, + "grad_norm": 1.9533539994792384, + "learning_rate": 7.375515257416973e-06, + "loss": 0.3458, + "step": 2605 + }, + { + "epoch": 0.36, + "grad_norm": 3.4712042306060513, + "learning_rate": 7.373533948775444e-06, + "loss": 0.3523, + "step": 2606 + }, + { + "epoch": 0.36, + "grad_norm": 2.4677862106969433, + "learning_rate": 7.371552158898322e-06, + "loss": 0.3641, + "step": 2607 + }, + { + "epoch": 0.36, + "grad_norm": 2.281253635827161, + "learning_rate": 7.369569888187418e-06, + "loss": 0.3309, + "step": 2608 + }, + { + "epoch": 0.36, + "grad_norm": 3.251204320719602, + "learning_rate": 7.3675871370446384e-06, + "loss": 0.3527, + "step": 2609 + }, + { + "epoch": 0.36, + "grad_norm": 2.2221567412070473, + "learning_rate": 7.365603905871988e-06, + "loss": 0.3611, + "step": 2610 + }, + { + "epoch": 0.36, + "grad_norm": 2.3181626720345827, + "learning_rate": 7.363620195071566e-06, + "loss": 0.3334, + "step": 2611 + }, + { + "epoch": 0.36, + "grad_norm": 3.0806870492638776, + "learning_rate": 7.361636005045571e-06, + "loss": 0.351, + "step": 2612 + }, + { + "epoch": 0.36, + "grad_norm": 2.832893801647765, + "learning_rate": 7.359651336196302e-06, + "loss": 0.3175, + "step": 2613 + }, + { + "epoch": 0.36, + "grad_norm": 2.0998820882358538, + "learning_rate": 7.3576661889261505e-06, + "loss": 0.3502, + "step": 2614 + }, + { + "epoch": 0.36, + "grad_norm": 2.537347450376591, + "learning_rate": 7.355680563637606e-06, + "loss": 0.3504, + "step": 2615 + }, + { + "epoch": 0.36, + "grad_norm": 2.6764807123158496, + "learning_rate": 7.353694460733254e-06, + "loss": 0.3238, + "step": 2616 + }, + { + "epoch": 0.36, + "grad_norm": 2.7430274410945867, + "learning_rate": 7.351707880615781e-06, + "loss": 0.3511, + "step": 2617 + }, + { + "epoch": 0.36, + "grad_norm": 2.8398446901177663, + "learning_rate": 7.349720823687964e-06, + "loss": 0.3263, + "step": 2618 + }, + { + "epoch": 0.36, + "grad_norm": 1.8960695978308304, + "learning_rate": 7.347733290352685e-06, + "loss": 0.3069, + "step": 2619 + }, + { + "epoch": 0.36, + "grad_norm": 2.920019116192516, + "learning_rate": 7.345745281012915e-06, + "loss": 0.3117, + "step": 2620 + }, + { + "epoch": 0.36, + "grad_norm": 2.128559668955168, + "learning_rate": 7.343756796071722e-06, + "loss": 0.3425, + "step": 2621 + }, + { + "epoch": 0.36, + "grad_norm": 2.4176690836301566, + "learning_rate": 7.341767835932277e-06, + "loss": 0.3312, + "step": 2622 + }, + { + "epoch": 0.36, + "grad_norm": 2.124723242736209, + "learning_rate": 7.339778400997838e-06, + "loss": 0.3392, + "step": 2623 + }, + { + "epoch": 0.36, + "grad_norm": 2.250128379370896, + "learning_rate": 7.337788491671769e-06, + "loss": 0.3433, + "step": 2624 + }, + { + "epoch": 0.36, + "grad_norm": 2.746276303633892, + "learning_rate": 7.335798108357522e-06, + "loss": 0.3656, + "step": 2625 + }, + { + "epoch": 0.37, + "grad_norm": 2.3758806623595703, + "learning_rate": 7.3338072514586515e-06, + "loss": 0.3392, + "step": 2626 + }, + { + "epoch": 0.37, + "grad_norm": 2.987930896711612, + "learning_rate": 7.3318159213788045e-06, + "loss": 0.3486, + "step": 2627 + }, + { + "epoch": 0.37, + "grad_norm": 2.5307781656982926, + "learning_rate": 7.329824118521721e-06, + "loss": 0.3114, + "step": 2628 + }, + { + "epoch": 0.37, + "grad_norm": 2.295594191469054, + "learning_rate": 7.327831843291245e-06, + "loss": 0.3472, + "step": 2629 + }, + { + "epoch": 0.37, + "grad_norm": 2.6771239590101263, + "learning_rate": 7.325839096091307e-06, + "loss": 0.3372, + "step": 2630 + }, + { + "epoch": 0.37, + "grad_norm": 6.163210661945434, + "learning_rate": 7.323845877325944e-06, + "loss": 0.3439, + "step": 2631 + }, + { + "epoch": 0.37, + "grad_norm": 2.334790769061295, + "learning_rate": 7.3218521873992765e-06, + "loss": 0.3354, + "step": 2632 + }, + { + "epoch": 0.37, + "grad_norm": 1.9916530318189398, + "learning_rate": 7.3198580267155285e-06, + "loss": 0.3506, + "step": 2633 + }, + { + "epoch": 0.37, + "grad_norm": 2.1601533375903093, + "learning_rate": 7.317863395679019e-06, + "loss": 0.33, + "step": 2634 + }, + { + "epoch": 0.37, + "grad_norm": 2.61598114392816, + "learning_rate": 7.31586829469416e-06, + "loss": 0.3227, + "step": 2635 + }, + { + "epoch": 0.37, + "grad_norm": 4.516730005951569, + "learning_rate": 7.313872724165459e-06, + "loss": 0.3221, + "step": 2636 + }, + { + "epoch": 0.37, + "grad_norm": 2.8023836011609005, + "learning_rate": 7.311876684497517e-06, + "loss": 0.3206, + "step": 2637 + }, + { + "epoch": 0.37, + "grad_norm": 2.605842834855539, + "learning_rate": 7.309880176095036e-06, + "loss": 0.3677, + "step": 2638 + }, + { + "epoch": 0.37, + "grad_norm": 2.178079142298736, + "learning_rate": 7.307883199362808e-06, + "loss": 0.3312, + "step": 2639 + }, + { + "epoch": 0.37, + "grad_norm": 2.605923411241691, + "learning_rate": 7.305885754705721e-06, + "loss": 0.328, + "step": 2640 + }, + { + "epoch": 0.37, + "grad_norm": 2.884296257182556, + "learning_rate": 7.3038878425287605e-06, + "loss": 0.3552, + "step": 2641 + }, + { + "epoch": 0.37, + "grad_norm": 2.119333418699514, + "learning_rate": 7.301889463237e-06, + "loss": 0.3531, + "step": 2642 + }, + { + "epoch": 0.37, + "grad_norm": 2.1473792214359664, + "learning_rate": 7.299890617235616e-06, + "loss": 0.3367, + "step": 2643 + }, + { + "epoch": 0.37, + "grad_norm": 2.3027517842281364, + "learning_rate": 7.297891304929876e-06, + "loss": 0.3492, + "step": 2644 + }, + { + "epoch": 0.37, + "grad_norm": 7.4317441762429075, + "learning_rate": 7.29589152672514e-06, + "loss": 0.3296, + "step": 2645 + }, + { + "epoch": 0.37, + "grad_norm": 3.2069809021829117, + "learning_rate": 7.293891283026865e-06, + "loss": 0.3363, + "step": 2646 + }, + { + "epoch": 0.37, + "grad_norm": 2.1529845981174547, + "learning_rate": 7.291890574240601e-06, + "loss": 0.3338, + "step": 2647 + }, + { + "epoch": 0.37, + "grad_norm": 2.2806054480905873, + "learning_rate": 7.289889400771996e-06, + "loss": 0.331, + "step": 2648 + }, + { + "epoch": 0.37, + "grad_norm": 2.0571721172655515, + "learning_rate": 7.287887763026786e-06, + "loss": 0.3621, + "step": 2649 + }, + { + "epoch": 0.37, + "grad_norm": 2.1457836908447705, + "learning_rate": 7.285885661410805e-06, + "loss": 0.3159, + "step": 2650 + }, + { + "epoch": 0.37, + "grad_norm": 1.8564133737411816, + "learning_rate": 7.28388309632998e-06, + "loss": 0.324, + "step": 2651 + }, + { + "epoch": 0.37, + "grad_norm": 2.911434746325424, + "learning_rate": 7.281880068190335e-06, + "loss": 0.3658, + "step": 2652 + }, + { + "epoch": 0.37, + "grad_norm": 2.165933679917731, + "learning_rate": 7.279876577397984e-06, + "loss": 0.3436, + "step": 2653 + }, + { + "epoch": 0.37, + "grad_norm": 3.9784554102349743, + "learning_rate": 7.277872624359134e-06, + "loss": 0.3234, + "step": 2654 + }, + { + "epoch": 0.37, + "grad_norm": 2.354783931398327, + "learning_rate": 7.275868209480092e-06, + "loss": 0.3418, + "step": 2655 + }, + { + "epoch": 0.37, + "grad_norm": 2.049438141459904, + "learning_rate": 7.273863333167251e-06, + "loss": 0.319, + "step": 2656 + }, + { + "epoch": 0.37, + "grad_norm": 1.8959108138922038, + "learning_rate": 7.2718579958271005e-06, + "loss": 0.3078, + "step": 2657 + }, + { + "epoch": 0.37, + "grad_norm": 3.4289723469009004, + "learning_rate": 7.269852197866227e-06, + "loss": 0.3279, + "step": 2658 + }, + { + "epoch": 0.37, + "grad_norm": 2.803317706448508, + "learning_rate": 7.267845939691305e-06, + "loss": 0.3838, + "step": 2659 + }, + { + "epoch": 0.37, + "grad_norm": 2.01132795985648, + "learning_rate": 7.265839221709105e-06, + "loss": 0.3003, + "step": 2660 + }, + { + "epoch": 0.37, + "grad_norm": 6.003665158215262, + "learning_rate": 7.2638320443264935e-06, + "loss": 0.343, + "step": 2661 + }, + { + "epoch": 0.37, + "grad_norm": 3.303491728623826, + "learning_rate": 7.261824407950423e-06, + "loss": 0.3421, + "step": 2662 + }, + { + "epoch": 0.37, + "grad_norm": 2.6966818519035125, + "learning_rate": 7.259816312987945e-06, + "loss": 0.3536, + "step": 2663 + }, + { + "epoch": 0.37, + "grad_norm": 2.253214384285918, + "learning_rate": 7.2578077598462e-06, + "loss": 0.3522, + "step": 2664 + }, + { + "epoch": 0.37, + "grad_norm": 3.121917248787243, + "learning_rate": 7.255798748932428e-06, + "loss": 0.3122, + "step": 2665 + }, + { + "epoch": 0.37, + "grad_norm": 3.0281932276045667, + "learning_rate": 7.253789280653953e-06, + "loss": 0.3518, + "step": 2666 + }, + { + "epoch": 0.37, + "grad_norm": 2.8863046108449057, + "learning_rate": 7.251779355418198e-06, + "loss": 0.3611, + "step": 2667 + }, + { + "epoch": 0.37, + "grad_norm": 1.8550726407262277, + "learning_rate": 7.249768973632675e-06, + "loss": 0.3226, + "step": 2668 + }, + { + "epoch": 0.37, + "grad_norm": 1.9512985603515125, + "learning_rate": 7.247758135704993e-06, + "loss": 0.3503, + "step": 2669 + }, + { + "epoch": 0.37, + "grad_norm": 2.6412755162105093, + "learning_rate": 7.245746842042849e-06, + "loss": 0.3521, + "step": 2670 + }, + { + "epoch": 0.37, + "grad_norm": 2.1596019214654683, + "learning_rate": 7.243735093054033e-06, + "loss": 0.361, + "step": 2671 + }, + { + "epoch": 0.37, + "grad_norm": 2.643089245881794, + "learning_rate": 7.241722889146431e-06, + "loss": 0.3422, + "step": 2672 + }, + { + "epoch": 0.37, + "grad_norm": 3.018663435055607, + "learning_rate": 7.239710230728017e-06, + "loss": 0.3255, + "step": 2673 + }, + { + "epoch": 0.37, + "grad_norm": 3.2377985225596646, + "learning_rate": 7.23769711820686e-06, + "loss": 0.3474, + "step": 2674 + }, + { + "epoch": 0.37, + "grad_norm": 2.3851163172148833, + "learning_rate": 7.235683551991119e-06, + "loss": 0.3171, + "step": 2675 + }, + { + "epoch": 0.37, + "grad_norm": 1.7431903795033536, + "learning_rate": 7.233669532489046e-06, + "loss": 0.3248, + "step": 2676 + }, + { + "epoch": 0.37, + "grad_norm": 2.933366200167332, + "learning_rate": 7.231655060108984e-06, + "loss": 0.3377, + "step": 2677 + }, + { + "epoch": 0.37, + "grad_norm": 2.570240719408544, + "learning_rate": 7.229640135259371e-06, + "loss": 0.3669, + "step": 2678 + }, + { + "epoch": 0.37, + "grad_norm": 2.1064852463576687, + "learning_rate": 7.227624758348732e-06, + "loss": 0.3345, + "step": 2679 + }, + { + "epoch": 0.37, + "grad_norm": 1.8232780133831519, + "learning_rate": 7.225608929785688e-06, + "loss": 0.3141, + "step": 2680 + }, + { + "epoch": 0.37, + "grad_norm": 5.978096348461096, + "learning_rate": 7.223592649978948e-06, + "loss": 0.3444, + "step": 2681 + }, + { + "epoch": 0.37, + "grad_norm": 2.5662259964013137, + "learning_rate": 7.221575919337314e-06, + "loss": 0.3283, + "step": 2682 + }, + { + "epoch": 0.37, + "grad_norm": 2.3965695565450904, + "learning_rate": 7.21955873826968e-06, + "loss": 0.3432, + "step": 2683 + }, + { + "epoch": 0.37, + "grad_norm": 2.3026206765486794, + "learning_rate": 7.217541107185032e-06, + "loss": 0.3447, + "step": 2684 + }, + { + "epoch": 0.37, + "grad_norm": 2.3875266345158477, + "learning_rate": 7.215523026492444e-06, + "loss": 0.3511, + "step": 2685 + }, + { + "epoch": 0.37, + "grad_norm": 2.1875575066601134, + "learning_rate": 7.213504496601085e-06, + "loss": 0.3394, + "step": 2686 + }, + { + "epoch": 0.37, + "grad_norm": 2.414120640072233, + "learning_rate": 7.211485517920212e-06, + "loss": 0.3615, + "step": 2687 + }, + { + "epoch": 0.37, + "grad_norm": 2.0469836775453008, + "learning_rate": 7.209466090859173e-06, + "loss": 0.3462, + "step": 2688 + }, + { + "epoch": 0.37, + "grad_norm": 2.3645737506161284, + "learning_rate": 7.207446215827412e-06, + "loss": 0.3686, + "step": 2689 + }, + { + "epoch": 0.37, + "grad_norm": 2.174987131446165, + "learning_rate": 7.2054258932344565e-06, + "loss": 0.3741, + "step": 2690 + }, + { + "epoch": 0.37, + "grad_norm": 1.7018571354005316, + "learning_rate": 7.20340512348993e-06, + "loss": 0.3421, + "step": 2691 + }, + { + "epoch": 0.37, + "grad_norm": 2.14809027782111, + "learning_rate": 7.201383907003545e-06, + "loss": 0.3387, + "step": 2692 + }, + { + "epoch": 0.37, + "grad_norm": 2.4954486364856616, + "learning_rate": 7.199362244185104e-06, + "loss": 0.365, + "step": 2693 + }, + { + "epoch": 0.37, + "grad_norm": 13.713829725312833, + "learning_rate": 7.1973401354445e-06, + "loss": 0.3285, + "step": 2694 + }, + { + "epoch": 0.37, + "grad_norm": 1.9095467525283811, + "learning_rate": 7.195317581191719e-06, + "loss": 0.3246, + "step": 2695 + }, + { + "epoch": 0.37, + "grad_norm": 2.830583442122394, + "learning_rate": 7.193294581836833e-06, + "loss": 0.3388, + "step": 2696 + }, + { + "epoch": 0.37, + "grad_norm": 4.60155440933869, + "learning_rate": 7.1912711377900055e-06, + "loss": 0.3562, + "step": 2697 + }, + { + "epoch": 0.38, + "grad_norm": 2.2399675372402745, + "learning_rate": 7.189247249461494e-06, + "loss": 0.3527, + "step": 2698 + }, + { + "epoch": 0.38, + "grad_norm": 1.866423561851487, + "learning_rate": 7.187222917261641e-06, + "loss": 0.3551, + "step": 2699 + }, + { + "epoch": 0.38, + "grad_norm": 1.966715014205407, + "learning_rate": 7.1851981416008845e-06, + "loss": 0.3407, + "step": 2700 + }, + { + "epoch": 0.38, + "grad_norm": 2.388790499921396, + "learning_rate": 7.183172922889745e-06, + "loss": 0.3605, + "step": 2701 + }, + { + "epoch": 0.38, + "grad_norm": 3.2457944691619067, + "learning_rate": 7.181147261538837e-06, + "loss": 0.3373, + "step": 2702 + }, + { + "epoch": 0.38, + "grad_norm": 2.2741141124588906, + "learning_rate": 7.179121157958869e-06, + "loss": 0.3352, + "step": 2703 + }, + { + "epoch": 0.38, + "grad_norm": 1.9311948327597575, + "learning_rate": 7.177094612560631e-06, + "loss": 0.3035, + "step": 2704 + }, + { + "epoch": 0.38, + "grad_norm": 3.4891616146835474, + "learning_rate": 7.17506762575501e-06, + "loss": 0.3437, + "step": 2705 + }, + { + "epoch": 0.38, + "grad_norm": 4.146840035505368, + "learning_rate": 7.173040197952974e-06, + "loss": 0.3489, + "step": 2706 + }, + { + "epoch": 0.38, + "grad_norm": 2.8746502763737696, + "learning_rate": 7.171012329565589e-06, + "loss": 0.3302, + "step": 2707 + }, + { + "epoch": 0.38, + "grad_norm": 2.187484928701156, + "learning_rate": 7.1689840210040055e-06, + "loss": 0.3591, + "step": 2708 + }, + { + "epoch": 0.38, + "grad_norm": 2.043962637988726, + "learning_rate": 7.166955272679464e-06, + "loss": 0.3178, + "step": 2709 + }, + { + "epoch": 0.38, + "grad_norm": 1.9845222916937983, + "learning_rate": 7.164926085003296e-06, + "loss": 0.3133, + "step": 2710 + }, + { + "epoch": 0.38, + "grad_norm": 2.5470334123765457, + "learning_rate": 7.162896458386919e-06, + "loss": 0.339, + "step": 2711 + }, + { + "epoch": 0.38, + "grad_norm": 2.044265478957731, + "learning_rate": 7.160866393241843e-06, + "loss": 0.3165, + "step": 2712 + }, + { + "epoch": 0.38, + "grad_norm": 2.5520767031660045, + "learning_rate": 7.158835889979663e-06, + "loss": 0.3204, + "step": 2713 + }, + { + "epoch": 0.38, + "grad_norm": 2.92946029069037, + "learning_rate": 7.156804949012066e-06, + "loss": 0.3333, + "step": 2714 + }, + { + "epoch": 0.38, + "grad_norm": 2.339487787980913, + "learning_rate": 7.154773570750826e-06, + "loss": 0.3179, + "step": 2715 + }, + { + "epoch": 0.38, + "grad_norm": 5.046560302440434, + "learning_rate": 7.152741755607807e-06, + "loss": 0.3268, + "step": 2716 + }, + { + "epoch": 0.38, + "grad_norm": 1.778220620375107, + "learning_rate": 7.1507095039949624e-06, + "loss": 0.3149, + "step": 2717 + }, + { + "epoch": 0.38, + "grad_norm": 3.1924129916390727, + "learning_rate": 7.148676816324329e-06, + "loss": 0.3543, + "step": 2718 + }, + { + "epoch": 0.38, + "grad_norm": 2.3372302115977606, + "learning_rate": 7.146643693008039e-06, + "loss": 0.3382, + "step": 2719 + }, + { + "epoch": 0.38, + "grad_norm": 2.3384091048878552, + "learning_rate": 7.1446101344583065e-06, + "loss": 0.342, + "step": 2720 + }, + { + "epoch": 0.38, + "grad_norm": 3.7018120727733477, + "learning_rate": 7.142576141087438e-06, + "loss": 0.3241, + "step": 2721 + }, + { + "epoch": 0.38, + "grad_norm": 2.347938890548561, + "learning_rate": 7.140541713307829e-06, + "loss": 0.3288, + "step": 2722 + }, + { + "epoch": 0.38, + "grad_norm": 3.944513301188907, + "learning_rate": 7.138506851531959e-06, + "loss": 0.3332, + "step": 2723 + }, + { + "epoch": 0.38, + "grad_norm": 2.432667341847325, + "learning_rate": 7.1364715561723965e-06, + "loss": 0.3297, + "step": 2724 + }, + { + "epoch": 0.38, + "grad_norm": 2.5216066175603045, + "learning_rate": 7.134435827641801e-06, + "loss": 0.3339, + "step": 2725 + }, + { + "epoch": 0.38, + "grad_norm": 2.453699017488778, + "learning_rate": 7.132399666352918e-06, + "loss": 0.3453, + "step": 2726 + }, + { + "epoch": 0.38, + "grad_norm": 2.3475035182231414, + "learning_rate": 7.13036307271858e-06, + "loss": 0.3333, + "step": 2727 + }, + { + "epoch": 0.38, + "grad_norm": 2.114379880393149, + "learning_rate": 7.1283260471517054e-06, + "loss": 0.3364, + "step": 2728 + }, + { + "epoch": 0.38, + "grad_norm": 2.5690061089361254, + "learning_rate": 7.126288590065309e-06, + "loss": 0.3471, + "step": 2729 + }, + { + "epoch": 0.38, + "grad_norm": 2.689952235420165, + "learning_rate": 7.124250701872479e-06, + "loss": 0.3414, + "step": 2730 + }, + { + "epoch": 0.38, + "grad_norm": 2.797134890065903, + "learning_rate": 7.1222123829864025e-06, + "loss": 0.3224, + "step": 2731 + }, + { + "epoch": 0.38, + "grad_norm": 2.2597578079372362, + "learning_rate": 7.12017363382035e-06, + "loss": 0.3103, + "step": 2732 + }, + { + "epoch": 0.38, + "grad_norm": 2.0722163221708185, + "learning_rate": 7.118134454787677e-06, + "loss": 0.3289, + "step": 2733 + }, + { + "epoch": 0.38, + "grad_norm": 3.090177583548337, + "learning_rate": 7.1160948463018305e-06, + "loss": 0.3112, + "step": 2734 + }, + { + "epoch": 0.38, + "grad_norm": 2.52116708205043, + "learning_rate": 7.114054808776343e-06, + "loss": 0.3374, + "step": 2735 + }, + { + "epoch": 0.38, + "grad_norm": 4.309356855080645, + "learning_rate": 7.112014342624834e-06, + "loss": 0.3283, + "step": 2736 + }, + { + "epoch": 0.38, + "grad_norm": 3.8873250801402017, + "learning_rate": 7.109973448261004e-06, + "loss": 0.3652, + "step": 2737 + }, + { + "epoch": 0.38, + "grad_norm": 1.988174886197088, + "learning_rate": 7.107932126098651e-06, + "loss": 0.3193, + "step": 2738 + }, + { + "epoch": 0.38, + "grad_norm": 2.2032180591424, + "learning_rate": 7.105890376551654e-06, + "loss": 0.3195, + "step": 2739 + }, + { + "epoch": 0.38, + "grad_norm": 2.57015939290619, + "learning_rate": 7.103848200033977e-06, + "loss": 0.3442, + "step": 2740 + }, + { + "epoch": 0.38, + "grad_norm": 2.7978314894036322, + "learning_rate": 7.101805596959675e-06, + "loss": 0.3434, + "step": 2741 + }, + { + "epoch": 0.38, + "grad_norm": 2.500529992021828, + "learning_rate": 7.099762567742884e-06, + "loss": 0.3381, + "step": 2742 + }, + { + "epoch": 0.38, + "grad_norm": 2.094292590928764, + "learning_rate": 7.097719112797833e-06, + "loss": 0.3257, + "step": 2743 + }, + { + "epoch": 0.38, + "grad_norm": 12.95518210546755, + "learning_rate": 7.095675232538831e-06, + "loss": 0.33, + "step": 2744 + }, + { + "epoch": 0.38, + "grad_norm": 3.2595629422516956, + "learning_rate": 7.093630927380277e-06, + "loss": 0.3364, + "step": 2745 + }, + { + "epoch": 0.38, + "grad_norm": 3.1310229326000325, + "learning_rate": 7.091586197736658e-06, + "loss": 0.3163, + "step": 2746 + }, + { + "epoch": 0.38, + "grad_norm": 2.162881835323285, + "learning_rate": 7.0895410440225386e-06, + "loss": 0.338, + "step": 2747 + }, + { + "epoch": 0.38, + "grad_norm": 2.0833111824275505, + "learning_rate": 7.087495466652581e-06, + "loss": 0.3397, + "step": 2748 + }, + { + "epoch": 0.38, + "grad_norm": 2.270557382500306, + "learning_rate": 7.085449466041523e-06, + "loss": 0.3414, + "step": 2749 + }, + { + "epoch": 0.38, + "grad_norm": 3.1016397630941226, + "learning_rate": 7.083403042604196e-06, + "loss": 0.3295, + "step": 2750 + }, + { + "epoch": 0.38, + "grad_norm": 2.560069420975474, + "learning_rate": 7.08135619675551e-06, + "loss": 0.3196, + "step": 2751 + }, + { + "epoch": 0.38, + "grad_norm": 2.4255744220172573, + "learning_rate": 7.079308928910467e-06, + "loss": 0.3414, + "step": 2752 + }, + { + "epoch": 0.38, + "grad_norm": 2.4823273975365305, + "learning_rate": 7.077261239484152e-06, + "loss": 0.3237, + "step": 2753 + }, + { + "epoch": 0.38, + "grad_norm": 3.7716889510014293, + "learning_rate": 7.075213128891733e-06, + "loss": 0.3563, + "step": 2754 + }, + { + "epoch": 0.38, + "grad_norm": 3.415401044393085, + "learning_rate": 7.07316459754847e-06, + "loss": 0.3339, + "step": 2755 + }, + { + "epoch": 0.38, + "grad_norm": 2.7605921052278393, + "learning_rate": 7.0711156458697e-06, + "loss": 0.3322, + "step": 2756 + }, + { + "epoch": 0.38, + "grad_norm": 2.338920249073949, + "learning_rate": 7.0690662742708515e-06, + "loss": 0.3536, + "step": 2757 + }, + { + "epoch": 0.38, + "grad_norm": 3.1819811722098588, + "learning_rate": 7.067016483167435e-06, + "loss": 0.3087, + "step": 2758 + }, + { + "epoch": 0.38, + "grad_norm": 2.881553506422599, + "learning_rate": 7.0649662729750465e-06, + "loss": 0.3421, + "step": 2759 + }, + { + "epoch": 0.38, + "grad_norm": 2.3600579486117184, + "learning_rate": 7.062915644109372e-06, + "loss": 0.3481, + "step": 2760 + }, + { + "epoch": 0.38, + "grad_norm": 2.1379928217234485, + "learning_rate": 7.06086459698617e-06, + "loss": 0.3484, + "step": 2761 + }, + { + "epoch": 0.38, + "grad_norm": 3.4598137130980717, + "learning_rate": 7.0588131320212985e-06, + "loss": 0.3387, + "step": 2762 + }, + { + "epoch": 0.38, + "grad_norm": 10.48666626839247, + "learning_rate": 7.056761249630688e-06, + "loss": 0.3294, + "step": 2763 + }, + { + "epoch": 0.38, + "grad_norm": 2.2995596795637794, + "learning_rate": 7.054708950230364e-06, + "loss": 0.3447, + "step": 2764 + }, + { + "epoch": 0.38, + "grad_norm": 2.329790849450386, + "learning_rate": 7.05265623423643e-06, + "loss": 0.3211, + "step": 2765 + }, + { + "epoch": 0.38, + "grad_norm": 2.977679372107594, + "learning_rate": 7.050603102065074e-06, + "loss": 0.3642, + "step": 2766 + }, + { + "epoch": 0.38, + "grad_norm": 3.740755734932413, + "learning_rate": 7.04854955413257e-06, + "loss": 0.3406, + "step": 2767 + }, + { + "epoch": 0.38, + "grad_norm": 2.280314280079698, + "learning_rate": 7.046495590855278e-06, + "loss": 0.3462, + "step": 2768 + }, + { + "epoch": 0.38, + "grad_norm": 3.7967033330505644, + "learning_rate": 7.0444412126496385e-06, + "loss": 0.3451, + "step": 2769 + }, + { + "epoch": 0.39, + "grad_norm": 2.2008435688347174, + "learning_rate": 7.0423864199321814e-06, + "loss": 0.3424, + "step": 2770 + }, + { + "epoch": 0.39, + "grad_norm": 3.3179546254835777, + "learning_rate": 7.040331213119511e-06, + "loss": 0.3259, + "step": 2771 + }, + { + "epoch": 0.39, + "grad_norm": 2.6789770089544502, + "learning_rate": 7.0382755926283295e-06, + "loss": 0.3466, + "step": 2772 + }, + { + "epoch": 0.39, + "grad_norm": 2.5431104308871397, + "learning_rate": 7.03621955887541e-06, + "loss": 0.3084, + "step": 2773 + }, + { + "epoch": 0.39, + "grad_norm": 2.274892657289413, + "learning_rate": 7.034163112277618e-06, + "loss": 0.347, + "step": 2774 + }, + { + "epoch": 0.39, + "grad_norm": 3.5677102139832146, + "learning_rate": 7.0321062532518954e-06, + "loss": 0.3473, + "step": 2775 + }, + { + "epoch": 0.39, + "grad_norm": 2.7970582353702134, + "learning_rate": 7.030048982215275e-06, + "loss": 0.3439, + "step": 2776 + }, + { + "epoch": 0.39, + "grad_norm": 2.019555545435735, + "learning_rate": 7.027991299584869e-06, + "loss": 0.3429, + "step": 2777 + }, + { + "epoch": 0.39, + "grad_norm": 2.092131435420124, + "learning_rate": 7.025933205777873e-06, + "loss": 0.3543, + "step": 2778 + }, + { + "epoch": 0.39, + "grad_norm": 3.1485783671492187, + "learning_rate": 7.02387470121157e-06, + "loss": 0.3505, + "step": 2779 + }, + { + "epoch": 0.39, + "grad_norm": 2.554184192209953, + "learning_rate": 7.021815786303318e-06, + "loss": 0.3408, + "step": 2780 + }, + { + "epoch": 0.39, + "grad_norm": 2.398640905656172, + "learning_rate": 7.019756461470568e-06, + "loss": 0.3546, + "step": 2781 + }, + { + "epoch": 0.39, + "grad_norm": 2.2725931768486176, + "learning_rate": 7.017696727130847e-06, + "loss": 0.3364, + "step": 2782 + }, + { + "epoch": 0.39, + "grad_norm": 3.9641874381551303, + "learning_rate": 7.015636583701769e-06, + "loss": 0.3324, + "step": 2783 + }, + { + "epoch": 0.39, + "grad_norm": 4.397372976366579, + "learning_rate": 7.013576031601029e-06, + "loss": 0.3363, + "step": 2784 + }, + { + "epoch": 0.39, + "grad_norm": 1.97007910826929, + "learning_rate": 7.011515071246404e-06, + "loss": 0.323, + "step": 2785 + }, + { + "epoch": 0.39, + "grad_norm": 2.4626982717695465, + "learning_rate": 7.009453703055758e-06, + "loss": 0.3556, + "step": 2786 + }, + { + "epoch": 0.39, + "grad_norm": 2.6045273183572224, + "learning_rate": 7.007391927447032e-06, + "loss": 0.3395, + "step": 2787 + }, + { + "epoch": 0.39, + "grad_norm": 2.3598987572730983, + "learning_rate": 7.005329744838253e-06, + "loss": 0.3443, + "step": 2788 + }, + { + "epoch": 0.39, + "grad_norm": 2.142118499029603, + "learning_rate": 7.00326715564753e-06, + "loss": 0.3281, + "step": 2789 + }, + { + "epoch": 0.39, + "grad_norm": 2.0229401710962738, + "learning_rate": 7.0012041602930554e-06, + "loss": 0.304, + "step": 2790 + }, + { + "epoch": 0.39, + "grad_norm": 2.530777518054452, + "learning_rate": 6.999140759193103e-06, + "loss": 0.3258, + "step": 2791 + }, + { + "epoch": 0.39, + "grad_norm": 2.2557827922616895, + "learning_rate": 6.997076952766025e-06, + "loss": 0.3398, + "step": 2792 + }, + { + "epoch": 0.39, + "grad_norm": 2.778931479058019, + "learning_rate": 6.995012741430263e-06, + "loss": 0.328, + "step": 2793 + }, + { + "epoch": 0.39, + "grad_norm": 2.219949907922741, + "learning_rate": 6.992948125604339e-06, + "loss": 0.3406, + "step": 2794 + }, + { + "epoch": 0.39, + "grad_norm": 4.979331295635959, + "learning_rate": 6.99088310570685e-06, + "loss": 0.3146, + "step": 2795 + }, + { + "epoch": 0.39, + "grad_norm": 2.5987902932165166, + "learning_rate": 6.9888176821564846e-06, + "loss": 0.3721, + "step": 2796 + }, + { + "epoch": 0.39, + "grad_norm": 2.1117952777321736, + "learning_rate": 6.986751855372006e-06, + "loss": 0.3412, + "step": 2797 + }, + { + "epoch": 0.39, + "grad_norm": 3.8750986996186803, + "learning_rate": 6.984685625772264e-06, + "loss": 0.3409, + "step": 2798 + }, + { + "epoch": 0.39, + "grad_norm": 4.283106531090739, + "learning_rate": 6.982618993776187e-06, + "loss": 0.3434, + "step": 2799 + }, + { + "epoch": 0.39, + "grad_norm": 2.7734438043492924, + "learning_rate": 6.980551959802786e-06, + "loss": 0.3256, + "step": 2800 + }, + { + "epoch": 0.39, + "grad_norm": 3.0167846446779736, + "learning_rate": 6.978484524271154e-06, + "loss": 0.3256, + "step": 2801 + }, + { + "epoch": 0.39, + "grad_norm": 2.0440936305680886, + "learning_rate": 6.976416687600463e-06, + "loss": 0.3379, + "step": 2802 + }, + { + "epoch": 0.39, + "grad_norm": 2.428464814002338, + "learning_rate": 6.974348450209972e-06, + "loss": 0.3613, + "step": 2803 + }, + { + "epoch": 0.39, + "grad_norm": 2.711796444701489, + "learning_rate": 6.972279812519015e-06, + "loss": 0.3242, + "step": 2804 + }, + { + "epoch": 0.39, + "grad_norm": 3.584759696495985, + "learning_rate": 6.970210774947011e-06, + "loss": 0.3268, + "step": 2805 + }, + { + "epoch": 0.39, + "grad_norm": 3.774654052002191, + "learning_rate": 6.9681413379134545e-06, + "loss": 0.32, + "step": 2806 + }, + { + "epoch": 0.39, + "grad_norm": 2.8471219640425507, + "learning_rate": 6.96607150183793e-06, + "loss": 0.3329, + "step": 2807 + }, + { + "epoch": 0.39, + "grad_norm": 2.7738586595302395, + "learning_rate": 6.964001267140098e-06, + "loss": 0.308, + "step": 2808 + }, + { + "epoch": 0.39, + "grad_norm": 3.0339615218291454, + "learning_rate": 6.961930634239698e-06, + "loss": 0.3185, + "step": 2809 + }, + { + "epoch": 0.39, + "grad_norm": 2.0774197890855497, + "learning_rate": 6.9598596035565525e-06, + "loss": 0.3372, + "step": 2810 + }, + { + "epoch": 0.39, + "grad_norm": 2.252318147097241, + "learning_rate": 6.957788175510565e-06, + "loss": 0.3268, + "step": 2811 + }, + { + "epoch": 0.39, + "grad_norm": 2.3450249376151517, + "learning_rate": 6.9557163505217185e-06, + "loss": 0.3353, + "step": 2812 + }, + { + "epoch": 0.39, + "grad_norm": 2.812610924397102, + "learning_rate": 6.953644129010077e-06, + "loss": 0.3393, + "step": 2813 + }, + { + "epoch": 0.39, + "grad_norm": 3.5679797130149438, + "learning_rate": 6.951571511395785e-06, + "loss": 0.3229, + "step": 2814 + }, + { + "epoch": 0.39, + "grad_norm": 3.695850281019974, + "learning_rate": 6.949498498099067e-06, + "loss": 0.3283, + "step": 2815 + }, + { + "epoch": 0.39, + "grad_norm": 2.541110624093472, + "learning_rate": 6.947425089540226e-06, + "loss": 0.3484, + "step": 2816 + }, + { + "epoch": 0.39, + "grad_norm": 2.542811391212058, + "learning_rate": 6.945351286139652e-06, + "loss": 0.3136, + "step": 2817 + }, + { + "epoch": 0.39, + "grad_norm": 2.662644247120839, + "learning_rate": 6.943277088317805e-06, + "loss": 0.3361, + "step": 2818 + }, + { + "epoch": 0.39, + "grad_norm": 2.7405323274978506, + "learning_rate": 6.941202496495231e-06, + "loss": 0.3353, + "step": 2819 + }, + { + "epoch": 0.39, + "grad_norm": 2.2798697704669952, + "learning_rate": 6.939127511092556e-06, + "loss": 0.3749, + "step": 2820 + }, + { + "epoch": 0.39, + "grad_norm": 2.363641722498714, + "learning_rate": 6.937052132530484e-06, + "loss": 0.3376, + "step": 2821 + }, + { + "epoch": 0.39, + "grad_norm": 3.5568215569516237, + "learning_rate": 6.9349763612298015e-06, + "loss": 0.3301, + "step": 2822 + }, + { + "epoch": 0.39, + "grad_norm": 2.2649023109099886, + "learning_rate": 6.932900197611369e-06, + "loss": 0.3349, + "step": 2823 + }, + { + "epoch": 0.39, + "grad_norm": 2.6392056377432267, + "learning_rate": 6.930823642096133e-06, + "loss": 0.3327, + "step": 2824 + }, + { + "epoch": 0.39, + "grad_norm": 2.5646120488026862, + "learning_rate": 6.928746695105114e-06, + "loss": 0.3385, + "step": 2825 + }, + { + "epoch": 0.39, + "grad_norm": 2.4479488875413136, + "learning_rate": 6.926669357059416e-06, + "loss": 0.3331, + "step": 2826 + }, + { + "epoch": 0.39, + "grad_norm": 2.389562440667786, + "learning_rate": 6.92459162838022e-06, + "loss": 0.3376, + "step": 2827 + }, + { + "epoch": 0.39, + "grad_norm": 2.25696653143524, + "learning_rate": 6.922513509488787e-06, + "loss": 0.3472, + "step": 2828 + }, + { + "epoch": 0.39, + "grad_norm": 2.4305257316613957, + "learning_rate": 6.920435000806459e-06, + "loss": 0.3398, + "step": 2829 + }, + { + "epoch": 0.39, + "grad_norm": 2.347913001114315, + "learning_rate": 6.918356102754652e-06, + "loss": 0.3212, + "step": 2830 + }, + { + "epoch": 0.39, + "grad_norm": 2.471371310998165, + "learning_rate": 6.916276815754865e-06, + "loss": 0.3141, + "step": 2831 + }, + { + "epoch": 0.39, + "grad_norm": 2.70427116123238, + "learning_rate": 6.914197140228675e-06, + "loss": 0.3223, + "step": 2832 + }, + { + "epoch": 0.39, + "grad_norm": 2.356677785995407, + "learning_rate": 6.912117076597738e-06, + "loss": 0.3326, + "step": 2833 + }, + { + "epoch": 0.39, + "grad_norm": 2.9454661337828485, + "learning_rate": 6.910036625283789e-06, + "loss": 0.3507, + "step": 2834 + }, + { + "epoch": 0.39, + "grad_norm": 2.371202871122713, + "learning_rate": 6.907955786708639e-06, + "loss": 0.3297, + "step": 2835 + }, + { + "epoch": 0.39, + "grad_norm": 2.997819044950302, + "learning_rate": 6.905874561294179e-06, + "loss": 0.3353, + "step": 2836 + }, + { + "epoch": 0.39, + "grad_norm": 3.5488239319165107, + "learning_rate": 6.903792949462381e-06, + "loss": 0.3326, + "step": 2837 + }, + { + "epoch": 0.39, + "grad_norm": 2.73420027538348, + "learning_rate": 6.9017109516352924e-06, + "loss": 0.3163, + "step": 2838 + }, + { + "epoch": 0.39, + "grad_norm": 4.1748551927682955, + "learning_rate": 6.899628568235039e-06, + "loss": 0.3435, + "step": 2839 + }, + { + "epoch": 0.39, + "grad_norm": 2.271056722354926, + "learning_rate": 6.897545799683826e-06, + "loss": 0.3498, + "step": 2840 + }, + { + "epoch": 0.39, + "grad_norm": 4.564685660491595, + "learning_rate": 6.895462646403936e-06, + "loss": 0.3444, + "step": 2841 + }, + { + "epoch": 0.4, + "grad_norm": 9.624340675155947, + "learning_rate": 6.893379108817731e-06, + "loss": 0.3245, + "step": 2842 + }, + { + "epoch": 0.4, + "grad_norm": 2.0956016021903467, + "learning_rate": 6.891295187347648e-06, + "loss": 0.329, + "step": 2843 + }, + { + "epoch": 0.4, + "grad_norm": 3.379268090538354, + "learning_rate": 6.889210882416202e-06, + "loss": 0.3238, + "step": 2844 + }, + { + "epoch": 0.4, + "grad_norm": 2.004942453777219, + "learning_rate": 6.8871261944459896e-06, + "loss": 0.3396, + "step": 2845 + }, + { + "epoch": 0.4, + "grad_norm": 2.7540460232143955, + "learning_rate": 6.885041123859684e-06, + "loss": 0.3237, + "step": 2846 + }, + { + "epoch": 0.4, + "grad_norm": 3.607058899389013, + "learning_rate": 6.882955671080031e-06, + "loss": 0.3564, + "step": 2847 + }, + { + "epoch": 0.4, + "grad_norm": 2.7411747894959855, + "learning_rate": 6.88086983652986e-06, + "loss": 0.3469, + "step": 2848 + }, + { + "epoch": 0.4, + "grad_norm": 2.1127152594550562, + "learning_rate": 6.878783620632073e-06, + "loss": 0.3518, + "step": 2849 + }, + { + "epoch": 0.4, + "grad_norm": 3.114804733941137, + "learning_rate": 6.876697023809656e-06, + "loss": 0.3367, + "step": 2850 + }, + { + "epoch": 0.4, + "grad_norm": 6.9784709766161, + "learning_rate": 6.874610046485664e-06, + "loss": 0.3533, + "step": 2851 + }, + { + "epoch": 0.4, + "grad_norm": 3.1469052624951126, + "learning_rate": 6.872522689083234e-06, + "loss": 0.3125, + "step": 2852 + }, + { + "epoch": 0.4, + "grad_norm": 3.9393853884957397, + "learning_rate": 6.87043495202558e-06, + "loss": 0.3648, + "step": 2853 + }, + { + "epoch": 0.4, + "grad_norm": 2.553494093180967, + "learning_rate": 6.868346835735991e-06, + "loss": 0.3208, + "step": 2854 + }, + { + "epoch": 0.4, + "grad_norm": 2.410150441670187, + "learning_rate": 6.866258340637836e-06, + "loss": 0.3421, + "step": 2855 + }, + { + "epoch": 0.4, + "grad_norm": 2.7811511774341975, + "learning_rate": 6.864169467154555e-06, + "loss": 0.3414, + "step": 2856 + }, + { + "epoch": 0.4, + "grad_norm": 2.121433278613242, + "learning_rate": 6.862080215709672e-06, + "loss": 0.329, + "step": 2857 + }, + { + "epoch": 0.4, + "grad_norm": 2.414711233221827, + "learning_rate": 6.859990586726782e-06, + "loss": 0.3189, + "step": 2858 + }, + { + "epoch": 0.4, + "grad_norm": 2.963875325628373, + "learning_rate": 6.85790058062956e-06, + "loss": 0.3247, + "step": 2859 + }, + { + "epoch": 0.4, + "grad_norm": 2.2278154937676513, + "learning_rate": 6.855810197841757e-06, + "loss": 0.3343, + "step": 2860 + }, + { + "epoch": 0.4, + "grad_norm": 2.7720602288990706, + "learning_rate": 6.8537194387871965e-06, + "loss": 0.3536, + "step": 2861 + }, + { + "epoch": 0.4, + "grad_norm": 2.901198885220056, + "learning_rate": 6.851628303889782e-06, + "loss": 0.3292, + "step": 2862 + }, + { + "epoch": 0.4, + "grad_norm": 2.6330798380786007, + "learning_rate": 6.849536793573495e-06, + "loss": 0.3713, + "step": 2863 + }, + { + "epoch": 0.4, + "grad_norm": 2.152900316178203, + "learning_rate": 6.8474449082623885e-06, + "loss": 0.3092, + "step": 2864 + }, + { + "epoch": 0.4, + "grad_norm": 3.2412618284917243, + "learning_rate": 6.8453526483805955e-06, + "loss": 0.3466, + "step": 2865 + }, + { + "epoch": 0.4, + "grad_norm": 5.981495740785992, + "learning_rate": 6.8432600143523185e-06, + "loss": 0.3299, + "step": 2866 + }, + { + "epoch": 0.4, + "grad_norm": 2.120943101914962, + "learning_rate": 6.841167006601848e-06, + "loss": 0.3411, + "step": 2867 + }, + { + "epoch": 0.4, + "grad_norm": 2.500157515389801, + "learning_rate": 6.839073625553535e-06, + "loss": 0.3186, + "step": 2868 + }, + { + "epoch": 0.4, + "grad_norm": 2.207450475392362, + "learning_rate": 6.836979871631819e-06, + "loss": 0.3238, + "step": 2869 + }, + { + "epoch": 0.4, + "grad_norm": 2.198209809050516, + "learning_rate": 6.834885745261209e-06, + "loss": 0.3365, + "step": 2870 + }, + { + "epoch": 0.4, + "grad_norm": 2.212284491351657, + "learning_rate": 6.8327912468662875e-06, + "loss": 0.3145, + "step": 2871 + }, + { + "epoch": 0.4, + "grad_norm": 2.3838404443623205, + "learning_rate": 6.83069637687172e-06, + "loss": 0.3329, + "step": 2872 + }, + { + "epoch": 0.4, + "grad_norm": 2.756732137330007, + "learning_rate": 6.82860113570224e-06, + "loss": 0.3533, + "step": 2873 + }, + { + "epoch": 0.4, + "grad_norm": 4.370932930113278, + "learning_rate": 6.82650552378266e-06, + "loss": 0.323, + "step": 2874 + }, + { + "epoch": 0.4, + "grad_norm": 2.522139497597163, + "learning_rate": 6.824409541537864e-06, + "loss": 0.3238, + "step": 2875 + }, + { + "epoch": 0.4, + "grad_norm": 3.9469382398371318, + "learning_rate": 6.822313189392818e-06, + "loss": 0.3395, + "step": 2876 + }, + { + "epoch": 0.4, + "grad_norm": 2.821300397506404, + "learning_rate": 6.820216467772558e-06, + "loss": 0.3479, + "step": 2877 + }, + { + "epoch": 0.4, + "grad_norm": 3.2571360703658088, + "learning_rate": 6.818119377102192e-06, + "loss": 0.3474, + "step": 2878 + }, + { + "epoch": 0.4, + "grad_norm": 2.2602650477041926, + "learning_rate": 6.816021917806911e-06, + "loss": 0.3206, + "step": 2879 + }, + { + "epoch": 0.4, + "grad_norm": 2.257031157251308, + "learning_rate": 6.813924090311974e-06, + "loss": 0.3332, + "step": 2880 + }, + { + "epoch": 0.4, + "grad_norm": 2.433693610057187, + "learning_rate": 6.811825895042719e-06, + "loss": 0.3235, + "step": 2881 + }, + { + "epoch": 0.4, + "grad_norm": 2.535311928153873, + "learning_rate": 6.809727332424553e-06, + "loss": 0.3164, + "step": 2882 + }, + { + "epoch": 0.4, + "grad_norm": 2.274013612112582, + "learning_rate": 6.807628402882963e-06, + "loss": 0.3171, + "step": 2883 + }, + { + "epoch": 0.4, + "grad_norm": 2.356272317414083, + "learning_rate": 6.805529106843508e-06, + "loss": 0.3214, + "step": 2884 + }, + { + "epoch": 0.4, + "grad_norm": 2.4749261936939924, + "learning_rate": 6.8034294447318225e-06, + "loss": 0.3244, + "step": 2885 + }, + { + "epoch": 0.4, + "grad_norm": 2.138957174763733, + "learning_rate": 6.801329416973614e-06, + "loss": 0.3692, + "step": 2886 + }, + { + "epoch": 0.4, + "grad_norm": 1.9976675907087953, + "learning_rate": 6.799229023994665e-06, + "loss": 0.3077, + "step": 2887 + }, + { + "epoch": 0.4, + "grad_norm": 2.6512837259415445, + "learning_rate": 6.7971282662208295e-06, + "loss": 0.3417, + "step": 2888 + }, + { + "epoch": 0.4, + "grad_norm": 3.130986811022379, + "learning_rate": 6.795027144078039e-06, + "loss": 0.3409, + "step": 2889 + }, + { + "epoch": 0.4, + "grad_norm": 2.4253160264341327, + "learning_rate": 6.792925657992297e-06, + "loss": 0.3335, + "step": 2890 + }, + { + "epoch": 0.4, + "grad_norm": 4.091495574018917, + "learning_rate": 6.790823808389684e-06, + "loss": 0.347, + "step": 2891 + }, + { + "epoch": 0.4, + "grad_norm": 3.2827177036991464, + "learning_rate": 6.788721595696346e-06, + "loss": 0.3172, + "step": 2892 + }, + { + "epoch": 0.4, + "grad_norm": 2.31800985979083, + "learning_rate": 6.786619020338513e-06, + "loss": 0.311, + "step": 2893 + }, + { + "epoch": 0.4, + "grad_norm": 2.6872050155910925, + "learning_rate": 6.784516082742481e-06, + "loss": 0.3419, + "step": 2894 + }, + { + "epoch": 0.4, + "grad_norm": 3.204350551878068, + "learning_rate": 6.782412783334622e-06, + "loss": 0.3739, + "step": 2895 + }, + { + "epoch": 0.4, + "grad_norm": 3.6919166228702083, + "learning_rate": 6.780309122541382e-06, + "loss": 0.3303, + "step": 2896 + }, + { + "epoch": 0.4, + "grad_norm": 2.340663967874866, + "learning_rate": 6.778205100789279e-06, + "loss": 0.3529, + "step": 2897 + }, + { + "epoch": 0.4, + "grad_norm": 2.5076561423283206, + "learning_rate": 6.7761007185049075e-06, + "loss": 0.3423, + "step": 2898 + }, + { + "epoch": 0.4, + "grad_norm": 3.2026431745449533, + "learning_rate": 6.7739959761149285e-06, + "loss": 0.33, + "step": 2899 + }, + { + "epoch": 0.4, + "grad_norm": 5.461237995721089, + "learning_rate": 6.771890874046081e-06, + "loss": 0.3338, + "step": 2900 + }, + { + "epoch": 0.4, + "grad_norm": 4.217047855352703, + "learning_rate": 6.769785412725178e-06, + "loss": 0.3053, + "step": 2901 + }, + { + "epoch": 0.4, + "grad_norm": 3.2621348373731003, + "learning_rate": 6.767679592579101e-06, + "loss": 0.3232, + "step": 2902 + }, + { + "epoch": 0.4, + "grad_norm": 2.632314193670778, + "learning_rate": 6.765573414034808e-06, + "loss": 0.3242, + "step": 2903 + }, + { + "epoch": 0.4, + "grad_norm": 2.8286283632930833, + "learning_rate": 6.763466877519327e-06, + "loss": 0.3429, + "step": 2904 + }, + { + "epoch": 0.4, + "grad_norm": 2.1479824211253096, + "learning_rate": 6.76135998345976e-06, + "loss": 0.3194, + "step": 2905 + }, + { + "epoch": 0.4, + "grad_norm": 2.845603117849009, + "learning_rate": 6.75925273228328e-06, + "loss": 0.3494, + "step": 2906 + }, + { + "epoch": 0.4, + "grad_norm": 2.6978865404143897, + "learning_rate": 6.757145124417135e-06, + "loss": 0.3606, + "step": 2907 + }, + { + "epoch": 0.4, + "grad_norm": 7.699562838370047, + "learning_rate": 6.755037160288646e-06, + "loss": 0.3101, + "step": 2908 + }, + { + "epoch": 0.4, + "grad_norm": 2.2111601503232685, + "learning_rate": 6.752928840325198e-06, + "loss": 0.3452, + "step": 2909 + }, + { + "epoch": 0.4, + "grad_norm": 2.3837584678661123, + "learning_rate": 6.750820164954262e-06, + "loss": 0.3179, + "step": 2910 + }, + { + "epoch": 0.4, + "grad_norm": 2.6049549489322166, + "learning_rate": 6.748711134603366e-06, + "loss": 0.3634, + "step": 2911 + }, + { + "epoch": 0.4, + "grad_norm": 2.566914837705339, + "learning_rate": 6.746601749700124e-06, + "loss": 0.3551, + "step": 2912 + }, + { + "epoch": 0.4, + "grad_norm": 2.92971007334313, + "learning_rate": 6.744492010672208e-06, + "loss": 0.3443, + "step": 2913 + }, + { + "epoch": 0.41, + "grad_norm": 3.7501046721056768, + "learning_rate": 6.742381917947374e-06, + "loss": 0.3375, + "step": 2914 + }, + { + "epoch": 0.41, + "grad_norm": 2.634153507272374, + "learning_rate": 6.740271471953446e-06, + "loss": 0.3149, + "step": 2915 + }, + { + "epoch": 0.41, + "grad_norm": 2.209074809049828, + "learning_rate": 6.738160673118313e-06, + "loss": 0.3487, + "step": 2916 + }, + { + "epoch": 0.41, + "grad_norm": 5.524092597520985, + "learning_rate": 6.736049521869946e-06, + "loss": 0.3364, + "step": 2917 + }, + { + "epoch": 0.41, + "grad_norm": 2.4946209545198976, + "learning_rate": 6.733938018636378e-06, + "loss": 0.3245, + "step": 2918 + }, + { + "epoch": 0.41, + "grad_norm": 1.885983088886619, + "learning_rate": 6.7318261638457215e-06, + "loss": 0.3127, + "step": 2919 + }, + { + "epoch": 0.41, + "grad_norm": 3.0249555580527394, + "learning_rate": 6.729713957926154e-06, + "loss": 0.3668, + "step": 2920 + }, + { + "epoch": 0.41, + "grad_norm": 6.004028846350347, + "learning_rate": 6.727601401305928e-06, + "loss": 0.3111, + "step": 2921 + }, + { + "epoch": 0.41, + "grad_norm": 2.254646992522527, + "learning_rate": 6.725488494413363e-06, + "loss": 0.3106, + "step": 2922 + }, + { + "epoch": 0.41, + "grad_norm": 2.086110606393287, + "learning_rate": 6.723375237676857e-06, + "loss": 0.3449, + "step": 2923 + }, + { + "epoch": 0.41, + "grad_norm": 2.7520051605460134, + "learning_rate": 6.721261631524871e-06, + "loss": 0.3452, + "step": 2924 + }, + { + "epoch": 0.41, + "grad_norm": 2.8240257438847727, + "learning_rate": 6.719147676385941e-06, + "loss": 0.3393, + "step": 2925 + }, + { + "epoch": 0.41, + "grad_norm": 2.033318463643835, + "learning_rate": 6.717033372688671e-06, + "loss": 0.3389, + "step": 2926 + }, + { + "epoch": 0.41, + "grad_norm": 2.9064738963087176, + "learning_rate": 6.714918720861741e-06, + "loss": 0.3235, + "step": 2927 + }, + { + "epoch": 0.41, + "grad_norm": 8.841661276924967, + "learning_rate": 6.712803721333895e-06, + "loss": 0.3276, + "step": 2928 + }, + { + "epoch": 0.41, + "grad_norm": 2.1964075747422838, + "learning_rate": 6.710688374533952e-06, + "loss": 0.3306, + "step": 2929 + }, + { + "epoch": 0.41, + "grad_norm": 2.4829674576923706, + "learning_rate": 6.7085726808908004e-06, + "loss": 0.3639, + "step": 2930 + }, + { + "epoch": 0.41, + "grad_norm": 5.69322616696734, + "learning_rate": 6.706456640833396e-06, + "loss": 0.3236, + "step": 2931 + }, + { + "epoch": 0.41, + "grad_norm": 4.088159794219506, + "learning_rate": 6.70434025479077e-06, + "loss": 0.3326, + "step": 2932 + }, + { + "epoch": 0.41, + "grad_norm": 2.713530178354204, + "learning_rate": 6.70222352319202e-06, + "loss": 0.3554, + "step": 2933 + }, + { + "epoch": 0.41, + "grad_norm": 2.476363857409744, + "learning_rate": 6.7001064464663165e-06, + "loss": 0.3283, + "step": 2934 + }, + { + "epoch": 0.41, + "grad_norm": 3.8684958662265068, + "learning_rate": 6.697989025042893e-06, + "loss": 0.3407, + "step": 2935 + }, + { + "epoch": 0.41, + "grad_norm": 2.6722435647251754, + "learning_rate": 6.695871259351065e-06, + "loss": 0.3305, + "step": 2936 + }, + { + "epoch": 0.41, + "grad_norm": 2.7178477162870984, + "learning_rate": 6.693753149820207e-06, + "loss": 0.3386, + "step": 2937 + }, + { + "epoch": 0.41, + "grad_norm": 2.833828420140715, + "learning_rate": 6.6916346968797675e-06, + "loss": 0.3431, + "step": 2938 + }, + { + "epoch": 0.41, + "grad_norm": 2.12297751478553, + "learning_rate": 6.6895159009592645e-06, + "loss": 0.3228, + "step": 2939 + }, + { + "epoch": 0.41, + "grad_norm": 1.9811545595326658, + "learning_rate": 6.6873967624882845e-06, + "loss": 0.3348, + "step": 2940 + }, + { + "epoch": 0.41, + "grad_norm": 2.5365679350309693, + "learning_rate": 6.685277281896486e-06, + "loss": 0.3479, + "step": 2941 + }, + { + "epoch": 0.41, + "grad_norm": 2.855639495380263, + "learning_rate": 6.683157459613594e-06, + "loss": 0.3538, + "step": 2942 + }, + { + "epoch": 0.41, + "grad_norm": 2.5166337245799086, + "learning_rate": 6.681037296069405e-06, + "loss": 0.3289, + "step": 2943 + }, + { + "epoch": 0.41, + "grad_norm": 2.5879450985940124, + "learning_rate": 6.6789167916937794e-06, + "loss": 0.3241, + "step": 2944 + }, + { + "epoch": 0.41, + "grad_norm": 2.0520659655498057, + "learning_rate": 6.676795946916656e-06, + "loss": 0.319, + "step": 2945 + }, + { + "epoch": 0.41, + "grad_norm": 2.2938775186835727, + "learning_rate": 6.6746747621680365e-06, + "loss": 0.3292, + "step": 2946 + }, + { + "epoch": 0.41, + "grad_norm": 2.2752589983701945, + "learning_rate": 6.67255323787799e-06, + "loss": 0.3239, + "step": 2947 + }, + { + "epoch": 0.41, + "grad_norm": 3.3258559470731504, + "learning_rate": 6.67043137447666e-06, + "loss": 0.3329, + "step": 2948 + }, + { + "epoch": 0.41, + "grad_norm": 2.347345412051781, + "learning_rate": 6.668309172394253e-06, + "loss": 0.3619, + "step": 2949 + }, + { + "epoch": 0.41, + "grad_norm": 2.2498337158060653, + "learning_rate": 6.6661866320610495e-06, + "loss": 0.3304, + "step": 2950 + }, + { + "epoch": 0.41, + "grad_norm": 2.342649769236426, + "learning_rate": 6.664063753907394e-06, + "loss": 0.3504, + "step": 2951 + }, + { + "epoch": 0.41, + "grad_norm": 3.148471600745826, + "learning_rate": 6.6619405383637015e-06, + "loss": 0.3123, + "step": 2952 + }, + { + "epoch": 0.41, + "grad_norm": 4.457911155549973, + "learning_rate": 6.659816985860457e-06, + "loss": 0.314, + "step": 2953 + }, + { + "epoch": 0.41, + "grad_norm": 3.9563352323055563, + "learning_rate": 6.6576930968282095e-06, + "loss": 0.3172, + "step": 2954 + }, + { + "epoch": 0.41, + "grad_norm": 3.086043790093396, + "learning_rate": 6.655568871697582e-06, + "loss": 0.3532, + "step": 2955 + }, + { + "epoch": 0.41, + "grad_norm": 2.1041463694016693, + "learning_rate": 6.653444310899261e-06, + "loss": 0.3244, + "step": 2956 + }, + { + "epoch": 0.41, + "grad_norm": 2.3225691266223807, + "learning_rate": 6.651319414864003e-06, + "loss": 0.3532, + "step": 2957 + }, + { + "epoch": 0.41, + "grad_norm": 3.122893232737489, + "learning_rate": 6.64919418402263e-06, + "loss": 0.3319, + "step": 2958 + }, + { + "epoch": 0.41, + "grad_norm": 2.822996494098619, + "learning_rate": 6.647068618806037e-06, + "loss": 0.3347, + "step": 2959 + }, + { + "epoch": 0.41, + "grad_norm": 7.605006524268733, + "learning_rate": 6.644942719645184e-06, + "loss": 0.3256, + "step": 2960 + }, + { + "epoch": 0.41, + "grad_norm": 2.8547119649659956, + "learning_rate": 6.6428164869710945e-06, + "loss": 0.3697, + "step": 2961 + }, + { + "epoch": 0.41, + "grad_norm": 2.2199837989205644, + "learning_rate": 6.640689921214868e-06, + "loss": 0.3383, + "step": 2962 + }, + { + "epoch": 0.41, + "grad_norm": 3.2952472025605344, + "learning_rate": 6.638563022807665e-06, + "loss": 0.3548, + "step": 2963 + }, + { + "epoch": 0.41, + "grad_norm": 2.5154576209119934, + "learning_rate": 6.636435792180714e-06, + "loss": 0.3221, + "step": 2964 + }, + { + "epoch": 0.41, + "grad_norm": 2.397189608408119, + "learning_rate": 6.634308229765316e-06, + "loss": 0.3146, + "step": 2965 + }, + { + "epoch": 0.41, + "grad_norm": 2.4258721485893187, + "learning_rate": 6.632180335992834e-06, + "loss": 0.3398, + "step": 2966 + }, + { + "epoch": 0.41, + "grad_norm": 3.179413839750228, + "learning_rate": 6.6300521112947004e-06, + "loss": 0.3478, + "step": 2967 + }, + { + "epoch": 0.41, + "grad_norm": 1.7716415593608874, + "learning_rate": 6.627923556102412e-06, + "loss": 0.3084, + "step": 2968 + }, + { + "epoch": 0.41, + "grad_norm": 6.965676832594838, + "learning_rate": 6.6257946708475374e-06, + "loss": 0.3352, + "step": 2969 + }, + { + "epoch": 0.41, + "grad_norm": 2.5019581494478014, + "learning_rate": 6.623665455961707e-06, + "loss": 0.3396, + "step": 2970 + }, + { + "epoch": 0.41, + "grad_norm": 3.7715865143737193, + "learning_rate": 6.6215359118766244e-06, + "loss": 0.3232, + "step": 2971 + }, + { + "epoch": 0.41, + "grad_norm": 2.3903327529861684, + "learning_rate": 6.619406039024053e-06, + "loss": 0.3526, + "step": 2972 + }, + { + "epoch": 0.41, + "grad_norm": 2.1599056580050804, + "learning_rate": 6.617275837835826e-06, + "loss": 0.3336, + "step": 2973 + }, + { + "epoch": 0.41, + "grad_norm": 2.8595405773760816, + "learning_rate": 6.615145308743844e-06, + "loss": 0.3193, + "step": 2974 + }, + { + "epoch": 0.41, + "grad_norm": 5.637554345282657, + "learning_rate": 6.613014452180073e-06, + "loss": 0.3563, + "step": 2975 + }, + { + "epoch": 0.41, + "grad_norm": 2.127027616906468, + "learning_rate": 6.610883268576545e-06, + "loss": 0.3096, + "step": 2976 + }, + { + "epoch": 0.41, + "grad_norm": 2.1398753768130034, + "learning_rate": 6.608751758365362e-06, + "loss": 0.3399, + "step": 2977 + }, + { + "epoch": 0.41, + "grad_norm": 2.578438960211693, + "learning_rate": 6.606619921978683e-06, + "loss": 0.3333, + "step": 2978 + }, + { + "epoch": 0.41, + "grad_norm": 3.8876416123337245, + "learning_rate": 6.604487759848746e-06, + "loss": 0.3257, + "step": 2979 + }, + { + "epoch": 0.41, + "grad_norm": 9.494069337736413, + "learning_rate": 6.602355272407843e-06, + "loss": 0.3073, + "step": 2980 + }, + { + "epoch": 0.41, + "grad_norm": 2.5729104016272943, + "learning_rate": 6.600222460088341e-06, + "loss": 0.3148, + "step": 2981 + }, + { + "epoch": 0.41, + "grad_norm": 3.3145584202703846, + "learning_rate": 6.598089323322666e-06, + "loss": 0.3432, + "step": 2982 + }, + { + "epoch": 0.41, + "grad_norm": 3.149126061434311, + "learning_rate": 6.595955862543314e-06, + "loss": 0.3327, + "step": 2983 + }, + { + "epoch": 0.41, + "grad_norm": 6.3127747381660795, + "learning_rate": 6.5938220781828474e-06, + "loss": 0.3105, + "step": 2984 + }, + { + "epoch": 0.41, + "grad_norm": 14.874202968756316, + "learning_rate": 6.59168797067389e-06, + "loss": 0.3483, + "step": 2985 + }, + { + "epoch": 0.42, + "grad_norm": 3.736260160917209, + "learning_rate": 6.589553540449136e-06, + "loss": 0.3552, + "step": 2986 + }, + { + "epoch": 0.42, + "grad_norm": 2.339403737592814, + "learning_rate": 6.58741878794134e-06, + "loss": 0.3463, + "step": 2987 + }, + { + "epoch": 0.42, + "grad_norm": 2.54999052410714, + "learning_rate": 6.585283713583326e-06, + "loss": 0.3424, + "step": 2988 + }, + { + "epoch": 0.42, + "grad_norm": 2.601689789574767, + "learning_rate": 6.583148317807982e-06, + "loss": 0.3679, + "step": 2989 + }, + { + "epoch": 0.42, + "grad_norm": 2.732277247643928, + "learning_rate": 6.5810126010482585e-06, + "loss": 0.3288, + "step": 2990 + }, + { + "epoch": 0.42, + "grad_norm": 1.9556171333683698, + "learning_rate": 6.578876563737176e-06, + "loss": 0.3173, + "step": 2991 + }, + { + "epoch": 0.42, + "grad_norm": 1.9844267366929798, + "learning_rate": 6.5767402063078164e-06, + "loss": 0.3009, + "step": 2992 + }, + { + "epoch": 0.42, + "grad_norm": 2.338373255001337, + "learning_rate": 6.5746035291933295e-06, + "loss": 0.3227, + "step": 2993 + }, + { + "epoch": 0.42, + "grad_norm": 2.9299255554461747, + "learning_rate": 6.5724665328269256e-06, + "loss": 0.3341, + "step": 2994 + }, + { + "epoch": 0.42, + "grad_norm": 3.6309858511644637, + "learning_rate": 6.5703292176418824e-06, + "loss": 0.3356, + "step": 2995 + }, + { + "epoch": 0.42, + "grad_norm": 4.1824427584176425, + "learning_rate": 6.568191584071543e-06, + "loss": 0.3935, + "step": 2996 + }, + { + "epoch": 0.42, + "grad_norm": 3.497630052093696, + "learning_rate": 6.566053632549313e-06, + "loss": 0.3171, + "step": 2997 + }, + { + "epoch": 0.42, + "grad_norm": 2.0529635497380885, + "learning_rate": 6.563915363508665e-06, + "loss": 0.3314, + "step": 2998 + }, + { + "epoch": 0.42, + "grad_norm": 2.1448734837265055, + "learning_rate": 6.561776777383133e-06, + "loss": 0.3546, + "step": 2999 + }, + { + "epoch": 0.42, + "grad_norm": 2.239472113629484, + "learning_rate": 6.559637874606317e-06, + "loss": 0.3328, + "step": 3000 + }, + { + "epoch": 0.42, + "grad_norm": 2.479064434808749, + "learning_rate": 6.5574986556118816e-06, + "loss": 0.3213, + "step": 3001 + }, + { + "epoch": 0.42, + "grad_norm": 2.5957029874014372, + "learning_rate": 6.555359120833554e-06, + "loss": 0.337, + "step": 3002 + }, + { + "epoch": 0.42, + "grad_norm": 3.2108615069702306, + "learning_rate": 6.553219270705129e-06, + "loss": 0.3143, + "step": 3003 + }, + { + "epoch": 0.42, + "grad_norm": 2.392078344748405, + "learning_rate": 6.5510791056604584e-06, + "loss": 0.3431, + "step": 3004 + }, + { + "epoch": 0.42, + "grad_norm": 2.7989469198557613, + "learning_rate": 6.5489386261334655e-06, + "loss": 0.3382, + "step": 3005 + }, + { + "epoch": 0.42, + "grad_norm": 2.1446107212607686, + "learning_rate": 6.5467978325581315e-06, + "loss": 0.306, + "step": 3006 + }, + { + "epoch": 0.42, + "grad_norm": 6.672919642823521, + "learning_rate": 6.544656725368506e-06, + "loss": 0.3421, + "step": 3007 + }, + { + "epoch": 0.42, + "grad_norm": 2.241832880035416, + "learning_rate": 6.5425153049986985e-06, + "loss": 0.3225, + "step": 3008 + }, + { + "epoch": 0.42, + "grad_norm": 3.190364041475798, + "learning_rate": 6.540373571882882e-06, + "loss": 0.3467, + "step": 3009 + }, + { + "epoch": 0.42, + "grad_norm": 2.464608707711113, + "learning_rate": 6.5382315264552986e-06, + "loss": 0.3406, + "step": 3010 + }, + { + "epoch": 0.42, + "grad_norm": 20.37608122909689, + "learning_rate": 6.536089169150246e-06, + "loss": 0.3339, + "step": 3011 + }, + { + "epoch": 0.42, + "grad_norm": 2.685122677984543, + "learning_rate": 6.533946500402088e-06, + "loss": 0.359, + "step": 3012 + }, + { + "epoch": 0.42, + "grad_norm": 2.544199116299748, + "learning_rate": 6.531803520645254e-06, + "loss": 0.3361, + "step": 3013 + }, + { + "epoch": 0.42, + "grad_norm": 2.578515725998569, + "learning_rate": 6.529660230314235e-06, + "loss": 0.346, + "step": 3014 + }, + { + "epoch": 0.42, + "grad_norm": 2.5127826299470493, + "learning_rate": 6.527516629843585e-06, + "loss": 0.3341, + "step": 3015 + }, + { + "epoch": 0.42, + "grad_norm": 2.461978942410529, + "learning_rate": 6.525372719667917e-06, + "loss": 0.3241, + "step": 3016 + }, + { + "epoch": 0.42, + "grad_norm": 3.6191727876109017, + "learning_rate": 6.5232285002219134e-06, + "loss": 0.3271, + "step": 3017 + }, + { + "epoch": 0.42, + "grad_norm": 3.2116248952786215, + "learning_rate": 6.521083971940314e-06, + "loss": 0.3373, + "step": 3018 + }, + { + "epoch": 0.42, + "grad_norm": 2.152455405912308, + "learning_rate": 6.518939135257926e-06, + "loss": 0.3345, + "step": 3019 + }, + { + "epoch": 0.42, + "grad_norm": 2.202977716454202, + "learning_rate": 6.5167939906096134e-06, + "loss": 0.3339, + "step": 3020 + }, + { + "epoch": 0.42, + "grad_norm": 2.3204693933615124, + "learning_rate": 6.514648538430308e-06, + "loss": 0.3213, + "step": 3021 + }, + { + "epoch": 0.42, + "grad_norm": 3.0716383535861183, + "learning_rate": 6.512502779155e-06, + "loss": 0.3347, + "step": 3022 + }, + { + "epoch": 0.42, + "grad_norm": 2.305255362914392, + "learning_rate": 6.510356713218744e-06, + "loss": 0.3203, + "step": 3023 + }, + { + "epoch": 0.42, + "grad_norm": 3.167207618760522, + "learning_rate": 6.508210341056659e-06, + "loss": 0.3426, + "step": 3024 + }, + { + "epoch": 0.42, + "grad_norm": 2.7896914114384774, + "learning_rate": 6.5060636631039185e-06, + "loss": 0.3404, + "step": 3025 + }, + { + "epoch": 0.42, + "grad_norm": 3.2439541808482026, + "learning_rate": 6.503916679795767e-06, + "loss": 0.3457, + "step": 3026 + }, + { + "epoch": 0.42, + "grad_norm": 2.8837911144611756, + "learning_rate": 6.501769391567504e-06, + "loss": 0.3358, + "step": 3027 + }, + { + "epoch": 0.42, + "grad_norm": 3.996180924513746, + "learning_rate": 6.499621798854495e-06, + "loss": 0.3523, + "step": 3028 + }, + { + "epoch": 0.42, + "grad_norm": 4.539685954424726, + "learning_rate": 6.497473902092167e-06, + "loss": 0.3432, + "step": 3029 + }, + { + "epoch": 0.42, + "grad_norm": 3.827441628797356, + "learning_rate": 6.495325701716004e-06, + "loss": 0.3392, + "step": 3030 + }, + { + "epoch": 0.42, + "grad_norm": 2.3124182538272833, + "learning_rate": 6.49317719816156e-06, + "loss": 0.334, + "step": 3031 + }, + { + "epoch": 0.42, + "grad_norm": 2.4366596889527754, + "learning_rate": 6.49102839186444e-06, + "loss": 0.3331, + "step": 3032 + }, + { + "epoch": 0.42, + "grad_norm": 2.2167484162628046, + "learning_rate": 6.4888792832603205e-06, + "loss": 0.339, + "step": 3033 + }, + { + "epoch": 0.42, + "grad_norm": 2.373919205405758, + "learning_rate": 6.486729872784933e-06, + "loss": 0.3616, + "step": 3034 + }, + { + "epoch": 0.42, + "grad_norm": 2.3134671168309997, + "learning_rate": 6.48458016087407e-06, + "loss": 0.3306, + "step": 3035 + }, + { + "epoch": 0.42, + "grad_norm": 4.4450725948697265, + "learning_rate": 6.482430147963592e-06, + "loss": 0.3126, + "step": 3036 + }, + { + "epoch": 0.42, + "grad_norm": 2.5666648044680134, + "learning_rate": 6.480279834489409e-06, + "loss": 0.3382, + "step": 3037 + }, + { + "epoch": 0.42, + "grad_norm": 2.5254143909352584, + "learning_rate": 6.478129220887504e-06, + "loss": 0.3535, + "step": 3038 + }, + { + "epoch": 0.42, + "grad_norm": 3.304193029678222, + "learning_rate": 6.475978307593912e-06, + "loss": 0.3163, + "step": 3039 + }, + { + "epoch": 0.42, + "grad_norm": 4.032177489047911, + "learning_rate": 6.473827095044734e-06, + "loss": 0.3376, + "step": 3040 + }, + { + "epoch": 0.42, + "grad_norm": 2.3813951875801522, + "learning_rate": 6.471675583676129e-06, + "loss": 0.3315, + "step": 3041 + }, + { + "epoch": 0.42, + "grad_norm": 2.743198601308202, + "learning_rate": 6.469523773924318e-06, + "loss": 0.3543, + "step": 3042 + }, + { + "epoch": 0.42, + "grad_norm": 2.330217240674918, + "learning_rate": 6.467371666225581e-06, + "loss": 0.3255, + "step": 3043 + }, + { + "epoch": 0.42, + "grad_norm": 2.4981205725555173, + "learning_rate": 6.465219261016259e-06, + "loss": 0.3542, + "step": 3044 + }, + { + "epoch": 0.42, + "grad_norm": 2.6763205255755707, + "learning_rate": 6.463066558732754e-06, + "loss": 0.3334, + "step": 3045 + }, + { + "epoch": 0.42, + "grad_norm": 2.4023008088395454, + "learning_rate": 6.46091355981153e-06, + "loss": 0.337, + "step": 3046 + }, + { + "epoch": 0.42, + "grad_norm": 2.240116103886775, + "learning_rate": 6.458760264689103e-06, + "loss": 0.3358, + "step": 3047 + }, + { + "epoch": 0.42, + "grad_norm": 3.0342043448766516, + "learning_rate": 6.456606673802064e-06, + "loss": 0.3096, + "step": 3048 + }, + { + "epoch": 0.42, + "grad_norm": 2.0251329114309584, + "learning_rate": 6.454452787587047e-06, + "loss": 0.3322, + "step": 3049 + }, + { + "epoch": 0.42, + "grad_norm": 2.461097287664761, + "learning_rate": 6.452298606480758e-06, + "loss": 0.3361, + "step": 3050 + }, + { + "epoch": 0.42, + "grad_norm": 2.7853773170417315, + "learning_rate": 6.4501441309199575e-06, + "loss": 0.3238, + "step": 3051 + }, + { + "epoch": 0.42, + "grad_norm": 2.622929639410463, + "learning_rate": 6.447989361341464e-06, + "loss": 0.3308, + "step": 3052 + }, + { + "epoch": 0.42, + "grad_norm": 2.421326204265296, + "learning_rate": 6.445834298182165e-06, + "loss": 0.3308, + "step": 3053 + }, + { + "epoch": 0.42, + "grad_norm": 2.340903692632448, + "learning_rate": 6.443678941878996e-06, + "loss": 0.3315, + "step": 3054 + }, + { + "epoch": 0.42, + "grad_norm": 2.304547057761992, + "learning_rate": 6.4415232928689595e-06, + "loss": 0.3282, + "step": 3055 + }, + { + "epoch": 0.42, + "grad_norm": 2.7010922344970374, + "learning_rate": 6.439367351589111e-06, + "loss": 0.3486, + "step": 3056 + }, + { + "epoch": 0.42, + "grad_norm": 2.491741445844739, + "learning_rate": 6.437211118476574e-06, + "loss": 0.3322, + "step": 3057 + }, + { + "epoch": 0.43, + "grad_norm": 3.76897041044929, + "learning_rate": 6.435054593968522e-06, + "loss": 0.3308, + "step": 3058 + }, + { + "epoch": 0.43, + "grad_norm": 5.898580555308823, + "learning_rate": 6.432897778502194e-06, + "loss": 0.3488, + "step": 3059 + }, + { + "epoch": 0.43, + "grad_norm": 2.2601076104311115, + "learning_rate": 6.430740672514886e-06, + "loss": 0.3222, + "step": 3060 + }, + { + "epoch": 0.43, + "grad_norm": 3.1587485501342343, + "learning_rate": 6.428583276443952e-06, + "loss": 0.3129, + "step": 3061 + }, + { + "epoch": 0.43, + "grad_norm": 2.9344562308510382, + "learning_rate": 6.426425590726806e-06, + "loss": 0.3034, + "step": 3062 + }, + { + "epoch": 0.43, + "grad_norm": 3.0740930508570568, + "learning_rate": 6.424267615800918e-06, + "loss": 0.335, + "step": 3063 + }, + { + "epoch": 0.43, + "grad_norm": 4.195470823457616, + "learning_rate": 6.422109352103822e-06, + "loss": 0.3191, + "step": 3064 + }, + { + "epoch": 0.43, + "grad_norm": 3.649511401818285, + "learning_rate": 6.419950800073105e-06, + "loss": 0.349, + "step": 3065 + }, + { + "epoch": 0.43, + "grad_norm": 2.6577654333129868, + "learning_rate": 6.417791960146416e-06, + "loss": 0.3131, + "step": 3066 + }, + { + "epoch": 0.43, + "grad_norm": 2.140320496679178, + "learning_rate": 6.415632832761463e-06, + "loss": 0.337, + "step": 3067 + }, + { + "epoch": 0.43, + "grad_norm": 2.2763408517450214, + "learning_rate": 6.413473418356007e-06, + "loss": 0.3725, + "step": 3068 + }, + { + "epoch": 0.43, + "grad_norm": 4.621220633521064, + "learning_rate": 6.411313717367872e-06, + "loss": 0.3306, + "step": 3069 + }, + { + "epoch": 0.43, + "grad_norm": 2.1546540556721223, + "learning_rate": 6.409153730234939e-06, + "loss": 0.3094, + "step": 3070 + }, + { + "epoch": 0.43, + "grad_norm": 2.084888333757691, + "learning_rate": 6.406993457395148e-06, + "loss": 0.3196, + "step": 3071 + }, + { + "epoch": 0.43, + "grad_norm": 2.3264453010218094, + "learning_rate": 6.404832899286495e-06, + "loss": 0.3348, + "step": 3072 + }, + { + "epoch": 0.43, + "grad_norm": 2.502818059850782, + "learning_rate": 6.402672056347033e-06, + "loss": 0.3125, + "step": 3073 + }, + { + "epoch": 0.43, + "grad_norm": 4.061940442307535, + "learning_rate": 6.400510929014876e-06, + "loss": 0.3256, + "step": 3074 + }, + { + "epoch": 0.43, + "grad_norm": 2.464273388770265, + "learning_rate": 6.398349517728193e-06, + "loss": 0.2971, + "step": 3075 + }, + { + "epoch": 0.43, + "grad_norm": 2.269568821842442, + "learning_rate": 6.396187822925214e-06, + "loss": 0.3308, + "step": 3076 + }, + { + "epoch": 0.43, + "grad_norm": 2.9935599510024944, + "learning_rate": 6.394025845044222e-06, + "loss": 0.3414, + "step": 3077 + }, + { + "epoch": 0.43, + "grad_norm": 2.116209559460788, + "learning_rate": 6.391863584523558e-06, + "loss": 0.3039, + "step": 3078 + }, + { + "epoch": 0.43, + "grad_norm": 2.157359219526196, + "learning_rate": 6.389701041801626e-06, + "loss": 0.3222, + "step": 3079 + }, + { + "epoch": 0.43, + "grad_norm": 2.4593181251056446, + "learning_rate": 6.387538217316879e-06, + "loss": 0.3469, + "step": 3080 + }, + { + "epoch": 0.43, + "grad_norm": 2.3351980363949214, + "learning_rate": 6.385375111507833e-06, + "loss": 0.3159, + "step": 3081 + }, + { + "epoch": 0.43, + "grad_norm": 2.8079496717133288, + "learning_rate": 6.383211724813058e-06, + "loss": 0.3313, + "step": 3082 + }, + { + "epoch": 0.43, + "grad_norm": 4.621201964866919, + "learning_rate": 6.381048057671184e-06, + "loss": 0.3199, + "step": 3083 + }, + { + "epoch": 0.43, + "grad_norm": 2.531904595411792, + "learning_rate": 6.378884110520896e-06, + "loss": 0.3126, + "step": 3084 + }, + { + "epoch": 0.43, + "grad_norm": 2.534831866044361, + "learning_rate": 6.376719883800933e-06, + "loss": 0.3129, + "step": 3085 + }, + { + "epoch": 0.43, + "grad_norm": 2.181127438333511, + "learning_rate": 6.374555377950093e-06, + "loss": 0.3378, + "step": 3086 + }, + { + "epoch": 0.43, + "grad_norm": 3.5756721552736526, + "learning_rate": 6.372390593407236e-06, + "loss": 0.3573, + "step": 3087 + }, + { + "epoch": 0.43, + "grad_norm": 2.7616814572807145, + "learning_rate": 6.37022553061127e-06, + "loss": 0.3464, + "step": 3088 + }, + { + "epoch": 0.43, + "grad_norm": 2.198584244998352, + "learning_rate": 6.368060190001163e-06, + "loss": 0.3047, + "step": 3089 + }, + { + "epoch": 0.43, + "grad_norm": 2.288814013150937, + "learning_rate": 6.365894572015939e-06, + "loss": 0.3261, + "step": 3090 + }, + { + "epoch": 0.43, + "grad_norm": 2.1190574468910643, + "learning_rate": 6.363728677094681e-06, + "loss": 0.3305, + "step": 3091 + }, + { + "epoch": 0.43, + "grad_norm": 2.5901775765442356, + "learning_rate": 6.361562505676522e-06, + "loss": 0.3309, + "step": 3092 + }, + { + "epoch": 0.43, + "grad_norm": 2.391331342511503, + "learning_rate": 6.359396058200657e-06, + "loss": 0.3352, + "step": 3093 + }, + { + "epoch": 0.43, + "grad_norm": 2.1662957896255985, + "learning_rate": 6.357229335106334e-06, + "loss": 0.3452, + "step": 3094 + }, + { + "epoch": 0.43, + "grad_norm": 2.874050187755096, + "learning_rate": 6.355062336832858e-06, + "loss": 0.3117, + "step": 3095 + }, + { + "epoch": 0.43, + "grad_norm": 2.1446298995884647, + "learning_rate": 6.352895063819589e-06, + "loss": 0.3444, + "step": 3096 + }, + { + "epoch": 0.43, + "grad_norm": 2.5696543297038037, + "learning_rate": 6.350727516505943e-06, + "loss": 0.3181, + "step": 3097 + }, + { + "epoch": 0.43, + "grad_norm": 1.9546530003186737, + "learning_rate": 6.348559695331394e-06, + "loss": 0.34, + "step": 3098 + }, + { + "epoch": 0.43, + "grad_norm": 4.306805974148148, + "learning_rate": 6.346391600735463e-06, + "loss": 0.3336, + "step": 3099 + }, + { + "epoch": 0.43, + "grad_norm": 2.5328478151103577, + "learning_rate": 6.344223233157739e-06, + "loss": 0.3339, + "step": 3100 + }, + { + "epoch": 0.43, + "grad_norm": 3.9803350395754, + "learning_rate": 6.342054593037857e-06, + "loss": 0.2925, + "step": 3101 + }, + { + "epoch": 0.43, + "grad_norm": 2.2783637419781675, + "learning_rate": 6.339885680815512e-06, + "loss": 0.3144, + "step": 3102 + }, + { + "epoch": 0.43, + "grad_norm": 2.1507090671104283, + "learning_rate": 6.33771649693045e-06, + "loss": 0.3421, + "step": 3103 + }, + { + "epoch": 0.43, + "grad_norm": 2.5557444347136413, + "learning_rate": 6.335547041822475e-06, + "loss": 0.3321, + "step": 3104 + }, + { + "epoch": 0.43, + "grad_norm": 2.6569703781045715, + "learning_rate": 6.3333773159314484e-06, + "loss": 0.3316, + "step": 3105 + }, + { + "epoch": 0.43, + "grad_norm": 2.1521597921193627, + "learning_rate": 6.331207319697279e-06, + "loss": 0.319, + "step": 3106 + }, + { + "epoch": 0.43, + "grad_norm": 3.734222229254037, + "learning_rate": 6.329037053559937e-06, + "loss": 0.3041, + "step": 3107 + }, + { + "epoch": 0.43, + "grad_norm": 2.278845808119172, + "learning_rate": 6.326866517959448e-06, + "loss": 0.2952, + "step": 3108 + }, + { + "epoch": 0.43, + "grad_norm": 2.615353554810849, + "learning_rate": 6.324695713335886e-06, + "loss": 0.3612, + "step": 3109 + }, + { + "epoch": 0.43, + "grad_norm": 2.322350683052351, + "learning_rate": 6.322524640129385e-06, + "loss": 0.3254, + "step": 3110 + }, + { + "epoch": 0.43, + "grad_norm": 2.181786622200561, + "learning_rate": 6.3203532987801295e-06, + "loss": 0.3252, + "step": 3111 + }, + { + "epoch": 0.43, + "grad_norm": 1.9653987330239016, + "learning_rate": 6.318181689728362e-06, + "loss": 0.3286, + "step": 3112 + }, + { + "epoch": 0.43, + "grad_norm": 2.5201548666438436, + "learning_rate": 6.316009813414378e-06, + "loss": 0.3331, + "step": 3113 + }, + { + "epoch": 0.43, + "grad_norm": 2.356994835753332, + "learning_rate": 6.313837670278526e-06, + "loss": 0.3273, + "step": 3114 + }, + { + "epoch": 0.43, + "grad_norm": 2.267849875780941, + "learning_rate": 6.3116652607612106e-06, + "loss": 0.3306, + "step": 3115 + }, + { + "epoch": 0.43, + "grad_norm": 3.326958432507613, + "learning_rate": 6.3094925853028854e-06, + "loss": 0.3107, + "step": 3116 + }, + { + "epoch": 0.43, + "grad_norm": 2.112630187355005, + "learning_rate": 6.307319644344068e-06, + "loss": 0.3377, + "step": 3117 + }, + { + "epoch": 0.43, + "grad_norm": 2.412430457165281, + "learning_rate": 6.305146438325319e-06, + "loss": 0.3555, + "step": 3118 + }, + { + "epoch": 0.43, + "grad_norm": 2.2456700099618394, + "learning_rate": 6.302972967687258e-06, + "loss": 0.357, + "step": 3119 + }, + { + "epoch": 0.43, + "grad_norm": 2.1806967198194105, + "learning_rate": 6.3007992328705615e-06, + "loss": 0.3363, + "step": 3120 + }, + { + "epoch": 0.43, + "grad_norm": 2.9875667844614964, + "learning_rate": 6.298625234315949e-06, + "loss": 0.3481, + "step": 3121 + }, + { + "epoch": 0.43, + "grad_norm": 2.3852728555579215, + "learning_rate": 6.2964509724642065e-06, + "loss": 0.3052, + "step": 3122 + }, + { + "epoch": 0.43, + "grad_norm": 2.2394013842668428, + "learning_rate": 6.294276447756162e-06, + "loss": 0.3095, + "step": 3123 + }, + { + "epoch": 0.43, + "grad_norm": 3.162007432785092, + "learning_rate": 6.2921016606327065e-06, + "loss": 0.331, + "step": 3124 + }, + { + "epoch": 0.43, + "grad_norm": 1.8042598197440092, + "learning_rate": 6.289926611534776e-06, + "loss": 0.3371, + "step": 3125 + }, + { + "epoch": 0.43, + "grad_norm": 3.461722040326216, + "learning_rate": 6.287751300903364e-06, + "loss": 0.3538, + "step": 3126 + }, + { + "epoch": 0.43, + "grad_norm": 1.8762957454377505, + "learning_rate": 6.2855757291795176e-06, + "loss": 0.3572, + "step": 3127 + }, + { + "epoch": 0.43, + "grad_norm": 2.450463954608125, + "learning_rate": 6.283399896804333e-06, + "loss": 0.3331, + "step": 3128 + }, + { + "epoch": 0.43, + "grad_norm": 2.81436001495674, + "learning_rate": 6.281223804218964e-06, + "loss": 0.3458, + "step": 3129 + }, + { + "epoch": 0.44, + "grad_norm": 2.7428166253360295, + "learning_rate": 6.279047451864613e-06, + "loss": 0.3272, + "step": 3130 + }, + { + "epoch": 0.44, + "grad_norm": 1.9412891983573102, + "learning_rate": 6.276870840182538e-06, + "loss": 0.3303, + "step": 3131 + }, + { + "epoch": 0.44, + "grad_norm": 3.49722442751255, + "learning_rate": 6.27469396961405e-06, + "loss": 0.3293, + "step": 3132 + }, + { + "epoch": 0.44, + "grad_norm": 2.0621276119217242, + "learning_rate": 6.272516840600506e-06, + "loss": 0.3613, + "step": 3133 + }, + { + "epoch": 0.44, + "grad_norm": 2.6611252533342644, + "learning_rate": 6.270339453583325e-06, + "loss": 0.3594, + "step": 3134 + }, + { + "epoch": 0.44, + "grad_norm": 2.383662104786115, + "learning_rate": 6.268161809003972e-06, + "loss": 0.342, + "step": 3135 + }, + { + "epoch": 0.44, + "grad_norm": 4.866625887106142, + "learning_rate": 6.265983907303967e-06, + "loss": 0.3364, + "step": 3136 + }, + { + "epoch": 0.44, + "grad_norm": 2.670338191851704, + "learning_rate": 6.26380574892488e-06, + "loss": 0.3257, + "step": 3137 + }, + { + "epoch": 0.44, + "grad_norm": 3.082935621334335, + "learning_rate": 6.261627334308333e-06, + "loss": 0.3203, + "step": 3138 + }, + { + "epoch": 0.44, + "grad_norm": 2.547955028577156, + "learning_rate": 6.259448663896001e-06, + "loss": 0.3314, + "step": 3139 + }, + { + "epoch": 0.44, + "grad_norm": 1.9636961580391858, + "learning_rate": 6.257269738129614e-06, + "loss": 0.3127, + "step": 3140 + }, + { + "epoch": 0.44, + "grad_norm": 2.008540266587715, + "learning_rate": 6.25509055745095e-06, + "loss": 0.3361, + "step": 3141 + }, + { + "epoch": 0.44, + "grad_norm": 2.06243618397378, + "learning_rate": 6.252911122301835e-06, + "loss": 0.3252, + "step": 3142 + }, + { + "epoch": 0.44, + "grad_norm": 3.538899556885099, + "learning_rate": 6.250731433124157e-06, + "loss": 0.3326, + "step": 3143 + }, + { + "epoch": 0.44, + "grad_norm": 2.13976969612692, + "learning_rate": 6.248551490359844e-06, + "loss": 0.3136, + "step": 3144 + }, + { + "epoch": 0.44, + "grad_norm": 2.3276487565570627, + "learning_rate": 6.246371294450884e-06, + "loss": 0.3404, + "step": 3145 + }, + { + "epoch": 0.44, + "grad_norm": 2.6955675221556428, + "learning_rate": 6.244190845839314e-06, + "loss": 0.3391, + "step": 3146 + }, + { + "epoch": 0.44, + "grad_norm": 1.9387160158637222, + "learning_rate": 6.242010144967218e-06, + "loss": 0.3118, + "step": 3147 + }, + { + "epoch": 0.44, + "grad_norm": 2.7892996060059536, + "learning_rate": 6.2398291922767375e-06, + "loss": 0.3179, + "step": 3148 + }, + { + "epoch": 0.44, + "grad_norm": 2.517053275393886, + "learning_rate": 6.23764798821006e-06, + "loss": 0.3558, + "step": 3149 + }, + { + "epoch": 0.44, + "grad_norm": 2.7687363593972845, + "learning_rate": 6.235466533209427e-06, + "loss": 0.3326, + "step": 3150 + }, + { + "epoch": 0.44, + "grad_norm": 2.8363543432625824, + "learning_rate": 6.233284827717131e-06, + "loss": 0.3171, + "step": 3151 + }, + { + "epoch": 0.44, + "grad_norm": 2.328397710010473, + "learning_rate": 6.231102872175512e-06, + "loss": 0.3297, + "step": 3152 + }, + { + "epoch": 0.44, + "grad_norm": 4.703251061577456, + "learning_rate": 6.2289206670269665e-06, + "loss": 0.3191, + "step": 3153 + }, + { + "epoch": 0.44, + "grad_norm": 2.5676465063710308, + "learning_rate": 6.226738212713935e-06, + "loss": 0.3249, + "step": 3154 + }, + { + "epoch": 0.44, + "grad_norm": 2.4033536651760925, + "learning_rate": 6.22455550967891e-06, + "loss": 0.3648, + "step": 3155 + }, + { + "epoch": 0.44, + "grad_norm": 2.06038108084366, + "learning_rate": 6.22237255836444e-06, + "loss": 0.3064, + "step": 3156 + }, + { + "epoch": 0.44, + "grad_norm": 2.2508214292475186, + "learning_rate": 6.220189359213118e-06, + "loss": 0.3296, + "step": 3157 + }, + { + "epoch": 0.44, + "grad_norm": 3.007605377842274, + "learning_rate": 6.2180059126675906e-06, + "loss": 0.3473, + "step": 3158 + }, + { + "epoch": 0.44, + "grad_norm": 3.1365943189996313, + "learning_rate": 6.215822219170551e-06, + "loss": 0.3115, + "step": 3159 + }, + { + "epoch": 0.44, + "grad_norm": 2.6762777898822563, + "learning_rate": 6.213638279164744e-06, + "loss": 0.3378, + "step": 3160 + }, + { + "epoch": 0.44, + "grad_norm": 3.527615495125093, + "learning_rate": 6.211454093092968e-06, + "loss": 0.3251, + "step": 3161 + }, + { + "epoch": 0.44, + "grad_norm": 2.151863409896485, + "learning_rate": 6.209269661398065e-06, + "loss": 0.3388, + "step": 3162 + }, + { + "epoch": 0.44, + "grad_norm": 2.3383432103740422, + "learning_rate": 6.207084984522932e-06, + "loss": 0.3121, + "step": 3163 + }, + { + "epoch": 0.44, + "grad_norm": 2.9414069490978942, + "learning_rate": 6.204900062910513e-06, + "loss": 0.3328, + "step": 3164 + }, + { + "epoch": 0.44, + "grad_norm": 3.236893930079336, + "learning_rate": 6.202714897003801e-06, + "loss": 0.3352, + "step": 3165 + }, + { + "epoch": 0.44, + "grad_norm": 2.25097336037832, + "learning_rate": 6.200529487245842e-06, + "loss": 0.3305, + "step": 3166 + }, + { + "epoch": 0.44, + "grad_norm": 2.0346939280583998, + "learning_rate": 6.1983438340797295e-06, + "loss": 0.3411, + "step": 3167 + }, + { + "epoch": 0.44, + "grad_norm": 2.2761654598229297, + "learning_rate": 6.1961579379486035e-06, + "loss": 0.3222, + "step": 3168 + }, + { + "epoch": 0.44, + "grad_norm": 1.738941681111378, + "learning_rate": 6.193971799295658e-06, + "loss": 0.3412, + "step": 3169 + }, + { + "epoch": 0.44, + "grad_norm": 2.5887323475734934, + "learning_rate": 6.191785418564133e-06, + "loss": 0.3184, + "step": 3170 + }, + { + "epoch": 0.44, + "grad_norm": 2.4955114335367465, + "learning_rate": 6.18959879619732e-06, + "loss": 0.3251, + "step": 3171 + }, + { + "epoch": 0.44, + "grad_norm": 2.406618944398195, + "learning_rate": 6.187411932638558e-06, + "loss": 0.3505, + "step": 3172 + }, + { + "epoch": 0.44, + "grad_norm": 3.167771454288522, + "learning_rate": 6.1852248283312334e-06, + "loss": 0.3521, + "step": 3173 + }, + { + "epoch": 0.44, + "grad_norm": 2.414140355964755, + "learning_rate": 6.183037483718784e-06, + "loss": 0.3337, + "step": 3174 + }, + { + "epoch": 0.44, + "grad_norm": 5.987230214715174, + "learning_rate": 6.180849899244696e-06, + "loss": 0.3203, + "step": 3175 + }, + { + "epoch": 0.44, + "grad_norm": 1.8653695488038275, + "learning_rate": 6.1786620753525025e-06, + "loss": 0.314, + "step": 3176 + }, + { + "epoch": 0.44, + "grad_norm": 1.9556613794098476, + "learning_rate": 6.176474012485787e-06, + "loss": 0.3231, + "step": 3177 + }, + { + "epoch": 0.44, + "grad_norm": 2.2458310164629185, + "learning_rate": 6.17428571108818e-06, + "loss": 0.3353, + "step": 3178 + }, + { + "epoch": 0.44, + "grad_norm": 4.13386855769598, + "learning_rate": 6.1720971716033625e-06, + "loss": 0.334, + "step": 3179 + }, + { + "epoch": 0.44, + "grad_norm": 2.2357076418621777, + "learning_rate": 6.1699083944750616e-06, + "loss": 0.3295, + "step": 3180 + }, + { + "epoch": 0.44, + "grad_norm": 2.1286932612889684, + "learning_rate": 6.1677193801470515e-06, + "loss": 0.3608, + "step": 3181 + }, + { + "epoch": 0.44, + "grad_norm": 2.499120649881049, + "learning_rate": 6.165530129063159e-06, + "loss": 0.3469, + "step": 3182 + }, + { + "epoch": 0.44, + "grad_norm": 2.590161505450227, + "learning_rate": 6.163340641667255e-06, + "loss": 0.339, + "step": 3183 + }, + { + "epoch": 0.44, + "grad_norm": 2.689647871040837, + "learning_rate": 6.16115091840326e-06, + "loss": 0.3271, + "step": 3184 + }, + { + "epoch": 0.44, + "grad_norm": 2.4374672696342863, + "learning_rate": 6.15896095971514e-06, + "loss": 0.326, + "step": 3185 + }, + { + "epoch": 0.44, + "grad_norm": 1.8958186879438241, + "learning_rate": 6.156770766046915e-06, + "loss": 0.308, + "step": 3186 + }, + { + "epoch": 0.44, + "grad_norm": 2.1081308987019507, + "learning_rate": 6.154580337842644e-06, + "loss": 0.3792, + "step": 3187 + }, + { + "epoch": 0.44, + "grad_norm": 2.1213509967235824, + "learning_rate": 6.152389675546438e-06, + "loss": 0.3185, + "step": 3188 + }, + { + "epoch": 0.44, + "grad_norm": 2.4079241656608295, + "learning_rate": 6.150198779602459e-06, + "loss": 0.345, + "step": 3189 + }, + { + "epoch": 0.44, + "grad_norm": 1.9260968554468698, + "learning_rate": 6.148007650454906e-06, + "loss": 0.3436, + "step": 3190 + }, + { + "epoch": 0.44, + "grad_norm": 2.313498559766399, + "learning_rate": 6.145816288548039e-06, + "loss": 0.3332, + "step": 3191 + }, + { + "epoch": 0.44, + "grad_norm": 2.19728453269538, + "learning_rate": 6.143624694326154e-06, + "loss": 0.3402, + "step": 3192 + }, + { + "epoch": 0.44, + "grad_norm": 2.180982352182766, + "learning_rate": 6.1414328682336e-06, + "loss": 0.3426, + "step": 3193 + }, + { + "epoch": 0.44, + "grad_norm": 4.410374853339398, + "learning_rate": 6.139240810714768e-06, + "loss": 0.3442, + "step": 3194 + }, + { + "epoch": 0.44, + "grad_norm": 2.18157931547772, + "learning_rate": 6.137048522214104e-06, + "loss": 0.317, + "step": 3195 + }, + { + "epoch": 0.44, + "grad_norm": 1.9043365720402659, + "learning_rate": 6.134856003176093e-06, + "loss": 0.357, + "step": 3196 + }, + { + "epoch": 0.44, + "grad_norm": 2.956823604287208, + "learning_rate": 6.13266325404527e-06, + "loss": 0.3147, + "step": 3197 + }, + { + "epoch": 0.44, + "grad_norm": 2.2913305065178426, + "learning_rate": 6.130470275266216e-06, + "loss": 0.3233, + "step": 3198 + }, + { + "epoch": 0.44, + "grad_norm": 1.9177953144857462, + "learning_rate": 6.128277067283559e-06, + "loss": 0.3309, + "step": 3199 + }, + { + "epoch": 0.44, + "grad_norm": 2.0608473715287854, + "learning_rate": 6.1260836305419755e-06, + "loss": 0.317, + "step": 3200 + }, + { + "epoch": 0.44, + "grad_norm": 1.941424585581053, + "learning_rate": 6.123889965486184e-06, + "loss": 0.3098, + "step": 3201 + }, + { + "epoch": 0.45, + "grad_norm": 1.7962442387928228, + "learning_rate": 6.121696072560951e-06, + "loss": 0.3507, + "step": 3202 + }, + { + "epoch": 0.45, + "grad_norm": 1.8832437350555442, + "learning_rate": 6.119501952211091e-06, + "loss": 0.3524, + "step": 3203 + }, + { + "epoch": 0.45, + "grad_norm": 1.8933634527434124, + "learning_rate": 6.117307604881464e-06, + "loss": 0.3202, + "step": 3204 + }, + { + "epoch": 0.45, + "grad_norm": 2.7243017918994656, + "learning_rate": 6.115113031016975e-06, + "loss": 0.3284, + "step": 3205 + }, + { + "epoch": 0.45, + "grad_norm": 2.3268587430567647, + "learning_rate": 6.112918231062575e-06, + "loss": 0.332, + "step": 3206 + }, + { + "epoch": 0.45, + "grad_norm": 2.2093683836251548, + "learning_rate": 6.110723205463259e-06, + "loss": 0.3145, + "step": 3207 + }, + { + "epoch": 0.45, + "grad_norm": 2.1541879269710034, + "learning_rate": 6.108527954664074e-06, + "loss": 0.364, + "step": 3208 + }, + { + "epoch": 0.45, + "grad_norm": 2.4997989551117032, + "learning_rate": 6.106332479110105e-06, + "loss": 0.2939, + "step": 3209 + }, + { + "epoch": 0.45, + "grad_norm": 2.487961883781983, + "learning_rate": 6.10413677924649e-06, + "loss": 0.3243, + "step": 3210 + }, + { + "epoch": 0.45, + "grad_norm": 3.4306501130658407, + "learning_rate": 6.101940855518403e-06, + "loss": 0.3205, + "step": 3211 + }, + { + "epoch": 0.45, + "grad_norm": 1.973660490697887, + "learning_rate": 6.099744708371076e-06, + "loss": 0.3377, + "step": 3212 + }, + { + "epoch": 0.45, + "grad_norm": 2.4298310646969057, + "learning_rate": 6.097548338249773e-06, + "loss": 0.345, + "step": 3213 + }, + { + "epoch": 0.45, + "grad_norm": 2.0048053959873617, + "learning_rate": 6.095351745599813e-06, + "loss": 0.3304, + "step": 3214 + }, + { + "epoch": 0.45, + "grad_norm": 2.440601715769709, + "learning_rate": 6.093154930866556e-06, + "loss": 0.3136, + "step": 3215 + }, + { + "epoch": 0.45, + "grad_norm": 1.996929049942869, + "learning_rate": 6.0909578944954064e-06, + "loss": 0.3157, + "step": 3216 + }, + { + "epoch": 0.45, + "grad_norm": 2.189125111149454, + "learning_rate": 6.088760636931817e-06, + "loss": 0.3084, + "step": 3217 + }, + { + "epoch": 0.45, + "grad_norm": 3.2918754427477883, + "learning_rate": 6.086563158621282e-06, + "loss": 0.3408, + "step": 3218 + }, + { + "epoch": 0.45, + "grad_norm": 3.5467894437223944, + "learning_rate": 6.0843654600093404e-06, + "loss": 0.3186, + "step": 3219 + }, + { + "epoch": 0.45, + "grad_norm": 2.5378710890124174, + "learning_rate": 6.082167541541578e-06, + "loss": 0.3336, + "step": 3220 + }, + { + "epoch": 0.45, + "grad_norm": 2.229366118755462, + "learning_rate": 6.079969403663626e-06, + "loss": 0.3478, + "step": 3221 + }, + { + "epoch": 0.45, + "grad_norm": 1.901767828790698, + "learning_rate": 6.077771046821157e-06, + "loss": 0.3436, + "step": 3222 + }, + { + "epoch": 0.45, + "grad_norm": 2.1389759741569865, + "learning_rate": 6.075572471459888e-06, + "loss": 0.3221, + "step": 3223 + }, + { + "epoch": 0.45, + "grad_norm": 2.113440835541529, + "learning_rate": 6.073373678025583e-06, + "loss": 0.3129, + "step": 3224 + }, + { + "epoch": 0.45, + "grad_norm": 2.395931608174629, + "learning_rate": 6.071174666964048e-06, + "loss": 0.3285, + "step": 3225 + }, + { + "epoch": 0.45, + "grad_norm": 2.3734997786420484, + "learning_rate": 6.068975438721135e-06, + "loss": 0.3336, + "step": 3226 + }, + { + "epoch": 0.45, + "grad_norm": 2.182909845224842, + "learning_rate": 6.066775993742739e-06, + "loss": 0.3246, + "step": 3227 + }, + { + "epoch": 0.45, + "grad_norm": 11.91788802244846, + "learning_rate": 6.064576332474799e-06, + "loss": 0.3533, + "step": 3228 + }, + { + "epoch": 0.45, + "grad_norm": 2.8308318336121507, + "learning_rate": 6.062376455363295e-06, + "loss": 0.3197, + "step": 3229 + }, + { + "epoch": 0.45, + "grad_norm": 2.4330290640856647, + "learning_rate": 6.060176362854259e-06, + "loss": 0.3269, + "step": 3230 + }, + { + "epoch": 0.45, + "grad_norm": 4.292152766595877, + "learning_rate": 6.057976055393757e-06, + "loss": 0.3178, + "step": 3231 + }, + { + "epoch": 0.45, + "grad_norm": 2.2564186386303704, + "learning_rate": 6.055775533427903e-06, + "loss": 0.3151, + "step": 3232 + }, + { + "epoch": 0.45, + "grad_norm": 2.5903796259985485, + "learning_rate": 6.053574797402855e-06, + "loss": 0.3168, + "step": 3233 + }, + { + "epoch": 0.45, + "grad_norm": 2.0969084236423363, + "learning_rate": 6.051373847764814e-06, + "loss": 0.3243, + "step": 3234 + }, + { + "epoch": 0.45, + "grad_norm": 2.042785160222756, + "learning_rate": 6.049172684960024e-06, + "loss": 0.3162, + "step": 3235 + }, + { + "epoch": 0.45, + "grad_norm": 2.218766300797483, + "learning_rate": 6.046971309434773e-06, + "loss": 0.3381, + "step": 3236 + }, + { + "epoch": 0.45, + "grad_norm": 2.4028876561547943, + "learning_rate": 6.044769721635388e-06, + "loss": 0.3049, + "step": 3237 + }, + { + "epoch": 0.45, + "grad_norm": 2.8945935607994073, + "learning_rate": 6.042567922008249e-06, + "loss": 0.303, + "step": 3238 + }, + { + "epoch": 0.45, + "grad_norm": 2.3193017018255766, + "learning_rate": 6.040365910999765e-06, + "loss": 0.3154, + "step": 3239 + }, + { + "epoch": 0.45, + "grad_norm": 2.0225626413614046, + "learning_rate": 6.038163689056399e-06, + "loss": 0.3343, + "step": 3240 + }, + { + "epoch": 0.45, + "grad_norm": 2.0377871104169514, + "learning_rate": 6.035961256624652e-06, + "loss": 0.3058, + "step": 3241 + }, + { + "epoch": 0.45, + "grad_norm": 1.92016902492025, + "learning_rate": 6.03375861415107e-06, + "loss": 0.3189, + "step": 3242 + }, + { + "epoch": 0.45, + "grad_norm": 15.130663902764443, + "learning_rate": 6.0315557620822395e-06, + "loss": 0.3378, + "step": 3243 + }, + { + "epoch": 0.45, + "grad_norm": 2.3081833244223193, + "learning_rate": 6.029352700864789e-06, + "loss": 0.3279, + "step": 3244 + }, + { + "epoch": 0.45, + "grad_norm": 4.165546530765966, + "learning_rate": 6.0271494309453915e-06, + "loss": 0.336, + "step": 3245 + }, + { + "epoch": 0.45, + "grad_norm": 2.662344467206245, + "learning_rate": 6.024945952770761e-06, + "loss": 0.3293, + "step": 3246 + }, + { + "epoch": 0.45, + "grad_norm": 2.50139460534969, + "learning_rate": 6.022742266787656e-06, + "loss": 0.3287, + "step": 3247 + }, + { + "epoch": 0.45, + "grad_norm": 4.22815363816888, + "learning_rate": 6.020538373442873e-06, + "loss": 0.3285, + "step": 3248 + }, + { + "epoch": 0.45, + "grad_norm": 2.425208011383801, + "learning_rate": 6.018334273183255e-06, + "loss": 0.334, + "step": 3249 + }, + { + "epoch": 0.45, + "grad_norm": 2.352027972020608, + "learning_rate": 6.016129966455683e-06, + "loss": 0.3351, + "step": 3250 + }, + { + "epoch": 0.45, + "grad_norm": 6.563652543275239, + "learning_rate": 6.013925453707082e-06, + "loss": 0.3332, + "step": 3251 + }, + { + "epoch": 0.45, + "grad_norm": 3.839875437832029, + "learning_rate": 6.011720735384419e-06, + "loss": 0.3653, + "step": 3252 + }, + { + "epoch": 0.45, + "grad_norm": 2.4922391816194156, + "learning_rate": 6.009515811934703e-06, + "loss": 0.3175, + "step": 3253 + }, + { + "epoch": 0.45, + "grad_norm": 3.006785053822283, + "learning_rate": 6.00731068380498e-06, + "loss": 0.324, + "step": 3254 + }, + { + "epoch": 0.45, + "grad_norm": 3.0304105593591197, + "learning_rate": 6.005105351442346e-06, + "loss": 0.3407, + "step": 3255 + }, + { + "epoch": 0.45, + "grad_norm": 1.8723889319163114, + "learning_rate": 6.00289981529393e-06, + "loss": 0.3374, + "step": 3256 + }, + { + "epoch": 0.45, + "grad_norm": 4.689620476142452, + "learning_rate": 6.000694075806908e-06, + "loss": 0.3257, + "step": 3257 + }, + { + "epoch": 0.45, + "grad_norm": 2.3489259072569433, + "learning_rate": 5.998488133428495e-06, + "loss": 0.3434, + "step": 3258 + }, + { + "epoch": 0.45, + "grad_norm": 2.2491164168272237, + "learning_rate": 5.996281988605945e-06, + "loss": 0.3298, + "step": 3259 + }, + { + "epoch": 0.45, + "grad_norm": 2.1616302651953276, + "learning_rate": 5.994075641786559e-06, + "loss": 0.3369, + "step": 3260 + }, + { + "epoch": 0.45, + "grad_norm": 2.082911263462624, + "learning_rate": 5.9918690934176726e-06, + "loss": 0.3257, + "step": 3261 + }, + { + "epoch": 0.45, + "grad_norm": 2.2150241900858396, + "learning_rate": 5.989662343946666e-06, + "loss": 0.3335, + "step": 3262 + }, + { + "epoch": 0.45, + "grad_norm": 2.0029866700970413, + "learning_rate": 5.987455393820958e-06, + "loss": 0.3461, + "step": 3263 + }, + { + "epoch": 0.45, + "grad_norm": 2.1313590214065945, + "learning_rate": 5.98524824348801e-06, + "loss": 0.3073, + "step": 3264 + }, + { + "epoch": 0.45, + "grad_norm": 2.6034638571438182, + "learning_rate": 5.983040893395324e-06, + "loss": 0.3245, + "step": 3265 + }, + { + "epoch": 0.45, + "grad_norm": 2.2714360810867786, + "learning_rate": 5.980833343990441e-06, + "loss": 0.3316, + "step": 3266 + }, + { + "epoch": 0.45, + "grad_norm": 2.0079684607692387, + "learning_rate": 5.978625595720942e-06, + "loss": 0.3268, + "step": 3267 + }, + { + "epoch": 0.45, + "grad_norm": 1.8006137821871189, + "learning_rate": 5.976417649034451e-06, + "loss": 0.3097, + "step": 3268 + }, + { + "epoch": 0.45, + "grad_norm": 2.4835155498031796, + "learning_rate": 5.974209504378632e-06, + "loss": 0.3232, + "step": 3269 + }, + { + "epoch": 0.45, + "grad_norm": 2.0876168666688306, + "learning_rate": 5.972001162201184e-06, + "loss": 0.3051, + "step": 3270 + }, + { + "epoch": 0.45, + "grad_norm": 2.1189145301122005, + "learning_rate": 5.9697926229498514e-06, + "loss": 0.3572, + "step": 3271 + }, + { + "epoch": 0.45, + "grad_norm": 2.1881621666476145, + "learning_rate": 5.967583887072419e-06, + "loss": 0.3058, + "step": 3272 + }, + { + "epoch": 0.45, + "grad_norm": 1.9267806769240103, + "learning_rate": 5.9653749550167064e-06, + "loss": 0.3501, + "step": 3273 + }, + { + "epoch": 0.46, + "grad_norm": 5.031481695535473, + "learning_rate": 5.963165827230579e-06, + "loss": 0.3277, + "step": 3274 + }, + { + "epoch": 0.46, + "grad_norm": 2.6635993544830963, + "learning_rate": 5.9609565041619366e-06, + "loss": 0.313, + "step": 3275 + }, + { + "epoch": 0.46, + "grad_norm": 1.9367704651491207, + "learning_rate": 5.958746986258722e-06, + "loss": 0.3648, + "step": 3276 + }, + { + "epoch": 0.46, + "grad_norm": 2.309324720487248, + "learning_rate": 5.956537273968916e-06, + "loss": 0.333, + "step": 3277 + }, + { + "epoch": 0.46, + "grad_norm": 2.201952415300351, + "learning_rate": 5.95432736774054e-06, + "loss": 0.3139, + "step": 3278 + }, + { + "epoch": 0.46, + "grad_norm": 2.0559156422121263, + "learning_rate": 5.952117268021653e-06, + "loss": 0.3306, + "step": 3279 + }, + { + "epoch": 0.46, + "grad_norm": 2.316191717237495, + "learning_rate": 5.949906975260353e-06, + "loss": 0.3485, + "step": 3280 + }, + { + "epoch": 0.46, + "grad_norm": 3.321455505781966, + "learning_rate": 5.947696489904784e-06, + "loss": 0.3322, + "step": 3281 + }, + { + "epoch": 0.46, + "grad_norm": 2.1549246100595325, + "learning_rate": 5.945485812403116e-06, + "loss": 0.3215, + "step": 3282 + }, + { + "epoch": 0.46, + "grad_norm": 2.20793521855317, + "learning_rate": 5.94327494320357e-06, + "loss": 0.3113, + "step": 3283 + }, + { + "epoch": 0.46, + "grad_norm": 2.3626104705468043, + "learning_rate": 5.941063882754401e-06, + "loss": 0.3361, + "step": 3284 + }, + { + "epoch": 0.46, + "grad_norm": 2.106021254046249, + "learning_rate": 5.938852631503899e-06, + "loss": 0.3316, + "step": 3285 + }, + { + "epoch": 0.46, + "grad_norm": 2.8789329893759597, + "learning_rate": 5.936641189900401e-06, + "loss": 0.3763, + "step": 3286 + }, + { + "epoch": 0.46, + "grad_norm": 2.595084333760474, + "learning_rate": 5.934429558392277e-06, + "loss": 0.3261, + "step": 3287 + }, + { + "epoch": 0.46, + "grad_norm": 4.08980601115616, + "learning_rate": 5.932217737427936e-06, + "loss": 0.3196, + "step": 3288 + }, + { + "epoch": 0.46, + "grad_norm": 1.859754967205828, + "learning_rate": 5.9300057274558275e-06, + "loss": 0.3395, + "step": 3289 + }, + { + "epoch": 0.46, + "grad_norm": 2.089891735875669, + "learning_rate": 5.927793528924436e-06, + "loss": 0.3339, + "step": 3290 + }, + { + "epoch": 0.46, + "grad_norm": 2.8291359908548603, + "learning_rate": 5.92558114228229e-06, + "loss": 0.3356, + "step": 3291 + }, + { + "epoch": 0.46, + "grad_norm": 2.463683149228752, + "learning_rate": 5.923368567977948e-06, + "loss": 0.3281, + "step": 3292 + }, + { + "epoch": 0.46, + "grad_norm": 2.939535860196919, + "learning_rate": 5.921155806460013e-06, + "loss": 0.3493, + "step": 3293 + }, + { + "epoch": 0.46, + "grad_norm": 3.166002488168272, + "learning_rate": 5.918942858177123e-06, + "loss": 0.3319, + "step": 3294 + }, + { + "epoch": 0.46, + "grad_norm": 2.6242296243580476, + "learning_rate": 5.916729723577955e-06, + "loss": 0.3488, + "step": 3295 + }, + { + "epoch": 0.46, + "grad_norm": 5.722238475759449, + "learning_rate": 5.914516403111226e-06, + "loss": 0.3168, + "step": 3296 + }, + { + "epoch": 0.46, + "grad_norm": 2.7490502282111624, + "learning_rate": 5.912302897225684e-06, + "loss": 0.3253, + "step": 3297 + }, + { + "epoch": 0.46, + "grad_norm": 2.2828584056799954, + "learning_rate": 5.9100892063701225e-06, + "loss": 0.3305, + "step": 3298 + }, + { + "epoch": 0.46, + "grad_norm": 2.6557033787416726, + "learning_rate": 5.907875330993365e-06, + "loss": 0.3398, + "step": 3299 + }, + { + "epoch": 0.46, + "grad_norm": 2.6828085682400507, + "learning_rate": 5.905661271544282e-06, + "loss": 0.3372, + "step": 3300 + }, + { + "epoch": 0.46, + "grad_norm": 3.010300426124929, + "learning_rate": 5.903447028471769e-06, + "loss": 0.3388, + "step": 3301 + }, + { + "epoch": 0.46, + "grad_norm": 3.206429021496195, + "learning_rate": 5.901232602224767e-06, + "loss": 0.3263, + "step": 3302 + }, + { + "epoch": 0.46, + "grad_norm": 2.24608954288412, + "learning_rate": 5.899017993252256e-06, + "loss": 0.3407, + "step": 3303 + }, + { + "epoch": 0.46, + "grad_norm": 2.1687520104730758, + "learning_rate": 5.896803202003247e-06, + "loss": 0.3213, + "step": 3304 + }, + { + "epoch": 0.46, + "grad_norm": 3.578514354674202, + "learning_rate": 5.894588228926791e-06, + "loss": 0.3202, + "step": 3305 + }, + { + "epoch": 0.46, + "grad_norm": 2.4506244761577864, + "learning_rate": 5.892373074471972e-06, + "loss": 0.3575, + "step": 3306 + }, + { + "epoch": 0.46, + "grad_norm": 2.8878375251207, + "learning_rate": 5.8901577390879195e-06, + "loss": 0.3431, + "step": 3307 + }, + { + "epoch": 0.46, + "grad_norm": 2.4996809897859373, + "learning_rate": 5.887942223223789e-06, + "loss": 0.3569, + "step": 3308 + }, + { + "epoch": 0.46, + "grad_norm": 2.164024930620495, + "learning_rate": 5.885726527328782e-06, + "loss": 0.349, + "step": 3309 + }, + { + "epoch": 0.46, + "grad_norm": 2.5204670684954187, + "learning_rate": 5.88351065185213e-06, + "loss": 0.3236, + "step": 3310 + }, + { + "epoch": 0.46, + "grad_norm": 2.6309989210851734, + "learning_rate": 5.881294597243103e-06, + "loss": 0.3369, + "step": 3311 + }, + { + "epoch": 0.46, + "grad_norm": 2.437563836044202, + "learning_rate": 5.879078363951011e-06, + "loss": 0.336, + "step": 3312 + }, + { + "epoch": 0.46, + "grad_norm": 3.293049518271743, + "learning_rate": 5.876861952425192e-06, + "loss": 0.3118, + "step": 3313 + }, + { + "epoch": 0.46, + "grad_norm": 2.630458218914031, + "learning_rate": 5.874645363115028e-06, + "loss": 0.3254, + "step": 3314 + }, + { + "epoch": 0.46, + "grad_norm": 3.6980838183265528, + "learning_rate": 5.872428596469932e-06, + "loss": 0.3247, + "step": 3315 + }, + { + "epoch": 0.46, + "grad_norm": 2.790713488996465, + "learning_rate": 5.8702116529393575e-06, + "loss": 0.3354, + "step": 3316 + }, + { + "epoch": 0.46, + "grad_norm": 1.7953555099492897, + "learning_rate": 5.867994532972789e-06, + "loss": 0.3349, + "step": 3317 + }, + { + "epoch": 0.46, + "grad_norm": 4.446652760743582, + "learning_rate": 5.8657772370197496e-06, + "loss": 0.3234, + "step": 3318 + }, + { + "epoch": 0.46, + "grad_norm": 2.136642362693186, + "learning_rate": 5.863559765529797e-06, + "loss": 0.3235, + "step": 3319 + }, + { + "epoch": 0.46, + "grad_norm": 2.2404303443994733, + "learning_rate": 5.8613421189525265e-06, + "loss": 0.3354, + "step": 3320 + }, + { + "epoch": 0.46, + "grad_norm": 2.395721586095779, + "learning_rate": 5.8591242977375665e-06, + "loss": 0.3503, + "step": 3321 + }, + { + "epoch": 0.46, + "grad_norm": 2.0920144331386306, + "learning_rate": 5.856906302334583e-06, + "loss": 0.3249, + "step": 3322 + }, + { + "epoch": 0.46, + "grad_norm": 2.330396968061422, + "learning_rate": 5.854688133193273e-06, + "loss": 0.3492, + "step": 3323 + }, + { + "epoch": 0.46, + "grad_norm": 2.0130256841973813, + "learning_rate": 5.852469790763375e-06, + "loss": 0.3228, + "step": 3324 + }, + { + "epoch": 0.46, + "grad_norm": 3.2707565539670145, + "learning_rate": 5.850251275494658e-06, + "loss": 0.3363, + "step": 3325 + }, + { + "epoch": 0.46, + "grad_norm": 2.6500494344770593, + "learning_rate": 5.848032587836927e-06, + "loss": 0.3307, + "step": 3326 + }, + { + "epoch": 0.46, + "grad_norm": 1.8919253494640673, + "learning_rate": 5.845813728240025e-06, + "loss": 0.3192, + "step": 3327 + }, + { + "epoch": 0.46, + "grad_norm": 2.534477105273033, + "learning_rate": 5.843594697153823e-06, + "loss": 0.3439, + "step": 3328 + }, + { + "epoch": 0.46, + "grad_norm": 2.0818037812752914, + "learning_rate": 5.841375495028234e-06, + "loss": 0.3268, + "step": 3329 + }, + { + "epoch": 0.46, + "grad_norm": 2.1403140942360186, + "learning_rate": 5.839156122313202e-06, + "loss": 0.3512, + "step": 3330 + }, + { + "epoch": 0.46, + "grad_norm": 3.4506085476302335, + "learning_rate": 5.836936579458707e-06, + "loss": 0.3489, + "step": 3331 + }, + { + "epoch": 0.46, + "grad_norm": 2.870558357971872, + "learning_rate": 5.83471686691476e-06, + "loss": 0.3215, + "step": 3332 + }, + { + "epoch": 0.46, + "grad_norm": 2.4241506030471616, + "learning_rate": 5.832496985131413e-06, + "loss": 0.3408, + "step": 3333 + }, + { + "epoch": 0.46, + "grad_norm": 2.358663304014426, + "learning_rate": 5.830276934558747e-06, + "loss": 0.3295, + "step": 3334 + }, + { + "epoch": 0.46, + "grad_norm": 2.7094667107590604, + "learning_rate": 5.828056715646878e-06, + "loss": 0.3073, + "step": 3335 + }, + { + "epoch": 0.46, + "grad_norm": 2.30170813386799, + "learning_rate": 5.825836328845957e-06, + "loss": 0.3079, + "step": 3336 + }, + { + "epoch": 0.46, + "grad_norm": 2.3012158403604737, + "learning_rate": 5.823615774606169e-06, + "loss": 0.3228, + "step": 3337 + }, + { + "epoch": 0.46, + "grad_norm": 2.295120194281977, + "learning_rate": 5.821395053377736e-06, + "loss": 0.3414, + "step": 3338 + }, + { + "epoch": 0.46, + "grad_norm": 2.7014241885560364, + "learning_rate": 5.819174165610905e-06, + "loss": 0.3296, + "step": 3339 + }, + { + "epoch": 0.46, + "grad_norm": 2.365180714092328, + "learning_rate": 5.816953111755966e-06, + "loss": 0.3229, + "step": 3340 + }, + { + "epoch": 0.46, + "grad_norm": 3.7497555581699507, + "learning_rate": 5.814731892263239e-06, + "loss": 0.3519, + "step": 3341 + }, + { + "epoch": 0.46, + "grad_norm": 3.8269528267965254, + "learning_rate": 5.812510507583075e-06, + "loss": 0.3304, + "step": 3342 + }, + { + "epoch": 0.46, + "grad_norm": 2.6517962356951865, + "learning_rate": 5.810288958165867e-06, + "loss": 0.3283, + "step": 3343 + }, + { + "epoch": 0.46, + "grad_norm": 3.091956384291614, + "learning_rate": 5.80806724446203e-06, + "loss": 0.344, + "step": 3344 + }, + { + "epoch": 0.47, + "grad_norm": 1.954080574935563, + "learning_rate": 5.805845366922021e-06, + "loss": 0.3492, + "step": 3345 + }, + { + "epoch": 0.47, + "grad_norm": 2.0626149054309195, + "learning_rate": 5.803623325996325e-06, + "loss": 0.3239, + "step": 3346 + }, + { + "epoch": 0.47, + "grad_norm": 2.463641218118619, + "learning_rate": 5.801401122135464e-06, + "loss": 0.3525, + "step": 3347 + }, + { + "epoch": 0.47, + "grad_norm": 2.0148019616684247, + "learning_rate": 5.799178755789992e-06, + "loss": 0.3539, + "step": 3348 + }, + { + "epoch": 0.47, + "grad_norm": 2.033436166858871, + "learning_rate": 5.796956227410491e-06, + "loss": 0.3396, + "step": 3349 + }, + { + "epoch": 0.47, + "grad_norm": 2.5162632849123496, + "learning_rate": 5.7947335374475855e-06, + "loss": 0.328, + "step": 3350 + }, + { + "epoch": 0.47, + "grad_norm": 2.105287915947888, + "learning_rate": 5.7925106863519244e-06, + "loss": 0.363, + "step": 3351 + }, + { + "epoch": 0.47, + "grad_norm": 2.203060307462363, + "learning_rate": 5.790287674574193e-06, + "loss": 0.3047, + "step": 3352 + }, + { + "epoch": 0.47, + "grad_norm": 2.130357280071207, + "learning_rate": 5.78806450256511e-06, + "loss": 0.3175, + "step": 3353 + }, + { + "epoch": 0.47, + "grad_norm": 2.2797697580096514, + "learning_rate": 5.7858411707754195e-06, + "loss": 0.3433, + "step": 3354 + }, + { + "epoch": 0.47, + "grad_norm": 1.9040785484194611, + "learning_rate": 5.78361767965591e-06, + "loss": 0.3204, + "step": 3355 + }, + { + "epoch": 0.47, + "grad_norm": 2.027450695585762, + "learning_rate": 5.781394029657392e-06, + "loss": 0.3084, + "step": 3356 + }, + { + "epoch": 0.47, + "grad_norm": 2.793631140729466, + "learning_rate": 5.779170221230713e-06, + "loss": 0.355, + "step": 3357 + }, + { + "epoch": 0.47, + "grad_norm": 2.2944649313577887, + "learning_rate": 5.776946254826753e-06, + "loss": 0.3333, + "step": 3358 + }, + { + "epoch": 0.47, + "grad_norm": 2.1208991656347673, + "learning_rate": 5.7747221308964204e-06, + "loss": 0.3084, + "step": 3359 + }, + { + "epoch": 0.47, + "grad_norm": 2.339889062361765, + "learning_rate": 5.772497849890661e-06, + "loss": 0.333, + "step": 3360 + }, + { + "epoch": 0.47, + "grad_norm": 2.491827615027172, + "learning_rate": 5.770273412260444e-06, + "loss": 0.3661, + "step": 3361 + }, + { + "epoch": 0.47, + "grad_norm": 2.5111885101522105, + "learning_rate": 5.76804881845678e-06, + "loss": 0.3454, + "step": 3362 + }, + { + "epoch": 0.47, + "grad_norm": 3.3890669243328935, + "learning_rate": 5.765824068930706e-06, + "loss": 0.347, + "step": 3363 + }, + { + "epoch": 0.47, + "grad_norm": 1.9968679160691045, + "learning_rate": 5.76359916413329e-06, + "loss": 0.3254, + "step": 3364 + }, + { + "epoch": 0.47, + "grad_norm": 2.1442197403996315, + "learning_rate": 5.761374104515636e-06, + "loss": 0.3248, + "step": 3365 + }, + { + "epoch": 0.47, + "grad_norm": 2.143544770346087, + "learning_rate": 5.759148890528871e-06, + "loss": 0.3306, + "step": 3366 + }, + { + "epoch": 0.47, + "grad_norm": 2.081342370634595, + "learning_rate": 5.7569235226241645e-06, + "loss": 0.3308, + "step": 3367 + }, + { + "epoch": 0.47, + "grad_norm": 3.2879659174701756, + "learning_rate": 5.754698001252708e-06, + "loss": 0.3248, + "step": 3368 + }, + { + "epoch": 0.47, + "grad_norm": 2.3584127358605174, + "learning_rate": 5.752472326865729e-06, + "loss": 0.316, + "step": 3369 + }, + { + "epoch": 0.47, + "grad_norm": 2.0558908854414772, + "learning_rate": 5.750246499914483e-06, + "loss": 0.3407, + "step": 3370 + }, + { + "epoch": 0.47, + "grad_norm": 2.129240085904539, + "learning_rate": 5.748020520850257e-06, + "loss": 0.3588, + "step": 3371 + }, + { + "epoch": 0.47, + "grad_norm": 2.190425532361071, + "learning_rate": 5.745794390124374e-06, + "loss": 0.3341, + "step": 3372 + }, + { + "epoch": 0.47, + "grad_norm": 2.2217078796853693, + "learning_rate": 5.743568108188179e-06, + "loss": 0.3123, + "step": 3373 + }, + { + "epoch": 0.47, + "grad_norm": 2.1404355024789052, + "learning_rate": 5.741341675493055e-06, + "loss": 0.3456, + "step": 3374 + }, + { + "epoch": 0.47, + "grad_norm": 2.153835598956077, + "learning_rate": 5.739115092490411e-06, + "loss": 0.3261, + "step": 3375 + }, + { + "epoch": 0.47, + "grad_norm": 1.9388558837937773, + "learning_rate": 5.736888359631689e-06, + "loss": 0.3402, + "step": 3376 + }, + { + "epoch": 0.47, + "grad_norm": 2.063666909227037, + "learning_rate": 5.7346614773683605e-06, + "loss": 0.3418, + "step": 3377 + }, + { + "epoch": 0.47, + "grad_norm": 2.480640560123854, + "learning_rate": 5.7324344461519266e-06, + "loss": 0.3314, + "step": 3378 + }, + { + "epoch": 0.47, + "grad_norm": 3.775575995258435, + "learning_rate": 5.73020726643392e-06, + "loss": 0.3263, + "step": 3379 + }, + { + "epoch": 0.47, + "grad_norm": 2.615710064472931, + "learning_rate": 5.727979938665904e-06, + "loss": 0.3334, + "step": 3380 + }, + { + "epoch": 0.47, + "grad_norm": 2.013575717679005, + "learning_rate": 5.72575246329947e-06, + "loss": 0.3383, + "step": 3381 + }, + { + "epoch": 0.47, + "grad_norm": 1.7434770376803694, + "learning_rate": 5.723524840786239e-06, + "loss": 0.3435, + "step": 3382 + }, + { + "epoch": 0.47, + "grad_norm": 3.8108215179878253, + "learning_rate": 5.7212970715778635e-06, + "loss": 0.3512, + "step": 3383 + }, + { + "epoch": 0.47, + "grad_norm": 2.337031538541087, + "learning_rate": 5.719069156126026e-06, + "loss": 0.3279, + "step": 3384 + }, + { + "epoch": 0.47, + "grad_norm": 2.1399199707393897, + "learning_rate": 5.7168410948824375e-06, + "loss": 0.3402, + "step": 3385 + }, + { + "epoch": 0.47, + "grad_norm": 2.1435601620278, + "learning_rate": 5.714612888298838e-06, + "loss": 0.3191, + "step": 3386 + }, + { + "epoch": 0.47, + "grad_norm": 2.6318961297738936, + "learning_rate": 5.712384536826999e-06, + "loss": 0.3236, + "step": 3387 + }, + { + "epoch": 0.47, + "grad_norm": 1.9515364669859154, + "learning_rate": 5.7101560409187205e-06, + "loss": 0.3485, + "step": 3388 + }, + { + "epoch": 0.47, + "grad_norm": 2.145166699962152, + "learning_rate": 5.70792740102583e-06, + "loss": 0.3341, + "step": 3389 + }, + { + "epoch": 0.47, + "grad_norm": 2.2363777568856795, + "learning_rate": 5.705698617600186e-06, + "loss": 0.3089, + "step": 3390 + }, + { + "epoch": 0.47, + "grad_norm": 1.7382668630606646, + "learning_rate": 5.703469691093678e-06, + "loss": 0.3333, + "step": 3391 + }, + { + "epoch": 0.47, + "grad_norm": 2.353628982351083, + "learning_rate": 5.701240621958219e-06, + "loss": 0.3252, + "step": 3392 + }, + { + "epoch": 0.47, + "grad_norm": 2.277832802887428, + "learning_rate": 5.699011410645758e-06, + "loss": 0.343, + "step": 3393 + }, + { + "epoch": 0.47, + "grad_norm": 3.0098684007543963, + "learning_rate": 5.696782057608265e-06, + "loss": 0.3076, + "step": 3394 + }, + { + "epoch": 0.47, + "grad_norm": 2.7777471187238363, + "learning_rate": 5.694552563297746e-06, + "loss": 0.3328, + "step": 3395 + }, + { + "epoch": 0.47, + "grad_norm": 3.195942162608622, + "learning_rate": 5.692322928166233e-06, + "loss": 0.3487, + "step": 3396 + }, + { + "epoch": 0.47, + "grad_norm": 2.125809766087196, + "learning_rate": 5.690093152665781e-06, + "loss": 0.339, + "step": 3397 + }, + { + "epoch": 0.47, + "grad_norm": 1.866082542663915, + "learning_rate": 5.6878632372484845e-06, + "loss": 0.3423, + "step": 3398 + }, + { + "epoch": 0.47, + "grad_norm": 3.0232090905788858, + "learning_rate": 5.685633182366457e-06, + "loss": 0.3197, + "step": 3399 + }, + { + "epoch": 0.47, + "grad_norm": 2.0672172012891683, + "learning_rate": 5.683402988471843e-06, + "loss": 0.3341, + "step": 3400 + }, + { + "epoch": 0.47, + "grad_norm": 2.4966531999520454, + "learning_rate": 5.681172656016818e-06, + "loss": 0.3423, + "step": 3401 + }, + { + "epoch": 0.47, + "grad_norm": 1.6604016760360913, + "learning_rate": 5.678942185453583e-06, + "loss": 0.314, + "step": 3402 + }, + { + "epoch": 0.47, + "grad_norm": 1.8917876924240895, + "learning_rate": 5.676711577234367e-06, + "loss": 0.338, + "step": 3403 + }, + { + "epoch": 0.47, + "grad_norm": 3.072053563404728, + "learning_rate": 5.674480831811426e-06, + "loss": 0.3091, + "step": 3404 + }, + { + "epoch": 0.47, + "grad_norm": 2.4178964958779767, + "learning_rate": 5.672249949637048e-06, + "loss": 0.3214, + "step": 3405 + }, + { + "epoch": 0.47, + "grad_norm": 4.952508027714412, + "learning_rate": 5.670018931163543e-06, + "loss": 0.3218, + "step": 3406 + }, + { + "epoch": 0.47, + "grad_norm": 3.022454584602335, + "learning_rate": 5.667787776843253e-06, + "loss": 0.3243, + "step": 3407 + }, + { + "epoch": 0.47, + "grad_norm": 2.4745822747445514, + "learning_rate": 5.665556487128545e-06, + "loss": 0.3089, + "step": 3408 + }, + { + "epoch": 0.47, + "grad_norm": 2.1008143995908144, + "learning_rate": 5.663325062471816e-06, + "loss": 0.3078, + "step": 3409 + }, + { + "epoch": 0.47, + "grad_norm": 2.115550595252085, + "learning_rate": 5.661093503325487e-06, + "loss": 0.3347, + "step": 3410 + }, + { + "epoch": 0.47, + "grad_norm": 2.0808211286171368, + "learning_rate": 5.6588618101420105e-06, + "loss": 0.347, + "step": 3411 + }, + { + "epoch": 0.47, + "grad_norm": 2.06873016204296, + "learning_rate": 5.656629983373863e-06, + "loss": 0.3261, + "step": 3412 + }, + { + "epoch": 0.47, + "grad_norm": 2.5695679906995696, + "learning_rate": 5.654398023473546e-06, + "loss": 0.3305, + "step": 3413 + }, + { + "epoch": 0.47, + "grad_norm": 2.656671613631957, + "learning_rate": 5.652165930893594e-06, + "loss": 0.3356, + "step": 3414 + }, + { + "epoch": 0.47, + "grad_norm": 2.294299046638888, + "learning_rate": 5.649933706086564e-06, + "loss": 0.3294, + "step": 3415 + }, + { + "epoch": 0.47, + "grad_norm": 2.565851129387244, + "learning_rate": 5.6477013495050416e-06, + "loss": 0.3417, + "step": 3416 + }, + { + "epoch": 0.48, + "grad_norm": 2.0031566665763663, + "learning_rate": 5.64546886160164e-06, + "loss": 0.3225, + "step": 3417 + }, + { + "epoch": 0.48, + "grad_norm": 2.25043545557839, + "learning_rate": 5.643236242828993e-06, + "loss": 0.319, + "step": 3418 + }, + { + "epoch": 0.48, + "grad_norm": 2.1113795520108196, + "learning_rate": 5.641003493639771e-06, + "loss": 0.3423, + "step": 3419 + }, + { + "epoch": 0.48, + "grad_norm": 2.7095252526818348, + "learning_rate": 5.638770614486662e-06, + "loss": 0.341, + "step": 3420 + }, + { + "epoch": 0.48, + "grad_norm": 2.1660778893466377, + "learning_rate": 5.636537605822385e-06, + "loss": 0.3216, + "step": 3421 + }, + { + "epoch": 0.48, + "grad_norm": 1.9241352215696368, + "learning_rate": 5.634304468099685e-06, + "loss": 0.347, + "step": 3422 + }, + { + "epoch": 0.48, + "grad_norm": 5.212597620913563, + "learning_rate": 5.632071201771328e-06, + "loss": 0.3145, + "step": 3423 + }, + { + "epoch": 0.48, + "grad_norm": 2.428914482061054, + "learning_rate": 5.629837807290116e-06, + "loss": 0.3253, + "step": 3424 + }, + { + "epoch": 0.48, + "grad_norm": 2.04813288000467, + "learning_rate": 5.627604285108866e-06, + "loss": 0.3386, + "step": 3425 + }, + { + "epoch": 0.48, + "grad_norm": 2.482805071635763, + "learning_rate": 5.625370635680429e-06, + "loss": 0.3275, + "step": 3426 + }, + { + "epoch": 0.48, + "grad_norm": 1.9274979865548685, + "learning_rate": 5.623136859457679e-06, + "loss": 0.3179, + "step": 3427 + }, + { + "epoch": 0.48, + "grad_norm": 2.5795974986976113, + "learning_rate": 5.620902956893515e-06, + "loss": 0.331, + "step": 3428 + }, + { + "epoch": 0.48, + "grad_norm": 2.799470770362533, + "learning_rate": 5.618668928440862e-06, + "loss": 0.3429, + "step": 3429 + }, + { + "epoch": 0.48, + "grad_norm": 2.1005729936828246, + "learning_rate": 5.6164347745526715e-06, + "loss": 0.345, + "step": 3430 + }, + { + "epoch": 0.48, + "grad_norm": 1.9925665110884045, + "learning_rate": 5.6142004956819184e-06, + "loss": 0.2974, + "step": 3431 + }, + { + "epoch": 0.48, + "grad_norm": 2.0543491062409145, + "learning_rate": 5.611966092281605e-06, + "loss": 0.3024, + "step": 3432 + }, + { + "epoch": 0.48, + "grad_norm": 2.282178737576564, + "learning_rate": 5.609731564804758e-06, + "loss": 0.3415, + "step": 3433 + }, + { + "epoch": 0.48, + "grad_norm": 2.231853697967703, + "learning_rate": 5.60749691370443e-06, + "loss": 0.3096, + "step": 3434 + }, + { + "epoch": 0.48, + "grad_norm": 1.9420453327583311, + "learning_rate": 5.605262139433696e-06, + "loss": 0.3243, + "step": 3435 + }, + { + "epoch": 0.48, + "grad_norm": 2.9375255872270207, + "learning_rate": 5.603027242445661e-06, + "loss": 0.3258, + "step": 3436 + }, + { + "epoch": 0.48, + "grad_norm": 2.0860013876454, + "learning_rate": 5.60079222319345e-06, + "loss": 0.3171, + "step": 3437 + }, + { + "epoch": 0.48, + "grad_norm": 3.1777750789811705, + "learning_rate": 5.598557082130216e-06, + "loss": 0.3154, + "step": 3438 + }, + { + "epoch": 0.48, + "grad_norm": 2.386585997851883, + "learning_rate": 5.5963218197091315e-06, + "loss": 0.3233, + "step": 3439 + }, + { + "epoch": 0.48, + "grad_norm": 2.4339757195383505, + "learning_rate": 5.594086436383402e-06, + "loss": 0.3264, + "step": 3440 + }, + { + "epoch": 0.48, + "grad_norm": 3.1578554696035175, + "learning_rate": 5.5918509326062495e-06, + "loss": 0.3401, + "step": 3441 + }, + { + "epoch": 0.48, + "grad_norm": 1.938406511040668, + "learning_rate": 5.5896153088309265e-06, + "loss": 0.3275, + "step": 3442 + }, + { + "epoch": 0.48, + "grad_norm": 1.9180757490353564, + "learning_rate": 5.587379565510708e-06, + "loss": 0.3129, + "step": 3443 + }, + { + "epoch": 0.48, + "grad_norm": 1.9584550759366972, + "learning_rate": 5.585143703098888e-06, + "loss": 0.3345, + "step": 3444 + }, + { + "epoch": 0.48, + "grad_norm": 2.2509252226531062, + "learning_rate": 5.5829077220487935e-06, + "loss": 0.3056, + "step": 3445 + }, + { + "epoch": 0.48, + "grad_norm": 3.0464223743540266, + "learning_rate": 5.580671622813768e-06, + "loss": 0.323, + "step": 3446 + }, + { + "epoch": 0.48, + "grad_norm": 2.2420680554787986, + "learning_rate": 5.578435405847185e-06, + "loss": 0.3111, + "step": 3447 + }, + { + "epoch": 0.48, + "grad_norm": 2.938078149140311, + "learning_rate": 5.576199071602436e-06, + "loss": 0.3245, + "step": 3448 + }, + { + "epoch": 0.48, + "grad_norm": 2.524173667431699, + "learning_rate": 5.573962620532941e-06, + "loss": 0.3018, + "step": 3449 + }, + { + "epoch": 0.48, + "grad_norm": 2.069890384590373, + "learning_rate": 5.5717260530921435e-06, + "loss": 0.3054, + "step": 3450 + }, + { + "epoch": 0.48, + "grad_norm": 2.978790173935361, + "learning_rate": 5.5694893697335065e-06, + "loss": 0.3219, + "step": 3451 + }, + { + "epoch": 0.48, + "grad_norm": 3.8839514556715313, + "learning_rate": 5.567252570910519e-06, + "loss": 0.3348, + "step": 3452 + }, + { + "epoch": 0.48, + "grad_norm": 2.8455716126985773, + "learning_rate": 5.565015657076693e-06, + "loss": 0.3253, + "step": 3453 + }, + { + "epoch": 0.48, + "grad_norm": 2.274533876794745, + "learning_rate": 5.562778628685568e-06, + "loss": 0.3323, + "step": 3454 + }, + { + "epoch": 0.48, + "grad_norm": 2.277809867694502, + "learning_rate": 5.5605414861907005e-06, + "loss": 0.3461, + "step": 3455 + }, + { + "epoch": 0.48, + "grad_norm": 2.1113216323966726, + "learning_rate": 5.558304230045671e-06, + "loss": 0.3252, + "step": 3456 + }, + { + "epoch": 0.48, + "grad_norm": 1.8996076041535137, + "learning_rate": 5.556066860704088e-06, + "loss": 0.2903, + "step": 3457 + }, + { + "epoch": 0.48, + "grad_norm": 1.718610430465354, + "learning_rate": 5.553829378619576e-06, + "loss": 0.3343, + "step": 3458 + }, + { + "epoch": 0.48, + "grad_norm": 2.4955328018060983, + "learning_rate": 5.551591784245789e-06, + "loss": 0.3342, + "step": 3459 + }, + { + "epoch": 0.48, + "grad_norm": 2.543071355008027, + "learning_rate": 5.549354078036401e-06, + "loss": 0.3156, + "step": 3460 + }, + { + "epoch": 0.48, + "grad_norm": 2.135846574143196, + "learning_rate": 5.547116260445105e-06, + "loss": 0.3416, + "step": 3461 + }, + { + "epoch": 0.48, + "grad_norm": 2.443942220466248, + "learning_rate": 5.5448783319256235e-06, + "loss": 0.3218, + "step": 3462 + }, + { + "epoch": 0.48, + "grad_norm": 2.0564463676299365, + "learning_rate": 5.542640292931696e-06, + "loss": 0.3141, + "step": 3463 + }, + { + "epoch": 0.48, + "grad_norm": 2.1418935968292963, + "learning_rate": 5.540402143917088e-06, + "loss": 0.3291, + "step": 3464 + }, + { + "epoch": 0.48, + "grad_norm": 2.4149418810688834, + "learning_rate": 5.538163885335584e-06, + "loss": 0.3409, + "step": 3465 + }, + { + "epoch": 0.48, + "grad_norm": 1.9100469188707967, + "learning_rate": 5.535925517640992e-06, + "loss": 0.3246, + "step": 3466 + }, + { + "epoch": 0.48, + "grad_norm": 3.4762989183460857, + "learning_rate": 5.533687041287146e-06, + "loss": 0.3442, + "step": 3467 + }, + { + "epoch": 0.48, + "grad_norm": 2.4012802856115507, + "learning_rate": 5.531448456727895e-06, + "loss": 0.3409, + "step": 3468 + }, + { + "epoch": 0.48, + "grad_norm": 1.9151239610578115, + "learning_rate": 5.529209764417115e-06, + "loss": 0.3477, + "step": 3469 + }, + { + "epoch": 0.48, + "grad_norm": 1.9870024496756313, + "learning_rate": 5.526970964808702e-06, + "loss": 0.3245, + "step": 3470 + }, + { + "epoch": 0.48, + "grad_norm": 2.1027994330955555, + "learning_rate": 5.5247320583565745e-06, + "loss": 0.3206, + "step": 3471 + }, + { + "epoch": 0.48, + "grad_norm": 2.1246504400403636, + "learning_rate": 5.522493045514672e-06, + "loss": 0.3327, + "step": 3472 + }, + { + "epoch": 0.48, + "grad_norm": 1.7484756894003095, + "learning_rate": 5.520253926736956e-06, + "loss": 0.3443, + "step": 3473 + }, + { + "epoch": 0.48, + "grad_norm": 2.878670035857508, + "learning_rate": 5.518014702477409e-06, + "loss": 0.3567, + "step": 3474 + }, + { + "epoch": 0.48, + "grad_norm": 2.506280742321759, + "learning_rate": 5.515775373190035e-06, + "loss": 0.3123, + "step": 3475 + }, + { + "epoch": 0.48, + "grad_norm": 2.1544975397441393, + "learning_rate": 5.513535939328861e-06, + "loss": 0.3275, + "step": 3476 + }, + { + "epoch": 0.48, + "grad_norm": 2.3759047142646224, + "learning_rate": 5.511296401347933e-06, + "loss": 0.3486, + "step": 3477 + }, + { + "epoch": 0.48, + "grad_norm": 2.887208408019372, + "learning_rate": 5.509056759701317e-06, + "loss": 0.2972, + "step": 3478 + }, + { + "epoch": 0.48, + "grad_norm": 2.255971931835604, + "learning_rate": 5.506817014843104e-06, + "loss": 0.3155, + "step": 3479 + }, + { + "epoch": 0.48, + "grad_norm": 1.9183675156213775, + "learning_rate": 5.504577167227403e-06, + "loss": 0.3187, + "step": 3480 + }, + { + "epoch": 0.48, + "grad_norm": 2.3734698746714713, + "learning_rate": 5.502337217308346e-06, + "loss": 0.3162, + "step": 3481 + }, + { + "epoch": 0.48, + "grad_norm": 1.9084221489823747, + "learning_rate": 5.500097165540082e-06, + "loss": 0.3388, + "step": 3482 + }, + { + "epoch": 0.48, + "grad_norm": 1.990474149071576, + "learning_rate": 5.497857012376783e-06, + "loss": 0.3044, + "step": 3483 + }, + { + "epoch": 0.48, + "grad_norm": 3.3628836365147996, + "learning_rate": 5.4956167582726425e-06, + "loss": 0.3319, + "step": 3484 + }, + { + "epoch": 0.48, + "grad_norm": 2.207044941528813, + "learning_rate": 5.493376403681874e-06, + "loss": 0.3174, + "step": 3485 + }, + { + "epoch": 0.48, + "grad_norm": 2.026746203966414, + "learning_rate": 5.491135949058713e-06, + "loss": 0.3636, + "step": 3486 + }, + { + "epoch": 0.48, + "grad_norm": 1.9601795241552362, + "learning_rate": 5.488895394857407e-06, + "loss": 0.3274, + "step": 3487 + }, + { + "epoch": 0.48, + "grad_norm": 2.3639891844094905, + "learning_rate": 5.486654741532234e-06, + "loss": 0.3267, + "step": 3488 + }, + { + "epoch": 0.49, + "grad_norm": 2.3585414379988077, + "learning_rate": 5.484413989537489e-06, + "loss": 0.3287, + "step": 3489 + }, + { + "epoch": 0.49, + "grad_norm": 2.4644990363952886, + "learning_rate": 5.482173139327483e-06, + "loss": 0.2878, + "step": 3490 + }, + { + "epoch": 0.49, + "grad_norm": 2.2291827125132815, + "learning_rate": 5.479932191356554e-06, + "loss": 0.3345, + "step": 3491 + }, + { + "epoch": 0.49, + "grad_norm": 2.8329063131825336, + "learning_rate": 5.477691146079049e-06, + "loss": 0.3048, + "step": 3492 + }, + { + "epoch": 0.49, + "grad_norm": 2.581427770805799, + "learning_rate": 5.475450003949349e-06, + "loss": 0.3363, + "step": 3493 + }, + { + "epoch": 0.49, + "grad_norm": 2.2100155986247865, + "learning_rate": 5.473208765421841e-06, + "loss": 0.3257, + "step": 3494 + }, + { + "epoch": 0.49, + "grad_norm": 2.360121488964415, + "learning_rate": 5.470967430950942e-06, + "loss": 0.3406, + "step": 3495 + }, + { + "epoch": 0.49, + "grad_norm": 5.809790125328825, + "learning_rate": 5.46872600099108e-06, + "loss": 0.3549, + "step": 3496 + }, + { + "epoch": 0.49, + "grad_norm": 1.8800176562963493, + "learning_rate": 5.466484475996709e-06, + "loss": 0.322, + "step": 3497 + }, + { + "epoch": 0.49, + "grad_norm": 3.034864717740237, + "learning_rate": 5.4642428564223016e-06, + "loss": 0.3327, + "step": 3498 + }, + { + "epoch": 0.49, + "grad_norm": 2.903819101105612, + "learning_rate": 5.462001142722344e-06, + "loss": 0.3148, + "step": 3499 + }, + { + "epoch": 0.49, + "grad_norm": 2.3701945419730883, + "learning_rate": 5.459759335351345e-06, + "loss": 0.3173, + "step": 3500 + }, + { + "epoch": 0.49, + "grad_norm": 2.278862041642003, + "learning_rate": 5.457517434763835e-06, + "loss": 0.3188, + "step": 3501 + }, + { + "epoch": 0.49, + "grad_norm": 2.390730907188571, + "learning_rate": 5.45527544141436e-06, + "loss": 0.3063, + "step": 3502 + }, + { + "epoch": 0.49, + "grad_norm": 2.254715927931301, + "learning_rate": 5.4530333557574864e-06, + "loss": 0.354, + "step": 3503 + }, + { + "epoch": 0.49, + "grad_norm": 3.089732243508109, + "learning_rate": 5.450791178247795e-06, + "loss": 0.3251, + "step": 3504 + }, + { + "epoch": 0.49, + "grad_norm": 3.4942470936532812, + "learning_rate": 5.448548909339895e-06, + "loss": 0.3134, + "step": 3505 + }, + { + "epoch": 0.49, + "grad_norm": 1.7476413382902536, + "learning_rate": 5.4463065494884015e-06, + "loss": 0.2954, + "step": 3506 + }, + { + "epoch": 0.49, + "grad_norm": 1.9979368076094872, + "learning_rate": 5.444064099147958e-06, + "loss": 0.2759, + "step": 3507 + }, + { + "epoch": 0.49, + "grad_norm": 4.387600111464807, + "learning_rate": 5.441821558773224e-06, + "loss": 0.3238, + "step": 3508 + }, + { + "epoch": 0.49, + "grad_norm": 3.0903299826323436, + "learning_rate": 5.43957892881887e-06, + "loss": 0.3288, + "step": 3509 + }, + { + "epoch": 0.49, + "grad_norm": 2.00745512663032, + "learning_rate": 5.437336209739599e-06, + "loss": 0.3312, + "step": 3510 + }, + { + "epoch": 0.49, + "grad_norm": 1.7598129522201322, + "learning_rate": 5.435093401990118e-06, + "loss": 0.3358, + "step": 3511 + }, + { + "epoch": 0.49, + "grad_norm": 2.431568165605301, + "learning_rate": 5.43285050602516e-06, + "loss": 0.313, + "step": 3512 + }, + { + "epoch": 0.49, + "grad_norm": 2.839316150382499, + "learning_rate": 5.430607522299471e-06, + "loss": 0.3424, + "step": 3513 + }, + { + "epoch": 0.49, + "grad_norm": 4.629755656462268, + "learning_rate": 5.428364451267823e-06, + "loss": 0.3531, + "step": 3514 + }, + { + "epoch": 0.49, + "grad_norm": 2.081756401492246, + "learning_rate": 5.426121293384994e-06, + "loss": 0.3575, + "step": 3515 + }, + { + "epoch": 0.49, + "grad_norm": 2.1939009602560273, + "learning_rate": 5.4238780491057895e-06, + "loss": 0.3368, + "step": 3516 + }, + { + "epoch": 0.49, + "grad_norm": 2.3796256999081025, + "learning_rate": 5.421634718885027e-06, + "loss": 0.3573, + "step": 3517 + }, + { + "epoch": 0.49, + "grad_norm": 3.1483160117248055, + "learning_rate": 5.4193913031775445e-06, + "loss": 0.341, + "step": 3518 + }, + { + "epoch": 0.49, + "grad_norm": 1.9095018247914608, + "learning_rate": 5.417147802438196e-06, + "loss": 0.354, + "step": 3519 + }, + { + "epoch": 0.49, + "grad_norm": 2.495019796505762, + "learning_rate": 5.414904217121852e-06, + "loss": 0.3478, + "step": 3520 + }, + { + "epoch": 0.49, + "grad_norm": 2.266655438892548, + "learning_rate": 5.4126605476834e-06, + "loss": 0.3269, + "step": 3521 + }, + { + "epoch": 0.49, + "grad_norm": 2.252610627171713, + "learning_rate": 5.410416794577747e-06, + "loss": 0.3365, + "step": 3522 + }, + { + "epoch": 0.49, + "grad_norm": 12.978904003279974, + "learning_rate": 5.408172958259815e-06, + "loss": 0.3564, + "step": 3523 + }, + { + "epoch": 0.49, + "grad_norm": 2.1277257006717365, + "learning_rate": 5.405929039184545e-06, + "loss": 0.3202, + "step": 3524 + }, + { + "epoch": 0.49, + "grad_norm": 3.0622391240580256, + "learning_rate": 5.40368503780689e-06, + "loss": 0.323, + "step": 3525 + }, + { + "epoch": 0.49, + "grad_norm": 1.9907702613753995, + "learning_rate": 5.401440954581825e-06, + "loss": 0.3414, + "step": 3526 + }, + { + "epoch": 0.49, + "grad_norm": 3.275207237315393, + "learning_rate": 5.399196789964337e-06, + "loss": 0.2905, + "step": 3527 + }, + { + "epoch": 0.49, + "grad_norm": 2.2146098832852354, + "learning_rate": 5.396952544409435e-06, + "loss": 0.3489, + "step": 3528 + }, + { + "epoch": 0.49, + "grad_norm": 2.5680488723829233, + "learning_rate": 5.39470821837214e-06, + "loss": 0.3449, + "step": 3529 + }, + { + "epoch": 0.49, + "grad_norm": 2.3933258522814884, + "learning_rate": 5.392463812307488e-06, + "loss": 0.3281, + "step": 3530 + }, + { + "epoch": 0.49, + "grad_norm": 1.8747066218276203, + "learning_rate": 5.39021932667054e-06, + "loss": 0.326, + "step": 3531 + }, + { + "epoch": 0.49, + "grad_norm": 2.135414970988149, + "learning_rate": 5.38797476191636e-06, + "loss": 0.3442, + "step": 3532 + }, + { + "epoch": 0.49, + "grad_norm": 2.733636255437045, + "learning_rate": 5.3857301185000384e-06, + "loss": 0.2999, + "step": 3533 + }, + { + "epoch": 0.49, + "grad_norm": 2.5767317533377314, + "learning_rate": 5.38348539687668e-06, + "loss": 0.3111, + "step": 3534 + }, + { + "epoch": 0.49, + "grad_norm": 2.76075089036852, + "learning_rate": 5.3812405975013995e-06, + "loss": 0.3135, + "step": 3535 + }, + { + "epoch": 0.49, + "grad_norm": 3.708104643606495, + "learning_rate": 5.378995720829334e-06, + "loss": 0.3167, + "step": 3536 + }, + { + "epoch": 0.49, + "grad_norm": 2.6830819462827904, + "learning_rate": 5.376750767315634e-06, + "loss": 0.3356, + "step": 3537 + }, + { + "epoch": 0.49, + "grad_norm": 2.6237559694369743, + "learning_rate": 5.374505737415463e-06, + "loss": 0.3278, + "step": 3538 + }, + { + "epoch": 0.49, + "grad_norm": 2.1396754426857414, + "learning_rate": 5.372260631584006e-06, + "loss": 0.3553, + "step": 3539 + }, + { + "epoch": 0.49, + "grad_norm": 2.169704712302133, + "learning_rate": 5.370015450276456e-06, + "loss": 0.3113, + "step": 3540 + }, + { + "epoch": 0.49, + "grad_norm": 3.994396648116526, + "learning_rate": 5.3677701939480275e-06, + "loss": 0.3189, + "step": 3541 + }, + { + "epoch": 0.49, + "grad_norm": 2.090310093389935, + "learning_rate": 5.365524863053946e-06, + "loss": 0.345, + "step": 3542 + }, + { + "epoch": 0.49, + "grad_norm": 2.7048998499672154, + "learning_rate": 5.363279458049454e-06, + "loss": 0.3536, + "step": 3543 + }, + { + "epoch": 0.49, + "grad_norm": 2.1586740471225383, + "learning_rate": 5.36103397938981e-06, + "loss": 0.3335, + "step": 3544 + }, + { + "epoch": 0.49, + "grad_norm": 1.994523046270407, + "learning_rate": 5.358788427530286e-06, + "loss": 0.3187, + "step": 3545 + }, + { + "epoch": 0.49, + "grad_norm": 2.272077319829334, + "learning_rate": 5.35654280292617e-06, + "loss": 0.3092, + "step": 3546 + }, + { + "epoch": 0.49, + "grad_norm": 2.7062887047434745, + "learning_rate": 5.354297106032762e-06, + "loss": 0.2921, + "step": 3547 + }, + { + "epoch": 0.49, + "grad_norm": 2.6889408137527657, + "learning_rate": 5.35205133730538e-06, + "loss": 0.3293, + "step": 3548 + }, + { + "epoch": 0.49, + "grad_norm": 3.5253452224985797, + "learning_rate": 5.3498054971993525e-06, + "loss": 0.3293, + "step": 3549 + }, + { + "epoch": 0.49, + "grad_norm": 2.5121601090279464, + "learning_rate": 5.347559586170029e-06, + "loss": 0.3106, + "step": 3550 + }, + { + "epoch": 0.49, + "grad_norm": 2.7922953460055178, + "learning_rate": 5.345313604672767e-06, + "loss": 0.3227, + "step": 3551 + }, + { + "epoch": 0.49, + "grad_norm": 2.87242145506072, + "learning_rate": 5.34306755316294e-06, + "loss": 0.3066, + "step": 3552 + }, + { + "epoch": 0.49, + "grad_norm": 1.9125716197753517, + "learning_rate": 5.34082143209594e-06, + "loss": 0.3019, + "step": 3553 + }, + { + "epoch": 0.49, + "grad_norm": 2.1331184166384736, + "learning_rate": 5.338575241927165e-06, + "loss": 0.3405, + "step": 3554 + }, + { + "epoch": 0.49, + "grad_norm": 8.09786558477203, + "learning_rate": 5.336328983112036e-06, + "loss": 0.327, + "step": 3555 + }, + { + "epoch": 0.49, + "grad_norm": 3.092273016397871, + "learning_rate": 5.334082656105979e-06, + "loss": 0.3184, + "step": 3556 + }, + { + "epoch": 0.49, + "grad_norm": 2.3555477604986255, + "learning_rate": 5.3318362613644415e-06, + "loss": 0.3239, + "step": 3557 + }, + { + "epoch": 0.49, + "grad_norm": 2.295066828214714, + "learning_rate": 5.329589799342881e-06, + "loss": 0.3256, + "step": 3558 + }, + { + "epoch": 0.49, + "grad_norm": 2.629124948925884, + "learning_rate": 5.327343270496767e-06, + "loss": 0.3235, + "step": 3559 + }, + { + "epoch": 0.49, + "grad_norm": 2.8328957051312, + "learning_rate": 5.32509667528159e-06, + "loss": 0.3455, + "step": 3560 + }, + { + "epoch": 0.5, + "grad_norm": 2.8724845875774565, + "learning_rate": 5.322850014152842e-06, + "loss": 0.3294, + "step": 3561 + }, + { + "epoch": 0.5, + "grad_norm": 2.553544458629406, + "learning_rate": 5.320603287566039e-06, + "loss": 0.343, + "step": 3562 + }, + { + "epoch": 0.5, + "grad_norm": 1.9781208846860199, + "learning_rate": 5.318356495976706e-06, + "loss": 0.3315, + "step": 3563 + }, + { + "epoch": 0.5, + "grad_norm": 1.9062324482150947, + "learning_rate": 5.316109639840381e-06, + "loss": 0.3351, + "step": 3564 + }, + { + "epoch": 0.5, + "grad_norm": 2.912945090141197, + "learning_rate": 5.313862719612616e-06, + "loss": 0.3314, + "step": 3565 + }, + { + "epoch": 0.5, + "grad_norm": 2.7132390475088277, + "learning_rate": 5.311615735748974e-06, + "loss": 0.3044, + "step": 3566 + }, + { + "epoch": 0.5, + "grad_norm": 2.1238502804486377, + "learning_rate": 5.309368688705036e-06, + "loss": 0.3372, + "step": 3567 + }, + { + "epoch": 0.5, + "grad_norm": 2.353099655966241, + "learning_rate": 5.307121578936387e-06, + "loss": 0.3498, + "step": 3568 + }, + { + "epoch": 0.5, + "grad_norm": 2.317041812295594, + "learning_rate": 5.3048744068986345e-06, + "loss": 0.3232, + "step": 3569 + }, + { + "epoch": 0.5, + "grad_norm": 2.1199709333044208, + "learning_rate": 5.302627173047392e-06, + "loss": 0.3385, + "step": 3570 + }, + { + "epoch": 0.5, + "grad_norm": 1.9039759411537935, + "learning_rate": 5.300379877838289e-06, + "loss": 0.3404, + "step": 3571 + }, + { + "epoch": 0.5, + "grad_norm": 6.327241500072996, + "learning_rate": 5.298132521726965e-06, + "loss": 0.2928, + "step": 3572 + }, + { + "epoch": 0.5, + "grad_norm": 2.0274445962781367, + "learning_rate": 5.295885105169072e-06, + "loss": 0.342, + "step": 3573 + }, + { + "epoch": 0.5, + "grad_norm": 3.7432058733519935, + "learning_rate": 5.293637628620278e-06, + "loss": 0.3473, + "step": 3574 + }, + { + "epoch": 0.5, + "grad_norm": 2.516454680577046, + "learning_rate": 5.291390092536259e-06, + "loss": 0.331, + "step": 3575 + }, + { + "epoch": 0.5, + "grad_norm": 2.6737298167446286, + "learning_rate": 5.289142497372703e-06, + "loss": 0.3183, + "step": 3576 + }, + { + "epoch": 0.5, + "grad_norm": 2.2731523198184584, + "learning_rate": 5.286894843585315e-06, + "loss": 0.3062, + "step": 3577 + }, + { + "epoch": 0.5, + "grad_norm": 3.0978018936462477, + "learning_rate": 5.284647131629803e-06, + "loss": 0.3255, + "step": 3578 + }, + { + "epoch": 0.5, + "grad_norm": 3.4090608382550047, + "learning_rate": 5.282399361961898e-06, + "loss": 0.3217, + "step": 3579 + }, + { + "epoch": 0.5, + "grad_norm": 2.0435633734513066, + "learning_rate": 5.280151535037334e-06, + "loss": 0.3359, + "step": 3580 + }, + { + "epoch": 0.5, + "grad_norm": 2.7304066755688643, + "learning_rate": 5.277903651311858e-06, + "loss": 0.3291, + "step": 3581 + }, + { + "epoch": 0.5, + "grad_norm": 2.2300684487112368, + "learning_rate": 5.27565571124123e-06, + "loss": 0.3427, + "step": 3582 + }, + { + "epoch": 0.5, + "grad_norm": 2.7321782289460628, + "learning_rate": 5.273407715281224e-06, + "loss": 0.3328, + "step": 3583 + }, + { + "epoch": 0.5, + "grad_norm": 1.9465144183453025, + "learning_rate": 5.2711596638876214e-06, + "loss": 0.3304, + "step": 3584 + }, + { + "epoch": 0.5, + "grad_norm": 2.715427310641643, + "learning_rate": 5.2689115575162166e-06, + "loss": 0.3115, + "step": 3585 + }, + { + "epoch": 0.5, + "grad_norm": 2.4256157296867045, + "learning_rate": 5.266663396622813e-06, + "loss": 0.322, + "step": 3586 + }, + { + "epoch": 0.5, + "grad_norm": 2.548796705814579, + "learning_rate": 5.264415181663228e-06, + "loss": 0.3119, + "step": 3587 + }, + { + "epoch": 0.5, + "grad_norm": 2.3514735465417616, + "learning_rate": 5.26216691309329e-06, + "loss": 0.3124, + "step": 3588 + }, + { + "epoch": 0.5, + "grad_norm": 2.341955336664152, + "learning_rate": 5.259918591368833e-06, + "loss": 0.3278, + "step": 3589 + }, + { + "epoch": 0.5, + "grad_norm": 2.1656921651426657, + "learning_rate": 5.25767021694571e-06, + "loss": 0.348, + "step": 3590 + }, + { + "epoch": 0.5, + "grad_norm": 2.263137650711426, + "learning_rate": 5.255421790279778e-06, + "loss": 0.345, + "step": 3591 + }, + { + "epoch": 0.5, + "grad_norm": 2.859610445810728, + "learning_rate": 5.253173311826907e-06, + "loss": 0.3367, + "step": 3592 + }, + { + "epoch": 0.5, + "grad_norm": 2.8200193071442308, + "learning_rate": 5.250924782042979e-06, + "loss": 0.3201, + "step": 3593 + }, + { + "epoch": 0.5, + "grad_norm": 1.8281607769481147, + "learning_rate": 5.248676201383884e-06, + "loss": 0.3287, + "step": 3594 + }, + { + "epoch": 0.5, + "grad_norm": 3.0761986186590105, + "learning_rate": 5.246427570305522e-06, + "loss": 0.3206, + "step": 3595 + }, + { + "epoch": 0.5, + "grad_norm": 2.2191847433446723, + "learning_rate": 5.244178889263806e-06, + "loss": 0.333, + "step": 3596 + }, + { + "epoch": 0.5, + "grad_norm": 2.707940796863809, + "learning_rate": 5.241930158714658e-06, + "loss": 0.3094, + "step": 3597 + }, + { + "epoch": 0.5, + "grad_norm": 2.6478399644659465, + "learning_rate": 5.23968137911401e-06, + "loss": 0.3218, + "step": 3598 + }, + { + "epoch": 0.5, + "grad_norm": 2.2906882154204773, + "learning_rate": 5.237432550917799e-06, + "loss": 0.3136, + "step": 3599 + }, + { + "epoch": 0.5, + "grad_norm": 2.846212347524546, + "learning_rate": 5.235183674581984e-06, + "loss": 0.3115, + "step": 3600 + }, + { + "epoch": 0.5, + "grad_norm": 2.5881632227401234, + "learning_rate": 5.232934750562519e-06, + "loss": 0.3224, + "step": 3601 + }, + { + "epoch": 0.5, + "grad_norm": 2.771867394882285, + "learning_rate": 5.23068577931538e-06, + "loss": 0.3248, + "step": 3602 + }, + { + "epoch": 0.5, + "grad_norm": 2.8131769946710006, + "learning_rate": 5.228436761296545e-06, + "loss": 0.3134, + "step": 3603 + }, + { + "epoch": 0.5, + "grad_norm": 10.047768713947276, + "learning_rate": 5.226187696962002e-06, + "loss": 0.3289, + "step": 3604 + }, + { + "epoch": 0.5, + "grad_norm": 2.2630578293890693, + "learning_rate": 5.223938586767755e-06, + "loss": 0.3855, + "step": 3605 + }, + { + "epoch": 0.5, + "grad_norm": 2.0454815449598898, + "learning_rate": 5.2216894311698096e-06, + "loss": 0.3256, + "step": 3606 + }, + { + "epoch": 0.5, + "grad_norm": 2.1968171519368846, + "learning_rate": 5.219440230624183e-06, + "loss": 0.3433, + "step": 3607 + }, + { + "epoch": 0.5, + "grad_norm": 11.474319727206364, + "learning_rate": 5.2171909855869045e-06, + "loss": 0.3376, + "step": 3608 + }, + { + "epoch": 0.5, + "grad_norm": 2.340093104106848, + "learning_rate": 5.214941696514008e-06, + "loss": 0.3295, + "step": 3609 + }, + { + "epoch": 0.5, + "grad_norm": 2.075089330073494, + "learning_rate": 5.21269236386154e-06, + "loss": 0.3149, + "step": 3610 + }, + { + "epoch": 0.5, + "grad_norm": 3.6662002719036146, + "learning_rate": 5.210442988085553e-06, + "loss": 0.3432, + "step": 3611 + }, + { + "epoch": 0.5, + "grad_norm": 2.6965489463938286, + "learning_rate": 5.208193569642108e-06, + "loss": 0.3389, + "step": 3612 + }, + { + "epoch": 0.5, + "grad_norm": 1.9925459052168117, + "learning_rate": 5.2059441089872795e-06, + "loss": 0.351, + "step": 3613 + }, + { + "epoch": 0.5, + "grad_norm": 2.5887199007846475, + "learning_rate": 5.203694606577145e-06, + "loss": 0.3229, + "step": 3614 + }, + { + "epoch": 0.5, + "grad_norm": 3.201050020320482, + "learning_rate": 5.201445062867794e-06, + "loss": 0.3021, + "step": 3615 + }, + { + "epoch": 0.5, + "grad_norm": 3.40875823374161, + "learning_rate": 5.199195478315321e-06, + "loss": 0.357, + "step": 3616 + }, + { + "epoch": 0.5, + "grad_norm": 2.901481154444252, + "learning_rate": 5.196945853375832e-06, + "loss": 0.3416, + "step": 3617 + }, + { + "epoch": 0.5, + "grad_norm": 3.556027557586053, + "learning_rate": 5.19469618850544e-06, + "loss": 0.3188, + "step": 3618 + }, + { + "epoch": 0.5, + "grad_norm": 2.05065057312118, + "learning_rate": 5.192446484160267e-06, + "loss": 0.3188, + "step": 3619 + }, + { + "epoch": 0.5, + "grad_norm": 2.00262156577329, + "learning_rate": 5.190196740796441e-06, + "loss": 0.3109, + "step": 3620 + }, + { + "epoch": 0.5, + "grad_norm": 2.340772944962844, + "learning_rate": 5.187946958870097e-06, + "loss": 0.3197, + "step": 3621 + }, + { + "epoch": 0.5, + "grad_norm": 3.651742828161224, + "learning_rate": 5.1856971388373824e-06, + "loss": 0.3245, + "step": 3622 + }, + { + "epoch": 0.5, + "grad_norm": 1.9704857770843751, + "learning_rate": 5.183447281154449e-06, + "loss": 0.3114, + "step": 3623 + }, + { + "epoch": 0.5, + "grad_norm": 2.6813928703220222, + "learning_rate": 5.181197386277458e-06, + "loss": 0.3363, + "step": 3624 + }, + { + "epoch": 0.5, + "grad_norm": 2.0464872909227374, + "learning_rate": 5.178947454662573e-06, + "loss": 0.333, + "step": 3625 + }, + { + "epoch": 0.5, + "grad_norm": 1.9654796649589101, + "learning_rate": 5.176697486765976e-06, + "loss": 0.3399, + "step": 3626 + }, + { + "epoch": 0.5, + "grad_norm": 1.9459252535694413, + "learning_rate": 5.174447483043843e-06, + "loss": 0.341, + "step": 3627 + }, + { + "epoch": 0.5, + "grad_norm": 2.086496594163488, + "learning_rate": 5.172197443952367e-06, + "loss": 0.2967, + "step": 3628 + }, + { + "epoch": 0.5, + "grad_norm": 2.5199975062886715, + "learning_rate": 5.1699473699477465e-06, + "loss": 0.3224, + "step": 3629 + }, + { + "epoch": 0.5, + "grad_norm": 2.7501817738282432, + "learning_rate": 5.16769726148618e-06, + "loss": 0.3349, + "step": 3630 + }, + { + "epoch": 0.5, + "grad_norm": 2.7150817237791864, + "learning_rate": 5.165447119023884e-06, + "loss": 0.3451, + "step": 3631 + }, + { + "epoch": 0.5, + "grad_norm": 2.5087350874534744, + "learning_rate": 5.163196943017074e-06, + "loss": 0.3162, + "step": 3632 + }, + { + "epoch": 0.51, + "grad_norm": 2.515226920939397, + "learning_rate": 5.160946733921974e-06, + "loss": 0.3354, + "step": 3633 + }, + { + "epoch": 0.51, + "grad_norm": 2.6083265775049815, + "learning_rate": 5.1586964921948165e-06, + "loss": 0.3405, + "step": 3634 + }, + { + "epoch": 0.51, + "grad_norm": 3.5509332737518395, + "learning_rate": 5.156446218291839e-06, + "loss": 0.35, + "step": 3635 + }, + { + "epoch": 0.51, + "grad_norm": 2.60637316634183, + "learning_rate": 5.154195912669288e-06, + "loss": 0.3556, + "step": 3636 + }, + { + "epoch": 0.51, + "grad_norm": 2.401849041797173, + "learning_rate": 5.151945575783413e-06, + "loss": 0.3144, + "step": 3637 + }, + { + "epoch": 0.51, + "grad_norm": 3.3910427778664936, + "learning_rate": 5.149695208090469e-06, + "loss": 0.2956, + "step": 3638 + }, + { + "epoch": 0.51, + "grad_norm": 2.151937730174569, + "learning_rate": 5.147444810046722e-06, + "loss": 0.3148, + "step": 3639 + }, + { + "epoch": 0.51, + "grad_norm": 2.209646066376099, + "learning_rate": 5.145194382108442e-06, + "loss": 0.3343, + "step": 3640 + }, + { + "epoch": 0.51, + "grad_norm": 1.8065561988173204, + "learning_rate": 5.142943924731904e-06, + "loss": 0.3129, + "step": 3641 + }, + { + "epoch": 0.51, + "grad_norm": 2.0136584816164853, + "learning_rate": 5.14069343837339e-06, + "loss": 0.3122, + "step": 3642 + }, + { + "epoch": 0.51, + "grad_norm": 2.6882188411766554, + "learning_rate": 5.138442923489188e-06, + "loss": 0.3326, + "step": 3643 + }, + { + "epoch": 0.51, + "grad_norm": 2.4927372944070205, + "learning_rate": 5.136192380535591e-06, + "loss": 0.3368, + "step": 3644 + }, + { + "epoch": 0.51, + "grad_norm": 2.314042877565598, + "learning_rate": 5.133941809968897e-06, + "loss": 0.3178, + "step": 3645 + }, + { + "epoch": 0.51, + "grad_norm": 2.345065959824308, + "learning_rate": 5.131691212245414e-06, + "loss": 0.3419, + "step": 3646 + }, + { + "epoch": 0.51, + "grad_norm": 2.9040285109913393, + "learning_rate": 5.129440587821446e-06, + "loss": 0.3364, + "step": 3647 + }, + { + "epoch": 0.51, + "grad_norm": 2.5314465066680394, + "learning_rate": 5.1271899371533155e-06, + "loss": 0.3022, + "step": 3648 + }, + { + "epoch": 0.51, + "grad_norm": 4.427430509814211, + "learning_rate": 5.124939260697338e-06, + "loss": 0.3405, + "step": 3649 + }, + { + "epoch": 0.51, + "grad_norm": 2.6089048155983354, + "learning_rate": 5.122688558909844e-06, + "loss": 0.322, + "step": 3650 + }, + { + "epoch": 0.51, + "grad_norm": 2.2816326659887185, + "learning_rate": 5.120437832247162e-06, + "loss": 0.3187, + "step": 3651 + }, + { + "epoch": 0.51, + "grad_norm": 8.038542884779579, + "learning_rate": 5.118187081165628e-06, + "loss": 0.3214, + "step": 3652 + }, + { + "epoch": 0.51, + "grad_norm": 1.9806041199312323, + "learning_rate": 5.115936306121587e-06, + "loss": 0.3143, + "step": 3653 + }, + { + "epoch": 0.51, + "grad_norm": 2.5392905619811224, + "learning_rate": 5.113685507571381e-06, + "loss": 0.312, + "step": 3654 + }, + { + "epoch": 0.51, + "grad_norm": 1.9701605884786335, + "learning_rate": 5.111434685971362e-06, + "loss": 0.3315, + "step": 3655 + }, + { + "epoch": 0.51, + "grad_norm": 2.251323385877451, + "learning_rate": 5.109183841777887e-06, + "loss": 0.3411, + "step": 3656 + }, + { + "epoch": 0.51, + "grad_norm": 1.826477424880107, + "learning_rate": 5.1069329754473165e-06, + "loss": 0.3378, + "step": 3657 + }, + { + "epoch": 0.51, + "grad_norm": 2.321461667857506, + "learning_rate": 5.104682087436011e-06, + "loss": 0.3308, + "step": 3658 + }, + { + "epoch": 0.51, + "grad_norm": 2.794982064388319, + "learning_rate": 5.102431178200344e-06, + "loss": 0.3505, + "step": 3659 + }, + { + "epoch": 0.51, + "grad_norm": 2.212192770072912, + "learning_rate": 5.100180248196688e-06, + "loss": 0.3171, + "step": 3660 + }, + { + "epoch": 0.51, + "grad_norm": 2.170622597827225, + "learning_rate": 5.097929297881419e-06, + "loss": 0.3352, + "step": 3661 + }, + { + "epoch": 0.51, + "grad_norm": 2.14082296625287, + "learning_rate": 5.09567832771092e-06, + "loss": 0.321, + "step": 3662 + }, + { + "epoch": 0.51, + "grad_norm": 2.3340649089285006, + "learning_rate": 5.093427338141576e-06, + "loss": 0.3184, + "step": 3663 + }, + { + "epoch": 0.51, + "grad_norm": 2.1124211296958237, + "learning_rate": 5.091176329629775e-06, + "loss": 0.3332, + "step": 3664 + }, + { + "epoch": 0.51, + "grad_norm": 2.1984851141425743, + "learning_rate": 5.088925302631914e-06, + "loss": 0.3344, + "step": 3665 + }, + { + "epoch": 0.51, + "grad_norm": 2.0668948342447733, + "learning_rate": 5.0866742576043885e-06, + "loss": 0.3096, + "step": 3666 + }, + { + "epoch": 0.51, + "grad_norm": 6.613460307253313, + "learning_rate": 5.0844231950036e-06, + "loss": 0.3137, + "step": 3667 + }, + { + "epoch": 0.51, + "grad_norm": 2.023254299322644, + "learning_rate": 5.082172115285951e-06, + "loss": 0.2909, + "step": 3668 + }, + { + "epoch": 0.51, + "grad_norm": 2.1187522027445347, + "learning_rate": 5.079921018907852e-06, + "loss": 0.3331, + "step": 3669 + }, + { + "epoch": 0.51, + "grad_norm": 1.934976461968143, + "learning_rate": 5.0776699063257125e-06, + "loss": 0.3432, + "step": 3670 + }, + { + "epoch": 0.51, + "grad_norm": 1.8938830750874074, + "learning_rate": 5.075418777995948e-06, + "loss": 0.352, + "step": 3671 + }, + { + "epoch": 0.51, + "grad_norm": 2.7088909784846735, + "learning_rate": 5.073167634374977e-06, + "loss": 0.3317, + "step": 3672 + }, + { + "epoch": 0.51, + "grad_norm": 2.1306298439807034, + "learning_rate": 5.070916475919217e-06, + "loss": 0.3368, + "step": 3673 + }, + { + "epoch": 0.51, + "grad_norm": 2.403698176127534, + "learning_rate": 5.068665303085095e-06, + "loss": 0.3086, + "step": 3674 + }, + { + "epoch": 0.51, + "grad_norm": 2.185463490677054, + "learning_rate": 5.066414116329038e-06, + "loss": 0.3136, + "step": 3675 + }, + { + "epoch": 0.51, + "grad_norm": 2.306786515438456, + "learning_rate": 5.0641629161074735e-06, + "loss": 0.3326, + "step": 3676 + }, + { + "epoch": 0.51, + "grad_norm": 2.1135487621585454, + "learning_rate": 5.061911702876835e-06, + "loss": 0.3286, + "step": 3677 + }, + { + "epoch": 0.51, + "grad_norm": 1.8931326297375488, + "learning_rate": 5.059660477093558e-06, + "loss": 0.3279, + "step": 3678 + }, + { + "epoch": 0.51, + "grad_norm": 2.4254447377014854, + "learning_rate": 5.057409239214079e-06, + "loss": 0.3321, + "step": 3679 + }, + { + "epoch": 0.51, + "grad_norm": 3.340887631347657, + "learning_rate": 5.055157989694839e-06, + "loss": 0.3365, + "step": 3680 + }, + { + "epoch": 0.51, + "grad_norm": 2.5738552203793508, + "learning_rate": 5.05290672899228e-06, + "loss": 0.3088, + "step": 3681 + }, + { + "epoch": 0.51, + "grad_norm": 2.2344144193542053, + "learning_rate": 5.050655457562845e-06, + "loss": 0.3516, + "step": 3682 + }, + { + "epoch": 0.51, + "grad_norm": 2.4246080074596947, + "learning_rate": 5.048404175862983e-06, + "loss": 0.3117, + "step": 3683 + }, + { + "epoch": 0.51, + "grad_norm": 2.7633497371838796, + "learning_rate": 5.046152884349143e-06, + "loss": 0.332, + "step": 3684 + }, + { + "epoch": 0.51, + "grad_norm": 4.564597119979726, + "learning_rate": 5.043901583477774e-06, + "loss": 0.3261, + "step": 3685 + }, + { + "epoch": 0.51, + "grad_norm": 2.1711729325868427, + "learning_rate": 5.0416502737053284e-06, + "loss": 0.3243, + "step": 3686 + }, + { + "epoch": 0.51, + "grad_norm": 2.1464923478948883, + "learning_rate": 5.039398955488264e-06, + "loss": 0.3509, + "step": 3687 + }, + { + "epoch": 0.51, + "grad_norm": 2.024220628187801, + "learning_rate": 5.037147629283035e-06, + "loss": 0.3251, + "step": 3688 + }, + { + "epoch": 0.51, + "grad_norm": 4.78592416297691, + "learning_rate": 5.034896295546098e-06, + "loss": 0.3058, + "step": 3689 + }, + { + "epoch": 0.51, + "grad_norm": 2.1319856769408645, + "learning_rate": 5.032644954733913e-06, + "loss": 0.349, + "step": 3690 + }, + { + "epoch": 0.51, + "grad_norm": 3.9795387414323384, + "learning_rate": 5.030393607302945e-06, + "loss": 0.3197, + "step": 3691 + }, + { + "epoch": 0.51, + "grad_norm": 1.9918426008259824, + "learning_rate": 5.02814225370965e-06, + "loss": 0.3213, + "step": 3692 + }, + { + "epoch": 0.51, + "grad_norm": 2.5223867781066236, + "learning_rate": 5.0258908944104965e-06, + "loss": 0.2883, + "step": 3693 + }, + { + "epoch": 0.51, + "grad_norm": 1.9690550999482004, + "learning_rate": 5.023639529861944e-06, + "loss": 0.3128, + "step": 3694 + }, + { + "epoch": 0.51, + "grad_norm": 5.236678295038263, + "learning_rate": 5.021388160520464e-06, + "loss": 0.3177, + "step": 3695 + }, + { + "epoch": 0.51, + "grad_norm": 2.1009418039662053, + "learning_rate": 5.019136786842519e-06, + "loss": 0.3268, + "step": 3696 + }, + { + "epoch": 0.51, + "grad_norm": 8.649623370377583, + "learning_rate": 5.016885409284576e-06, + "loss": 0.3299, + "step": 3697 + }, + { + "epoch": 0.51, + "grad_norm": 2.0913172364288806, + "learning_rate": 5.014634028303108e-06, + "loss": 0.3034, + "step": 3698 + }, + { + "epoch": 0.51, + "grad_norm": 4.232349446349175, + "learning_rate": 5.012382644354579e-06, + "loss": 0.3068, + "step": 3699 + }, + { + "epoch": 0.51, + "grad_norm": 2.5652536768835135, + "learning_rate": 5.010131257895462e-06, + "loss": 0.3354, + "step": 3700 + }, + { + "epoch": 0.51, + "grad_norm": 2.407002832500989, + "learning_rate": 5.007879869382224e-06, + "loss": 0.3283, + "step": 3701 + }, + { + "epoch": 0.51, + "grad_norm": 3.7525160437946283, + "learning_rate": 5.005628479271338e-06, + "loss": 0.3189, + "step": 3702 + }, + { + "epoch": 0.51, + "grad_norm": 2.0802385877049367, + "learning_rate": 5.003377088019274e-06, + "loss": 0.2941, + "step": 3703 + }, + { + "epoch": 0.51, + "grad_norm": 2.6879946839130353, + "learning_rate": 5.001125696082504e-06, + "loss": 0.3388, + "step": 3704 + }, + { + "epoch": 0.52, + "grad_norm": 2.95651665050192, + "learning_rate": 4.998874303917497e-06, + "loss": 0.3245, + "step": 3705 + }, + { + "epoch": 0.52, + "grad_norm": 2.3819663422152253, + "learning_rate": 4.996622911980726e-06, + "loss": 0.3183, + "step": 3706 + }, + { + "epoch": 0.52, + "grad_norm": 2.1644722006356005, + "learning_rate": 4.994371520728663e-06, + "loss": 0.3266, + "step": 3707 + }, + { + "epoch": 0.52, + "grad_norm": 2.2694112976048193, + "learning_rate": 4.992120130617777e-06, + "loss": 0.3236, + "step": 3708 + }, + { + "epoch": 0.52, + "grad_norm": 2.1494765344742937, + "learning_rate": 4.98986874210454e-06, + "loss": 0.3432, + "step": 3709 + }, + { + "epoch": 0.52, + "grad_norm": 2.793861113149543, + "learning_rate": 4.987617355645423e-06, + "loss": 0.3674, + "step": 3710 + }, + { + "epoch": 0.52, + "grad_norm": 1.8923846290446171, + "learning_rate": 4.985365971696895e-06, + "loss": 0.3156, + "step": 3711 + }, + { + "epoch": 0.52, + "grad_norm": 2.185139821059049, + "learning_rate": 4.983114590715425e-06, + "loss": 0.3135, + "step": 3712 + }, + { + "epoch": 0.52, + "grad_norm": 2.1441691492485844, + "learning_rate": 4.980863213157484e-06, + "loss": 0.2975, + "step": 3713 + }, + { + "epoch": 0.52, + "grad_norm": 15.324815308614214, + "learning_rate": 4.978611839479537e-06, + "loss": 0.3254, + "step": 3714 + }, + { + "epoch": 0.52, + "grad_norm": 1.728244287844285, + "learning_rate": 4.9763604701380565e-06, + "loss": 0.3082, + "step": 3715 + }, + { + "epoch": 0.52, + "grad_norm": 3.785166774823947, + "learning_rate": 4.974109105589505e-06, + "loss": 0.311, + "step": 3716 + }, + { + "epoch": 0.52, + "grad_norm": 2.1672363981922462, + "learning_rate": 4.9718577462903515e-06, + "loss": 0.3062, + "step": 3717 + }, + { + "epoch": 0.52, + "grad_norm": 2.1658964814392068, + "learning_rate": 4.9696063926970574e-06, + "loss": 0.3182, + "step": 3718 + }, + { + "epoch": 0.52, + "grad_norm": 2.339885126226148, + "learning_rate": 4.967355045266088e-06, + "loss": 0.3345, + "step": 3719 + }, + { + "epoch": 0.52, + "grad_norm": 3.877350986928417, + "learning_rate": 4.965103704453904e-06, + "loss": 0.3203, + "step": 3720 + }, + { + "epoch": 0.52, + "grad_norm": 2.020223992281288, + "learning_rate": 4.962852370716968e-06, + "loss": 0.296, + "step": 3721 + }, + { + "epoch": 0.52, + "grad_norm": 2.6893232702163807, + "learning_rate": 4.960601044511737e-06, + "loss": 0.3272, + "step": 3722 + }, + { + "epoch": 0.52, + "grad_norm": 2.1221743134513833, + "learning_rate": 4.9583497262946715e-06, + "loss": 0.2784, + "step": 3723 + }, + { + "epoch": 0.52, + "grad_norm": 3.304871719237621, + "learning_rate": 4.956098416522228e-06, + "loss": 0.317, + "step": 3724 + }, + { + "epoch": 0.52, + "grad_norm": 1.817191947985422, + "learning_rate": 4.953847115650859e-06, + "loss": 0.3551, + "step": 3725 + }, + { + "epoch": 0.52, + "grad_norm": 2.999706157135499, + "learning_rate": 4.951595824137019e-06, + "loss": 0.2824, + "step": 3726 + }, + { + "epoch": 0.52, + "grad_norm": 2.3943988029961623, + "learning_rate": 4.9493445424371564e-06, + "loss": 0.3312, + "step": 3727 + }, + { + "epoch": 0.52, + "grad_norm": 1.8722954203964466, + "learning_rate": 4.947093271007723e-06, + "loss": 0.3421, + "step": 3728 + }, + { + "epoch": 0.52, + "grad_norm": 2.22458599574498, + "learning_rate": 4.944842010305163e-06, + "loss": 0.3228, + "step": 3729 + }, + { + "epoch": 0.52, + "grad_norm": 1.9555614829145145, + "learning_rate": 4.942590760785923e-06, + "loss": 0.3076, + "step": 3730 + }, + { + "epoch": 0.52, + "grad_norm": 2.958793476138508, + "learning_rate": 4.940339522906443e-06, + "loss": 0.3476, + "step": 3731 + }, + { + "epoch": 0.52, + "grad_norm": 2.8086876654787547, + "learning_rate": 4.938088297123165e-06, + "loss": 0.3025, + "step": 3732 + }, + { + "epoch": 0.52, + "grad_norm": 2.0195651175152376, + "learning_rate": 4.935837083892528e-06, + "loss": 0.3491, + "step": 3733 + }, + { + "epoch": 0.52, + "grad_norm": 1.7855442349225201, + "learning_rate": 4.933585883670964e-06, + "loss": 0.3277, + "step": 3734 + }, + { + "epoch": 0.52, + "grad_norm": 2.2476212454012363, + "learning_rate": 4.931334696914906e-06, + "loss": 0.3119, + "step": 3735 + }, + { + "epoch": 0.52, + "grad_norm": 1.7109370153341783, + "learning_rate": 4.929083524080785e-06, + "loss": 0.3452, + "step": 3736 + }, + { + "epoch": 0.52, + "grad_norm": 2.9921953711829947, + "learning_rate": 4.926832365625026e-06, + "loss": 0.326, + "step": 3737 + }, + { + "epoch": 0.52, + "grad_norm": 1.9270904136855975, + "learning_rate": 4.924581222004053e-06, + "loss": 0.3292, + "step": 3738 + }, + { + "epoch": 0.52, + "grad_norm": 1.7246046519851355, + "learning_rate": 4.9223300936742874e-06, + "loss": 0.3107, + "step": 3739 + }, + { + "epoch": 0.52, + "grad_norm": 2.146638534355419, + "learning_rate": 4.920078981092149e-06, + "loss": 0.3245, + "step": 3740 + }, + { + "epoch": 0.52, + "grad_norm": 1.9574099285456468, + "learning_rate": 4.917827884714049e-06, + "loss": 0.3387, + "step": 3741 + }, + { + "epoch": 0.52, + "grad_norm": 2.0227546134258443, + "learning_rate": 4.9155768049964006e-06, + "loss": 0.319, + "step": 3742 + }, + { + "epoch": 0.52, + "grad_norm": 2.440638276183815, + "learning_rate": 4.913325742395612e-06, + "loss": 0.3258, + "step": 3743 + }, + { + "epoch": 0.52, + "grad_norm": 8.347677111147293, + "learning_rate": 4.911074697368087e-06, + "loss": 0.3128, + "step": 3744 + }, + { + "epoch": 0.52, + "grad_norm": 2.7919104826827668, + "learning_rate": 4.908823670370225e-06, + "loss": 0.3267, + "step": 3745 + }, + { + "epoch": 0.52, + "grad_norm": 3.1389049730783767, + "learning_rate": 4.906572661858428e-06, + "loss": 0.3181, + "step": 3746 + }, + { + "epoch": 0.52, + "grad_norm": 2.3553479719588433, + "learning_rate": 4.904321672289083e-06, + "loss": 0.3426, + "step": 3747 + }, + { + "epoch": 0.52, + "grad_norm": 2.528060430685631, + "learning_rate": 4.902070702118582e-06, + "loss": 0.3355, + "step": 3748 + }, + { + "epoch": 0.52, + "grad_norm": 2.447016392284756, + "learning_rate": 4.899819751803313e-06, + "loss": 0.32, + "step": 3749 + }, + { + "epoch": 0.52, + "grad_norm": 2.067288744664771, + "learning_rate": 4.897568821799657e-06, + "loss": 0.2971, + "step": 3750 + }, + { + "epoch": 0.52, + "grad_norm": 2.1623245028723166, + "learning_rate": 4.895317912563989e-06, + "loss": 0.3074, + "step": 3751 + }, + { + "epoch": 0.52, + "grad_norm": 2.3719480067100345, + "learning_rate": 4.893067024552685e-06, + "loss": 0.3217, + "step": 3752 + }, + { + "epoch": 0.52, + "grad_norm": 2.2934127689832353, + "learning_rate": 4.890816158222114e-06, + "loss": 0.3399, + "step": 3753 + }, + { + "epoch": 0.52, + "grad_norm": 2.103118160127973, + "learning_rate": 4.8885653140286384e-06, + "loss": 0.351, + "step": 3754 + }, + { + "epoch": 0.52, + "grad_norm": 1.7436674234414276, + "learning_rate": 4.886314492428621e-06, + "loss": 0.3243, + "step": 3755 + }, + { + "epoch": 0.52, + "grad_norm": 4.317913578057225, + "learning_rate": 4.884063693878416e-06, + "loss": 0.3356, + "step": 3756 + }, + { + "epoch": 0.52, + "grad_norm": 2.7175664132569954, + "learning_rate": 4.881812918834372e-06, + "loss": 0.3268, + "step": 3757 + }, + { + "epoch": 0.52, + "grad_norm": 3.2351920991585175, + "learning_rate": 4.879562167752839e-06, + "loss": 0.3474, + "step": 3758 + }, + { + "epoch": 0.52, + "grad_norm": 2.325697959779574, + "learning_rate": 4.877311441090157e-06, + "loss": 0.3297, + "step": 3759 + }, + { + "epoch": 0.52, + "grad_norm": 1.8676601873525158, + "learning_rate": 4.875060739302664e-06, + "loss": 0.2977, + "step": 3760 + }, + { + "epoch": 0.52, + "grad_norm": 2.9050276519932074, + "learning_rate": 4.872810062846687e-06, + "loss": 0.3232, + "step": 3761 + }, + { + "epoch": 0.52, + "grad_norm": 2.4880381883320126, + "learning_rate": 4.870559412178556e-06, + "loss": 0.3348, + "step": 3762 + }, + { + "epoch": 0.52, + "grad_norm": 2.726015438313076, + "learning_rate": 4.86830878775459e-06, + "loss": 0.3418, + "step": 3763 + }, + { + "epoch": 0.52, + "grad_norm": 1.991425987306654, + "learning_rate": 4.866058190031104e-06, + "loss": 0.3245, + "step": 3764 + }, + { + "epoch": 0.52, + "grad_norm": 4.1448169421740015, + "learning_rate": 4.863807619464409e-06, + "loss": 0.3229, + "step": 3765 + }, + { + "epoch": 0.52, + "grad_norm": 2.20165152328002, + "learning_rate": 4.8615570765108124e-06, + "loss": 0.3334, + "step": 3766 + }, + { + "epoch": 0.52, + "grad_norm": 2.170750747050972, + "learning_rate": 4.85930656162661e-06, + "loss": 0.3256, + "step": 3767 + }, + { + "epoch": 0.52, + "grad_norm": 2.1190073224214228, + "learning_rate": 4.8570560752680966e-06, + "loss": 0.3461, + "step": 3768 + }, + { + "epoch": 0.52, + "grad_norm": 2.404311415265855, + "learning_rate": 4.854805617891559e-06, + "loss": 0.325, + "step": 3769 + }, + { + "epoch": 0.52, + "grad_norm": 2.5556235324238044, + "learning_rate": 4.852555189953279e-06, + "loss": 0.3277, + "step": 3770 + }, + { + "epoch": 0.52, + "grad_norm": 1.961144289943782, + "learning_rate": 4.850304791909532e-06, + "loss": 0.3324, + "step": 3771 + }, + { + "epoch": 0.52, + "grad_norm": 2.119084889614373, + "learning_rate": 4.848054424216591e-06, + "loss": 0.3413, + "step": 3772 + }, + { + "epoch": 0.52, + "grad_norm": 1.9872491230773357, + "learning_rate": 4.845804087330714e-06, + "loss": 0.3489, + "step": 3773 + }, + { + "epoch": 0.52, + "grad_norm": 2.2918295649509, + "learning_rate": 4.843553781708161e-06, + "loss": 0.3061, + "step": 3774 + }, + { + "epoch": 0.52, + "grad_norm": 2.3903909723111667, + "learning_rate": 4.841303507805184e-06, + "loss": 0.3191, + "step": 3775 + }, + { + "epoch": 0.52, + "grad_norm": 2.3567560775729905, + "learning_rate": 4.8390532660780275e-06, + "loss": 0.3308, + "step": 3776 + }, + { + "epoch": 0.53, + "grad_norm": 3.7418802247621414, + "learning_rate": 4.836803056982928e-06, + "loss": 0.3329, + "step": 3777 + }, + { + "epoch": 0.53, + "grad_norm": 2.078500196589512, + "learning_rate": 4.834552880976117e-06, + "loss": 0.3227, + "step": 3778 + }, + { + "epoch": 0.53, + "grad_norm": 2.547158703955047, + "learning_rate": 4.832302738513822e-06, + "loss": 0.3027, + "step": 3779 + }, + { + "epoch": 0.53, + "grad_norm": 4.659522638351441, + "learning_rate": 4.830052630052257e-06, + "loss": 0.3203, + "step": 3780 + }, + { + "epoch": 0.53, + "grad_norm": 1.9891127819894645, + "learning_rate": 4.827802556047634e-06, + "loss": 0.3287, + "step": 3781 + }, + { + "epoch": 0.53, + "grad_norm": 2.296317862620574, + "learning_rate": 4.825552516956159e-06, + "loss": 0.3289, + "step": 3782 + }, + { + "epoch": 0.53, + "grad_norm": 3.884020083839976, + "learning_rate": 4.823302513234026e-06, + "loss": 0.3425, + "step": 3783 + }, + { + "epoch": 0.53, + "grad_norm": 2.09488715863694, + "learning_rate": 4.821052545337427e-06, + "loss": 0.337, + "step": 3784 + }, + { + "epoch": 0.53, + "grad_norm": 1.7307125348867034, + "learning_rate": 4.818802613722544e-06, + "loss": 0.3158, + "step": 3785 + }, + { + "epoch": 0.53, + "grad_norm": 3.2866552873908748, + "learning_rate": 4.816552718845553e-06, + "loss": 0.3133, + "step": 3786 + }, + { + "epoch": 0.53, + "grad_norm": 2.191523197216044, + "learning_rate": 4.814302861162619e-06, + "loss": 0.2949, + "step": 3787 + }, + { + "epoch": 0.53, + "grad_norm": 1.8812162046510557, + "learning_rate": 4.812053041129905e-06, + "loss": 0.3358, + "step": 3788 + }, + { + "epoch": 0.53, + "grad_norm": 1.9711129276622181, + "learning_rate": 4.8098032592035625e-06, + "loss": 0.3243, + "step": 3789 + }, + { + "epoch": 0.53, + "grad_norm": 2.377104582330529, + "learning_rate": 4.807553515839735e-06, + "loss": 0.326, + "step": 3790 + }, + { + "epoch": 0.53, + "grad_norm": 2.323810481238308, + "learning_rate": 4.80530381149456e-06, + "loss": 0.3228, + "step": 3791 + }, + { + "epoch": 0.53, + "grad_norm": 2.119391015988449, + "learning_rate": 4.803054146624168e-06, + "loss": 0.3294, + "step": 3792 + }, + { + "epoch": 0.53, + "grad_norm": 2.53016175910288, + "learning_rate": 4.80080452168468e-06, + "loss": 0.3128, + "step": 3793 + }, + { + "epoch": 0.53, + "grad_norm": 1.7341794983134566, + "learning_rate": 4.798554937132207e-06, + "loss": 0.32, + "step": 3794 + }, + { + "epoch": 0.53, + "grad_norm": 2.9970836434511647, + "learning_rate": 4.796305393422856e-06, + "loss": 0.3468, + "step": 3795 + }, + { + "epoch": 0.53, + "grad_norm": 2.2626376558892147, + "learning_rate": 4.794055891012722e-06, + "loss": 0.313, + "step": 3796 + }, + { + "epoch": 0.53, + "grad_norm": 1.9375544217617904, + "learning_rate": 4.791806430357893e-06, + "loss": 0.3217, + "step": 3797 + }, + { + "epoch": 0.53, + "grad_norm": 1.8682758816704095, + "learning_rate": 4.78955701191445e-06, + "loss": 0.3045, + "step": 3798 + }, + { + "epoch": 0.53, + "grad_norm": 2.935662424094154, + "learning_rate": 4.787307636138463e-06, + "loss": 0.3163, + "step": 3799 + }, + { + "epoch": 0.53, + "grad_norm": 2.480249993391537, + "learning_rate": 4.785058303485993e-06, + "loss": 0.308, + "step": 3800 + }, + { + "epoch": 0.53, + "grad_norm": 2.40614089196602, + "learning_rate": 4.782809014413096e-06, + "loss": 0.3196, + "step": 3801 + }, + { + "epoch": 0.53, + "grad_norm": 2.1757901802117, + "learning_rate": 4.780559769375818e-06, + "loss": 0.3012, + "step": 3802 + }, + { + "epoch": 0.53, + "grad_norm": 2.5989647588513973, + "learning_rate": 4.778310568830192e-06, + "loss": 0.3334, + "step": 3803 + }, + { + "epoch": 0.53, + "grad_norm": 3.466247869856953, + "learning_rate": 4.776061413232247e-06, + "loss": 0.3216, + "step": 3804 + }, + { + "epoch": 0.53, + "grad_norm": 2.246751317197353, + "learning_rate": 4.773812303038e-06, + "loss": 0.3148, + "step": 3805 + }, + { + "epoch": 0.53, + "grad_norm": 9.931552226321953, + "learning_rate": 4.7715632387034575e-06, + "loss": 0.3159, + "step": 3806 + }, + { + "epoch": 0.53, + "grad_norm": 2.435523891682708, + "learning_rate": 4.769314220684622e-06, + "loss": 0.3379, + "step": 3807 + }, + { + "epoch": 0.53, + "grad_norm": 2.6407859997836094, + "learning_rate": 4.767065249437481e-06, + "loss": 0.3091, + "step": 3808 + }, + { + "epoch": 0.53, + "grad_norm": 4.098057078510977, + "learning_rate": 4.764816325418018e-06, + "loss": 0.3376, + "step": 3809 + }, + { + "epoch": 0.53, + "grad_norm": 1.8320344510179338, + "learning_rate": 4.7625674490822e-06, + "loss": 0.3219, + "step": 3810 + }, + { + "epoch": 0.53, + "grad_norm": 2.0848381996189156, + "learning_rate": 4.7603186208859916e-06, + "loss": 0.3158, + "step": 3811 + }, + { + "epoch": 0.53, + "grad_norm": 2.3829758532360863, + "learning_rate": 4.758069841285343e-06, + "loss": 0.3133, + "step": 3812 + }, + { + "epoch": 0.53, + "grad_norm": 1.6618454335240296, + "learning_rate": 4.755821110736195e-06, + "loss": 0.2949, + "step": 3813 + }, + { + "epoch": 0.53, + "grad_norm": 1.8260403797144735, + "learning_rate": 4.753572429694479e-06, + "loss": 0.2908, + "step": 3814 + }, + { + "epoch": 0.53, + "grad_norm": 2.0165262754423257, + "learning_rate": 4.751323798616119e-06, + "loss": 0.3248, + "step": 3815 + }, + { + "epoch": 0.53, + "grad_norm": 1.9976599035013407, + "learning_rate": 4.749075217957024e-06, + "loss": 0.3131, + "step": 3816 + }, + { + "epoch": 0.53, + "grad_norm": 1.9166085774639534, + "learning_rate": 4.746826688173094e-06, + "loss": 0.3215, + "step": 3817 + }, + { + "epoch": 0.53, + "grad_norm": 1.9752146039439609, + "learning_rate": 4.744578209720223e-06, + "loss": 0.2996, + "step": 3818 + }, + { + "epoch": 0.53, + "grad_norm": 2.2198529483726688, + "learning_rate": 4.742329783054292e-06, + "loss": 0.3367, + "step": 3819 + }, + { + "epoch": 0.53, + "grad_norm": 2.536027933051609, + "learning_rate": 4.740081408631168e-06, + "loss": 0.3259, + "step": 3820 + }, + { + "epoch": 0.53, + "grad_norm": 2.1248353593176876, + "learning_rate": 4.737833086906712e-06, + "loss": 0.3031, + "step": 3821 + }, + { + "epoch": 0.53, + "grad_norm": 2.407312782365516, + "learning_rate": 4.735584818336773e-06, + "loss": 0.3405, + "step": 3822 + }, + { + "epoch": 0.53, + "grad_norm": 3.205014556139283, + "learning_rate": 4.733336603377189e-06, + "loss": 0.3174, + "step": 3823 + }, + { + "epoch": 0.53, + "grad_norm": 2.155497725120881, + "learning_rate": 4.731088442483787e-06, + "loss": 0.3336, + "step": 3824 + }, + { + "epoch": 0.53, + "grad_norm": 2.3451441361641234, + "learning_rate": 4.728840336112381e-06, + "loss": 0.2976, + "step": 3825 + }, + { + "epoch": 0.53, + "grad_norm": 2.5642869115627445, + "learning_rate": 4.726592284718777e-06, + "loss": 0.314, + "step": 3826 + }, + { + "epoch": 0.53, + "grad_norm": 1.8715333239006549, + "learning_rate": 4.724344288758771e-06, + "loss": 0.3179, + "step": 3827 + }, + { + "epoch": 0.53, + "grad_norm": 3.7429819943269886, + "learning_rate": 4.722096348688143e-06, + "loss": 0.341, + "step": 3828 + }, + { + "epoch": 0.53, + "grad_norm": 3.2642315628563905, + "learning_rate": 4.719848464962669e-06, + "loss": 0.328, + "step": 3829 + }, + { + "epoch": 0.53, + "grad_norm": 2.1954589968131106, + "learning_rate": 4.717600638038103e-06, + "loss": 0.3093, + "step": 3830 + }, + { + "epoch": 0.53, + "grad_norm": 2.422194963678918, + "learning_rate": 4.715352868370198e-06, + "loss": 0.2949, + "step": 3831 + }, + { + "epoch": 0.53, + "grad_norm": 2.6677871882014763, + "learning_rate": 4.713105156414687e-06, + "loss": 0.3335, + "step": 3832 + }, + { + "epoch": 0.53, + "grad_norm": 2.218207702567347, + "learning_rate": 4.710857502627298e-06, + "loss": 0.3296, + "step": 3833 + }, + { + "epoch": 0.53, + "grad_norm": 2.5509890202450256, + "learning_rate": 4.708609907463742e-06, + "loss": 0.349, + "step": 3834 + }, + { + "epoch": 0.53, + "grad_norm": 2.4103018812927646, + "learning_rate": 4.706362371379722e-06, + "loss": 0.3352, + "step": 3835 + }, + { + "epoch": 0.53, + "grad_norm": 2.5383836919204157, + "learning_rate": 4.704114894830929e-06, + "loss": 0.3437, + "step": 3836 + }, + { + "epoch": 0.53, + "grad_norm": 5.438403935149058, + "learning_rate": 4.701867478273036e-06, + "loss": 0.3292, + "step": 3837 + }, + { + "epoch": 0.53, + "grad_norm": 2.114672308505817, + "learning_rate": 4.699620122161714e-06, + "loss": 0.3276, + "step": 3838 + }, + { + "epoch": 0.53, + "grad_norm": 2.241976575295351, + "learning_rate": 4.697372826952609e-06, + "loss": 0.3111, + "step": 3839 + }, + { + "epoch": 0.53, + "grad_norm": 1.8430113685775693, + "learning_rate": 4.695125593101367e-06, + "loss": 0.3112, + "step": 3840 + }, + { + "epoch": 0.53, + "grad_norm": 2.1561773421854875, + "learning_rate": 4.6928784210636155e-06, + "loss": 0.3035, + "step": 3841 + }, + { + "epoch": 0.53, + "grad_norm": 5.781420162824833, + "learning_rate": 4.690631311294967e-06, + "loss": 0.327, + "step": 3842 + }, + { + "epoch": 0.53, + "grad_norm": 2.5251067681514594, + "learning_rate": 4.688384264251026e-06, + "loss": 0.3318, + "step": 3843 + }, + { + "epoch": 0.53, + "grad_norm": 2.2996549494819227, + "learning_rate": 4.686137280387384e-06, + "loss": 0.353, + "step": 3844 + }, + { + "epoch": 0.53, + "grad_norm": 2.290496361674193, + "learning_rate": 4.68389036015962e-06, + "loss": 0.3303, + "step": 3845 + }, + { + "epoch": 0.53, + "grad_norm": 1.9701705580036755, + "learning_rate": 4.681643504023295e-06, + "loss": 0.3236, + "step": 3846 + }, + { + "epoch": 0.53, + "grad_norm": 3.2499936312092075, + "learning_rate": 4.679396712433962e-06, + "loss": 0.3277, + "step": 3847 + }, + { + "epoch": 0.53, + "grad_norm": 2.777660878629059, + "learning_rate": 4.67714998584716e-06, + "loss": 0.3561, + "step": 3848 + }, + { + "epoch": 0.54, + "grad_norm": 2.3657817877731895, + "learning_rate": 4.674903324718412e-06, + "loss": 0.3575, + "step": 3849 + }, + { + "epoch": 0.54, + "grad_norm": 2.673371195856261, + "learning_rate": 4.6726567295032335e-06, + "loss": 0.3245, + "step": 3850 + }, + { + "epoch": 0.54, + "grad_norm": 2.345111671095542, + "learning_rate": 4.670410200657121e-06, + "loss": 0.3097, + "step": 3851 + }, + { + "epoch": 0.54, + "grad_norm": 1.8332703653185807, + "learning_rate": 4.668163738635559e-06, + "loss": 0.3385, + "step": 3852 + }, + { + "epoch": 0.54, + "grad_norm": 1.9828474552662452, + "learning_rate": 4.665917343894022e-06, + "loss": 0.3051, + "step": 3853 + }, + { + "epoch": 0.54, + "grad_norm": 2.290276220632295, + "learning_rate": 4.663671016887966e-06, + "loss": 0.2861, + "step": 3854 + }, + { + "epoch": 0.54, + "grad_norm": 2.601978881295107, + "learning_rate": 4.661424758072836e-06, + "loss": 0.3046, + "step": 3855 + }, + { + "epoch": 0.54, + "grad_norm": 2.3862281233246, + "learning_rate": 4.659178567904062e-06, + "loss": 0.3068, + "step": 3856 + }, + { + "epoch": 0.54, + "grad_norm": 2.6362952573984955, + "learning_rate": 4.65693244683706e-06, + "loss": 0.3519, + "step": 3857 + }, + { + "epoch": 0.54, + "grad_norm": 2.103243559384056, + "learning_rate": 4.6546863953272355e-06, + "loss": 0.3455, + "step": 3858 + }, + { + "epoch": 0.54, + "grad_norm": 2.239831832957138, + "learning_rate": 4.6524404138299725e-06, + "loss": 0.3021, + "step": 3859 + }, + { + "epoch": 0.54, + "grad_norm": 2.4075784773604645, + "learning_rate": 4.6501945028006474e-06, + "loss": 0.3299, + "step": 3860 + }, + { + "epoch": 0.54, + "grad_norm": 2.8801526925126213, + "learning_rate": 4.647948662694621e-06, + "loss": 0.3255, + "step": 3861 + }, + { + "epoch": 0.54, + "grad_norm": 2.0545039205601805, + "learning_rate": 4.645702893967239e-06, + "loss": 0.3498, + "step": 3862 + }, + { + "epoch": 0.54, + "grad_norm": 2.625400790852262, + "learning_rate": 4.643457197073831e-06, + "loss": 0.3424, + "step": 3863 + }, + { + "epoch": 0.54, + "grad_norm": 2.7147948708080594, + "learning_rate": 4.641211572469715e-06, + "loss": 0.3302, + "step": 3864 + }, + { + "epoch": 0.54, + "grad_norm": 1.9997116630442864, + "learning_rate": 4.6389660206101915e-06, + "loss": 0.3165, + "step": 3865 + }, + { + "epoch": 0.54, + "grad_norm": 2.0531488951635355, + "learning_rate": 4.636720541950547e-06, + "loss": 0.3135, + "step": 3866 + }, + { + "epoch": 0.54, + "grad_norm": 2.434571265788574, + "learning_rate": 4.634475136946057e-06, + "loss": 0.3163, + "step": 3867 + }, + { + "epoch": 0.54, + "grad_norm": 2.0868370145060124, + "learning_rate": 4.632229806051976e-06, + "loss": 0.328, + "step": 3868 + }, + { + "epoch": 0.54, + "grad_norm": 1.8138653837090315, + "learning_rate": 4.629984549723545e-06, + "loss": 0.3183, + "step": 3869 + }, + { + "epoch": 0.54, + "grad_norm": 1.8015090087463115, + "learning_rate": 4.627739368415995e-06, + "loss": 0.332, + "step": 3870 + }, + { + "epoch": 0.54, + "grad_norm": 3.520866049441887, + "learning_rate": 4.625494262584538e-06, + "loss": 0.3307, + "step": 3871 + }, + { + "epoch": 0.54, + "grad_norm": 2.320164268273901, + "learning_rate": 4.623249232684368e-06, + "loss": 0.326, + "step": 3872 + }, + { + "epoch": 0.54, + "grad_norm": 2.2889914742789483, + "learning_rate": 4.621004279170667e-06, + "loss": 0.3382, + "step": 3873 + }, + { + "epoch": 0.54, + "grad_norm": 1.8072127599405217, + "learning_rate": 4.618759402498603e-06, + "loss": 0.3138, + "step": 3874 + }, + { + "epoch": 0.54, + "grad_norm": 2.1099807566657556, + "learning_rate": 4.616514603123322e-06, + "loss": 0.3481, + "step": 3875 + }, + { + "epoch": 0.54, + "grad_norm": 2.020932723755262, + "learning_rate": 4.614269881499962e-06, + "loss": 0.327, + "step": 3876 + }, + { + "epoch": 0.54, + "grad_norm": 1.9062849766878833, + "learning_rate": 4.612025238083641e-06, + "loss": 0.3076, + "step": 3877 + }, + { + "epoch": 0.54, + "grad_norm": 2.243293695968906, + "learning_rate": 4.609780673329463e-06, + "loss": 0.3302, + "step": 3878 + }, + { + "epoch": 0.54, + "grad_norm": 2.0410284178766758, + "learning_rate": 4.6075361876925126e-06, + "loss": 0.3193, + "step": 3879 + }, + { + "epoch": 0.54, + "grad_norm": 2.3345243613998776, + "learning_rate": 4.605291781627862e-06, + "loss": 0.296, + "step": 3880 + }, + { + "epoch": 0.54, + "grad_norm": 2.3098215639938426, + "learning_rate": 4.603047455590567e-06, + "loss": 0.3282, + "step": 3881 + }, + { + "epoch": 0.54, + "grad_norm": 1.632047417520254, + "learning_rate": 4.600803210035664e-06, + "loss": 0.3341, + "step": 3882 + }, + { + "epoch": 0.54, + "grad_norm": 2.487572671926806, + "learning_rate": 4.5985590454181775e-06, + "loss": 0.3262, + "step": 3883 + }, + { + "epoch": 0.54, + "grad_norm": 2.283023792531121, + "learning_rate": 4.596314962193113e-06, + "loss": 0.3591, + "step": 3884 + }, + { + "epoch": 0.54, + "grad_norm": 3.221725782656233, + "learning_rate": 4.5940709608154576e-06, + "loss": 0.3172, + "step": 3885 + }, + { + "epoch": 0.54, + "grad_norm": 1.7784506711870445, + "learning_rate": 4.591827041740185e-06, + "loss": 0.2967, + "step": 3886 + }, + { + "epoch": 0.54, + "grad_norm": 2.1252205462769265, + "learning_rate": 4.5895832054222536e-06, + "loss": 0.3292, + "step": 3887 + }, + { + "epoch": 0.54, + "grad_norm": 2.4053449427005904, + "learning_rate": 4.587339452316601e-06, + "loss": 0.3338, + "step": 3888 + }, + { + "epoch": 0.54, + "grad_norm": 1.8405957607645436, + "learning_rate": 4.58509578287815e-06, + "loss": 0.3238, + "step": 3889 + }, + { + "epoch": 0.54, + "grad_norm": 6.455808483863978, + "learning_rate": 4.5828521975618055e-06, + "loss": 0.2897, + "step": 3890 + }, + { + "epoch": 0.54, + "grad_norm": 3.054950262336092, + "learning_rate": 4.580608696822457e-06, + "loss": 0.318, + "step": 3891 + }, + { + "epoch": 0.54, + "grad_norm": 2.1819447960229232, + "learning_rate": 4.578365281114974e-06, + "loss": 0.3434, + "step": 3892 + }, + { + "epoch": 0.54, + "grad_norm": 1.8469940925146966, + "learning_rate": 4.576121950894213e-06, + "loss": 0.3212, + "step": 3893 + }, + { + "epoch": 0.54, + "grad_norm": 2.3688706439329086, + "learning_rate": 4.573878706615009e-06, + "loss": 0.3411, + "step": 3894 + }, + { + "epoch": 0.54, + "grad_norm": 2.4437274246232485, + "learning_rate": 4.571635548732179e-06, + "loss": 0.3448, + "step": 3895 + }, + { + "epoch": 0.54, + "grad_norm": 2.585801175201017, + "learning_rate": 4.569392477700529e-06, + "loss": 0.3613, + "step": 3896 + }, + { + "epoch": 0.54, + "grad_norm": 4.528777812483612, + "learning_rate": 4.5671494939748415e-06, + "loss": 0.318, + "step": 3897 + }, + { + "epoch": 0.54, + "grad_norm": 2.2985072081851063, + "learning_rate": 4.564906598009884e-06, + "loss": 0.3393, + "step": 3898 + }, + { + "epoch": 0.54, + "grad_norm": 1.850542435956144, + "learning_rate": 4.562663790260403e-06, + "loss": 0.3156, + "step": 3899 + }, + { + "epoch": 0.54, + "grad_norm": 2.063836514384173, + "learning_rate": 4.560421071181131e-06, + "loss": 0.3206, + "step": 3900 + }, + { + "epoch": 0.54, + "grad_norm": 2.158632804613971, + "learning_rate": 4.558178441226779e-06, + "loss": 0.3177, + "step": 3901 + }, + { + "epoch": 0.54, + "grad_norm": 2.470188709107252, + "learning_rate": 4.555935900852043e-06, + "loss": 0.3068, + "step": 3902 + }, + { + "epoch": 0.54, + "grad_norm": 4.126490153055333, + "learning_rate": 4.5536934505115985e-06, + "loss": 0.3175, + "step": 3903 + }, + { + "epoch": 0.54, + "grad_norm": 2.6271386638703467, + "learning_rate": 4.551451090660106e-06, + "loss": 0.333, + "step": 3904 + }, + { + "epoch": 0.54, + "grad_norm": 3.8466431307175335, + "learning_rate": 4.549208821752205e-06, + "loss": 0.2957, + "step": 3905 + }, + { + "epoch": 0.54, + "grad_norm": 6.316117835700089, + "learning_rate": 4.546966644242515e-06, + "loss": 0.3246, + "step": 3906 + }, + { + "epoch": 0.54, + "grad_norm": 1.9653413512839488, + "learning_rate": 4.544724558585641e-06, + "loss": 0.3416, + "step": 3907 + }, + { + "epoch": 0.54, + "grad_norm": 2.0432589388372073, + "learning_rate": 4.542482565236166e-06, + "loss": 0.3327, + "step": 3908 + }, + { + "epoch": 0.54, + "grad_norm": 3.8160965925151373, + "learning_rate": 4.540240664648656e-06, + "loss": 0.3108, + "step": 3909 + }, + { + "epoch": 0.54, + "grad_norm": 2.634825868159181, + "learning_rate": 4.537998857277659e-06, + "loss": 0.3307, + "step": 3910 + }, + { + "epoch": 0.54, + "grad_norm": 3.7887161788992065, + "learning_rate": 4.535757143577701e-06, + "loss": 0.333, + "step": 3911 + }, + { + "epoch": 0.54, + "grad_norm": 2.3189303612610286, + "learning_rate": 4.533515524003291e-06, + "loss": 0.3296, + "step": 3912 + }, + { + "epoch": 0.54, + "grad_norm": 2.3413506906623094, + "learning_rate": 4.53127399900892e-06, + "loss": 0.2924, + "step": 3913 + }, + { + "epoch": 0.54, + "grad_norm": 3.225743035307694, + "learning_rate": 4.52903256904906e-06, + "loss": 0.3114, + "step": 3914 + }, + { + "epoch": 0.54, + "grad_norm": 2.0150529826309977, + "learning_rate": 4.52679123457816e-06, + "loss": 0.308, + "step": 3915 + }, + { + "epoch": 0.54, + "grad_norm": 2.055731117385913, + "learning_rate": 4.524549996050652e-06, + "loss": 0.3494, + "step": 3916 + }, + { + "epoch": 0.54, + "grad_norm": 2.1678057181703028, + "learning_rate": 4.522308853920952e-06, + "loss": 0.3093, + "step": 3917 + }, + { + "epoch": 0.54, + "grad_norm": 1.9960059186877204, + "learning_rate": 4.520067808643449e-06, + "loss": 0.3362, + "step": 3918 + }, + { + "epoch": 0.54, + "grad_norm": 1.8158050008527173, + "learning_rate": 4.517826860672518e-06, + "loss": 0.3101, + "step": 3919 + }, + { + "epoch": 0.54, + "grad_norm": 3.7594273312540927, + "learning_rate": 4.515586010462513e-06, + "loss": 0.3265, + "step": 3920 + }, + { + "epoch": 0.55, + "grad_norm": 2.8150868092641614, + "learning_rate": 4.5133452584677664e-06, + "loss": 0.3048, + "step": 3921 + }, + { + "epoch": 0.55, + "grad_norm": 2.179707587709834, + "learning_rate": 4.511104605142594e-06, + "loss": 0.323, + "step": 3922 + }, + { + "epoch": 0.55, + "grad_norm": 3.224369806087855, + "learning_rate": 4.50886405094129e-06, + "loss": 0.3055, + "step": 3923 + }, + { + "epoch": 0.55, + "grad_norm": 2.4048166389336725, + "learning_rate": 4.506623596318127e-06, + "loss": 0.3243, + "step": 3924 + }, + { + "epoch": 0.55, + "grad_norm": 2.8054766971793534, + "learning_rate": 4.504383241727359e-06, + "loss": 0.3302, + "step": 3925 + }, + { + "epoch": 0.55, + "grad_norm": 2.3064053080412132, + "learning_rate": 4.5021429876232196e-06, + "loss": 0.3329, + "step": 3926 + }, + { + "epoch": 0.55, + "grad_norm": 2.2233513365303614, + "learning_rate": 4.499902834459922e-06, + "loss": 0.3329, + "step": 3927 + }, + { + "epoch": 0.55, + "grad_norm": 2.32811140128498, + "learning_rate": 4.4976627826916575e-06, + "loss": 0.3025, + "step": 3928 + }, + { + "epoch": 0.55, + "grad_norm": 1.9735585929319654, + "learning_rate": 4.495422832772597e-06, + "loss": 0.3315, + "step": 3929 + }, + { + "epoch": 0.55, + "grad_norm": 2.5003210408543337, + "learning_rate": 4.493182985156897e-06, + "loss": 0.3065, + "step": 3930 + }, + { + "epoch": 0.55, + "grad_norm": 1.9564963778939535, + "learning_rate": 4.490943240298684e-06, + "loss": 0.3548, + "step": 3931 + }, + { + "epoch": 0.55, + "grad_norm": 2.1904857286065997, + "learning_rate": 4.4887035986520684e-06, + "loss": 0.3205, + "step": 3932 + }, + { + "epoch": 0.55, + "grad_norm": 2.681809454531757, + "learning_rate": 4.48646406067114e-06, + "loss": 0.3228, + "step": 3933 + }, + { + "epoch": 0.55, + "grad_norm": 1.8495183246766547, + "learning_rate": 4.484224626809966e-06, + "loss": 0.2893, + "step": 3934 + }, + { + "epoch": 0.55, + "grad_norm": 1.9060195303685266, + "learning_rate": 4.481985297522592e-06, + "loss": 0.3266, + "step": 3935 + }, + { + "epoch": 0.55, + "grad_norm": 4.4919339279936885, + "learning_rate": 4.479746073263046e-06, + "loss": 0.336, + "step": 3936 + }, + { + "epoch": 0.55, + "grad_norm": 2.275771100378637, + "learning_rate": 4.47750695448533e-06, + "loss": 0.3437, + "step": 3937 + }, + { + "epoch": 0.55, + "grad_norm": 2.2914466173188237, + "learning_rate": 4.475267941643426e-06, + "loss": 0.289, + "step": 3938 + }, + { + "epoch": 0.55, + "grad_norm": 2.3868082579012877, + "learning_rate": 4.473029035191299e-06, + "loss": 0.3252, + "step": 3939 + }, + { + "epoch": 0.55, + "grad_norm": 3.4669534371134296, + "learning_rate": 4.470790235582887e-06, + "loss": 0.3121, + "step": 3940 + }, + { + "epoch": 0.55, + "grad_norm": 2.517838070820561, + "learning_rate": 4.468551543272107e-06, + "loss": 0.3038, + "step": 3941 + }, + { + "epoch": 0.55, + "grad_norm": 2.2870216362477214, + "learning_rate": 4.466312958712856e-06, + "loss": 0.2919, + "step": 3942 + }, + { + "epoch": 0.55, + "grad_norm": 2.282407497217109, + "learning_rate": 4.4640744823590096e-06, + "loss": 0.324, + "step": 3943 + }, + { + "epoch": 0.55, + "grad_norm": 2.0556315114576873, + "learning_rate": 4.461836114664418e-06, + "loss": 0.3432, + "step": 3944 + }, + { + "epoch": 0.55, + "grad_norm": 2.1470795997819465, + "learning_rate": 4.459597856082915e-06, + "loss": 0.3359, + "step": 3945 + }, + { + "epoch": 0.55, + "grad_norm": 2.644990067660358, + "learning_rate": 4.457359707068304e-06, + "loss": 0.3132, + "step": 3946 + }, + { + "epoch": 0.55, + "grad_norm": 3.417653174658862, + "learning_rate": 4.455121668074378e-06, + "loss": 0.3113, + "step": 3947 + }, + { + "epoch": 0.55, + "grad_norm": 3.9065263527782994, + "learning_rate": 4.452883739554896e-06, + "loss": 0.3356, + "step": 3948 + }, + { + "epoch": 0.55, + "grad_norm": 1.820669487245421, + "learning_rate": 4.4506459219636e-06, + "loss": 0.31, + "step": 3949 + }, + { + "epoch": 0.55, + "grad_norm": 2.004589426028121, + "learning_rate": 4.448408215754212e-06, + "loss": 0.3233, + "step": 3950 + }, + { + "epoch": 0.55, + "grad_norm": 2.0236855699125447, + "learning_rate": 4.446170621380425e-06, + "loss": 0.3013, + "step": 3951 + }, + { + "epoch": 0.55, + "grad_norm": 2.005749861812051, + "learning_rate": 4.443933139295914e-06, + "loss": 0.3107, + "step": 3952 + }, + { + "epoch": 0.55, + "grad_norm": 2.3760357086218375, + "learning_rate": 4.441695769954331e-06, + "loss": 0.3175, + "step": 3953 + }, + { + "epoch": 0.55, + "grad_norm": 2.398178758678999, + "learning_rate": 4.439458513809302e-06, + "loss": 0.3595, + "step": 3954 + }, + { + "epoch": 0.55, + "grad_norm": 1.9655041157493576, + "learning_rate": 4.437221371314432e-06, + "loss": 0.3112, + "step": 3955 + }, + { + "epoch": 0.55, + "grad_norm": 2.318091060122488, + "learning_rate": 4.434984342923307e-06, + "loss": 0.2881, + "step": 3956 + }, + { + "epoch": 0.55, + "grad_norm": 2.3018228425099005, + "learning_rate": 4.432747429089483e-06, + "loss": 0.3436, + "step": 3957 + }, + { + "epoch": 0.55, + "grad_norm": 3.378895009065065, + "learning_rate": 4.430510630266496e-06, + "loss": 0.327, + "step": 3958 + }, + { + "epoch": 0.55, + "grad_norm": 2.1209391037791803, + "learning_rate": 4.428273946907858e-06, + "loss": 0.314, + "step": 3959 + }, + { + "epoch": 0.55, + "grad_norm": 2.2914767664291538, + "learning_rate": 4.4260373794670606e-06, + "loss": 0.3054, + "step": 3960 + }, + { + "epoch": 0.55, + "grad_norm": 3.8118515558843877, + "learning_rate": 4.423800928397566e-06, + "loss": 0.3314, + "step": 3961 + }, + { + "epoch": 0.55, + "grad_norm": 2.175408359790447, + "learning_rate": 4.4215645941528186e-06, + "loss": 0.3096, + "step": 3962 + }, + { + "epoch": 0.55, + "grad_norm": 2.4212959564523198, + "learning_rate": 4.419328377186235e-06, + "loss": 0.3318, + "step": 3963 + }, + { + "epoch": 0.55, + "grad_norm": 2.4719608070036383, + "learning_rate": 4.417092277951208e-06, + "loss": 0.3163, + "step": 3964 + }, + { + "epoch": 0.55, + "grad_norm": 2.610578175544118, + "learning_rate": 4.414856296901112e-06, + "loss": 0.3396, + "step": 3965 + }, + { + "epoch": 0.55, + "grad_norm": 1.8308121365101964, + "learning_rate": 4.412620434489294e-06, + "loss": 0.2966, + "step": 3966 + }, + { + "epoch": 0.55, + "grad_norm": 1.9431876833024473, + "learning_rate": 4.410384691169074e-06, + "loss": 0.2971, + "step": 3967 + }, + { + "epoch": 0.55, + "grad_norm": 2.068682517568742, + "learning_rate": 4.408149067393751e-06, + "loss": 0.327, + "step": 3968 + }, + { + "epoch": 0.55, + "grad_norm": 3.8327661289480233, + "learning_rate": 4.405913563616601e-06, + "loss": 0.3237, + "step": 3969 + }, + { + "epoch": 0.55, + "grad_norm": 2.011986461036353, + "learning_rate": 4.403678180290871e-06, + "loss": 0.3239, + "step": 3970 + }, + { + "epoch": 0.55, + "grad_norm": 2.0599987460412046, + "learning_rate": 4.401442917869787e-06, + "loss": 0.321, + "step": 3971 + }, + { + "epoch": 0.55, + "grad_norm": 2.014143489257982, + "learning_rate": 4.39920777680655e-06, + "loss": 0.3081, + "step": 3972 + }, + { + "epoch": 0.55, + "grad_norm": 2.711686769888963, + "learning_rate": 4.396972757554339e-06, + "loss": 0.3074, + "step": 3973 + }, + { + "epoch": 0.55, + "grad_norm": 2.8897503086502074, + "learning_rate": 4.3947378605663035e-06, + "loss": 0.2994, + "step": 3974 + }, + { + "epoch": 0.55, + "grad_norm": 2.85624235629734, + "learning_rate": 4.392503086295571e-06, + "loss": 0.3181, + "step": 3975 + }, + { + "epoch": 0.55, + "grad_norm": 1.903983622920009, + "learning_rate": 4.390268435195244e-06, + "loss": 0.3349, + "step": 3976 + }, + { + "epoch": 0.55, + "grad_norm": 2.2815711555300533, + "learning_rate": 4.388033907718396e-06, + "loss": 0.2775, + "step": 3977 + }, + { + "epoch": 0.55, + "grad_norm": 2.5024019352112354, + "learning_rate": 4.385799504318083e-06, + "loss": 0.3094, + "step": 3978 + }, + { + "epoch": 0.55, + "grad_norm": 2.813896965326014, + "learning_rate": 4.383565225447331e-06, + "loss": 0.3275, + "step": 3979 + }, + { + "epoch": 0.55, + "grad_norm": 4.374204993129532, + "learning_rate": 4.3813310715591405e-06, + "loss": 0.3294, + "step": 3980 + }, + { + "epoch": 0.55, + "grad_norm": 1.9960805813539166, + "learning_rate": 4.379097043106485e-06, + "loss": 0.313, + "step": 3981 + }, + { + "epoch": 0.55, + "grad_norm": 5.056173519952342, + "learning_rate": 4.376863140542321e-06, + "loss": 0.3166, + "step": 3982 + }, + { + "epoch": 0.55, + "grad_norm": 2.2337246816813385, + "learning_rate": 4.374629364319572e-06, + "loss": 0.3274, + "step": 3983 + }, + { + "epoch": 0.55, + "grad_norm": 2.882590074365306, + "learning_rate": 4.372395714891135e-06, + "loss": 0.3724, + "step": 3984 + }, + { + "epoch": 0.55, + "grad_norm": 5.193397026234418, + "learning_rate": 4.370162192709886e-06, + "loss": 0.2968, + "step": 3985 + }, + { + "epoch": 0.55, + "grad_norm": 2.1750211633606473, + "learning_rate": 4.367928798228673e-06, + "loss": 0.3375, + "step": 3986 + }, + { + "epoch": 0.55, + "grad_norm": 2.09869874865467, + "learning_rate": 4.365695531900318e-06, + "loss": 0.3185, + "step": 3987 + }, + { + "epoch": 0.55, + "grad_norm": 2.424774698078086, + "learning_rate": 4.3634623941776164e-06, + "loss": 0.3227, + "step": 3988 + }, + { + "epoch": 0.55, + "grad_norm": 3.0639134530781234, + "learning_rate": 4.36122938551334e-06, + "loss": 0.3062, + "step": 3989 + }, + { + "epoch": 0.55, + "grad_norm": 2.7144995641036056, + "learning_rate": 4.35899650636023e-06, + "loss": 0.324, + "step": 3990 + }, + { + "epoch": 0.55, + "grad_norm": 2.2675056193443095, + "learning_rate": 4.356763757171007e-06, + "loss": 0.3234, + "step": 3991 + }, + { + "epoch": 0.55, + "grad_norm": 2.0521621729038646, + "learning_rate": 4.354531138398361e-06, + "loss": 0.3055, + "step": 3992 + }, + { + "epoch": 0.56, + "grad_norm": 3.778613145076412, + "learning_rate": 4.352298650494959e-06, + "loss": 0.3054, + "step": 3993 + }, + { + "epoch": 0.56, + "grad_norm": 2.453300426533258, + "learning_rate": 4.3500662939134375e-06, + "loss": 0.2992, + "step": 3994 + }, + { + "epoch": 0.56, + "grad_norm": 2.3643681119418622, + "learning_rate": 4.347834069106408e-06, + "loss": 0.2945, + "step": 3995 + }, + { + "epoch": 0.56, + "grad_norm": 2.4893554707855365, + "learning_rate": 4.345601976526456e-06, + "loss": 0.3165, + "step": 3996 + }, + { + "epoch": 0.56, + "grad_norm": 2.636175688348806, + "learning_rate": 4.34337001662614e-06, + "loss": 0.3378, + "step": 3997 + }, + { + "epoch": 0.56, + "grad_norm": 2.2597907610390373, + "learning_rate": 4.34113818985799e-06, + "loss": 0.2993, + "step": 3998 + }, + { + "epoch": 0.56, + "grad_norm": 2.2493005108806776, + "learning_rate": 4.338906496674513e-06, + "loss": 0.3306, + "step": 3999 + }, + { + "epoch": 0.56, + "grad_norm": 2.5951357114970968, + "learning_rate": 4.3366749375281856e-06, + "loss": 0.3369, + "step": 4000 + }, + { + "epoch": 0.56, + "grad_norm": 3.027319994724848, + "learning_rate": 4.334443512871456e-06, + "loss": 0.3288, + "step": 4001 + }, + { + "epoch": 0.56, + "grad_norm": 1.8884598394641114, + "learning_rate": 4.332212223156748e-06, + "loss": 0.3048, + "step": 4002 + }, + { + "epoch": 0.56, + "grad_norm": 1.9982514748980795, + "learning_rate": 4.329981068836458e-06, + "loss": 0.3358, + "step": 4003 + }, + { + "epoch": 0.56, + "grad_norm": 3.108795165557873, + "learning_rate": 4.327750050362954e-06, + "loss": 0.3329, + "step": 4004 + }, + { + "epoch": 0.56, + "grad_norm": 3.56096010142109, + "learning_rate": 4.325519168188575e-06, + "loss": 0.3206, + "step": 4005 + }, + { + "epoch": 0.56, + "grad_norm": 3.146002284533565, + "learning_rate": 4.323288422765635e-06, + "loss": 0.3197, + "step": 4006 + }, + { + "epoch": 0.56, + "grad_norm": 2.155891946928914, + "learning_rate": 4.321057814546418e-06, + "loss": 0.3286, + "step": 4007 + }, + { + "epoch": 0.56, + "grad_norm": 7.904845791330232, + "learning_rate": 4.3188273439831815e-06, + "loss": 0.3294, + "step": 4008 + }, + { + "epoch": 0.56, + "grad_norm": 2.210979370175521, + "learning_rate": 4.316597011528158e-06, + "loss": 0.3233, + "step": 4009 + }, + { + "epoch": 0.56, + "grad_norm": 2.932177915849577, + "learning_rate": 4.314366817633545e-06, + "loss": 0.3164, + "step": 4010 + }, + { + "epoch": 0.56, + "grad_norm": 1.890723094987123, + "learning_rate": 4.312136762751517e-06, + "loss": 0.3162, + "step": 4011 + }, + { + "epoch": 0.56, + "grad_norm": 2.0179220846534567, + "learning_rate": 4.3099068473342205e-06, + "loss": 0.3087, + "step": 4012 + }, + { + "epoch": 0.56, + "grad_norm": 2.607863953350475, + "learning_rate": 4.30767707183377e-06, + "loss": 0.3193, + "step": 4013 + }, + { + "epoch": 0.56, + "grad_norm": 1.9510734226413717, + "learning_rate": 4.305447436702255e-06, + "loss": 0.3196, + "step": 4014 + }, + { + "epoch": 0.56, + "grad_norm": 2.0335404915028823, + "learning_rate": 4.303217942391735e-06, + "loss": 0.3172, + "step": 4015 + }, + { + "epoch": 0.56, + "grad_norm": 2.0057424512127593, + "learning_rate": 4.3009885893542436e-06, + "loss": 0.3101, + "step": 4016 + }, + { + "epoch": 0.56, + "grad_norm": 2.0111298500935932, + "learning_rate": 4.298759378041782e-06, + "loss": 0.3226, + "step": 4017 + }, + { + "epoch": 0.56, + "grad_norm": 3.1852364657167125, + "learning_rate": 4.296530308906323e-06, + "loss": 0.3075, + "step": 4018 + }, + { + "epoch": 0.56, + "grad_norm": 4.61396124386514, + "learning_rate": 4.2943013823998155e-06, + "loss": 0.3008, + "step": 4019 + }, + { + "epoch": 0.56, + "grad_norm": 2.2401863489641793, + "learning_rate": 4.292072598974172e-06, + "loss": 0.3403, + "step": 4020 + }, + { + "epoch": 0.56, + "grad_norm": 3.1560181015042206, + "learning_rate": 4.289843959081281e-06, + "loss": 0.3069, + "step": 4021 + }, + { + "epoch": 0.56, + "grad_norm": 2.4122250372906326, + "learning_rate": 4.287615463173002e-06, + "loss": 0.2997, + "step": 4022 + }, + { + "epoch": 0.56, + "grad_norm": 2.2711882268997723, + "learning_rate": 4.285387111701164e-06, + "loss": 0.3177, + "step": 4023 + }, + { + "epoch": 0.56, + "grad_norm": 4.084615288548753, + "learning_rate": 4.283158905117563e-06, + "loss": 0.3086, + "step": 4024 + }, + { + "epoch": 0.56, + "grad_norm": 2.6883952777795477, + "learning_rate": 4.280930843873975e-06, + "loss": 0.3133, + "step": 4025 + }, + { + "epoch": 0.56, + "grad_norm": 1.8961359978576333, + "learning_rate": 4.278702928422137e-06, + "loss": 0.2975, + "step": 4026 + }, + { + "epoch": 0.56, + "grad_norm": 4.452458350729629, + "learning_rate": 4.276475159213762e-06, + "loss": 0.3308, + "step": 4027 + }, + { + "epoch": 0.56, + "grad_norm": 2.2918933084256077, + "learning_rate": 4.274247536700531e-06, + "loss": 0.3025, + "step": 4028 + }, + { + "epoch": 0.56, + "grad_norm": 3.027728987223265, + "learning_rate": 4.2720200613340975e-06, + "loss": 0.3185, + "step": 4029 + }, + { + "epoch": 0.56, + "grad_norm": 3.2114171848187985, + "learning_rate": 4.2697927335660806e-06, + "loss": 0.3086, + "step": 4030 + }, + { + "epoch": 0.56, + "grad_norm": 3.2847985672182283, + "learning_rate": 4.267565553848075e-06, + "loss": 0.3185, + "step": 4031 + }, + { + "epoch": 0.56, + "grad_norm": 8.97291014731387, + "learning_rate": 4.265338522631642e-06, + "loss": 0.3236, + "step": 4032 + }, + { + "epoch": 0.56, + "grad_norm": 1.8948471006783991, + "learning_rate": 4.263111640368312e-06, + "loss": 0.3158, + "step": 4033 + }, + { + "epoch": 0.56, + "grad_norm": 2.3211542499145237, + "learning_rate": 4.26088490750959e-06, + "loss": 0.3194, + "step": 4034 + }, + { + "epoch": 0.56, + "grad_norm": 4.455328112863972, + "learning_rate": 4.258658324506945e-06, + "loss": 0.3175, + "step": 4035 + }, + { + "epoch": 0.56, + "grad_norm": 2.13619208660058, + "learning_rate": 4.256431891811823e-06, + "loss": 0.3174, + "step": 4036 + }, + { + "epoch": 0.56, + "grad_norm": 2.176773290366292, + "learning_rate": 4.254205609875628e-06, + "loss": 0.3096, + "step": 4037 + }, + { + "epoch": 0.56, + "grad_norm": 2.3795010253528415, + "learning_rate": 4.251979479149744e-06, + "loss": 0.3266, + "step": 4038 + }, + { + "epoch": 0.56, + "grad_norm": 9.867716774118007, + "learning_rate": 4.24975350008552e-06, + "loss": 0.3114, + "step": 4039 + }, + { + "epoch": 0.56, + "grad_norm": 2.133615310635956, + "learning_rate": 4.247527673134274e-06, + "loss": 0.3145, + "step": 4040 + }, + { + "epoch": 0.56, + "grad_norm": 2.200757345281006, + "learning_rate": 4.245301998747292e-06, + "loss": 0.3046, + "step": 4041 + }, + { + "epoch": 0.56, + "grad_norm": 2.6781863258813283, + "learning_rate": 4.243076477375836e-06, + "loss": 0.329, + "step": 4042 + }, + { + "epoch": 0.56, + "grad_norm": 2.8084652180417495, + "learning_rate": 4.24085110947113e-06, + "loss": 0.3269, + "step": 4043 + }, + { + "epoch": 0.56, + "grad_norm": 2.39923756747221, + "learning_rate": 4.2386258954843664e-06, + "loss": 0.3092, + "step": 4044 + }, + { + "epoch": 0.56, + "grad_norm": 1.753067574894478, + "learning_rate": 4.236400835866712e-06, + "loss": 0.2979, + "step": 4045 + }, + { + "epoch": 0.56, + "grad_norm": 4.781510337152967, + "learning_rate": 4.234175931069297e-06, + "loss": 0.3001, + "step": 4046 + }, + { + "epoch": 0.56, + "grad_norm": 2.7430332129135553, + "learning_rate": 4.231951181543221e-06, + "loss": 0.3133, + "step": 4047 + }, + { + "epoch": 0.56, + "grad_norm": 2.8369093145295463, + "learning_rate": 4.229726587739559e-06, + "loss": 0.3173, + "step": 4048 + }, + { + "epoch": 0.56, + "grad_norm": 1.8224473249153736, + "learning_rate": 4.227502150109343e-06, + "loss": 0.3011, + "step": 4049 + }, + { + "epoch": 0.56, + "grad_norm": 2.227499065368728, + "learning_rate": 4.22527786910358e-06, + "loss": 0.3289, + "step": 4050 + }, + { + "epoch": 0.56, + "grad_norm": 2.806634137682211, + "learning_rate": 4.2230537451732476e-06, + "loss": 0.3301, + "step": 4051 + }, + { + "epoch": 0.56, + "grad_norm": 3.2166042717495875, + "learning_rate": 4.220829778769288e-06, + "loss": 0.3304, + "step": 4052 + }, + { + "epoch": 0.56, + "grad_norm": 2.7559342917708998, + "learning_rate": 4.218605970342609e-06, + "loss": 0.3325, + "step": 4053 + }, + { + "epoch": 0.56, + "grad_norm": 2.190791610461451, + "learning_rate": 4.216382320344091e-06, + "loss": 0.3072, + "step": 4054 + }, + { + "epoch": 0.56, + "grad_norm": 2.12227284273334, + "learning_rate": 4.214158829224582e-06, + "loss": 0.3318, + "step": 4055 + }, + { + "epoch": 0.56, + "grad_norm": 4.680072386423659, + "learning_rate": 4.211935497434894e-06, + "loss": 0.3423, + "step": 4056 + }, + { + "epoch": 0.56, + "grad_norm": 2.625463526053145, + "learning_rate": 4.209712325425808e-06, + "loss": 0.3042, + "step": 4057 + }, + { + "epoch": 0.56, + "grad_norm": 2.7461550303473556, + "learning_rate": 4.207489313648078e-06, + "loss": 0.3097, + "step": 4058 + }, + { + "epoch": 0.56, + "grad_norm": 2.3467908762149547, + "learning_rate": 4.205266462552415e-06, + "loss": 0.3185, + "step": 4059 + }, + { + "epoch": 0.56, + "grad_norm": 3.2135947208147604, + "learning_rate": 4.203043772589509e-06, + "loss": 0.331, + "step": 4060 + }, + { + "epoch": 0.56, + "grad_norm": 2.116686656450943, + "learning_rate": 4.20082124421001e-06, + "loss": 0.3256, + "step": 4061 + }, + { + "epoch": 0.56, + "grad_norm": 2.3603317865332984, + "learning_rate": 4.198598877864538e-06, + "loss": 0.329, + "step": 4062 + }, + { + "epoch": 0.56, + "grad_norm": 2.0760915810160245, + "learning_rate": 4.196376674003676e-06, + "loss": 0.322, + "step": 4063 + }, + { + "epoch": 0.56, + "grad_norm": 5.131429302376609, + "learning_rate": 4.194154633077981e-06, + "loss": 0.3314, + "step": 4064 + }, + { + "epoch": 0.57, + "grad_norm": 2.5978922082936053, + "learning_rate": 4.191932755537972e-06, + "loss": 0.3327, + "step": 4065 + }, + { + "epoch": 0.57, + "grad_norm": 2.2389206675729567, + "learning_rate": 4.189711041834136e-06, + "loss": 0.3045, + "step": 4066 + }, + { + "epoch": 0.57, + "grad_norm": 2.7267465231752506, + "learning_rate": 4.187489492416924e-06, + "loss": 0.3219, + "step": 4067 + }, + { + "epoch": 0.57, + "grad_norm": 2.1876903412609927, + "learning_rate": 4.185268107736762e-06, + "loss": 0.302, + "step": 4068 + }, + { + "epoch": 0.57, + "grad_norm": 2.811036825251275, + "learning_rate": 4.1830468882440355e-06, + "loss": 0.3256, + "step": 4069 + }, + { + "epoch": 0.57, + "grad_norm": 2.444658313396258, + "learning_rate": 4.180825834389096e-06, + "loss": 0.3247, + "step": 4070 + }, + { + "epoch": 0.57, + "grad_norm": 2.37857125114916, + "learning_rate": 4.178604946622265e-06, + "loss": 0.3178, + "step": 4071 + }, + { + "epoch": 0.57, + "grad_norm": 2.5408244044936468, + "learning_rate": 4.1763842253938316e-06, + "loss": 0.3268, + "step": 4072 + }, + { + "epoch": 0.57, + "grad_norm": 2.0835131734837025, + "learning_rate": 4.174163671154044e-06, + "loss": 0.2895, + "step": 4073 + }, + { + "epoch": 0.57, + "grad_norm": 2.7680859903612727, + "learning_rate": 4.171943284353124e-06, + "loss": 0.3064, + "step": 4074 + }, + { + "epoch": 0.57, + "grad_norm": 2.1199703521757534, + "learning_rate": 4.169723065441255e-06, + "loss": 0.2959, + "step": 4075 + }, + { + "epoch": 0.57, + "grad_norm": 2.224824933501814, + "learning_rate": 4.167503014868588e-06, + "loss": 0.3319, + "step": 4076 + }, + { + "epoch": 0.57, + "grad_norm": 2.2713586076206314, + "learning_rate": 4.16528313308524e-06, + "loss": 0.3178, + "step": 4077 + }, + { + "epoch": 0.57, + "grad_norm": 1.9570294307758813, + "learning_rate": 4.163063420541294e-06, + "loss": 0.3287, + "step": 4078 + }, + { + "epoch": 0.57, + "grad_norm": 2.3680554359295782, + "learning_rate": 4.1608438776868e-06, + "loss": 0.3282, + "step": 4079 + }, + { + "epoch": 0.57, + "grad_norm": 2.115089060404997, + "learning_rate": 4.158624504971767e-06, + "loss": 0.327, + "step": 4080 + }, + { + "epoch": 0.57, + "grad_norm": 30.650938500170522, + "learning_rate": 4.15640530284618e-06, + "loss": 0.3196, + "step": 4081 + }, + { + "epoch": 0.57, + "grad_norm": 2.746762044124129, + "learning_rate": 4.154186271759978e-06, + "loss": 0.3112, + "step": 4082 + }, + { + "epoch": 0.57, + "grad_norm": 3.064375145412914, + "learning_rate": 4.1519674121630735e-06, + "loss": 0.3086, + "step": 4083 + }, + { + "epoch": 0.57, + "grad_norm": 2.8614139921230834, + "learning_rate": 4.149748724505342e-06, + "loss": 0.3323, + "step": 4084 + }, + { + "epoch": 0.57, + "grad_norm": 2.6924910421998645, + "learning_rate": 4.147530209236626e-06, + "loss": 0.3052, + "step": 4085 + }, + { + "epoch": 0.57, + "grad_norm": 2.950526180933524, + "learning_rate": 4.145311866806727e-06, + "loss": 0.3509, + "step": 4086 + }, + { + "epoch": 0.57, + "grad_norm": 2.410671064733369, + "learning_rate": 4.143093697665418e-06, + "loss": 0.3031, + "step": 4087 + }, + { + "epoch": 0.57, + "grad_norm": 2.4360735577099275, + "learning_rate": 4.140875702262434e-06, + "loss": 0.3023, + "step": 4088 + }, + { + "epoch": 0.57, + "grad_norm": 2.5688839572954154, + "learning_rate": 4.138657881047475e-06, + "loss": 0.2933, + "step": 4089 + }, + { + "epoch": 0.57, + "grad_norm": 2.700689242438908, + "learning_rate": 4.136440234470204e-06, + "loss": 0.3274, + "step": 4090 + }, + { + "epoch": 0.57, + "grad_norm": 2.0309720597926866, + "learning_rate": 4.134222762980254e-06, + "loss": 0.3196, + "step": 4091 + }, + { + "epoch": 0.57, + "grad_norm": 4.54605640435938, + "learning_rate": 4.1320054670272135e-06, + "loss": 0.3189, + "step": 4092 + }, + { + "epoch": 0.57, + "grad_norm": 2.047579933800053, + "learning_rate": 4.129788347060643e-06, + "loss": 0.3134, + "step": 4093 + }, + { + "epoch": 0.57, + "grad_norm": 2.348804065727382, + "learning_rate": 4.127571403530068e-06, + "loss": 0.3281, + "step": 4094 + }, + { + "epoch": 0.57, + "grad_norm": 2.044343426233728, + "learning_rate": 4.125354636884974e-06, + "loss": 0.3058, + "step": 4095 + }, + { + "epoch": 0.57, + "grad_norm": 2.463340426106452, + "learning_rate": 4.123138047574809e-06, + "loss": 0.3473, + "step": 4096 + }, + { + "epoch": 0.57, + "grad_norm": 2.4382880415788617, + "learning_rate": 4.12092163604899e-06, + "loss": 0.3018, + "step": 4097 + }, + { + "epoch": 0.57, + "grad_norm": 2.177601460157521, + "learning_rate": 4.118705402756898e-06, + "loss": 0.3233, + "step": 4098 + }, + { + "epoch": 0.57, + "grad_norm": 3.9886984296603236, + "learning_rate": 4.116489348147871e-06, + "loss": 0.3265, + "step": 4099 + }, + { + "epoch": 0.57, + "grad_norm": 2.6471610611735312, + "learning_rate": 4.1142734726712195e-06, + "loss": 0.3185, + "step": 4100 + }, + { + "epoch": 0.57, + "grad_norm": 3.1751941591165824, + "learning_rate": 4.112057776776213e-06, + "loss": 0.3487, + "step": 4101 + }, + { + "epoch": 0.57, + "grad_norm": 2.6086640990414383, + "learning_rate": 4.109842260912082e-06, + "loss": 0.3285, + "step": 4102 + }, + { + "epoch": 0.57, + "grad_norm": 3.2720888288818686, + "learning_rate": 4.107626925528028e-06, + "loss": 0.3336, + "step": 4103 + }, + { + "epoch": 0.57, + "grad_norm": 2.3129833094378855, + "learning_rate": 4.105411771073211e-06, + "loss": 0.3422, + "step": 4104 + }, + { + "epoch": 0.57, + "grad_norm": 3.8949160693925027, + "learning_rate": 4.1031967979967546e-06, + "loss": 0.3251, + "step": 4105 + }, + { + "epoch": 0.57, + "grad_norm": 1.9838045395831538, + "learning_rate": 4.100982006747745e-06, + "loss": 0.3337, + "step": 4106 + }, + { + "epoch": 0.57, + "grad_norm": 2.3591000793336456, + "learning_rate": 4.098767397775234e-06, + "loss": 0.3282, + "step": 4107 + }, + { + "epoch": 0.57, + "grad_norm": 2.0571196623942365, + "learning_rate": 4.096552971528234e-06, + "loss": 0.2991, + "step": 4108 + }, + { + "epoch": 0.57, + "grad_norm": 2.266446036550651, + "learning_rate": 4.094338728455721e-06, + "loss": 0.344, + "step": 4109 + }, + { + "epoch": 0.57, + "grad_norm": 2.3549363359315354, + "learning_rate": 4.092124669006635e-06, + "loss": 0.3566, + "step": 4110 + }, + { + "epoch": 0.57, + "grad_norm": 2.0866966315882993, + "learning_rate": 4.089910793629878e-06, + "loss": 0.3183, + "step": 4111 + }, + { + "epoch": 0.57, + "grad_norm": 1.9828394560579068, + "learning_rate": 4.0876971027743166e-06, + "loss": 0.3188, + "step": 4112 + }, + { + "epoch": 0.57, + "grad_norm": 2.938605139133012, + "learning_rate": 4.085483596888776e-06, + "loss": 0.3051, + "step": 4113 + }, + { + "epoch": 0.57, + "grad_norm": 2.668934674558454, + "learning_rate": 4.083270276422046e-06, + "loss": 0.3236, + "step": 4114 + }, + { + "epoch": 0.57, + "grad_norm": 3.6960731720057978, + "learning_rate": 4.081057141822879e-06, + "loss": 0.3513, + "step": 4115 + }, + { + "epoch": 0.57, + "grad_norm": 1.9098568608722177, + "learning_rate": 4.078844193539989e-06, + "loss": 0.3096, + "step": 4116 + }, + { + "epoch": 0.57, + "grad_norm": 2.4040281449638106, + "learning_rate": 4.076631432022055e-06, + "loss": 0.3531, + "step": 4117 + }, + { + "epoch": 0.57, + "grad_norm": 2.8878440401838374, + "learning_rate": 4.074418857717713e-06, + "loss": 0.3164, + "step": 4118 + }, + { + "epoch": 0.57, + "grad_norm": 2.34403709536733, + "learning_rate": 4.072206471075564e-06, + "loss": 0.283, + "step": 4119 + }, + { + "epoch": 0.57, + "grad_norm": 1.9742002794052833, + "learning_rate": 4.069994272544173e-06, + "loss": 0.3429, + "step": 4120 + }, + { + "epoch": 0.57, + "grad_norm": 2.2399490793546897, + "learning_rate": 4.067782262572065e-06, + "loss": 0.2973, + "step": 4121 + }, + { + "epoch": 0.57, + "grad_norm": 2.785179050800626, + "learning_rate": 4.0655704416077244e-06, + "loss": 0.2903, + "step": 4122 + }, + { + "epoch": 0.57, + "grad_norm": 2.8236690397662607, + "learning_rate": 4.0633588100996e-06, + "loss": 0.317, + "step": 4123 + }, + { + "epoch": 0.57, + "grad_norm": 2.1941463135823898, + "learning_rate": 4.061147368496103e-06, + "loss": 0.3275, + "step": 4124 + }, + { + "epoch": 0.57, + "grad_norm": 2.24290893634486, + "learning_rate": 4.058936117245602e-06, + "loss": 0.332, + "step": 4125 + }, + { + "epoch": 0.57, + "grad_norm": 1.7083686445431767, + "learning_rate": 4.056725056796432e-06, + "loss": 0.3301, + "step": 4126 + }, + { + "epoch": 0.57, + "grad_norm": 2.2791144965853385, + "learning_rate": 4.054514187596886e-06, + "loss": 0.3167, + "step": 4127 + }, + { + "epoch": 0.57, + "grad_norm": 2.2450431407597438, + "learning_rate": 4.052303510095218e-06, + "loss": 0.3137, + "step": 4128 + }, + { + "epoch": 0.57, + "grad_norm": 2.0335800680235527, + "learning_rate": 4.050093024739647e-06, + "loss": 0.303, + "step": 4129 + }, + { + "epoch": 0.57, + "grad_norm": 1.9746000304097246, + "learning_rate": 4.0478827319783484e-06, + "loss": 0.3217, + "step": 4130 + }, + { + "epoch": 0.57, + "grad_norm": 2.1078644548507866, + "learning_rate": 4.045672632259462e-06, + "loss": 0.3142, + "step": 4131 + }, + { + "epoch": 0.57, + "grad_norm": 2.4212340749857804, + "learning_rate": 4.043462726031085e-06, + "loss": 0.3276, + "step": 4132 + }, + { + "epoch": 0.57, + "grad_norm": 2.123387771339042, + "learning_rate": 4.0412530137412795e-06, + "loss": 0.3253, + "step": 4133 + }, + { + "epoch": 0.57, + "grad_norm": 2.774302283575276, + "learning_rate": 4.039043495838066e-06, + "loss": 0.2929, + "step": 4134 + }, + { + "epoch": 0.57, + "grad_norm": 2.276367697619921, + "learning_rate": 4.0368341727694235e-06, + "loss": 0.3069, + "step": 4135 + }, + { + "epoch": 0.57, + "grad_norm": 3.30727922515471, + "learning_rate": 4.0346250449832935e-06, + "loss": 0.3231, + "step": 4136 + }, + { + "epoch": 0.58, + "grad_norm": 3.1174149324362617, + "learning_rate": 4.032416112927582e-06, + "loss": 0.2957, + "step": 4137 + }, + { + "epoch": 0.58, + "grad_norm": 2.099618058662372, + "learning_rate": 4.030207377050149e-06, + "loss": 0.3364, + "step": 4138 + }, + { + "epoch": 0.58, + "grad_norm": 2.1339233933189443, + "learning_rate": 4.0279988377988174e-06, + "loss": 0.3032, + "step": 4139 + }, + { + "epoch": 0.58, + "grad_norm": 2.7843635034480885, + "learning_rate": 4.025790495621369e-06, + "loss": 0.3182, + "step": 4140 + }, + { + "epoch": 0.58, + "grad_norm": 2.6323548285859055, + "learning_rate": 4.02358235096555e-06, + "loss": 0.3302, + "step": 4141 + }, + { + "epoch": 0.58, + "grad_norm": 2.116898597761089, + "learning_rate": 4.021374404279059e-06, + "loss": 0.285, + "step": 4142 + }, + { + "epoch": 0.58, + "grad_norm": 2.0760633542731433, + "learning_rate": 4.019166656009562e-06, + "loss": 0.3161, + "step": 4143 + }, + { + "epoch": 0.58, + "grad_norm": 2.135422955323538, + "learning_rate": 4.0169591066046785e-06, + "loss": 0.3258, + "step": 4144 + }, + { + "epoch": 0.58, + "grad_norm": 5.92208465278019, + "learning_rate": 4.014751756511991e-06, + "loss": 0.3036, + "step": 4145 + }, + { + "epoch": 0.58, + "grad_norm": 2.883136898166131, + "learning_rate": 4.012544606179043e-06, + "loss": 0.315, + "step": 4146 + }, + { + "epoch": 0.58, + "grad_norm": 2.35931928322298, + "learning_rate": 4.010337656053335e-06, + "loss": 0.312, + "step": 4147 + }, + { + "epoch": 0.58, + "grad_norm": 2.128083778125683, + "learning_rate": 4.008130906582329e-06, + "loss": 0.3027, + "step": 4148 + }, + { + "epoch": 0.58, + "grad_norm": 2.4116842183730394, + "learning_rate": 4.005924358213442e-06, + "loss": 0.3196, + "step": 4149 + }, + { + "epoch": 0.58, + "grad_norm": 3.3428519868059037, + "learning_rate": 4.003718011394056e-06, + "loss": 0.3206, + "step": 4150 + }, + { + "epoch": 0.58, + "grad_norm": 3.48897688846685, + "learning_rate": 4.001511866571507e-06, + "loss": 0.3323, + "step": 4151 + }, + { + "epoch": 0.58, + "grad_norm": 1.9078445864106892, + "learning_rate": 3.999305924193093e-06, + "loss": 0.3243, + "step": 4152 + }, + { + "epoch": 0.58, + "grad_norm": 2.3418402051524736, + "learning_rate": 3.99710018470607e-06, + "loss": 0.3121, + "step": 4153 + }, + { + "epoch": 0.58, + "grad_norm": 2.2175154771643766, + "learning_rate": 3.994894648557655e-06, + "loss": 0.316, + "step": 4154 + }, + { + "epoch": 0.58, + "grad_norm": 3.1446387143047856, + "learning_rate": 3.9926893161950206e-06, + "loss": 0.317, + "step": 4155 + }, + { + "epoch": 0.58, + "grad_norm": 4.665033417192698, + "learning_rate": 3.990484188065299e-06, + "loss": 0.2906, + "step": 4156 + }, + { + "epoch": 0.58, + "grad_norm": 1.827700155539771, + "learning_rate": 3.988279264615583e-06, + "loss": 0.3215, + "step": 4157 + }, + { + "epoch": 0.58, + "grad_norm": 10.396412220718492, + "learning_rate": 3.98607454629292e-06, + "loss": 0.3157, + "step": 4158 + }, + { + "epoch": 0.58, + "grad_norm": 2.0181949421241483, + "learning_rate": 3.983870033544319e-06, + "loss": 0.3144, + "step": 4159 + }, + { + "epoch": 0.58, + "grad_norm": 2.888575026198183, + "learning_rate": 3.981665726816747e-06, + "loss": 0.2834, + "step": 4160 + }, + { + "epoch": 0.58, + "grad_norm": 2.709977445698788, + "learning_rate": 3.979461626557128e-06, + "loss": 0.3247, + "step": 4161 + }, + { + "epoch": 0.58, + "grad_norm": 1.986572961056301, + "learning_rate": 3.977257733212344e-06, + "loss": 0.2794, + "step": 4162 + }, + { + "epoch": 0.58, + "grad_norm": 2.1781547926852096, + "learning_rate": 3.9750540472292394e-06, + "loss": 0.3287, + "step": 4163 + }, + { + "epoch": 0.58, + "grad_norm": 2.133713114961178, + "learning_rate": 3.97285056905461e-06, + "loss": 0.323, + "step": 4164 + }, + { + "epoch": 0.58, + "grad_norm": 1.710131006825012, + "learning_rate": 3.970647299135213e-06, + "loss": 0.3182, + "step": 4165 + }, + { + "epoch": 0.58, + "grad_norm": 9.016872758603892, + "learning_rate": 3.968444237917762e-06, + "loss": 0.3214, + "step": 4166 + }, + { + "epoch": 0.58, + "grad_norm": 3.394258284789461, + "learning_rate": 3.966241385848932e-06, + "loss": 0.3233, + "step": 4167 + }, + { + "epoch": 0.58, + "grad_norm": 2.301858014234091, + "learning_rate": 3.964038743375349e-06, + "loss": 0.3133, + "step": 4168 + }, + { + "epoch": 0.58, + "grad_norm": 2.044591382523394, + "learning_rate": 3.961836310943603e-06, + "loss": 0.3309, + "step": 4169 + }, + { + "epoch": 0.58, + "grad_norm": 2.6491236855232003, + "learning_rate": 3.959634089000238e-06, + "loss": 0.3228, + "step": 4170 + }, + { + "epoch": 0.58, + "grad_norm": 3.197911613122868, + "learning_rate": 3.957432077991753e-06, + "loss": 0.3205, + "step": 4171 + }, + { + "epoch": 0.58, + "grad_norm": 2.031304581869512, + "learning_rate": 3.955230278364611e-06, + "loss": 0.3211, + "step": 4172 + }, + { + "epoch": 0.58, + "grad_norm": 2.7069107854093706, + "learning_rate": 3.953028690565228e-06, + "loss": 0.3194, + "step": 4173 + }, + { + "epoch": 0.58, + "grad_norm": 2.1818192804548007, + "learning_rate": 3.950827315039977e-06, + "loss": 0.3215, + "step": 4174 + }, + { + "epoch": 0.58, + "grad_norm": 3.6384247468123303, + "learning_rate": 3.948626152235187e-06, + "loss": 0.3452, + "step": 4175 + }, + { + "epoch": 0.58, + "grad_norm": 2.804918896071445, + "learning_rate": 3.946425202597147e-06, + "loss": 0.3199, + "step": 4176 + }, + { + "epoch": 0.58, + "grad_norm": 1.908739690791918, + "learning_rate": 3.9442244665720995e-06, + "loss": 0.3182, + "step": 4177 + }, + { + "epoch": 0.58, + "grad_norm": 3.1670678226457714, + "learning_rate": 3.942023944606246e-06, + "loss": 0.3186, + "step": 4178 + }, + { + "epoch": 0.58, + "grad_norm": 2.340690697262893, + "learning_rate": 3.939823637145742e-06, + "loss": 0.3112, + "step": 4179 + }, + { + "epoch": 0.58, + "grad_norm": 2.8932682657438042, + "learning_rate": 3.937623544636704e-06, + "loss": 0.3114, + "step": 4180 + }, + { + "epoch": 0.58, + "grad_norm": 1.9181864573799647, + "learning_rate": 3.935423667525202e-06, + "loss": 0.3078, + "step": 4181 + }, + { + "epoch": 0.58, + "grad_norm": 2.1912929206766627, + "learning_rate": 3.933224006257262e-06, + "loss": 0.3, + "step": 4182 + }, + { + "epoch": 0.58, + "grad_norm": 3.037500451620543, + "learning_rate": 3.931024561278867e-06, + "loss": 0.3079, + "step": 4183 + }, + { + "epoch": 0.58, + "grad_norm": 4.212883010148041, + "learning_rate": 3.928825333035953e-06, + "loss": 0.3056, + "step": 4184 + }, + { + "epoch": 0.58, + "grad_norm": 2.972505348412662, + "learning_rate": 3.926626321974419e-06, + "loss": 0.3287, + "step": 4185 + }, + { + "epoch": 0.58, + "grad_norm": 1.9220716122804866, + "learning_rate": 3.924427528540115e-06, + "loss": 0.3122, + "step": 4186 + }, + { + "epoch": 0.58, + "grad_norm": 1.832696141732382, + "learning_rate": 3.922228953178846e-06, + "loss": 0.3041, + "step": 4187 + }, + { + "epoch": 0.58, + "grad_norm": 3.175441368363064, + "learning_rate": 3.920030596336375e-06, + "loss": 0.328, + "step": 4188 + }, + { + "epoch": 0.58, + "grad_norm": 2.4307317067616863, + "learning_rate": 3.9178324584584225e-06, + "loss": 0.3053, + "step": 4189 + }, + { + "epoch": 0.58, + "grad_norm": 1.8839560564128721, + "learning_rate": 3.915634539990661e-06, + "loss": 0.3211, + "step": 4190 + }, + { + "epoch": 0.58, + "grad_norm": 1.878136153650511, + "learning_rate": 3.91343684137872e-06, + "loss": 0.3063, + "step": 4191 + }, + { + "epoch": 0.58, + "grad_norm": 3.935477682125078, + "learning_rate": 3.911239363068184e-06, + "loss": 0.3258, + "step": 4192 + }, + { + "epoch": 0.58, + "grad_norm": 2.154761974448189, + "learning_rate": 3.909042105504596e-06, + "loss": 0.3327, + "step": 4193 + }, + { + "epoch": 0.58, + "grad_norm": 2.43455876855334, + "learning_rate": 3.906845069133446e-06, + "loss": 0.3397, + "step": 4194 + }, + { + "epoch": 0.58, + "grad_norm": 12.314286995556046, + "learning_rate": 3.904648254400188e-06, + "loss": 0.3232, + "step": 4195 + }, + { + "epoch": 0.58, + "grad_norm": 1.9981689451808702, + "learning_rate": 3.902451661750229e-06, + "loss": 0.3401, + "step": 4196 + }, + { + "epoch": 0.58, + "grad_norm": 2.3763371464235785, + "learning_rate": 3.900255291628926e-06, + "loss": 0.3123, + "step": 4197 + }, + { + "epoch": 0.58, + "grad_norm": 2.2091321011397134, + "learning_rate": 3.898059144481597e-06, + "loss": 0.3015, + "step": 4198 + }, + { + "epoch": 0.58, + "grad_norm": 1.9588011025384824, + "learning_rate": 3.895863220753512e-06, + "loss": 0.3208, + "step": 4199 + }, + { + "epoch": 0.58, + "grad_norm": 3.137734923770493, + "learning_rate": 3.893667520889896e-06, + "loss": 0.3255, + "step": 4200 + }, + { + "epoch": 0.58, + "grad_norm": 4.195468785305381, + "learning_rate": 3.891472045335927e-06, + "loss": 0.2954, + "step": 4201 + }, + { + "epoch": 0.58, + "grad_norm": 3.205512773731775, + "learning_rate": 3.889276794536742e-06, + "loss": 0.3064, + "step": 4202 + }, + { + "epoch": 0.58, + "grad_norm": 2.564953177128413, + "learning_rate": 3.887081768937429e-06, + "loss": 0.3027, + "step": 4203 + }, + { + "epoch": 0.58, + "grad_norm": 2.545050336673661, + "learning_rate": 3.8848869689830275e-06, + "loss": 0.3195, + "step": 4204 + }, + { + "epoch": 0.58, + "grad_norm": 2.510734962773391, + "learning_rate": 3.882692395118536e-06, + "loss": 0.3045, + "step": 4205 + }, + { + "epoch": 0.58, + "grad_norm": 2.444913711408594, + "learning_rate": 3.8804980477889084e-06, + "loss": 0.3285, + "step": 4206 + }, + { + "epoch": 0.58, + "grad_norm": 4.071070028646231, + "learning_rate": 3.87830392743905e-06, + "loss": 0.3229, + "step": 4207 + }, + { + "epoch": 0.58, + "grad_norm": 2.1724180176779835, + "learning_rate": 3.8761100345138185e-06, + "loss": 0.3207, + "step": 4208 + }, + { + "epoch": 0.59, + "grad_norm": 3.323284597954652, + "learning_rate": 3.873916369458025e-06, + "loss": 0.314, + "step": 4209 + }, + { + "epoch": 0.59, + "grad_norm": 2.8230326595851882, + "learning_rate": 3.871722932716442e-06, + "loss": 0.32, + "step": 4210 + }, + { + "epoch": 0.59, + "grad_norm": 2.4437793486692185, + "learning_rate": 3.869529724733785e-06, + "loss": 0.3102, + "step": 4211 + }, + { + "epoch": 0.59, + "grad_norm": 5.217538996343241, + "learning_rate": 3.867336745954733e-06, + "loss": 0.3014, + "step": 4212 + }, + { + "epoch": 0.59, + "grad_norm": 4.159702736455695, + "learning_rate": 3.865143996823909e-06, + "loss": 0.3287, + "step": 4213 + }, + { + "epoch": 0.59, + "grad_norm": 2.3953679148049205, + "learning_rate": 3.862951477785897e-06, + "loss": 0.305, + "step": 4214 + }, + { + "epoch": 0.59, + "grad_norm": 3.269999534545128, + "learning_rate": 3.860759189285231e-06, + "loss": 0.3126, + "step": 4215 + }, + { + "epoch": 0.59, + "grad_norm": 2.0914900581880835, + "learning_rate": 3.858567131766401e-06, + "loss": 0.3079, + "step": 4216 + }, + { + "epoch": 0.59, + "grad_norm": 2.8772278597715726, + "learning_rate": 3.8563753056738475e-06, + "loss": 0.3371, + "step": 4217 + }, + { + "epoch": 0.59, + "grad_norm": 3.1714135212328114, + "learning_rate": 3.854183711451962e-06, + "loss": 0.2947, + "step": 4218 + }, + { + "epoch": 0.59, + "grad_norm": 2.0005233745290214, + "learning_rate": 3.851992349545095e-06, + "loss": 0.3248, + "step": 4219 + }, + { + "epoch": 0.59, + "grad_norm": 2.2314984879192727, + "learning_rate": 3.8498012203975445e-06, + "loss": 0.3161, + "step": 4220 + }, + { + "epoch": 0.59, + "grad_norm": 1.8955977196237597, + "learning_rate": 3.847610324453563e-06, + "loss": 0.306, + "step": 4221 + }, + { + "epoch": 0.59, + "grad_norm": 2.911109928048865, + "learning_rate": 3.845419662157357e-06, + "loss": 0.2972, + "step": 4222 + }, + { + "epoch": 0.59, + "grad_norm": 2.0463967271214156, + "learning_rate": 3.843229233953085e-06, + "loss": 0.3051, + "step": 4223 + }, + { + "epoch": 0.59, + "grad_norm": 2.184156599587968, + "learning_rate": 3.84103904028486e-06, + "loss": 0.3015, + "step": 4224 + }, + { + "epoch": 0.59, + "grad_norm": 2.1756895629993913, + "learning_rate": 3.838849081596741e-06, + "loss": 0.3116, + "step": 4225 + }, + { + "epoch": 0.59, + "grad_norm": 2.403179347020367, + "learning_rate": 3.836659358332747e-06, + "loss": 0.2965, + "step": 4226 + }, + { + "epoch": 0.59, + "grad_norm": 2.296714246105574, + "learning_rate": 3.834469870936842e-06, + "loss": 0.3105, + "step": 4227 + }, + { + "epoch": 0.59, + "grad_norm": 2.0948126718729814, + "learning_rate": 3.832280619852949e-06, + "loss": 0.3095, + "step": 4228 + }, + { + "epoch": 0.59, + "grad_norm": 2.028495108580485, + "learning_rate": 3.830091605524942e-06, + "loss": 0.2984, + "step": 4229 + }, + { + "epoch": 0.59, + "grad_norm": 2.7299671072985343, + "learning_rate": 3.82790282839664e-06, + "loss": 0.3099, + "step": 4230 + }, + { + "epoch": 0.59, + "grad_norm": 1.9206814980078364, + "learning_rate": 3.825714288911821e-06, + "loss": 0.3226, + "step": 4231 + }, + { + "epoch": 0.59, + "grad_norm": 4.391394186878069, + "learning_rate": 3.823525987514214e-06, + "loss": 0.3168, + "step": 4232 + }, + { + "epoch": 0.59, + "grad_norm": 2.774982283716798, + "learning_rate": 3.821337924647499e-06, + "loss": 0.33, + "step": 4233 + }, + { + "epoch": 0.59, + "grad_norm": 1.8235467515728816, + "learning_rate": 3.819150100755306e-06, + "loss": 0.3146, + "step": 4234 + }, + { + "epoch": 0.59, + "grad_norm": 2.152919803646103, + "learning_rate": 3.816962516281217e-06, + "loss": 0.3064, + "step": 4235 + }, + { + "epoch": 0.59, + "grad_norm": 2.7935537891385778, + "learning_rate": 3.814775171668769e-06, + "loss": 0.2996, + "step": 4236 + }, + { + "epoch": 0.59, + "grad_norm": 6.330974671339897, + "learning_rate": 3.8125880673614435e-06, + "loss": 0.3273, + "step": 4237 + }, + { + "epoch": 0.59, + "grad_norm": 1.9994192873047114, + "learning_rate": 3.8104012038026804e-06, + "loss": 0.3055, + "step": 4238 + }, + { + "epoch": 0.59, + "grad_norm": 1.974443305479644, + "learning_rate": 3.8082145814358684e-06, + "loss": 0.3071, + "step": 4239 + }, + { + "epoch": 0.59, + "grad_norm": 2.0785082271039435, + "learning_rate": 3.8060282007043425e-06, + "loss": 0.2929, + "step": 4240 + }, + { + "epoch": 0.59, + "grad_norm": 2.4466494469212057, + "learning_rate": 3.8038420620513973e-06, + "loss": 0.3292, + "step": 4241 + }, + { + "epoch": 0.59, + "grad_norm": 2.007566545437742, + "learning_rate": 3.8016561659202717e-06, + "loss": 0.3036, + "step": 4242 + }, + { + "epoch": 0.59, + "grad_norm": 2.268071399830023, + "learning_rate": 3.799470512754159e-06, + "loss": 0.3189, + "step": 4243 + }, + { + "epoch": 0.59, + "grad_norm": 2.467110918538254, + "learning_rate": 3.7972851029962e-06, + "loss": 0.3044, + "step": 4244 + }, + { + "epoch": 0.59, + "grad_norm": 2.018738585997095, + "learning_rate": 3.7950999370894886e-06, + "loss": 0.3147, + "step": 4245 + }, + { + "epoch": 0.59, + "grad_norm": 2.166798504641617, + "learning_rate": 3.7929150154770706e-06, + "loss": 0.3183, + "step": 4246 + }, + { + "epoch": 0.59, + "grad_norm": 3.7939781030708977, + "learning_rate": 3.790730338601937e-06, + "loss": 0.3225, + "step": 4247 + }, + { + "epoch": 0.59, + "grad_norm": 2.0314601939306343, + "learning_rate": 3.788545906907033e-06, + "loss": 0.314, + "step": 4248 + }, + { + "epoch": 0.59, + "grad_norm": 2.444018932422656, + "learning_rate": 3.786361720835256e-06, + "loss": 0.3188, + "step": 4249 + }, + { + "epoch": 0.59, + "grad_norm": 2.157131488827354, + "learning_rate": 3.7841777808294506e-06, + "loss": 0.304, + "step": 4250 + }, + { + "epoch": 0.59, + "grad_norm": 2.056192765939265, + "learning_rate": 3.7819940873324107e-06, + "loss": 0.3233, + "step": 4251 + }, + { + "epoch": 0.59, + "grad_norm": 2.033025145495622, + "learning_rate": 3.779810640786883e-06, + "loss": 0.2972, + "step": 4252 + }, + { + "epoch": 0.59, + "grad_norm": 2.2707591858534246, + "learning_rate": 3.7776274416355614e-06, + "loss": 0.2934, + "step": 4253 + }, + { + "epoch": 0.59, + "grad_norm": 2.1091296384319276, + "learning_rate": 3.775444490321091e-06, + "loss": 0.2962, + "step": 4254 + }, + { + "epoch": 0.59, + "grad_norm": 3.833667701667341, + "learning_rate": 3.7732617872860688e-06, + "loss": 0.3081, + "step": 4255 + }, + { + "epoch": 0.59, + "grad_norm": 3.467314107733532, + "learning_rate": 3.7710793329730365e-06, + "loss": 0.3065, + "step": 4256 + }, + { + "epoch": 0.59, + "grad_norm": 2.725350951719964, + "learning_rate": 3.7688971278244878e-06, + "loss": 0.3052, + "step": 4257 + }, + { + "epoch": 0.59, + "grad_norm": 2.65756600782525, + "learning_rate": 3.76671517228287e-06, + "loss": 0.3325, + "step": 4258 + }, + { + "epoch": 0.59, + "grad_norm": 3.838848132826358, + "learning_rate": 3.764533466790574e-06, + "loss": 0.2988, + "step": 4259 + }, + { + "epoch": 0.59, + "grad_norm": 2.4954669577448776, + "learning_rate": 3.7623520117899415e-06, + "loss": 0.3241, + "step": 4260 + }, + { + "epoch": 0.59, + "grad_norm": 11.10902485178655, + "learning_rate": 3.7601708077232646e-06, + "loss": 0.3103, + "step": 4261 + }, + { + "epoch": 0.59, + "grad_norm": 2.1734663309510323, + "learning_rate": 3.7579898550327844e-06, + "loss": 0.3055, + "step": 4262 + }, + { + "epoch": 0.59, + "grad_norm": 2.27157173710974, + "learning_rate": 3.7558091541606886e-06, + "loss": 0.3142, + "step": 4263 + }, + { + "epoch": 0.59, + "grad_norm": 2.544138178789773, + "learning_rate": 3.7536287055491164e-06, + "loss": 0.3301, + "step": 4264 + }, + { + "epoch": 0.59, + "grad_norm": 2.2555623886645506, + "learning_rate": 3.7514485096401577e-06, + "loss": 0.3356, + "step": 4265 + }, + { + "epoch": 0.59, + "grad_norm": 2.101383240231242, + "learning_rate": 3.7492685668758443e-06, + "loss": 0.3097, + "step": 4266 + }, + { + "epoch": 0.59, + "grad_norm": 3.502570758552127, + "learning_rate": 3.747088877698165e-06, + "loss": 0.3206, + "step": 4267 + }, + { + "epoch": 0.59, + "grad_norm": 2.3516766570066423, + "learning_rate": 3.744909442549051e-06, + "loss": 0.3231, + "step": 4268 + }, + { + "epoch": 0.59, + "grad_norm": 2.9678462878036607, + "learning_rate": 3.7427302618703866e-06, + "loss": 0.3282, + "step": 4269 + }, + { + "epoch": 0.59, + "grad_norm": 2.2637425489716394, + "learning_rate": 3.740551336103999e-06, + "loss": 0.3192, + "step": 4270 + }, + { + "epoch": 0.59, + "grad_norm": 2.338510415626582, + "learning_rate": 3.7383726656916684e-06, + "loss": 0.311, + "step": 4271 + }, + { + "epoch": 0.59, + "grad_norm": 2.772806176116788, + "learning_rate": 3.7361942510751233e-06, + "loss": 0.3023, + "step": 4272 + }, + { + "epoch": 0.59, + "grad_norm": 2.141259498903598, + "learning_rate": 3.7340160926960354e-06, + "loss": 0.3114, + "step": 4273 + }, + { + "epoch": 0.59, + "grad_norm": 2.6396198439147853, + "learning_rate": 3.7318381909960284e-06, + "loss": 0.3154, + "step": 4274 + }, + { + "epoch": 0.59, + "grad_norm": 2.7668351162739415, + "learning_rate": 3.7296605464166753e-06, + "loss": 0.3132, + "step": 4275 + }, + { + "epoch": 0.59, + "grad_norm": 2.298642059105651, + "learning_rate": 3.7274831593994954e-06, + "loss": 0.3117, + "step": 4276 + }, + { + "epoch": 0.59, + "grad_norm": 2.3873894013509265, + "learning_rate": 3.7253060303859525e-06, + "loss": 0.2927, + "step": 4277 + }, + { + "epoch": 0.59, + "grad_norm": 2.188915248207724, + "learning_rate": 3.723129159817462e-06, + "loss": 0.2942, + "step": 4278 + }, + { + "epoch": 0.59, + "grad_norm": 3.1945841292174713, + "learning_rate": 3.7209525481353887e-06, + "loss": 0.3144, + "step": 4279 + }, + { + "epoch": 0.59, + "grad_norm": 2.5275586154654364, + "learning_rate": 3.7187761957810377e-06, + "loss": 0.3219, + "step": 4280 + }, + { + "epoch": 0.6, + "grad_norm": 3.23614756255516, + "learning_rate": 3.7166001031956695e-06, + "loss": 0.3326, + "step": 4281 + }, + { + "epoch": 0.6, + "grad_norm": 2.7022191053672096, + "learning_rate": 3.7144242708204854e-06, + "loss": 0.3279, + "step": 4282 + }, + { + "epoch": 0.6, + "grad_norm": 2.3650509023100845, + "learning_rate": 3.7122486990966367e-06, + "loss": 0.3422, + "step": 4283 + }, + { + "epoch": 0.6, + "grad_norm": 2.197694243304614, + "learning_rate": 3.710073388465225e-06, + "loss": 0.3345, + "step": 4284 + }, + { + "epoch": 0.6, + "grad_norm": 2.0990657421636856, + "learning_rate": 3.707898339367294e-06, + "loss": 0.2939, + "step": 4285 + }, + { + "epoch": 0.6, + "grad_norm": 3.1796214425844807, + "learning_rate": 3.7057235522438386e-06, + "loss": 0.3103, + "step": 4286 + }, + { + "epoch": 0.6, + "grad_norm": 2.0651992648367736, + "learning_rate": 3.7035490275357948e-06, + "loss": 0.3027, + "step": 4287 + }, + { + "epoch": 0.6, + "grad_norm": 2.38390023963287, + "learning_rate": 3.7013747656840525e-06, + "loss": 0.3265, + "step": 4288 + }, + { + "epoch": 0.6, + "grad_norm": 2.281941184685513, + "learning_rate": 3.699200767129441e-06, + "loss": 0.3188, + "step": 4289 + }, + { + "epoch": 0.6, + "grad_norm": 2.228740488370659, + "learning_rate": 3.697027032312742e-06, + "loss": 0.2964, + "step": 4290 + }, + { + "epoch": 0.6, + "grad_norm": 2.6943183739688723, + "learning_rate": 3.6948535616746816e-06, + "loss": 0.3439, + "step": 4291 + }, + { + "epoch": 0.6, + "grad_norm": 2.8970054575437776, + "learning_rate": 3.692680355655932e-06, + "loss": 0.331, + "step": 4292 + }, + { + "epoch": 0.6, + "grad_norm": 2.9728412652413887, + "learning_rate": 3.690507414697115e-06, + "loss": 0.3137, + "step": 4293 + }, + { + "epoch": 0.6, + "grad_norm": 2.1461152403533, + "learning_rate": 3.6883347392387915e-06, + "loss": 0.3052, + "step": 4294 + }, + { + "epoch": 0.6, + "grad_norm": 2.7200574380678235, + "learning_rate": 3.686162329721476e-06, + "loss": 0.3244, + "step": 4295 + }, + { + "epoch": 0.6, + "grad_norm": 2.301730313706705, + "learning_rate": 3.6839901865856237e-06, + "loss": 0.3151, + "step": 4296 + }, + { + "epoch": 0.6, + "grad_norm": 2.3660914984667625, + "learning_rate": 3.6818183102716386e-06, + "loss": 0.2903, + "step": 4297 + }, + { + "epoch": 0.6, + "grad_norm": 2.011922209060022, + "learning_rate": 3.6796467012198726e-06, + "loss": 0.3092, + "step": 4298 + }, + { + "epoch": 0.6, + "grad_norm": 2.8516550788064334, + "learning_rate": 3.6774753598706173e-06, + "loss": 0.3105, + "step": 4299 + }, + { + "epoch": 0.6, + "grad_norm": 2.3145706323259168, + "learning_rate": 3.6753042866641148e-06, + "loss": 0.334, + "step": 4300 + }, + { + "epoch": 0.6, + "grad_norm": 8.926102924874227, + "learning_rate": 3.6731334820405527e-06, + "loss": 0.358, + "step": 4301 + }, + { + "epoch": 0.6, + "grad_norm": 2.223761245232349, + "learning_rate": 3.670962946440063e-06, + "loss": 0.296, + "step": 4302 + }, + { + "epoch": 0.6, + "grad_norm": 2.0871870203524963, + "learning_rate": 3.6687926803027224e-06, + "loss": 0.3275, + "step": 4303 + }, + { + "epoch": 0.6, + "grad_norm": 2.080794625004391, + "learning_rate": 3.6666226840685536e-06, + "loss": 0.3189, + "step": 4304 + }, + { + "epoch": 0.6, + "grad_norm": 2.4077088859962488, + "learning_rate": 3.6644529581775267e-06, + "loss": 0.317, + "step": 4305 + }, + { + "epoch": 0.6, + "grad_norm": 2.1088157702317734, + "learning_rate": 3.6622835030695525e-06, + "loss": 0.3224, + "step": 4306 + }, + { + "epoch": 0.6, + "grad_norm": 2.2096575799101004, + "learning_rate": 3.66011431918449e-06, + "loss": 0.3084, + "step": 4307 + }, + { + "epoch": 0.6, + "grad_norm": 2.4688014147179502, + "learning_rate": 3.657945406962145e-06, + "loss": 0.3349, + "step": 4308 + }, + { + "epoch": 0.6, + "grad_norm": 3.5469462693728935, + "learning_rate": 3.6557767668422616e-06, + "loss": 0.2931, + "step": 4309 + }, + { + "epoch": 0.6, + "grad_norm": 2.215928415969946, + "learning_rate": 3.653608399264538e-06, + "loss": 0.2907, + "step": 4310 + }, + { + "epoch": 0.6, + "grad_norm": 2.089394753378041, + "learning_rate": 3.651440304668608e-06, + "loss": 0.3056, + "step": 4311 + }, + { + "epoch": 0.6, + "grad_norm": 2.3401411294686127, + "learning_rate": 3.649272483494058e-06, + "loss": 0.3441, + "step": 4312 + }, + { + "epoch": 0.6, + "grad_norm": 1.9088165164387692, + "learning_rate": 3.6471049361804124e-06, + "loss": 0.3307, + "step": 4313 + }, + { + "epoch": 0.6, + "grad_norm": 3.532066865149563, + "learning_rate": 3.644937663167143e-06, + "loss": 0.3382, + "step": 4314 + }, + { + "epoch": 0.6, + "grad_norm": 2.7694264144563054, + "learning_rate": 3.642770664893668e-06, + "loss": 0.2958, + "step": 4315 + }, + { + "epoch": 0.6, + "grad_norm": 2.5405966839167924, + "learning_rate": 3.640603941799345e-06, + "loss": 0.3249, + "step": 4316 + }, + { + "epoch": 0.6, + "grad_norm": 1.804289838579934, + "learning_rate": 3.638437494323479e-06, + "loss": 0.2999, + "step": 4317 + }, + { + "epoch": 0.6, + "grad_norm": 2.704652094013011, + "learning_rate": 3.6362713229053204e-06, + "loss": 0.3154, + "step": 4318 + }, + { + "epoch": 0.6, + "grad_norm": 3.0941457189956085, + "learning_rate": 3.6341054279840617e-06, + "loss": 0.3279, + "step": 4319 + }, + { + "epoch": 0.6, + "grad_norm": 18.755269205241664, + "learning_rate": 3.631939809998838e-06, + "loss": 0.3191, + "step": 4320 + }, + { + "epoch": 0.6, + "grad_norm": 2.0483498032828398, + "learning_rate": 3.629774469388731e-06, + "loss": 0.3255, + "step": 4321 + }, + { + "epoch": 0.6, + "grad_norm": 1.907068817099314, + "learning_rate": 3.6276094065927653e-06, + "loss": 0.3119, + "step": 4322 + }, + { + "epoch": 0.6, + "grad_norm": 2.584722052009604, + "learning_rate": 3.6254446220499075e-06, + "loss": 0.3045, + "step": 4323 + }, + { + "epoch": 0.6, + "grad_norm": 2.359233105715773, + "learning_rate": 3.6232801161990705e-06, + "loss": 0.3351, + "step": 4324 + }, + { + "epoch": 0.6, + "grad_norm": 1.8566675903902914, + "learning_rate": 3.621115889479107e-06, + "loss": 0.3542, + "step": 4325 + }, + { + "epoch": 0.6, + "grad_norm": 1.8946281384108392, + "learning_rate": 3.6189519423288164e-06, + "loss": 0.3232, + "step": 4326 + }, + { + "epoch": 0.6, + "grad_norm": 1.7828631505082122, + "learning_rate": 3.616788275186942e-06, + "loss": 0.2964, + "step": 4327 + }, + { + "epoch": 0.6, + "grad_norm": 2.38419539731054, + "learning_rate": 3.6146248884921687e-06, + "loss": 0.3056, + "step": 4328 + }, + { + "epoch": 0.6, + "grad_norm": 3.3685446878868355, + "learning_rate": 3.612461782683122e-06, + "loss": 0.3252, + "step": 4329 + }, + { + "epoch": 0.6, + "grad_norm": 2.5748718461661757, + "learning_rate": 3.6102989581983753e-06, + "loss": 0.3277, + "step": 4330 + }, + { + "epoch": 0.6, + "grad_norm": 2.2635218605807426, + "learning_rate": 3.608136415476443e-06, + "loss": 0.3387, + "step": 4331 + }, + { + "epoch": 0.6, + "grad_norm": 2.467048450613862, + "learning_rate": 3.6059741549557805e-06, + "loss": 0.3243, + "step": 4332 + }, + { + "epoch": 0.6, + "grad_norm": 4.459660233501713, + "learning_rate": 3.603812177074788e-06, + "loss": 0.3055, + "step": 4333 + }, + { + "epoch": 0.6, + "grad_norm": 2.387468292445309, + "learning_rate": 3.601650482271809e-06, + "loss": 0.3228, + "step": 4334 + }, + { + "epoch": 0.6, + "grad_norm": 2.186319535703766, + "learning_rate": 3.5994890709851255e-06, + "loss": 0.3133, + "step": 4335 + }, + { + "epoch": 0.6, + "grad_norm": 2.263773910208385, + "learning_rate": 3.5973279436529685e-06, + "loss": 0.2992, + "step": 4336 + }, + { + "epoch": 0.6, + "grad_norm": 2.183559035057324, + "learning_rate": 3.595167100713507e-06, + "loss": 0.3031, + "step": 4337 + }, + { + "epoch": 0.6, + "grad_norm": 3.623259131130104, + "learning_rate": 3.5930065426048543e-06, + "loss": 0.3256, + "step": 4338 + }, + { + "epoch": 0.6, + "grad_norm": 2.43487069380839, + "learning_rate": 3.5908462697650626e-06, + "loss": 0.2962, + "step": 4339 + }, + { + "epoch": 0.6, + "grad_norm": 3.9121023447063465, + "learning_rate": 3.58868628263213e-06, + "loss": 0.3099, + "step": 4340 + }, + { + "epoch": 0.6, + "grad_norm": 2.030894361686682, + "learning_rate": 3.586526581643996e-06, + "loss": 0.3036, + "step": 4341 + }, + { + "epoch": 0.6, + "grad_norm": 2.3170556056351472, + "learning_rate": 3.58436716723854e-06, + "loss": 0.3255, + "step": 4342 + }, + { + "epoch": 0.6, + "grad_norm": 2.078031013752107, + "learning_rate": 3.582208039853584e-06, + "loss": 0.3299, + "step": 4343 + }, + { + "epoch": 0.6, + "grad_norm": 2.1275935863133295, + "learning_rate": 3.580049199926895e-06, + "loss": 0.2973, + "step": 4344 + }, + { + "epoch": 0.6, + "grad_norm": 4.347772785518005, + "learning_rate": 3.5778906478961796e-06, + "loss": 0.3361, + "step": 4345 + }, + { + "epoch": 0.6, + "grad_norm": 4.294689192133297, + "learning_rate": 3.5757323841990833e-06, + "loss": 0.2948, + "step": 4346 + }, + { + "epoch": 0.6, + "grad_norm": 2.168970058606711, + "learning_rate": 3.5735744092731963e-06, + "loss": 0.3477, + "step": 4347 + }, + { + "epoch": 0.6, + "grad_norm": 2.7889803936964084, + "learning_rate": 3.57141672355605e-06, + "loss": 0.3097, + "step": 4348 + }, + { + "epoch": 0.6, + "grad_norm": 2.1911056508272515, + "learning_rate": 3.569259327485115e-06, + "loss": 0.3223, + "step": 4349 + }, + { + "epoch": 0.6, + "grad_norm": 2.240402577768887, + "learning_rate": 3.567102221497808e-06, + "loss": 0.3074, + "step": 4350 + }, + { + "epoch": 0.6, + "grad_norm": 1.8843370939867108, + "learning_rate": 3.56494540603148e-06, + "loss": 0.29, + "step": 4351 + }, + { + "epoch": 0.6, + "grad_norm": 2.2711098530948175, + "learning_rate": 3.562788881523428e-06, + "loss": 0.3171, + "step": 4352 + }, + { + "epoch": 0.61, + "grad_norm": 1.9986808115751211, + "learning_rate": 3.5606326484108894e-06, + "loss": 0.3123, + "step": 4353 + }, + { + "epoch": 0.61, + "grad_norm": 3.1075498120127123, + "learning_rate": 3.558476707131042e-06, + "loss": 0.2951, + "step": 4354 + }, + { + "epoch": 0.61, + "grad_norm": 1.8243792289729257, + "learning_rate": 3.556321058121005e-06, + "loss": 0.3013, + "step": 4355 + }, + { + "epoch": 0.61, + "grad_norm": 2.188556698654581, + "learning_rate": 3.5541657018178368e-06, + "loss": 0.3167, + "step": 4356 + }, + { + "epoch": 0.61, + "grad_norm": 2.0358138368666614, + "learning_rate": 3.5520106386585367e-06, + "loss": 0.3176, + "step": 4357 + }, + { + "epoch": 0.61, + "grad_norm": 2.4164490082567975, + "learning_rate": 3.5498558690800455e-06, + "loss": 0.327, + "step": 4358 + }, + { + "epoch": 0.61, + "grad_norm": 1.8085246389861283, + "learning_rate": 3.547701393519244e-06, + "loss": 0.3091, + "step": 4359 + }, + { + "epoch": 0.61, + "grad_norm": 2.2855758506615604, + "learning_rate": 3.5455472124129535e-06, + "loss": 0.314, + "step": 4360 + }, + { + "epoch": 0.61, + "grad_norm": 2.7317914928363884, + "learning_rate": 3.543393326197937e-06, + "loss": 0.3218, + "step": 4361 + }, + { + "epoch": 0.61, + "grad_norm": 2.138012079991209, + "learning_rate": 3.541239735310896e-06, + "loss": 0.3139, + "step": 4362 + }, + { + "epoch": 0.61, + "grad_norm": 2.1548756930626194, + "learning_rate": 3.5390864401884717e-06, + "loss": 0.3408, + "step": 4363 + }, + { + "epoch": 0.61, + "grad_norm": 3.4758167487604945, + "learning_rate": 3.5369334412672474e-06, + "loss": 0.3122, + "step": 4364 + }, + { + "epoch": 0.61, + "grad_norm": 1.8060701381627502, + "learning_rate": 3.534780738983743e-06, + "loss": 0.302, + "step": 4365 + }, + { + "epoch": 0.61, + "grad_norm": 3.7137060621484217, + "learning_rate": 3.5326283337744207e-06, + "loss": 0.3041, + "step": 4366 + }, + { + "epoch": 0.61, + "grad_norm": 3.223701619638287, + "learning_rate": 3.5304762260756844e-06, + "loss": 0.3033, + "step": 4367 + }, + { + "epoch": 0.61, + "grad_norm": 1.981531258391877, + "learning_rate": 3.5283244163238726e-06, + "loss": 0.3148, + "step": 4368 + }, + { + "epoch": 0.61, + "grad_norm": 1.8974303777774297, + "learning_rate": 3.526172904955266e-06, + "loss": 0.3057, + "step": 4369 + }, + { + "epoch": 0.61, + "grad_norm": 2.1353396560786284, + "learning_rate": 3.5240216924060883e-06, + "loss": 0.3083, + "step": 4370 + }, + { + "epoch": 0.61, + "grad_norm": 4.58032795753704, + "learning_rate": 3.5218707791124975e-06, + "loss": 0.3051, + "step": 4371 + }, + { + "epoch": 0.61, + "grad_norm": 2.2274538079764175, + "learning_rate": 3.5197201655105916e-06, + "loss": 0.3202, + "step": 4372 + }, + { + "epoch": 0.61, + "grad_norm": 1.9302969083751387, + "learning_rate": 3.51756985203641e-06, + "loss": 0.2838, + "step": 4373 + }, + { + "epoch": 0.61, + "grad_norm": 2.443404757748088, + "learning_rate": 3.515419839125931e-06, + "loss": 0.3196, + "step": 4374 + }, + { + "epoch": 0.61, + "grad_norm": 2.5512066531837556, + "learning_rate": 3.5132701272150693e-06, + "loss": 0.3468, + "step": 4375 + }, + { + "epoch": 0.61, + "grad_norm": 1.9128688090831636, + "learning_rate": 3.511120716739681e-06, + "loss": 0.3107, + "step": 4376 + }, + { + "epoch": 0.61, + "grad_norm": 2.285980930457886, + "learning_rate": 3.508971608135562e-06, + "loss": 0.3027, + "step": 4377 + }, + { + "epoch": 0.61, + "grad_norm": 2.081918869171168, + "learning_rate": 3.506822801838442e-06, + "loss": 0.3382, + "step": 4378 + }, + { + "epoch": 0.61, + "grad_norm": 3.452864166859561, + "learning_rate": 3.5046742982839964e-06, + "loss": 0.2962, + "step": 4379 + }, + { + "epoch": 0.61, + "grad_norm": 2.1532507667102516, + "learning_rate": 3.5025260979078347e-06, + "loss": 0.3009, + "step": 4380 + }, + { + "epoch": 0.61, + "grad_norm": 2.1777622722653898, + "learning_rate": 3.5003782011455063e-06, + "loss": 0.3184, + "step": 4381 + }, + { + "epoch": 0.61, + "grad_norm": 3.0140462930098306, + "learning_rate": 3.4982306084324974e-06, + "loss": 0.3161, + "step": 4382 + }, + { + "epoch": 0.61, + "grad_norm": 2.042603032013137, + "learning_rate": 3.496083320204235e-06, + "loss": 0.3191, + "step": 4383 + }, + { + "epoch": 0.61, + "grad_norm": 2.309467932702232, + "learning_rate": 3.493936336896083e-06, + "loss": 0.3182, + "step": 4384 + }, + { + "epoch": 0.61, + "grad_norm": 2.781703755282522, + "learning_rate": 3.4917896589433437e-06, + "loss": 0.312, + "step": 4385 + }, + { + "epoch": 0.61, + "grad_norm": 2.0897338120718723, + "learning_rate": 3.489643286781256e-06, + "loss": 0.3225, + "step": 4386 + }, + { + "epoch": 0.61, + "grad_norm": 2.6019022142530095, + "learning_rate": 3.4874972208450002e-06, + "loss": 0.3206, + "step": 4387 + }, + { + "epoch": 0.61, + "grad_norm": 2.2311407174499016, + "learning_rate": 3.4853514615696937e-06, + "loss": 0.3117, + "step": 4388 + }, + { + "epoch": 0.61, + "grad_norm": 1.829575706562002, + "learning_rate": 3.4832060093903874e-06, + "loss": 0.3118, + "step": 4389 + }, + { + "epoch": 0.61, + "grad_norm": 2.078605645281586, + "learning_rate": 3.4810608647420753e-06, + "loss": 0.3337, + "step": 4390 + }, + { + "epoch": 0.61, + "grad_norm": 1.9947462240255454, + "learning_rate": 3.4789160280596866e-06, + "loss": 0.2908, + "step": 4391 + }, + { + "epoch": 0.61, + "grad_norm": 2.3151795641027544, + "learning_rate": 3.4767714997780882e-06, + "loss": 0.3157, + "step": 4392 + }, + { + "epoch": 0.61, + "grad_norm": 2.319376480337022, + "learning_rate": 3.4746272803320848e-06, + "loss": 0.3025, + "step": 4393 + }, + { + "epoch": 0.61, + "grad_norm": 1.9999953181139194, + "learning_rate": 3.4724833701564176e-06, + "loss": 0.3093, + "step": 4394 + }, + { + "epoch": 0.61, + "grad_norm": 2.422682017488054, + "learning_rate": 3.4703397696857644e-06, + "loss": 0.3095, + "step": 4395 + }, + { + "epoch": 0.61, + "grad_norm": 2.0231439084577736, + "learning_rate": 3.468196479354745e-06, + "loss": 0.3106, + "step": 4396 + }, + { + "epoch": 0.61, + "grad_norm": 2.029572327565453, + "learning_rate": 3.4660534995979126e-06, + "loss": 0.3167, + "step": 4397 + }, + { + "epoch": 0.61, + "grad_norm": 2.142623552816673, + "learning_rate": 3.463910830849756e-06, + "loss": 0.3179, + "step": 4398 + }, + { + "epoch": 0.61, + "grad_norm": 2.232747020811696, + "learning_rate": 3.461768473544703e-06, + "loss": 0.2833, + "step": 4399 + }, + { + "epoch": 0.61, + "grad_norm": 2.857067753923206, + "learning_rate": 3.459626428117119e-06, + "loss": 0.3042, + "step": 4400 + }, + { + "epoch": 0.61, + "grad_norm": 2.6629883282993334, + "learning_rate": 3.457484695001304e-06, + "loss": 0.327, + "step": 4401 + }, + { + "epoch": 0.61, + "grad_norm": 3.022053632986339, + "learning_rate": 3.4553432746314963e-06, + "loss": 0.3112, + "step": 4402 + }, + { + "epoch": 0.61, + "grad_norm": 2.5882512732627236, + "learning_rate": 3.453202167441871e-06, + "loss": 0.3133, + "step": 4403 + }, + { + "epoch": 0.61, + "grad_norm": 2.22445410752946, + "learning_rate": 3.451061373866536e-06, + "loss": 0.2835, + "step": 4404 + }, + { + "epoch": 0.61, + "grad_norm": 2.281243508759865, + "learning_rate": 3.4489208943395424e-06, + "loss": 0.3089, + "step": 4405 + }, + { + "epoch": 0.61, + "grad_norm": 2.1749982001114456, + "learning_rate": 3.4467807292948726e-06, + "loss": 0.3217, + "step": 4406 + }, + { + "epoch": 0.61, + "grad_norm": 2.0928450167135053, + "learning_rate": 3.4446408791664464e-06, + "loss": 0.3371, + "step": 4407 + }, + { + "epoch": 0.61, + "grad_norm": 1.8983961595326884, + "learning_rate": 3.4425013443881193e-06, + "loss": 0.3242, + "step": 4408 + }, + { + "epoch": 0.61, + "grad_norm": 2.101605756427566, + "learning_rate": 3.4403621253936842e-06, + "loss": 0.3325, + "step": 4409 + }, + { + "epoch": 0.61, + "grad_norm": 2.5263253910394026, + "learning_rate": 3.4382232226168692e-06, + "loss": 0.3034, + "step": 4410 + }, + { + "epoch": 0.61, + "grad_norm": 4.334547540468409, + "learning_rate": 3.4360846364913373e-06, + "loss": 0.2879, + "step": 4411 + }, + { + "epoch": 0.61, + "grad_norm": 3.4551296648497223, + "learning_rate": 3.433946367450688e-06, + "loss": 0.3048, + "step": 4412 + }, + { + "epoch": 0.61, + "grad_norm": 1.9660776518028422, + "learning_rate": 3.4318084159284583e-06, + "loss": 0.2941, + "step": 4413 + }, + { + "epoch": 0.61, + "grad_norm": 1.9689733222100414, + "learning_rate": 3.4296707823581192e-06, + "loss": 0.3048, + "step": 4414 + }, + { + "epoch": 0.61, + "grad_norm": 3.708056846171895, + "learning_rate": 3.4275334671730765e-06, + "loss": 0.3243, + "step": 4415 + }, + { + "epoch": 0.61, + "grad_norm": 2.565572554392359, + "learning_rate": 3.425396470806672e-06, + "loss": 0.3187, + "step": 4416 + }, + { + "epoch": 0.61, + "grad_norm": 2.3139093573454868, + "learning_rate": 3.423259793692185e-06, + "loss": 0.3124, + "step": 4417 + }, + { + "epoch": 0.61, + "grad_norm": 2.2145200942775465, + "learning_rate": 3.4211234362628255e-06, + "loss": 0.3042, + "step": 4418 + }, + { + "epoch": 0.61, + "grad_norm": 2.006182209165535, + "learning_rate": 3.418987398951744e-06, + "loss": 0.3193, + "step": 4419 + }, + { + "epoch": 0.61, + "grad_norm": 1.9572420767486758, + "learning_rate": 3.4168516821920212e-06, + "loss": 0.3098, + "step": 4420 + }, + { + "epoch": 0.61, + "grad_norm": 1.937405195323466, + "learning_rate": 3.414716286416675e-06, + "loss": 0.2957, + "step": 4421 + }, + { + "epoch": 0.61, + "grad_norm": 2.5304550915422634, + "learning_rate": 3.412581212058661e-06, + "loss": 0.3101, + "step": 4422 + }, + { + "epoch": 0.61, + "grad_norm": 2.177693365894482, + "learning_rate": 3.4104464595508646e-06, + "loss": 0.335, + "step": 4423 + }, + { + "epoch": 0.61, + "grad_norm": 1.9441645589584349, + "learning_rate": 3.4083120293261106e-06, + "loss": 0.3226, + "step": 4424 + }, + { + "epoch": 0.62, + "grad_norm": 2.247242221110447, + "learning_rate": 3.406177921817154e-06, + "loss": 0.3303, + "step": 4425 + }, + { + "epoch": 0.62, + "grad_norm": 2.2767323714345715, + "learning_rate": 3.404044137456688e-06, + "loss": 0.3182, + "step": 4426 + }, + { + "epoch": 0.62, + "grad_norm": 2.574148822814514, + "learning_rate": 3.4019106766773364e-06, + "loss": 0.3223, + "step": 4427 + }, + { + "epoch": 0.62, + "grad_norm": 2.2035705057765047, + "learning_rate": 3.3997775399116615e-06, + "loss": 0.3086, + "step": 4428 + }, + { + "epoch": 0.62, + "grad_norm": 2.223687727133311, + "learning_rate": 3.3976447275921574e-06, + "loss": 0.3063, + "step": 4429 + }, + { + "epoch": 0.62, + "grad_norm": 3.0197780364296976, + "learning_rate": 3.3955122401512554e-06, + "loss": 0.3229, + "step": 4430 + }, + { + "epoch": 0.62, + "grad_norm": 2.6245746421096303, + "learning_rate": 3.3933800780213176e-06, + "loss": 0.3087, + "step": 4431 + }, + { + "epoch": 0.62, + "grad_norm": 1.951988509603891, + "learning_rate": 3.3912482416346405e-06, + "loss": 0.3004, + "step": 4432 + }, + { + "epoch": 0.62, + "grad_norm": 2.082289412142825, + "learning_rate": 3.3891167314234564e-06, + "loss": 0.3346, + "step": 4433 + }, + { + "epoch": 0.62, + "grad_norm": 2.1952783821348887, + "learning_rate": 3.386985547819929e-06, + "loss": 0.3361, + "step": 4434 + }, + { + "epoch": 0.62, + "grad_norm": 1.9502553025849432, + "learning_rate": 3.3848546912561576e-06, + "loss": 0.3202, + "step": 4435 + }, + { + "epoch": 0.62, + "grad_norm": 5.5967252814074415, + "learning_rate": 3.382724162164176e-06, + "loss": 0.3076, + "step": 4436 + }, + { + "epoch": 0.62, + "grad_norm": 3.0871939671293167, + "learning_rate": 3.3805939609759497e-06, + "loss": 0.3091, + "step": 4437 + }, + { + "epoch": 0.62, + "grad_norm": 2.616553167541666, + "learning_rate": 3.3784640881233764e-06, + "loss": 0.3115, + "step": 4438 + }, + { + "epoch": 0.62, + "grad_norm": 1.8486415121406814, + "learning_rate": 3.3763345440382924e-06, + "loss": 0.3279, + "step": 4439 + }, + { + "epoch": 0.62, + "grad_norm": 2.040793120656189, + "learning_rate": 3.3742053291524646e-06, + "loss": 0.3315, + "step": 4440 + }, + { + "epoch": 0.62, + "grad_norm": 1.9815686682500226, + "learning_rate": 3.3720764438975895e-06, + "loss": 0.2931, + "step": 4441 + }, + { + "epoch": 0.62, + "grad_norm": 2.0096556461264368, + "learning_rate": 3.3699478887053017e-06, + "loss": 0.3158, + "step": 4442 + }, + { + "epoch": 0.62, + "grad_norm": 2.1953430082721708, + "learning_rate": 3.367819664007168e-06, + "loss": 0.2841, + "step": 4443 + }, + { + "epoch": 0.62, + "grad_norm": 2.3515872254077945, + "learning_rate": 3.365691770234685e-06, + "loss": 0.3126, + "step": 4444 + }, + { + "epoch": 0.62, + "grad_norm": 1.9915533482902685, + "learning_rate": 3.3635642078192864e-06, + "loss": 0.316, + "step": 4445 + }, + { + "epoch": 0.62, + "grad_norm": 2.2625542368496556, + "learning_rate": 3.361436977192338e-06, + "loss": 0.3096, + "step": 4446 + }, + { + "epoch": 0.62, + "grad_norm": 2.6473157601813964, + "learning_rate": 3.3593100787851336e-06, + "loss": 0.3146, + "step": 4447 + }, + { + "epoch": 0.62, + "grad_norm": 2.1189429646698565, + "learning_rate": 3.357183513028906e-06, + "loss": 0.3158, + "step": 4448 + }, + { + "epoch": 0.62, + "grad_norm": 5.523608255042339, + "learning_rate": 3.3550572803548178e-06, + "loss": 0.3014, + "step": 4449 + }, + { + "epoch": 0.62, + "grad_norm": 1.9147274497014402, + "learning_rate": 3.352931381193964e-06, + "loss": 0.3162, + "step": 4450 + }, + { + "epoch": 0.62, + "grad_norm": 1.792447107377393, + "learning_rate": 3.350805815977371e-06, + "loss": 0.3049, + "step": 4451 + }, + { + "epoch": 0.62, + "grad_norm": 2.1832204653868703, + "learning_rate": 3.348680585135999e-06, + "loss": 0.3138, + "step": 4452 + }, + { + "epoch": 0.62, + "grad_norm": 2.5204722949656775, + "learning_rate": 3.346555689100741e-06, + "loss": 0.3109, + "step": 4453 + }, + { + "epoch": 0.62, + "grad_norm": 1.8496694210494569, + "learning_rate": 3.34443112830242e-06, + "loss": 0.303, + "step": 4454 + }, + { + "epoch": 0.62, + "grad_norm": 1.7537991963811215, + "learning_rate": 3.342306903171791e-06, + "loss": 0.3005, + "step": 4455 + }, + { + "epoch": 0.62, + "grad_norm": 2.287981485524274, + "learning_rate": 3.3401830141395442e-06, + "loss": 0.2972, + "step": 4456 + }, + { + "epoch": 0.62, + "grad_norm": 2.6732322207048407, + "learning_rate": 3.3380594616362997e-06, + "loss": 0.3058, + "step": 4457 + }, + { + "epoch": 0.62, + "grad_norm": 2.302032360859747, + "learning_rate": 3.3359362460926075e-06, + "loss": 0.305, + "step": 4458 + }, + { + "epoch": 0.62, + "grad_norm": 2.7490467927389495, + "learning_rate": 3.3338133679389517e-06, + "loss": 0.3142, + "step": 4459 + }, + { + "epoch": 0.62, + "grad_norm": 2.015606583947409, + "learning_rate": 3.331690827605748e-06, + "loss": 0.3268, + "step": 4460 + }, + { + "epoch": 0.62, + "grad_norm": 2.2296478746793564, + "learning_rate": 3.3295686255233417e-06, + "loss": 0.305, + "step": 4461 + }, + { + "epoch": 0.62, + "grad_norm": 2.2908642564655275, + "learning_rate": 3.327446762122012e-06, + "loss": 0.3247, + "step": 4462 + }, + { + "epoch": 0.62, + "grad_norm": 2.263834924884472, + "learning_rate": 3.3253252378319656e-06, + "loss": 0.3383, + "step": 4463 + }, + { + "epoch": 0.62, + "grad_norm": 2.7366030004211797, + "learning_rate": 3.3232040530833442e-06, + "loss": 0.3099, + "step": 4464 + }, + { + "epoch": 0.62, + "grad_norm": 2.718816260786483, + "learning_rate": 3.3210832083062205e-06, + "loss": 0.3262, + "step": 4465 + }, + { + "epoch": 0.62, + "grad_norm": 2.6232215543646342, + "learning_rate": 3.3189627039305972e-06, + "loss": 0.333, + "step": 4466 + }, + { + "epoch": 0.62, + "grad_norm": 2.202842086103821, + "learning_rate": 3.316842540386408e-06, + "loss": 0.2987, + "step": 4467 + }, + { + "epoch": 0.62, + "grad_norm": 2.4049841283697115, + "learning_rate": 3.3147227181035153e-06, + "loss": 0.3313, + "step": 4468 + }, + { + "epoch": 0.62, + "grad_norm": 1.8911860925407205, + "learning_rate": 3.3126032375117172e-06, + "loss": 0.2932, + "step": 4469 + }, + { + "epoch": 0.62, + "grad_norm": 2.1484651830871484, + "learning_rate": 3.3104840990407375e-06, + "loss": 0.3104, + "step": 4470 + }, + { + "epoch": 0.62, + "grad_norm": 2.423816698076463, + "learning_rate": 3.3083653031202346e-06, + "loss": 0.3265, + "step": 4471 + }, + { + "epoch": 0.62, + "grad_norm": 3.1562746146964633, + "learning_rate": 3.3062468501797956e-06, + "loss": 0.3171, + "step": 4472 + }, + { + "epoch": 0.62, + "grad_norm": 3.2185710726488574, + "learning_rate": 3.304128740648936e-06, + "loss": 0.3106, + "step": 4473 + }, + { + "epoch": 0.62, + "grad_norm": 2.3244007748816853, + "learning_rate": 3.3020109749571067e-06, + "loss": 0.3152, + "step": 4474 + }, + { + "epoch": 0.62, + "grad_norm": 2.2114840977016765, + "learning_rate": 3.299893553533685e-06, + "loss": 0.2987, + "step": 4475 + }, + { + "epoch": 0.62, + "grad_norm": 2.512115880264203, + "learning_rate": 3.297776476807981e-06, + "loss": 0.3017, + "step": 4476 + }, + { + "epoch": 0.62, + "grad_norm": 1.8093870863147465, + "learning_rate": 3.2956597452092307e-06, + "loss": 0.2978, + "step": 4477 + }, + { + "epoch": 0.62, + "grad_norm": 2.632347995526148, + "learning_rate": 3.2935433591666048e-06, + "loss": 0.3244, + "step": 4478 + }, + { + "epoch": 0.62, + "grad_norm": 2.197937339864178, + "learning_rate": 3.291427319109203e-06, + "loss": 0.3218, + "step": 4479 + }, + { + "epoch": 0.62, + "grad_norm": 4.710448096455717, + "learning_rate": 3.2893116254660495e-06, + "loss": 0.3233, + "step": 4480 + }, + { + "epoch": 0.62, + "grad_norm": 1.789596596463041, + "learning_rate": 3.2871962786661053e-06, + "loss": 0.2963, + "step": 4481 + }, + { + "epoch": 0.62, + "grad_norm": 2.356445329116527, + "learning_rate": 3.2850812791382596e-06, + "loss": 0.2996, + "step": 4482 + }, + { + "epoch": 0.62, + "grad_norm": 6.180403637124853, + "learning_rate": 3.28296662731133e-06, + "loss": 0.3132, + "step": 4483 + }, + { + "epoch": 0.62, + "grad_norm": 1.917782355045646, + "learning_rate": 3.280852323614061e-06, + "loss": 0.3206, + "step": 4484 + }, + { + "epoch": 0.62, + "grad_norm": 1.8180034198765023, + "learning_rate": 3.27873836847513e-06, + "loss": 0.3143, + "step": 4485 + }, + { + "epoch": 0.62, + "grad_norm": 9.340632072539782, + "learning_rate": 3.276624762323145e-06, + "loss": 0.3283, + "step": 4486 + }, + { + "epoch": 0.62, + "grad_norm": 2.051723117196589, + "learning_rate": 3.274511505586638e-06, + "loss": 0.3004, + "step": 4487 + }, + { + "epoch": 0.62, + "grad_norm": 2.6631912737627155, + "learning_rate": 3.2723985986940745e-06, + "loss": 0.3401, + "step": 4488 + }, + { + "epoch": 0.62, + "grad_norm": 2.444708231278984, + "learning_rate": 3.2702860420738484e-06, + "loss": 0.3317, + "step": 4489 + }, + { + "epoch": 0.62, + "grad_norm": 1.8548414115639253, + "learning_rate": 3.26817383615428e-06, + "loss": 0.2902, + "step": 4490 + }, + { + "epoch": 0.62, + "grad_norm": 2.0783018243083076, + "learning_rate": 3.2660619813636228e-06, + "loss": 0.3254, + "step": 4491 + }, + { + "epoch": 0.62, + "grad_norm": 2.1091812327040618, + "learning_rate": 3.2639504781300557e-06, + "loss": 0.336, + "step": 4492 + }, + { + "epoch": 0.62, + "grad_norm": 3.1342511983768118, + "learning_rate": 3.2618393268816884e-06, + "loss": 0.3082, + "step": 4493 + }, + { + "epoch": 0.62, + "grad_norm": 1.83361398164513, + "learning_rate": 3.2597285280465564e-06, + "loss": 0.2956, + "step": 4494 + }, + { + "epoch": 0.62, + "grad_norm": 2.674191395334611, + "learning_rate": 3.257618082052627e-06, + "loss": 0.3353, + "step": 4495 + }, + { + "epoch": 0.63, + "grad_norm": 2.052552606918373, + "learning_rate": 3.255507989327794e-06, + "loss": 0.2948, + "step": 4496 + }, + { + "epoch": 0.63, + "grad_norm": 3.0448698405342864, + "learning_rate": 3.2533982502998796e-06, + "loss": 0.3288, + "step": 4497 + }, + { + "epoch": 0.63, + "grad_norm": 2.6194736472043756, + "learning_rate": 3.2512888653966345e-06, + "loss": 0.302, + "step": 4498 + }, + { + "epoch": 0.63, + "grad_norm": 2.7959516540149836, + "learning_rate": 3.2491798350457393e-06, + "loss": 0.3297, + "step": 4499 + }, + { + "epoch": 0.63, + "grad_norm": 1.948279011434138, + "learning_rate": 3.2470711596748022e-06, + "loss": 0.3128, + "step": 4500 + }, + { + "epoch": 0.63, + "grad_norm": 2.1361086104117124, + "learning_rate": 3.244962839711356e-06, + "loss": 0.3305, + "step": 4501 + }, + { + "epoch": 0.63, + "grad_norm": 1.9997122493679742, + "learning_rate": 3.242854875582866e-06, + "loss": 0.3174, + "step": 4502 + }, + { + "epoch": 0.63, + "grad_norm": 1.9301739690994808, + "learning_rate": 3.2407472677167214e-06, + "loss": 0.3282, + "step": 4503 + }, + { + "epoch": 0.63, + "grad_norm": 2.30064860688609, + "learning_rate": 3.238640016540242e-06, + "loss": 0.3042, + "step": 4504 + }, + { + "epoch": 0.63, + "grad_norm": 3.093929237859183, + "learning_rate": 3.236533122480675e-06, + "loss": 0.2941, + "step": 4505 + }, + { + "epoch": 0.63, + "grad_norm": 2.2120486386790645, + "learning_rate": 3.234426585965194e-06, + "loss": 0.3335, + "step": 4506 + }, + { + "epoch": 0.63, + "grad_norm": 1.9233911994957262, + "learning_rate": 3.232320407420899e-06, + "loss": 0.3199, + "step": 4507 + }, + { + "epoch": 0.63, + "grad_norm": 3.0039833594357424, + "learning_rate": 3.2302145872748225e-06, + "loss": 0.2912, + "step": 4508 + }, + { + "epoch": 0.63, + "grad_norm": 4.354086843929722, + "learning_rate": 3.22810912595392e-06, + "loss": 0.3343, + "step": 4509 + }, + { + "epoch": 0.63, + "grad_norm": 2.3028291660284683, + "learning_rate": 3.226004023885073e-06, + "loss": 0.3426, + "step": 4510 + }, + { + "epoch": 0.63, + "grad_norm": 2.403037748270645, + "learning_rate": 3.223899281495094e-06, + "loss": 0.3107, + "step": 4511 + }, + { + "epoch": 0.63, + "grad_norm": 2.7504483992753537, + "learning_rate": 3.2217948992107224e-06, + "loss": 0.2941, + "step": 4512 + }, + { + "epoch": 0.63, + "grad_norm": 2.1074226362817092, + "learning_rate": 3.21969087745862e-06, + "loss": 0.3022, + "step": 4513 + }, + { + "epoch": 0.63, + "grad_norm": 39.2127481941327, + "learning_rate": 3.21758721666538e-06, + "loss": 0.3042, + "step": 4514 + }, + { + "epoch": 0.63, + "grad_norm": 2.159488995140221, + "learning_rate": 3.2154839172575225e-06, + "loss": 0.3329, + "step": 4515 + }, + { + "epoch": 0.63, + "grad_norm": 2.2986837839851493, + "learning_rate": 3.213380979661489e-06, + "loss": 0.3103, + "step": 4516 + }, + { + "epoch": 0.63, + "grad_norm": 2.489568383609296, + "learning_rate": 3.211278404303655e-06, + "loss": 0.3084, + "step": 4517 + }, + { + "epoch": 0.63, + "grad_norm": 3.59821839454926, + "learning_rate": 3.209176191610318e-06, + "loss": 0.3303, + "step": 4518 + }, + { + "epoch": 0.63, + "grad_norm": 2.2042282148338055, + "learning_rate": 3.207074342007704e-06, + "loss": 0.2957, + "step": 4519 + }, + { + "epoch": 0.63, + "grad_norm": 4.464244991561058, + "learning_rate": 3.2049728559219625e-06, + "loss": 0.3067, + "step": 4520 + }, + { + "epoch": 0.63, + "grad_norm": 2.2730102294350947, + "learning_rate": 3.202871733779172e-06, + "loss": 0.3276, + "step": 4521 + }, + { + "epoch": 0.63, + "grad_norm": 19.966403242386654, + "learning_rate": 3.200770976005338e-06, + "loss": 0.3091, + "step": 4522 + }, + { + "epoch": 0.63, + "grad_norm": 4.538853354137062, + "learning_rate": 3.1986705830263877e-06, + "loss": 0.3368, + "step": 4523 + }, + { + "epoch": 0.63, + "grad_norm": 2.0916686748528335, + "learning_rate": 3.196570555268178e-06, + "loss": 0.3, + "step": 4524 + }, + { + "epoch": 0.63, + "grad_norm": 1.8992819211923493, + "learning_rate": 3.1944708931564917e-06, + "loss": 0.3126, + "step": 4525 + }, + { + "epoch": 0.63, + "grad_norm": 3.0473293204298693, + "learning_rate": 3.1923715971170383e-06, + "loss": 0.3236, + "step": 4526 + }, + { + "epoch": 0.63, + "grad_norm": 2.096006984268341, + "learning_rate": 3.190272667575448e-06, + "loss": 0.3075, + "step": 4527 + }, + { + "epoch": 0.63, + "grad_norm": 2.381462068782995, + "learning_rate": 3.1881741049572822e-06, + "loss": 0.3258, + "step": 4528 + }, + { + "epoch": 0.63, + "grad_norm": 2.0807801577036584, + "learning_rate": 3.186075909688027e-06, + "loss": 0.3148, + "step": 4529 + }, + { + "epoch": 0.63, + "grad_norm": 2.3406535823173247, + "learning_rate": 3.18397808219309e-06, + "loss": 0.3051, + "step": 4530 + }, + { + "epoch": 0.63, + "grad_norm": 1.9906973636587206, + "learning_rate": 3.1818806228978094e-06, + "loss": 0.3049, + "step": 4531 + }, + { + "epoch": 0.63, + "grad_norm": 1.8054554904162308, + "learning_rate": 3.179783532227445e-06, + "loss": 0.3129, + "step": 4532 + }, + { + "epoch": 0.63, + "grad_norm": 2.9184270653309357, + "learning_rate": 3.1776868106071823e-06, + "loss": 0.3075, + "step": 4533 + }, + { + "epoch": 0.63, + "grad_norm": 2.4916666096566384, + "learning_rate": 3.1755904584621366e-06, + "loss": 0.3159, + "step": 4534 + }, + { + "epoch": 0.63, + "grad_norm": 2.4318808393860576, + "learning_rate": 3.173494476217342e-06, + "loss": 0.3099, + "step": 4535 + }, + { + "epoch": 0.63, + "grad_norm": 5.590373050564791, + "learning_rate": 3.171398864297762e-06, + "loss": 0.3005, + "step": 4536 + }, + { + "epoch": 0.63, + "grad_norm": 2.1845568800932496, + "learning_rate": 3.1693036231282814e-06, + "loss": 0.29, + "step": 4537 + }, + { + "epoch": 0.63, + "grad_norm": 2.453839601943302, + "learning_rate": 3.167208753133714e-06, + "loss": 0.3076, + "step": 4538 + }, + { + "epoch": 0.63, + "grad_norm": 2.251899525730206, + "learning_rate": 3.165114254738794e-06, + "loss": 0.3225, + "step": 4539 + }, + { + "epoch": 0.63, + "grad_norm": 1.9503914245408016, + "learning_rate": 3.1630201283681826e-06, + "loss": 0.3117, + "step": 4540 + }, + { + "epoch": 0.63, + "grad_norm": 2.9264386018965296, + "learning_rate": 3.160926374446467e-06, + "loss": 0.3278, + "step": 4541 + }, + { + "epoch": 0.63, + "grad_norm": 7.428439426212502, + "learning_rate": 3.158832993398153e-06, + "loss": 0.3367, + "step": 4542 + }, + { + "epoch": 0.63, + "grad_norm": 2.154009077634014, + "learning_rate": 3.156739985647681e-06, + "loss": 0.322, + "step": 4543 + }, + { + "epoch": 0.63, + "grad_norm": 1.898011112615613, + "learning_rate": 3.1546473516194066e-06, + "loss": 0.3074, + "step": 4544 + }, + { + "epoch": 0.63, + "grad_norm": 2.3630256471191555, + "learning_rate": 3.152555091737613e-06, + "loss": 0.3295, + "step": 4545 + }, + { + "epoch": 0.63, + "grad_norm": 3.2986918634778095, + "learning_rate": 3.150463206426506e-06, + "loss": 0.325, + "step": 4546 + }, + { + "epoch": 0.63, + "grad_norm": 2.6255363727485275, + "learning_rate": 3.148371696110219e-06, + "loss": 0.3153, + "step": 4547 + }, + { + "epoch": 0.63, + "grad_norm": 2.6168693139317525, + "learning_rate": 3.146280561212807e-06, + "loss": 0.3271, + "step": 4548 + }, + { + "epoch": 0.63, + "grad_norm": 2.4887651554419294, + "learning_rate": 3.144189802158246e-06, + "loss": 0.2947, + "step": 4549 + }, + { + "epoch": 0.63, + "grad_norm": 1.9185001441970475, + "learning_rate": 3.14209941937044e-06, + "loss": 0.3382, + "step": 4550 + }, + { + "epoch": 0.63, + "grad_norm": 2.335983224493277, + "learning_rate": 3.1400094132732183e-06, + "loss": 0.3078, + "step": 4551 + }, + { + "epoch": 0.63, + "grad_norm": 1.8436827008543066, + "learning_rate": 3.1379197842903297e-06, + "loss": 0.3294, + "step": 4552 + }, + { + "epoch": 0.63, + "grad_norm": 3.9637846564329617, + "learning_rate": 3.1358305328454464e-06, + "loss": 0.3174, + "step": 4553 + }, + { + "epoch": 0.63, + "grad_norm": 2.2610300585905105, + "learning_rate": 3.133741659362166e-06, + "loss": 0.3154, + "step": 4554 + }, + { + "epoch": 0.63, + "grad_norm": 2.1704754363195478, + "learning_rate": 3.1316531642640104e-06, + "loss": 0.2862, + "step": 4555 + }, + { + "epoch": 0.63, + "grad_norm": 5.932662359090514, + "learning_rate": 3.1295650479744212e-06, + "loss": 0.3032, + "step": 4556 + }, + { + "epoch": 0.63, + "grad_norm": 2.6420609192927267, + "learning_rate": 3.127477310916767e-06, + "loss": 0.3032, + "step": 4557 + }, + { + "epoch": 0.63, + "grad_norm": 2.656510232205265, + "learning_rate": 3.1253899535143384e-06, + "loss": 0.2929, + "step": 4558 + }, + { + "epoch": 0.63, + "grad_norm": 1.6969056154363475, + "learning_rate": 3.123302976190345e-06, + "loss": 0.3268, + "step": 4559 + }, + { + "epoch": 0.63, + "grad_norm": 1.9909268034152745, + "learning_rate": 3.1212163793679266e-06, + "loss": 0.3007, + "step": 4560 + }, + { + "epoch": 0.63, + "grad_norm": 2.713888147831308, + "learning_rate": 3.119130163470141e-06, + "loss": 0.3267, + "step": 4561 + }, + { + "epoch": 0.63, + "grad_norm": 2.063606714039725, + "learning_rate": 3.117044328919971e-06, + "loss": 0.293, + "step": 4562 + }, + { + "epoch": 0.63, + "grad_norm": 2.8279976627302936, + "learning_rate": 3.1149588761403184e-06, + "loss": 0.2995, + "step": 4563 + }, + { + "epoch": 0.63, + "grad_norm": 2.290073669025982, + "learning_rate": 3.112873805554012e-06, + "loss": 0.3178, + "step": 4564 + }, + { + "epoch": 0.63, + "grad_norm": 2.2176808702342066, + "learning_rate": 3.1107891175838003e-06, + "loss": 0.3021, + "step": 4565 + }, + { + "epoch": 0.63, + "grad_norm": 2.6020253349250284, + "learning_rate": 3.108704812652355e-06, + "loss": 0.2948, + "step": 4566 + }, + { + "epoch": 0.63, + "grad_norm": 1.894588231328132, + "learning_rate": 3.1066208911822706e-06, + "loss": 0.2973, + "step": 4567 + }, + { + "epoch": 0.64, + "grad_norm": 3.7130902766050142, + "learning_rate": 3.104537353596064e-06, + "loss": 0.3077, + "step": 4568 + }, + { + "epoch": 0.64, + "grad_norm": 2.461478944218948, + "learning_rate": 3.102454200316175e-06, + "loss": 0.3259, + "step": 4569 + }, + { + "epoch": 0.64, + "grad_norm": 2.4206287100017074, + "learning_rate": 3.100371431764961e-06, + "loss": 0.3094, + "step": 4570 + }, + { + "epoch": 0.64, + "grad_norm": 2.4787214553838357, + "learning_rate": 3.0982890483647092e-06, + "loss": 0.3209, + "step": 4571 + }, + { + "epoch": 0.64, + "grad_norm": 2.980553755476087, + "learning_rate": 3.0962070505376208e-06, + "loss": 0.3234, + "step": 4572 + }, + { + "epoch": 0.64, + "grad_norm": 2.0912631211083474, + "learning_rate": 3.094125438705822e-06, + "loss": 0.2689, + "step": 4573 + }, + { + "epoch": 0.64, + "grad_norm": 2.3010830268099594, + "learning_rate": 3.092044213291364e-06, + "loss": 0.3244, + "step": 4574 + }, + { + "epoch": 0.64, + "grad_norm": 2.479278830893024, + "learning_rate": 3.0899633747162133e-06, + "loss": 0.3203, + "step": 4575 + }, + { + "epoch": 0.64, + "grad_norm": 2.0267509160740684, + "learning_rate": 3.0878829234022624e-06, + "loss": 0.3162, + "step": 4576 + }, + { + "epoch": 0.64, + "grad_norm": 2.3840336155810884, + "learning_rate": 3.0858028597713248e-06, + "loss": 0.3156, + "step": 4577 + }, + { + "epoch": 0.64, + "grad_norm": 1.8393506191227251, + "learning_rate": 3.083723184245136e-06, + "loss": 0.3205, + "step": 4578 + }, + { + "epoch": 0.64, + "grad_norm": 2.3941446331288074, + "learning_rate": 3.0816438972453488e-06, + "loss": 0.3189, + "step": 4579 + }, + { + "epoch": 0.64, + "grad_norm": 2.609243207742619, + "learning_rate": 3.079564999193542e-06, + "loss": 0.3076, + "step": 4580 + }, + { + "epoch": 0.64, + "grad_norm": 2.2761977170335626, + "learning_rate": 3.0774864905112133e-06, + "loss": 0.2999, + "step": 4581 + }, + { + "epoch": 0.64, + "grad_norm": 2.286034836073993, + "learning_rate": 3.0754083716197814e-06, + "loss": 0.3209, + "step": 4582 + }, + { + "epoch": 0.64, + "grad_norm": 2.1199232753306894, + "learning_rate": 3.073330642940586e-06, + "loss": 0.3157, + "step": 4583 + }, + { + "epoch": 0.64, + "grad_norm": 3.982045590453636, + "learning_rate": 3.0712533048948888e-06, + "loss": 0.2957, + "step": 4584 + }, + { + "epoch": 0.64, + "grad_norm": 2.08482114387501, + "learning_rate": 3.069176357903869e-06, + "loss": 0.3025, + "step": 4585 + }, + { + "epoch": 0.64, + "grad_norm": 2.278038802451649, + "learning_rate": 3.0670998023886324e-06, + "loss": 0.3044, + "step": 4586 + }, + { + "epoch": 0.64, + "grad_norm": 1.9538482267563144, + "learning_rate": 3.0650236387701997e-06, + "loss": 0.2956, + "step": 4587 + }, + { + "epoch": 0.64, + "grad_norm": 3.337672647539919, + "learning_rate": 3.0629478674695167e-06, + "loss": 0.3231, + "step": 4588 + }, + { + "epoch": 0.64, + "grad_norm": 2.034855809161207, + "learning_rate": 3.0608724889074454e-06, + "loss": 0.3142, + "step": 4589 + }, + { + "epoch": 0.64, + "grad_norm": 6.647200441789763, + "learning_rate": 3.05879750350477e-06, + "loss": 0.3316, + "step": 4590 + }, + { + "epoch": 0.64, + "grad_norm": 2.5907593328944865, + "learning_rate": 3.0567229116821983e-06, + "loss": 0.2989, + "step": 4591 + }, + { + "epoch": 0.64, + "grad_norm": 1.9815180860786288, + "learning_rate": 3.0546487138603508e-06, + "loss": 0.3331, + "step": 4592 + }, + { + "epoch": 0.64, + "grad_norm": 1.7534426093061604, + "learning_rate": 3.0525749104597733e-06, + "loss": 0.332, + "step": 4593 + }, + { + "epoch": 0.64, + "grad_norm": 2.6325139716820463, + "learning_rate": 3.0505015019009344e-06, + "loss": 0.322, + "step": 4594 + }, + { + "epoch": 0.64, + "grad_norm": 2.0806685505406777, + "learning_rate": 3.048428488604217e-06, + "loss": 0.306, + "step": 4595 + }, + { + "epoch": 0.64, + "grad_norm": 2.151569780299858, + "learning_rate": 3.0463558709899244e-06, + "loss": 0.3414, + "step": 4596 + }, + { + "epoch": 0.64, + "grad_norm": 3.4257268203228537, + "learning_rate": 3.0442836494782823e-06, + "loss": 0.3476, + "step": 4597 + }, + { + "epoch": 0.64, + "grad_norm": 3.7171127453642914, + "learning_rate": 3.042211824489437e-06, + "loss": 0.2991, + "step": 4598 + }, + { + "epoch": 0.64, + "grad_norm": 2.0940847503685, + "learning_rate": 3.0401403964434496e-06, + "loss": 0.3413, + "step": 4599 + }, + { + "epoch": 0.64, + "grad_norm": 2.407474693584764, + "learning_rate": 3.0380693657603043e-06, + "loss": 0.3105, + "step": 4600 + }, + { + "epoch": 0.64, + "grad_norm": 2.0080067040043668, + "learning_rate": 3.035998732859904e-06, + "loss": 0.303, + "step": 4601 + }, + { + "epoch": 0.64, + "grad_norm": 2.779111722170203, + "learning_rate": 3.0339284981620697e-06, + "loss": 0.3412, + "step": 4602 + }, + { + "epoch": 0.64, + "grad_norm": 1.8032638652970514, + "learning_rate": 3.031858662086546e-06, + "loss": 0.3216, + "step": 4603 + }, + { + "epoch": 0.64, + "grad_norm": 1.9077274881682165, + "learning_rate": 3.0297892250529916e-06, + "loss": 0.316, + "step": 4604 + }, + { + "epoch": 0.64, + "grad_norm": 2.472091368054302, + "learning_rate": 3.027720187480987e-06, + "loss": 0.3023, + "step": 4605 + }, + { + "epoch": 0.64, + "grad_norm": 2.6004272014209686, + "learning_rate": 3.025651549790029e-06, + "loss": 0.3059, + "step": 4606 + }, + { + "epoch": 0.64, + "grad_norm": 1.7956469027662483, + "learning_rate": 3.023583312399538e-06, + "loss": 0.286, + "step": 4607 + }, + { + "epoch": 0.64, + "grad_norm": 3.7939213997107863, + "learning_rate": 3.021515475728848e-06, + "loss": 0.3482, + "step": 4608 + }, + { + "epoch": 0.64, + "grad_norm": 3.813697035293974, + "learning_rate": 3.019448040197216e-06, + "loss": 0.2936, + "step": 4609 + }, + { + "epoch": 0.64, + "grad_norm": 2.5585201885515994, + "learning_rate": 3.017381006223815e-06, + "loss": 0.3144, + "step": 4610 + }, + { + "epoch": 0.64, + "grad_norm": 1.825486451999484, + "learning_rate": 3.015314374227736e-06, + "loss": 0.2963, + "step": 4611 + }, + { + "epoch": 0.64, + "grad_norm": 2.2686333121115565, + "learning_rate": 3.013248144627995e-06, + "loss": 0.3033, + "step": 4612 + }, + { + "epoch": 0.64, + "grad_norm": 2.338282664212865, + "learning_rate": 3.0111823178435163e-06, + "loss": 0.3114, + "step": 4613 + }, + { + "epoch": 0.64, + "grad_norm": 2.827760655319645, + "learning_rate": 3.0091168942931515e-06, + "loss": 0.3159, + "step": 4614 + }, + { + "epoch": 0.64, + "grad_norm": 1.8733810527767238, + "learning_rate": 3.007051874395663e-06, + "loss": 0.323, + "step": 4615 + }, + { + "epoch": 0.64, + "grad_norm": 1.7909283921665031, + "learning_rate": 3.004987258569737e-06, + "loss": 0.3063, + "step": 4616 + }, + { + "epoch": 0.64, + "grad_norm": 2.2301820366106666, + "learning_rate": 3.0029230472339767e-06, + "loss": 0.3433, + "step": 4617 + }, + { + "epoch": 0.64, + "grad_norm": 3.6378660162948155, + "learning_rate": 3.0008592408069004e-06, + "loss": 0.3307, + "step": 4618 + }, + { + "epoch": 0.64, + "grad_norm": 2.3593778691356717, + "learning_rate": 2.998795839706945e-06, + "loss": 0.2993, + "step": 4619 + }, + { + "epoch": 0.64, + "grad_norm": 3.0629339264562487, + "learning_rate": 2.99673284435247e-06, + "loss": 0.3227, + "step": 4620 + }, + { + "epoch": 0.64, + "grad_norm": 1.9178776889854101, + "learning_rate": 2.994670255161748e-06, + "loss": 0.3126, + "step": 4621 + }, + { + "epoch": 0.64, + "grad_norm": 2.3857996117919664, + "learning_rate": 2.9926080725529694e-06, + "loss": 0.3027, + "step": 4622 + }, + { + "epoch": 0.64, + "grad_norm": 2.4518189967759945, + "learning_rate": 2.9905462969442433e-06, + "loss": 0.3284, + "step": 4623 + }, + { + "epoch": 0.64, + "grad_norm": 2.338750809448397, + "learning_rate": 2.9884849287535966e-06, + "loss": 0.32, + "step": 4624 + }, + { + "epoch": 0.64, + "grad_norm": 2.5069954456700003, + "learning_rate": 2.9864239683989724e-06, + "loss": 0.3207, + "step": 4625 + }, + { + "epoch": 0.64, + "grad_norm": 1.938172386011638, + "learning_rate": 2.9843634162982314e-06, + "loss": 0.3035, + "step": 4626 + }, + { + "epoch": 0.64, + "grad_norm": 2.3554829487811397, + "learning_rate": 2.9823032728691547e-06, + "loss": 0.3008, + "step": 4627 + }, + { + "epoch": 0.64, + "grad_norm": 5.154562650582727, + "learning_rate": 2.9802435385294325e-06, + "loss": 0.3036, + "step": 4628 + }, + { + "epoch": 0.64, + "grad_norm": 2.0562175601319566, + "learning_rate": 2.978184213696683e-06, + "loss": 0.31, + "step": 4629 + }, + { + "epoch": 0.64, + "grad_norm": 3.3947266897164816, + "learning_rate": 2.976125298788432e-06, + "loss": 0.3023, + "step": 4630 + }, + { + "epoch": 0.64, + "grad_norm": 1.958428046788897, + "learning_rate": 2.974066794222128e-06, + "loss": 0.3118, + "step": 4631 + }, + { + "epoch": 0.64, + "grad_norm": 2.2379624416555424, + "learning_rate": 2.9720087004151333e-06, + "loss": 0.3121, + "step": 4632 + }, + { + "epoch": 0.64, + "grad_norm": 2.2473274345204617, + "learning_rate": 2.969951017784727e-06, + "loss": 0.3, + "step": 4633 + }, + { + "epoch": 0.64, + "grad_norm": 2.484152461231096, + "learning_rate": 2.967893746748107e-06, + "loss": 0.3328, + "step": 4634 + }, + { + "epoch": 0.64, + "grad_norm": 2.5194717333955494, + "learning_rate": 2.9658368877223854e-06, + "loss": 0.3093, + "step": 4635 + }, + { + "epoch": 0.64, + "grad_norm": 2.209947892678851, + "learning_rate": 2.96378044112459e-06, + "loss": 0.2995, + "step": 4636 + }, + { + "epoch": 0.64, + "grad_norm": 1.8746983512916227, + "learning_rate": 2.961724407371671e-06, + "loss": 0.3222, + "step": 4637 + }, + { + "epoch": 0.64, + "grad_norm": 2.275390695927457, + "learning_rate": 2.9596687868804885e-06, + "loss": 0.3147, + "step": 4638 + }, + { + "epoch": 0.64, + "grad_norm": 2.0849404158082394, + "learning_rate": 2.9576135800678206e-06, + "loss": 0.3184, + "step": 4639 + }, + { + "epoch": 0.65, + "grad_norm": 2.286767813737265, + "learning_rate": 2.9555587873503623e-06, + "loss": 0.3167, + "step": 4640 + }, + { + "epoch": 0.65, + "grad_norm": 2.377476198997735, + "learning_rate": 2.9535044091447236e-06, + "loss": 0.3051, + "step": 4641 + }, + { + "epoch": 0.65, + "grad_norm": 1.997256817092078, + "learning_rate": 2.951450445867431e-06, + "loss": 0.3332, + "step": 4642 + }, + { + "epoch": 0.65, + "grad_norm": 1.9906715575193528, + "learning_rate": 2.9493968979349284e-06, + "loss": 0.326, + "step": 4643 + }, + { + "epoch": 0.65, + "grad_norm": 2.517184889640441, + "learning_rate": 2.9473437657635726e-06, + "loss": 0.2997, + "step": 4644 + }, + { + "epoch": 0.65, + "grad_norm": 2.7456303470153456, + "learning_rate": 2.945291049769636e-06, + "loss": 0.3049, + "step": 4645 + }, + { + "epoch": 0.65, + "grad_norm": 1.7497197842993173, + "learning_rate": 2.9432387503693117e-06, + "loss": 0.3112, + "step": 4646 + }, + { + "epoch": 0.65, + "grad_norm": 1.6665126448739016, + "learning_rate": 2.9411868679787036e-06, + "loss": 0.3098, + "step": 4647 + }, + { + "epoch": 0.65, + "grad_norm": 2.4620217214149984, + "learning_rate": 2.939135403013831e-06, + "loss": 0.3173, + "step": 4648 + }, + { + "epoch": 0.65, + "grad_norm": 1.816052800432161, + "learning_rate": 2.9370843558906305e-06, + "loss": 0.309, + "step": 4649 + }, + { + "epoch": 0.65, + "grad_norm": 1.8879206485270155, + "learning_rate": 2.935033727024954e-06, + "loss": 0.298, + "step": 4650 + }, + { + "epoch": 0.65, + "grad_norm": 2.1709398993369615, + "learning_rate": 2.932983516832566e-06, + "loss": 0.3109, + "step": 4651 + }, + { + "epoch": 0.65, + "grad_norm": 2.684086826711619, + "learning_rate": 2.9309337257291497e-06, + "loss": 0.3328, + "step": 4652 + }, + { + "epoch": 0.65, + "grad_norm": 2.135674060320824, + "learning_rate": 2.928884354130302e-06, + "loss": 0.3125, + "step": 4653 + }, + { + "epoch": 0.65, + "grad_norm": 2.3792841721723126, + "learning_rate": 2.926835402451531e-06, + "loss": 0.3218, + "step": 4654 + }, + { + "epoch": 0.65, + "grad_norm": 1.9139763904352614, + "learning_rate": 2.9247868711082668e-06, + "loss": 0.3055, + "step": 4655 + }, + { + "epoch": 0.65, + "grad_norm": 2.1406844426532348, + "learning_rate": 2.9227387605158486e-06, + "loss": 0.327, + "step": 4656 + }, + { + "epoch": 0.65, + "grad_norm": 1.9308747919115734, + "learning_rate": 2.920691071089534e-06, + "loss": 0.2853, + "step": 4657 + }, + { + "epoch": 0.65, + "grad_norm": 3.3069081439036587, + "learning_rate": 2.9186438032444915e-06, + "loss": 0.3079, + "step": 4658 + }, + { + "epoch": 0.65, + "grad_norm": 2.055297630226992, + "learning_rate": 2.916596957395807e-06, + "loss": 0.3126, + "step": 4659 + }, + { + "epoch": 0.65, + "grad_norm": 2.5306736691619203, + "learning_rate": 2.914550533958478e-06, + "loss": 0.3261, + "step": 4660 + }, + { + "epoch": 0.65, + "grad_norm": 1.9569341639880435, + "learning_rate": 2.912504533347421e-06, + "loss": 0.3231, + "step": 4661 + }, + { + "epoch": 0.65, + "grad_norm": 1.944935555639051, + "learning_rate": 2.9104589559774614e-06, + "loss": 0.3057, + "step": 4662 + }, + { + "epoch": 0.65, + "grad_norm": 2.1361453985381615, + "learning_rate": 2.9084138022633445e-06, + "loss": 0.3187, + "step": 4663 + }, + { + "epoch": 0.65, + "grad_norm": 3.581755705847279, + "learning_rate": 2.9063690726197226e-06, + "loss": 0.3158, + "step": 4664 + }, + { + "epoch": 0.65, + "grad_norm": 2.4011199668170833, + "learning_rate": 2.90432476746117e-06, + "loss": 0.3203, + "step": 4665 + }, + { + "epoch": 0.65, + "grad_norm": 1.9390924579676256, + "learning_rate": 2.9022808872021698e-06, + "loss": 0.3209, + "step": 4666 + }, + { + "epoch": 0.65, + "grad_norm": 2.4317784570032392, + "learning_rate": 2.900237432257117e-06, + "loss": 0.3086, + "step": 4667 + }, + { + "epoch": 0.65, + "grad_norm": 2.830242072721304, + "learning_rate": 2.8981944030403274e-06, + "loss": 0.3113, + "step": 4668 + }, + { + "epoch": 0.65, + "grad_norm": 2.212082146281652, + "learning_rate": 2.896151799966025e-06, + "loss": 0.3071, + "step": 4669 + }, + { + "epoch": 0.65, + "grad_norm": 2.460604975769443, + "learning_rate": 2.8941096234483475e-06, + "loss": 0.2922, + "step": 4670 + }, + { + "epoch": 0.65, + "grad_norm": 2.713889510179953, + "learning_rate": 2.8920678739013486e-06, + "loss": 0.3196, + "step": 4671 + }, + { + "epoch": 0.65, + "grad_norm": 2.4629880563252677, + "learning_rate": 2.890026551738997e-06, + "loss": 0.3203, + "step": 4672 + }, + { + "epoch": 0.65, + "grad_norm": 2.208505015389508, + "learning_rate": 2.8879856573751696e-06, + "loss": 0.3288, + "step": 4673 + }, + { + "epoch": 0.65, + "grad_norm": 2.0535483830993813, + "learning_rate": 2.885945191223657e-06, + "loss": 0.3029, + "step": 4674 + }, + { + "epoch": 0.65, + "grad_norm": 3.04839157449731, + "learning_rate": 2.88390515369817e-06, + "loss": 0.2921, + "step": 4675 + }, + { + "epoch": 0.65, + "grad_norm": 2.751646478802645, + "learning_rate": 2.8818655452123246e-06, + "loss": 0.2917, + "step": 4676 + }, + { + "epoch": 0.65, + "grad_norm": 2.152919564312792, + "learning_rate": 2.8798263661796515e-06, + "loss": 0.3103, + "step": 4677 + }, + { + "epoch": 0.65, + "grad_norm": 2.1153898969309277, + "learning_rate": 2.877787617013599e-06, + "loss": 0.305, + "step": 4678 + }, + { + "epoch": 0.65, + "grad_norm": 2.6049338727241915, + "learning_rate": 2.8757492981275235e-06, + "loss": 0.2933, + "step": 4679 + }, + { + "epoch": 0.65, + "grad_norm": 2.261398038070693, + "learning_rate": 2.873711409934693e-06, + "loss": 0.313, + "step": 4680 + }, + { + "epoch": 0.65, + "grad_norm": 3.0017218569886537, + "learning_rate": 2.871673952848293e-06, + "loss": 0.2943, + "step": 4681 + }, + { + "epoch": 0.65, + "grad_norm": 2.177781731336542, + "learning_rate": 2.8696369272814206e-06, + "loss": 0.3219, + "step": 4682 + }, + { + "epoch": 0.65, + "grad_norm": 2.7417509369886304, + "learning_rate": 2.8676003336470837e-06, + "loss": 0.3379, + "step": 4683 + }, + { + "epoch": 0.65, + "grad_norm": 2.2385955597657943, + "learning_rate": 2.865564172358199e-06, + "loss": 0.315, + "step": 4684 + }, + { + "epoch": 0.65, + "grad_norm": 1.9059762077032834, + "learning_rate": 2.8635284438276047e-06, + "loss": 0.3141, + "step": 4685 + }, + { + "epoch": 0.65, + "grad_norm": 2.2084790421306524, + "learning_rate": 2.8614931484680443e-06, + "loss": 0.2937, + "step": 4686 + }, + { + "epoch": 0.65, + "grad_norm": 1.912043815088991, + "learning_rate": 2.8594582866921746e-06, + "loss": 0.3057, + "step": 4687 + }, + { + "epoch": 0.65, + "grad_norm": 2.078232931890187, + "learning_rate": 2.8574238589125614e-06, + "loss": 0.2951, + "step": 4688 + }, + { + "epoch": 0.65, + "grad_norm": 2.834395313978144, + "learning_rate": 2.8553898655416947e-06, + "loss": 0.3179, + "step": 4689 + }, + { + "epoch": 0.65, + "grad_norm": 2.096401715516184, + "learning_rate": 2.8533563069919634e-06, + "loss": 0.3055, + "step": 4690 + }, + { + "epoch": 0.65, + "grad_norm": 2.3390189678249556, + "learning_rate": 2.8513231836756706e-06, + "loss": 0.3446, + "step": 4691 + }, + { + "epoch": 0.65, + "grad_norm": 1.89480451478971, + "learning_rate": 2.849290496005039e-06, + "loss": 0.3105, + "step": 4692 + }, + { + "epoch": 0.65, + "grad_norm": 2.315276132347376, + "learning_rate": 2.8472582443921936e-06, + "loss": 0.3112, + "step": 4693 + }, + { + "epoch": 0.65, + "grad_norm": 3.484412940365291, + "learning_rate": 2.845226429249176e-06, + "loss": 0.3281, + "step": 4694 + }, + { + "epoch": 0.65, + "grad_norm": 2.5790396892480487, + "learning_rate": 2.843195050987936e-06, + "loss": 0.2987, + "step": 4695 + }, + { + "epoch": 0.65, + "grad_norm": 3.2505279131698535, + "learning_rate": 2.84116411002034e-06, + "loss": 0.3121, + "step": 4696 + }, + { + "epoch": 0.65, + "grad_norm": 2.0064898370334925, + "learning_rate": 2.8391336067581586e-06, + "loss": 0.3241, + "step": 4697 + }, + { + "epoch": 0.65, + "grad_norm": 2.1165748272467853, + "learning_rate": 2.837103541613081e-06, + "loss": 0.3219, + "step": 4698 + }, + { + "epoch": 0.65, + "grad_norm": 1.9132480467880635, + "learning_rate": 2.8350739149967047e-06, + "loss": 0.3032, + "step": 4699 + }, + { + "epoch": 0.65, + "grad_norm": 2.019352576332769, + "learning_rate": 2.8330447273205376e-06, + "loss": 0.303, + "step": 4700 + }, + { + "epoch": 0.65, + "grad_norm": 4.666014067411772, + "learning_rate": 2.8310159789959966e-06, + "loss": 0.3316, + "step": 4701 + }, + { + "epoch": 0.65, + "grad_norm": 2.2915376854633163, + "learning_rate": 2.828987670434412e-06, + "loss": 0.322, + "step": 4702 + }, + { + "epoch": 0.65, + "grad_norm": 2.3440324633442433, + "learning_rate": 2.826959802047028e-06, + "loss": 0.2809, + "step": 4703 + }, + { + "epoch": 0.65, + "grad_norm": 2.379872976103527, + "learning_rate": 2.8249323742449934e-06, + "loss": 0.3112, + "step": 4704 + }, + { + "epoch": 0.65, + "grad_norm": 2.083698593445902, + "learning_rate": 2.82290538743937e-06, + "loss": 0.3096, + "step": 4705 + }, + { + "epoch": 0.65, + "grad_norm": 2.3861733536768033, + "learning_rate": 2.8208788420411307e-06, + "loss": 0.3262, + "step": 4706 + }, + { + "epoch": 0.65, + "grad_norm": 2.0793271468681955, + "learning_rate": 2.818852738461163e-06, + "loss": 0.3036, + "step": 4707 + }, + { + "epoch": 0.65, + "grad_norm": 3.333571019268716, + "learning_rate": 2.816827077110258e-06, + "loss": 0.3015, + "step": 4708 + }, + { + "epoch": 0.65, + "grad_norm": 2.665038376935829, + "learning_rate": 2.814801858399117e-06, + "loss": 0.3079, + "step": 4709 + }, + { + "epoch": 0.65, + "grad_norm": 2.437539316464912, + "learning_rate": 2.81277708273836e-06, + "loss": 0.3158, + "step": 4710 + }, + { + "epoch": 0.65, + "grad_norm": 2.0831289228274747, + "learning_rate": 2.8107527505385086e-06, + "loss": 0.3031, + "step": 4711 + }, + { + "epoch": 0.66, + "grad_norm": 1.9768630610058946, + "learning_rate": 2.8087288622099958e-06, + "loss": 0.3163, + "step": 4712 + }, + { + "epoch": 0.66, + "grad_norm": 2.4238848124530192, + "learning_rate": 2.8067054181631704e-06, + "loss": 0.3503, + "step": 4713 + }, + { + "epoch": 0.66, + "grad_norm": 2.677446739098037, + "learning_rate": 2.804682418808282e-06, + "loss": 0.3101, + "step": 4714 + }, + { + "epoch": 0.66, + "grad_norm": 1.8579454455746314, + "learning_rate": 2.8026598645555007e-06, + "loss": 0.3241, + "step": 4715 + }, + { + "epoch": 0.66, + "grad_norm": 2.2974642499780273, + "learning_rate": 2.8006377558148963e-06, + "loss": 0.2816, + "step": 4716 + }, + { + "epoch": 0.66, + "grad_norm": 2.292989463115239, + "learning_rate": 2.7986160929964556e-06, + "loss": 0.2871, + "step": 4717 + }, + { + "epoch": 0.66, + "grad_norm": 2.00204540813448, + "learning_rate": 2.7965948765100715e-06, + "loss": 0.3258, + "step": 4718 + }, + { + "epoch": 0.66, + "grad_norm": 1.9596359980109377, + "learning_rate": 2.794574106765544e-06, + "loss": 0.299, + "step": 4719 + }, + { + "epoch": 0.66, + "grad_norm": 1.9719629487263368, + "learning_rate": 2.7925537841725904e-06, + "loss": 0.3056, + "step": 4720 + }, + { + "epoch": 0.66, + "grad_norm": 1.754939942090652, + "learning_rate": 2.790533909140829e-06, + "loss": 0.3214, + "step": 4721 + }, + { + "epoch": 0.66, + "grad_norm": 3.022167979983147, + "learning_rate": 2.7885144820797904e-06, + "loss": 0.3162, + "step": 4722 + }, + { + "epoch": 0.66, + "grad_norm": 2.5061116874911606, + "learning_rate": 2.786495503398916e-06, + "loss": 0.3263, + "step": 4723 + }, + { + "epoch": 0.66, + "grad_norm": 1.9938057711892867, + "learning_rate": 2.7844769735075574e-06, + "loss": 0.3191, + "step": 4724 + }, + { + "epoch": 0.66, + "grad_norm": 5.17302931204813, + "learning_rate": 2.7824588928149698e-06, + "loss": 0.3168, + "step": 4725 + }, + { + "epoch": 0.66, + "grad_norm": 1.8970900253315373, + "learning_rate": 2.78044126173032e-06, + "loss": 0.2963, + "step": 4726 + }, + { + "epoch": 0.66, + "grad_norm": 10.50318094881357, + "learning_rate": 2.7784240806626873e-06, + "loss": 0.3179, + "step": 4727 + }, + { + "epoch": 0.66, + "grad_norm": 1.9699176929973914, + "learning_rate": 2.7764073500210543e-06, + "loss": 0.2926, + "step": 4728 + }, + { + "epoch": 0.66, + "grad_norm": 2.3842925179679697, + "learning_rate": 2.774391070214313e-06, + "loss": 0.2841, + "step": 4729 + }, + { + "epoch": 0.66, + "grad_norm": 3.3279046019058507, + "learning_rate": 2.7723752416512694e-06, + "loss": 0.3057, + "step": 4730 + }, + { + "epoch": 0.66, + "grad_norm": 2.5599795859705843, + "learning_rate": 2.7703598647406294e-06, + "loss": 0.3023, + "step": 4731 + }, + { + "epoch": 0.66, + "grad_norm": 3.776220239346771, + "learning_rate": 2.7683449398910167e-06, + "loss": 0.2986, + "step": 4732 + }, + { + "epoch": 0.66, + "grad_norm": 1.9600717106336103, + "learning_rate": 2.7663304675109546e-06, + "loss": 0.2933, + "step": 4733 + }, + { + "epoch": 0.66, + "grad_norm": 1.7180131282240816, + "learning_rate": 2.7643164480088823e-06, + "loss": 0.3194, + "step": 4734 + }, + { + "epoch": 0.66, + "grad_norm": 1.6720511090698176, + "learning_rate": 2.7623028817931417e-06, + "loss": 0.3073, + "step": 4735 + }, + { + "epoch": 0.66, + "grad_norm": 2.2751160275983557, + "learning_rate": 2.7602897692719832e-06, + "loss": 0.3061, + "step": 4736 + }, + { + "epoch": 0.66, + "grad_norm": 2.3804498623028545, + "learning_rate": 2.7582771108535705e-06, + "loss": 0.316, + "step": 4737 + }, + { + "epoch": 0.66, + "grad_norm": 1.930343045714823, + "learning_rate": 2.756264906945969e-06, + "loss": 0.2974, + "step": 4738 + }, + { + "epoch": 0.66, + "grad_norm": 2.122300322764968, + "learning_rate": 2.754253157957153e-06, + "loss": 0.2985, + "step": 4739 + }, + { + "epoch": 0.66, + "grad_norm": 1.8825424747543595, + "learning_rate": 2.752241864295008e-06, + "loss": 0.3252, + "step": 4740 + }, + { + "epoch": 0.66, + "grad_norm": 2.2932417132742633, + "learning_rate": 2.750231026367326e-06, + "loss": 0.3248, + "step": 4741 + }, + { + "epoch": 0.66, + "grad_norm": 2.5878289329576534, + "learning_rate": 2.748220644581805e-06, + "loss": 0.2939, + "step": 4742 + }, + { + "epoch": 0.66, + "grad_norm": 2.0571857510660694, + "learning_rate": 2.746210719346048e-06, + "loss": 0.3057, + "step": 4743 + }, + { + "epoch": 0.66, + "grad_norm": 2.068749069757806, + "learning_rate": 2.744201251067574e-06, + "loss": 0.3055, + "step": 4744 + }, + { + "epoch": 0.66, + "grad_norm": 2.5893536541882205, + "learning_rate": 2.7421922401538013e-06, + "loss": 0.2891, + "step": 4745 + }, + { + "epoch": 0.66, + "grad_norm": 2.001337680235996, + "learning_rate": 2.740183687012057e-06, + "loss": 0.3021, + "step": 4746 + }, + { + "epoch": 0.66, + "grad_norm": 2.2960251161839516, + "learning_rate": 2.738175592049579e-06, + "loss": 0.3059, + "step": 4747 + }, + { + "epoch": 0.66, + "grad_norm": 3.0683563915327268, + "learning_rate": 2.7361679556735094e-06, + "loss": 0.3031, + "step": 4748 + }, + { + "epoch": 0.66, + "grad_norm": 1.8225363997853539, + "learning_rate": 2.734160778290895e-06, + "loss": 0.3017, + "step": 4749 + }, + { + "epoch": 0.66, + "grad_norm": 2.487076953666619, + "learning_rate": 2.732154060308695e-06, + "loss": 0.3001, + "step": 4750 + }, + { + "epoch": 0.66, + "grad_norm": 1.6494284224055824, + "learning_rate": 2.7301478021337748e-06, + "loss": 0.2755, + "step": 4751 + }, + { + "epoch": 0.66, + "grad_norm": 2.6529343454207424, + "learning_rate": 2.728142004172901e-06, + "loss": 0.3222, + "step": 4752 + }, + { + "epoch": 0.66, + "grad_norm": 2.1113625491293093, + "learning_rate": 2.7261366668327505e-06, + "loss": 0.3128, + "step": 4753 + }, + { + "epoch": 0.66, + "grad_norm": 2.504963339107979, + "learning_rate": 2.72413179051991e-06, + "loss": 0.27, + "step": 4754 + }, + { + "epoch": 0.66, + "grad_norm": 2.4605865912946454, + "learning_rate": 2.7221273756408673e-06, + "loss": 0.3027, + "step": 4755 + }, + { + "epoch": 0.66, + "grad_norm": 2.158796111046285, + "learning_rate": 2.720123422602019e-06, + "loss": 0.3046, + "step": 4756 + }, + { + "epoch": 0.66, + "grad_norm": 2.200283588694543, + "learning_rate": 2.7181199318096642e-06, + "loss": 0.3052, + "step": 4757 + }, + { + "epoch": 0.66, + "grad_norm": 2.1611113152925467, + "learning_rate": 2.7161169036700197e-06, + "loss": 0.2879, + "step": 4758 + }, + { + "epoch": 0.66, + "grad_norm": 2.5746709151535567, + "learning_rate": 2.7141143385891967e-06, + "loss": 0.2828, + "step": 4759 + }, + { + "epoch": 0.66, + "grad_norm": 4.49151506498218, + "learning_rate": 2.712112236973215e-06, + "loss": 0.2946, + "step": 4760 + }, + { + "epoch": 0.66, + "grad_norm": 2.797663376783368, + "learning_rate": 2.7101105992280053e-06, + "loss": 0.3048, + "step": 4761 + }, + { + "epoch": 0.66, + "grad_norm": 2.601680364575405, + "learning_rate": 2.7081094257594e-06, + "loss": 0.3247, + "step": 4762 + }, + { + "epoch": 0.66, + "grad_norm": 1.8059778572806167, + "learning_rate": 2.7061087169731376e-06, + "loss": 0.3008, + "step": 4763 + }, + { + "epoch": 0.66, + "grad_norm": 2.673890430644996, + "learning_rate": 2.704108473274861e-06, + "loss": 0.3097, + "step": 4764 + }, + { + "epoch": 0.66, + "grad_norm": 2.713321975651211, + "learning_rate": 2.7021086950701257e-06, + "loss": 0.3047, + "step": 4765 + }, + { + "epoch": 0.66, + "grad_norm": 1.8858694953220896, + "learning_rate": 2.7001093827643843e-06, + "loss": 0.3175, + "step": 4766 + }, + { + "epoch": 0.66, + "grad_norm": 3.010083368703733, + "learning_rate": 2.6981105367629994e-06, + "loss": 0.2917, + "step": 4767 + }, + { + "epoch": 0.66, + "grad_norm": 3.133135180385427, + "learning_rate": 2.696112157471241e-06, + "loss": 0.31, + "step": 4768 + }, + { + "epoch": 0.66, + "grad_norm": 3.0284858528812375, + "learning_rate": 2.6941142452942805e-06, + "loss": 0.3201, + "step": 4769 + }, + { + "epoch": 0.66, + "grad_norm": 1.9972391325487884, + "learning_rate": 2.6921168006371945e-06, + "loss": 0.3097, + "step": 4770 + }, + { + "epoch": 0.66, + "grad_norm": 2.1007584058552022, + "learning_rate": 2.6901198239049652e-06, + "loss": 0.3074, + "step": 4771 + }, + { + "epoch": 0.66, + "grad_norm": 2.987230524010475, + "learning_rate": 2.6881233155024855e-06, + "loss": 0.3112, + "step": 4772 + }, + { + "epoch": 0.66, + "grad_norm": 2.421646325815111, + "learning_rate": 2.686127275834546e-06, + "loss": 0.3072, + "step": 4773 + }, + { + "epoch": 0.66, + "grad_norm": 1.9635175958427582, + "learning_rate": 2.6841317053058424e-06, + "loss": 0.32, + "step": 4774 + }, + { + "epoch": 0.66, + "grad_norm": 2.0143724414790576, + "learning_rate": 2.6821366043209814e-06, + "loss": 0.3115, + "step": 4775 + }, + { + "epoch": 0.66, + "grad_norm": 2.0926169106145918, + "learning_rate": 2.6801419732844723e-06, + "loss": 0.2975, + "step": 4776 + }, + { + "epoch": 0.66, + "grad_norm": 2.4458524368343237, + "learning_rate": 2.6781478126007255e-06, + "loss": 0.3259, + "step": 4777 + }, + { + "epoch": 0.66, + "grad_norm": 2.3462432910840776, + "learning_rate": 2.676154122674057e-06, + "loss": 0.3153, + "step": 4778 + }, + { + "epoch": 0.66, + "grad_norm": 2.464377763369349, + "learning_rate": 2.6741609039086935e-06, + "loss": 0.306, + "step": 4779 + }, + { + "epoch": 0.66, + "grad_norm": 1.9639284074974115, + "learning_rate": 2.6721681567087583e-06, + "loss": 0.2929, + "step": 4780 + }, + { + "epoch": 0.66, + "grad_norm": 2.3162748558512996, + "learning_rate": 2.6701758814782796e-06, + "loss": 0.3247, + "step": 4781 + }, + { + "epoch": 0.66, + "grad_norm": 1.924437149768121, + "learning_rate": 2.668184078621199e-06, + "loss": 0.3099, + "step": 4782 + }, + { + "epoch": 0.66, + "grad_norm": 3.1192974556390007, + "learning_rate": 2.6661927485413493e-06, + "loss": 0.3062, + "step": 4783 + }, + { + "epoch": 0.67, + "grad_norm": 2.2980537664982026, + "learning_rate": 2.6642018916424788e-06, + "loss": 0.3149, + "step": 4784 + }, + { + "epoch": 0.67, + "grad_norm": 2.276285275946601, + "learning_rate": 2.662211508328232e-06, + "loss": 0.2918, + "step": 4785 + }, + { + "epoch": 0.67, + "grad_norm": 2.682437070535472, + "learning_rate": 2.660221599002163e-06, + "loss": 0.3007, + "step": 4786 + }, + { + "epoch": 0.67, + "grad_norm": 2.7264650932427323, + "learning_rate": 2.6582321640677267e-06, + "loss": 0.3172, + "step": 4787 + }, + { + "epoch": 0.67, + "grad_norm": 2.638622685101641, + "learning_rate": 2.6562432039282793e-06, + "loss": 0.3285, + "step": 4788 + }, + { + "epoch": 0.67, + "grad_norm": 3.233315208827434, + "learning_rate": 2.6542547189870877e-06, + "loss": 0.3235, + "step": 4789 + }, + { + "epoch": 0.67, + "grad_norm": 2.126031988620262, + "learning_rate": 2.6522667096473177e-06, + "loss": 0.2995, + "step": 4790 + }, + { + "epoch": 0.67, + "grad_norm": 2.2923639384510466, + "learning_rate": 2.650279176312036e-06, + "loss": 0.313, + "step": 4791 + }, + { + "epoch": 0.67, + "grad_norm": 2.4366861943864753, + "learning_rate": 2.6482921193842192e-06, + "loss": 0.3196, + "step": 4792 + }, + { + "epoch": 0.67, + "grad_norm": 3.2789722916568484, + "learning_rate": 2.646305539266747e-06, + "loss": 0.3074, + "step": 4793 + }, + { + "epoch": 0.67, + "grad_norm": 3.7877088117362114, + "learning_rate": 2.644319436362397e-06, + "loss": 0.2935, + "step": 4794 + }, + { + "epoch": 0.67, + "grad_norm": 1.7959808683973524, + "learning_rate": 2.6423338110738507e-06, + "loss": 0.3156, + "step": 4795 + }, + { + "epoch": 0.67, + "grad_norm": 2.066895632343376, + "learning_rate": 2.640348663803699e-06, + "loss": 0.3034, + "step": 4796 + }, + { + "epoch": 0.67, + "grad_norm": 2.1718536243653785, + "learning_rate": 2.6383639949544304e-06, + "loss": 0.3139, + "step": 4797 + }, + { + "epoch": 0.67, + "grad_norm": 2.228298413403453, + "learning_rate": 2.636379804928436e-06, + "loss": 0.3133, + "step": 4798 + }, + { + "epoch": 0.67, + "grad_norm": 2.0836009075304442, + "learning_rate": 2.634396094128015e-06, + "loss": 0.3281, + "step": 4799 + }, + { + "epoch": 0.67, + "grad_norm": 2.569534642623685, + "learning_rate": 2.6324128629553624e-06, + "loss": 0.3256, + "step": 4800 + }, + { + "epoch": 0.67, + "grad_norm": 2.046520722235411, + "learning_rate": 2.630430111812583e-06, + "loss": 0.3024, + "step": 4801 + }, + { + "epoch": 0.67, + "grad_norm": 2.2497006142994453, + "learning_rate": 2.6284478411016786e-06, + "loss": 0.3067, + "step": 4802 + }, + { + "epoch": 0.67, + "grad_norm": 1.7578217516084154, + "learning_rate": 2.6264660512245577e-06, + "loss": 0.3027, + "step": 4803 + }, + { + "epoch": 0.67, + "grad_norm": 2.531668143162156, + "learning_rate": 2.624484742583029e-06, + "loss": 0.3131, + "step": 4804 + }, + { + "epoch": 0.67, + "grad_norm": 2.2240868866690184, + "learning_rate": 2.622503915578802e-06, + "loss": 0.3, + "step": 4805 + }, + { + "epoch": 0.67, + "grad_norm": 2.4191312875273683, + "learning_rate": 2.6205235706134936e-06, + "loss": 0.3028, + "step": 4806 + }, + { + "epoch": 0.67, + "grad_norm": 1.8300492606492609, + "learning_rate": 2.61854370808862e-06, + "loss": 0.3433, + "step": 4807 + }, + { + "epoch": 0.67, + "grad_norm": 1.8634231340093157, + "learning_rate": 2.6165643284055956e-06, + "loss": 0.321, + "step": 4808 + }, + { + "epoch": 0.67, + "grad_norm": 2.944514764500616, + "learning_rate": 2.6145854319657443e-06, + "loss": 0.3016, + "step": 4809 + }, + { + "epoch": 0.67, + "grad_norm": 1.8143647499276412, + "learning_rate": 2.61260701917029e-06, + "loss": 0.3065, + "step": 4810 + }, + { + "epoch": 0.67, + "grad_norm": 1.624701698359546, + "learning_rate": 2.610629090420356e-06, + "loss": 0.3124, + "step": 4811 + }, + { + "epoch": 0.67, + "grad_norm": 2.0551162260832383, + "learning_rate": 2.6086516461169666e-06, + "loss": 0.3207, + "step": 4812 + }, + { + "epoch": 0.67, + "grad_norm": 2.3545103981322613, + "learning_rate": 2.6066746866610527e-06, + "loss": 0.3028, + "step": 4813 + }, + { + "epoch": 0.67, + "grad_norm": 2.9328476780782, + "learning_rate": 2.6046982124534427e-06, + "loss": 0.3448, + "step": 4814 + }, + { + "epoch": 0.67, + "grad_norm": 1.9941410880463522, + "learning_rate": 2.602722223894867e-06, + "loss": 0.296, + "step": 4815 + }, + { + "epoch": 0.67, + "grad_norm": 2.2348735335257612, + "learning_rate": 2.6007467213859615e-06, + "loss": 0.2946, + "step": 4816 + }, + { + "epoch": 0.67, + "grad_norm": 2.2650093886024045, + "learning_rate": 2.5987717053272594e-06, + "loss": 0.3265, + "step": 4817 + }, + { + "epoch": 0.67, + "grad_norm": 2.430378778531038, + "learning_rate": 2.5967971761191947e-06, + "loss": 0.3055, + "step": 4818 + }, + { + "epoch": 0.67, + "grad_norm": 2.5276784645626145, + "learning_rate": 2.594823134162105e-06, + "loss": 0.3292, + "step": 4819 + }, + { + "epoch": 0.67, + "grad_norm": 3.1470501456135658, + "learning_rate": 2.592849579856232e-06, + "loss": 0.3152, + "step": 4820 + }, + { + "epoch": 0.67, + "grad_norm": 2.350757158766003, + "learning_rate": 2.5908765136017126e-06, + "loss": 0.3132, + "step": 4821 + }, + { + "epoch": 0.67, + "grad_norm": 2.296260776012958, + "learning_rate": 2.5889039357985855e-06, + "loss": 0.3183, + "step": 4822 + }, + { + "epoch": 0.67, + "grad_norm": 8.719719119565385, + "learning_rate": 2.586931846846797e-06, + "loss": 0.2814, + "step": 4823 + }, + { + "epoch": 0.67, + "grad_norm": 3.6596769963724256, + "learning_rate": 2.584960247146186e-06, + "loss": 0.3122, + "step": 4824 + }, + { + "epoch": 0.67, + "grad_norm": 2.289405255036066, + "learning_rate": 2.5829891370964965e-06, + "loss": 0.3319, + "step": 4825 + }, + { + "epoch": 0.67, + "grad_norm": 3.774220269691544, + "learning_rate": 2.581018517097368e-06, + "loss": 0.3157, + "step": 4826 + }, + { + "epoch": 0.67, + "grad_norm": 2.5720073259852665, + "learning_rate": 2.579048387548353e-06, + "loss": 0.2973, + "step": 4827 + }, + { + "epoch": 0.67, + "grad_norm": 2.9766224553280733, + "learning_rate": 2.5770787488488925e-06, + "loss": 0.3026, + "step": 4828 + }, + { + "epoch": 0.67, + "grad_norm": 2.306918830268566, + "learning_rate": 2.5751096013983313e-06, + "loss": 0.3214, + "step": 4829 + }, + { + "epoch": 0.67, + "grad_norm": 3.488356964901616, + "learning_rate": 2.573140945595917e-06, + "loss": 0.3187, + "step": 4830 + }, + { + "epoch": 0.67, + "grad_norm": 2.1979060478023817, + "learning_rate": 2.571172781840796e-06, + "loss": 0.3275, + "step": 4831 + }, + { + "epoch": 0.67, + "grad_norm": 7.101356693041764, + "learning_rate": 2.569205110532014e-06, + "loss": 0.3304, + "step": 4832 + }, + { + "epoch": 0.67, + "grad_norm": 2.177873305157263, + "learning_rate": 2.5672379320685158e-06, + "loss": 0.319, + "step": 4833 + }, + { + "epoch": 0.67, + "grad_norm": 1.8865324967156396, + "learning_rate": 2.5652712468491526e-06, + "loss": 0.2881, + "step": 4834 + }, + { + "epoch": 0.67, + "grad_norm": 1.8471463814782842, + "learning_rate": 2.5633050552726657e-06, + "loss": 0.325, + "step": 4835 + }, + { + "epoch": 0.67, + "grad_norm": 1.9470581645333351, + "learning_rate": 2.5613393577377054e-06, + "loss": 0.3087, + "step": 4836 + }, + { + "epoch": 0.67, + "grad_norm": 1.9455478158949282, + "learning_rate": 2.5593741546428184e-06, + "loss": 0.307, + "step": 4837 + }, + { + "epoch": 0.67, + "grad_norm": 2.142488676941894, + "learning_rate": 2.5574094463864508e-06, + "loss": 0.3186, + "step": 4838 + }, + { + "epoch": 0.67, + "grad_norm": 2.2473531482360243, + "learning_rate": 2.5554452333669476e-06, + "loss": 0.3175, + "step": 4839 + }, + { + "epoch": 0.67, + "grad_norm": 1.9290189486964422, + "learning_rate": 2.553481515982553e-06, + "loss": 0.3017, + "step": 4840 + }, + { + "epoch": 0.67, + "grad_norm": 2.036893373664227, + "learning_rate": 2.5515182946314145e-06, + "loss": 0.2951, + "step": 4841 + }, + { + "epoch": 0.67, + "grad_norm": 2.559479433540967, + "learning_rate": 2.549555569711576e-06, + "loss": 0.2777, + "step": 4842 + }, + { + "epoch": 0.67, + "grad_norm": 2.1568449826093126, + "learning_rate": 2.5475933416209762e-06, + "loss": 0.3246, + "step": 4843 + }, + { + "epoch": 0.67, + "grad_norm": 2.1638569273373305, + "learning_rate": 2.5456316107574676e-06, + "loss": 0.3125, + "step": 4844 + }, + { + "epoch": 0.67, + "grad_norm": 2.456940157007387, + "learning_rate": 2.5436703775187864e-06, + "loss": 0.2908, + "step": 4845 + }, + { + "epoch": 0.67, + "grad_norm": 1.8447665582302488, + "learning_rate": 2.5417096423025756e-06, + "loss": 0.3053, + "step": 4846 + }, + { + "epoch": 0.67, + "grad_norm": 1.9519213916404254, + "learning_rate": 2.539749405506373e-06, + "loss": 0.2944, + "step": 4847 + }, + { + "epoch": 0.67, + "grad_norm": 2.926568459180128, + "learning_rate": 2.5377896675276227e-06, + "loss": 0.3266, + "step": 4848 + }, + { + "epoch": 0.67, + "grad_norm": 2.0361850683637184, + "learning_rate": 2.5358304287636593e-06, + "loss": 0.2996, + "step": 4849 + }, + { + "epoch": 0.67, + "grad_norm": 2.542087997656609, + "learning_rate": 2.533871689611719e-06, + "loss": 0.2945, + "step": 4850 + }, + { + "epoch": 0.67, + "grad_norm": 2.3062944947797153, + "learning_rate": 2.53191345046894e-06, + "loss": 0.2993, + "step": 4851 + }, + { + "epoch": 0.67, + "grad_norm": 2.920849736290677, + "learning_rate": 2.529955711732355e-06, + "loss": 0.3136, + "step": 4852 + }, + { + "epoch": 0.67, + "grad_norm": 2.060347708452848, + "learning_rate": 2.5279984737988995e-06, + "loss": 0.3423, + "step": 4853 + }, + { + "epoch": 0.67, + "grad_norm": 2.7981984841406446, + "learning_rate": 2.5260417370654e-06, + "loss": 0.3094, + "step": 4854 + }, + { + "epoch": 0.67, + "grad_norm": 2.1495104865608634, + "learning_rate": 2.5240855019285916e-06, + "loss": 0.3124, + "step": 4855 + }, + { + "epoch": 0.68, + "grad_norm": 2.0922478197705834, + "learning_rate": 2.5221297687851e-06, + "loss": 0.3169, + "step": 4856 + }, + { + "epoch": 0.68, + "grad_norm": 2.3360187008062345, + "learning_rate": 2.5201745380314484e-06, + "loss": 0.3087, + "step": 4857 + }, + { + "epoch": 0.68, + "grad_norm": 2.605497564702712, + "learning_rate": 2.5182198100640657e-06, + "loss": 0.2989, + "step": 4858 + }, + { + "epoch": 0.68, + "grad_norm": 2.707947609563874, + "learning_rate": 2.5162655852792726e-06, + "loss": 0.3057, + "step": 4859 + }, + { + "epoch": 0.68, + "grad_norm": 5.113721017382816, + "learning_rate": 2.514311864073288e-06, + "loss": 0.275, + "step": 4860 + }, + { + "epoch": 0.68, + "grad_norm": 2.8479632192514184, + "learning_rate": 2.512358646842231e-06, + "loss": 0.3211, + "step": 4861 + }, + { + "epoch": 0.68, + "grad_norm": 4.577716629402094, + "learning_rate": 2.5104059339821203e-06, + "loss": 0.2908, + "step": 4862 + }, + { + "epoch": 0.68, + "grad_norm": 2.7006894696496646, + "learning_rate": 2.5084537258888674e-06, + "loss": 0.3198, + "step": 4863 + }, + { + "epoch": 0.68, + "grad_norm": 2.2500718281110963, + "learning_rate": 2.5065020229582814e-06, + "loss": 0.3028, + "step": 4864 + }, + { + "epoch": 0.68, + "grad_norm": 1.6999729084693318, + "learning_rate": 2.5045508255860773e-06, + "loss": 0.3027, + "step": 4865 + }, + { + "epoch": 0.68, + "grad_norm": 2.789077177815489, + "learning_rate": 2.5026001341678576e-06, + "loss": 0.3115, + "step": 4866 + }, + { + "epoch": 0.68, + "grad_norm": 1.8906045164726013, + "learning_rate": 2.500649949099125e-06, + "loss": 0.2903, + "step": 4867 + }, + { + "epoch": 0.68, + "grad_norm": 2.451324820820239, + "learning_rate": 2.4987002707752844e-06, + "loss": 0.3334, + "step": 4868 + }, + { + "epoch": 0.68, + "grad_norm": 4.574045898956562, + "learning_rate": 2.4967510995916315e-06, + "loss": 0.3317, + "step": 4869 + }, + { + "epoch": 0.68, + "grad_norm": 2.467184957361848, + "learning_rate": 2.4948024359433654e-06, + "loss": 0.3103, + "step": 4870 + }, + { + "epoch": 0.68, + "grad_norm": 2.263334073318673, + "learning_rate": 2.492854280225575e-06, + "loss": 0.2985, + "step": 4871 + }, + { + "epoch": 0.68, + "grad_norm": 2.2401810717772275, + "learning_rate": 2.490906632833254e-06, + "loss": 0.3179, + "step": 4872 + }, + { + "epoch": 0.68, + "grad_norm": 2.1843204219237697, + "learning_rate": 2.488959494161287e-06, + "loss": 0.3309, + "step": 4873 + }, + { + "epoch": 0.68, + "grad_norm": 2.4242402950232185, + "learning_rate": 2.487012864604456e-06, + "loss": 0.3077, + "step": 4874 + }, + { + "epoch": 0.68, + "grad_norm": 5.537050186826648, + "learning_rate": 2.4850667445574457e-06, + "loss": 0.303, + "step": 4875 + }, + { + "epoch": 0.68, + "grad_norm": 3.233765077835204, + "learning_rate": 2.4831211344148308e-06, + "loss": 0.3128, + "step": 4876 + }, + { + "epoch": 0.68, + "grad_norm": 3.5356930025797246, + "learning_rate": 2.4811760345710833e-06, + "loss": 0.2952, + "step": 4877 + }, + { + "epoch": 0.68, + "grad_norm": 2.492074350167361, + "learning_rate": 2.4792314454205755e-06, + "loss": 0.3023, + "step": 4878 + }, + { + "epoch": 0.68, + "grad_norm": 2.820965870937632, + "learning_rate": 2.4772873673575767e-06, + "loss": 0.3221, + "step": 4879 + }, + { + "epoch": 0.68, + "grad_norm": 2.0952821546416707, + "learning_rate": 2.4753438007762463e-06, + "loss": 0.3071, + "step": 4880 + }, + { + "epoch": 0.68, + "grad_norm": 2.8073512370279223, + "learning_rate": 2.4734007460706432e-06, + "loss": 0.2754, + "step": 4881 + }, + { + "epoch": 0.68, + "grad_norm": 2.341314920311504, + "learning_rate": 2.4714582036347268e-06, + "loss": 0.2942, + "step": 4882 + }, + { + "epoch": 0.68, + "grad_norm": 2.7875341273598067, + "learning_rate": 2.4695161738623464e-06, + "loss": 0.3487, + "step": 4883 + }, + { + "epoch": 0.68, + "grad_norm": 2.9401330073732614, + "learning_rate": 2.467574657147249e-06, + "loss": 0.3137, + "step": 4884 + }, + { + "epoch": 0.68, + "grad_norm": 2.170622161333336, + "learning_rate": 2.4656336538830807e-06, + "loss": 0.2952, + "step": 4885 + }, + { + "epoch": 0.68, + "grad_norm": 2.2138918710210373, + "learning_rate": 2.46369316446338e-06, + "loss": 0.3099, + "step": 4886 + }, + { + "epoch": 0.68, + "grad_norm": 2.6836329760925732, + "learning_rate": 2.4617531892815807e-06, + "loss": 0.3149, + "step": 4887 + }, + { + "epoch": 0.68, + "grad_norm": 2.1525471857755654, + "learning_rate": 2.4598137287310157e-06, + "loss": 0.305, + "step": 4888 + }, + { + "epoch": 0.68, + "grad_norm": 2.1577961891914024, + "learning_rate": 2.4578747832049145e-06, + "loss": 0.29, + "step": 4889 + }, + { + "epoch": 0.68, + "grad_norm": 2.4455165927567517, + "learning_rate": 2.455936353096397e-06, + "loss": 0.3439, + "step": 4890 + }, + { + "epoch": 0.68, + "grad_norm": 1.8042915617450153, + "learning_rate": 2.4539984387984795e-06, + "loss": 0.3183, + "step": 4891 + }, + { + "epoch": 0.68, + "grad_norm": 2.6182882966821115, + "learning_rate": 2.452061040704079e-06, + "loss": 0.3273, + "step": 4892 + }, + { + "epoch": 0.68, + "grad_norm": 2.4681652426519656, + "learning_rate": 2.4501241592060028e-06, + "loss": 0.3071, + "step": 4893 + }, + { + "epoch": 0.68, + "grad_norm": 4.139151028766597, + "learning_rate": 2.4481877946969545e-06, + "loss": 0.3133, + "step": 4894 + }, + { + "epoch": 0.68, + "grad_norm": 2.241878229267646, + "learning_rate": 2.4462519475695298e-06, + "loss": 0.2979, + "step": 4895 + }, + { + "epoch": 0.68, + "grad_norm": 2.097208317969886, + "learning_rate": 2.4443166182162305e-06, + "loss": 0.3063, + "step": 4896 + }, + { + "epoch": 0.68, + "grad_norm": 2.8028933186649647, + "learning_rate": 2.4423818070294415e-06, + "loss": 0.2676, + "step": 4897 + }, + { + "epoch": 0.68, + "grad_norm": 2.1102892874862293, + "learning_rate": 2.440447514401446e-06, + "loss": 0.2891, + "step": 4898 + }, + { + "epoch": 0.68, + "grad_norm": 2.374901699846415, + "learning_rate": 2.438513740724427e-06, + "loss": 0.3124, + "step": 4899 + }, + { + "epoch": 0.68, + "grad_norm": 1.8466864090693333, + "learning_rate": 2.4365804863904558e-06, + "loss": 0.3245, + "step": 4900 + }, + { + "epoch": 0.68, + "grad_norm": 2.036308082196111, + "learning_rate": 2.434647751791502e-06, + "loss": 0.3055, + "step": 4901 + }, + { + "epoch": 0.68, + "grad_norm": 3.3728892101296637, + "learning_rate": 2.4327155373194254e-06, + "loss": 0.32, + "step": 4902 + }, + { + "epoch": 0.68, + "grad_norm": 2.636448173723957, + "learning_rate": 2.430783843365989e-06, + "loss": 0.3192, + "step": 4903 + }, + { + "epoch": 0.68, + "grad_norm": 2.4056546778208348, + "learning_rate": 2.4288526703228405e-06, + "loss": 0.2999, + "step": 4904 + }, + { + "epoch": 0.68, + "grad_norm": 4.539116458421203, + "learning_rate": 2.426922018581529e-06, + "loss": 0.3317, + "step": 4905 + }, + { + "epoch": 0.68, + "grad_norm": 2.005888138388463, + "learning_rate": 2.424991888533497e-06, + "loss": 0.3106, + "step": 4906 + }, + { + "epoch": 0.68, + "grad_norm": 2.4278936368508086, + "learning_rate": 2.423062280570077e-06, + "loss": 0.3049, + "step": 4907 + }, + { + "epoch": 0.68, + "grad_norm": 2.5823652997685147, + "learning_rate": 2.4211331950824996e-06, + "loss": 0.2813, + "step": 4908 + }, + { + "epoch": 0.68, + "grad_norm": 2.1945128250858676, + "learning_rate": 2.419204632461885e-06, + "loss": 0.2995, + "step": 4909 + }, + { + "epoch": 0.68, + "grad_norm": 2.470785933348538, + "learning_rate": 2.417276593099256e-06, + "loss": 0.3004, + "step": 4910 + }, + { + "epoch": 0.68, + "grad_norm": 2.098117959520975, + "learning_rate": 2.4153490773855203e-06, + "loss": 0.309, + "step": 4911 + }, + { + "epoch": 0.68, + "grad_norm": 2.364639541223015, + "learning_rate": 2.4134220857114806e-06, + "loss": 0.3065, + "step": 4912 + }, + { + "epoch": 0.68, + "grad_norm": 5.043769161315669, + "learning_rate": 2.411495618467842e-06, + "loss": 0.3079, + "step": 4913 + }, + { + "epoch": 0.68, + "grad_norm": 1.983192253084533, + "learning_rate": 2.4095696760451937e-06, + "loss": 0.3043, + "step": 4914 + }, + { + "epoch": 0.68, + "grad_norm": 7.533272533314358, + "learning_rate": 2.407644258834023e-06, + "loss": 0.3094, + "step": 4915 + }, + { + "epoch": 0.68, + "grad_norm": 9.966620817051345, + "learning_rate": 2.4057193672247053e-06, + "loss": 0.2795, + "step": 4916 + }, + { + "epoch": 0.68, + "grad_norm": 2.354273925558096, + "learning_rate": 2.4037950016075197e-06, + "loss": 0.3196, + "step": 4917 + }, + { + "epoch": 0.68, + "grad_norm": 2.3189076256951835, + "learning_rate": 2.4018711623726293e-06, + "loss": 0.2972, + "step": 4918 + }, + { + "epoch": 0.68, + "grad_norm": 1.8736690040518507, + "learning_rate": 2.3999478499100924e-06, + "loss": 0.2781, + "step": 4919 + }, + { + "epoch": 0.68, + "grad_norm": 3.1441698999139254, + "learning_rate": 2.3980250646098663e-06, + "loss": 0.3071, + "step": 4920 + }, + { + "epoch": 0.68, + "grad_norm": 2.0298856233556632, + "learning_rate": 2.396102806861792e-06, + "loss": 0.2945, + "step": 4921 + }, + { + "epoch": 0.68, + "grad_norm": 2.141087141521005, + "learning_rate": 2.394181077055613e-06, + "loss": 0.3058, + "step": 4922 + }, + { + "epoch": 0.68, + "grad_norm": 2.1311476646757996, + "learning_rate": 2.3922598755809572e-06, + "loss": 0.3023, + "step": 4923 + }, + { + "epoch": 0.68, + "grad_norm": 2.629060648446314, + "learning_rate": 2.3903392028273537e-06, + "loss": 0.2993, + "step": 4924 + }, + { + "epoch": 0.68, + "grad_norm": 3.776951066834377, + "learning_rate": 2.3884190591842183e-06, + "loss": 0.3437, + "step": 4925 + }, + { + "epoch": 0.68, + "grad_norm": 1.96525342963026, + "learning_rate": 2.386499445040859e-06, + "loss": 0.3216, + "step": 4926 + }, + { + "epoch": 0.68, + "grad_norm": 5.325478861606673, + "learning_rate": 2.3845803607864827e-06, + "loss": 0.3148, + "step": 4927 + }, + { + "epoch": 0.69, + "grad_norm": 2.159114758075934, + "learning_rate": 2.382661806810184e-06, + "loss": 0.3164, + "step": 4928 + }, + { + "epoch": 0.69, + "grad_norm": 2.650373236988752, + "learning_rate": 2.380743783500948e-06, + "loss": 0.3201, + "step": 4929 + }, + { + "epoch": 0.69, + "grad_norm": 2.3140251201821607, + "learning_rate": 2.3788262912476574e-06, + "loss": 0.3218, + "step": 4930 + }, + { + "epoch": 0.69, + "grad_norm": 2.344603686120145, + "learning_rate": 2.376909330439087e-06, + "loss": 0.3038, + "step": 4931 + }, + { + "epoch": 0.69, + "grad_norm": 2.1758732450663114, + "learning_rate": 2.3749929014639e-06, + "loss": 0.3101, + "step": 4932 + }, + { + "epoch": 0.69, + "grad_norm": 2.082265296159366, + "learning_rate": 2.373077004710652e-06, + "loss": 0.31, + "step": 4933 + }, + { + "epoch": 0.69, + "grad_norm": 2.2997108514099045, + "learning_rate": 2.3711616405677963e-06, + "loss": 0.328, + "step": 4934 + }, + { + "epoch": 0.69, + "grad_norm": 2.1190159715484476, + "learning_rate": 2.3692468094236715e-06, + "loss": 0.2833, + "step": 4935 + }, + { + "epoch": 0.69, + "grad_norm": 2.012454013309936, + "learning_rate": 2.3673325116665095e-06, + "loss": 0.2901, + "step": 4936 + }, + { + "epoch": 0.69, + "grad_norm": 1.963331270491337, + "learning_rate": 2.3654187476844392e-06, + "loss": 0.3094, + "step": 4937 + }, + { + "epoch": 0.69, + "grad_norm": 2.0569742364957717, + "learning_rate": 2.363505517865474e-06, + "loss": 0.2809, + "step": 4938 + }, + { + "epoch": 0.69, + "grad_norm": 3.6719038239284405, + "learning_rate": 2.3615928225975267e-06, + "loss": 0.3078, + "step": 4939 + }, + { + "epoch": 0.69, + "grad_norm": 2.2927700981209616, + "learning_rate": 2.359680662268392e-06, + "loss": 0.3222, + "step": 4940 + }, + { + "epoch": 0.69, + "grad_norm": 2.0705993306198898, + "learning_rate": 2.3577690372657675e-06, + "loss": 0.3004, + "step": 4941 + }, + { + "epoch": 0.69, + "grad_norm": 2.067021439676367, + "learning_rate": 2.355857947977233e-06, + "loss": 0.3214, + "step": 4942 + }, + { + "epoch": 0.69, + "grad_norm": 2.537743923860405, + "learning_rate": 2.353947394790262e-06, + "loss": 0.3053, + "step": 4943 + }, + { + "epoch": 0.69, + "grad_norm": 2.8359930991645266, + "learning_rate": 2.3520373780922245e-06, + "loss": 0.3263, + "step": 4944 + }, + { + "epoch": 0.69, + "grad_norm": 2.042807095733499, + "learning_rate": 2.350127898270375e-06, + "loss": 0.3051, + "step": 4945 + }, + { + "epoch": 0.69, + "grad_norm": 2.158475462107335, + "learning_rate": 2.3482189557118605e-06, + "loss": 0.3132, + "step": 4946 + }, + { + "epoch": 0.69, + "grad_norm": 2.2963236729333283, + "learning_rate": 2.346310550803722e-06, + "loss": 0.2967, + "step": 4947 + }, + { + "epoch": 0.69, + "grad_norm": 2.199499282305439, + "learning_rate": 2.344402683932892e-06, + "loss": 0.2963, + "step": 4948 + }, + { + "epoch": 0.69, + "grad_norm": 2.316192236358476, + "learning_rate": 2.3424953554861895e-06, + "loss": 0.3209, + "step": 4949 + }, + { + "epoch": 0.69, + "grad_norm": 2.063666093032625, + "learning_rate": 2.3405885658503254e-06, + "loss": 0.3038, + "step": 4950 + }, + { + "epoch": 0.69, + "grad_norm": 3.8457584589024916, + "learning_rate": 2.338682315411905e-06, + "loss": 0.3337, + "step": 4951 + }, + { + "epoch": 0.69, + "grad_norm": 2.311476283698372, + "learning_rate": 2.3367766045574214e-06, + "loss": 0.3043, + "step": 4952 + }, + { + "epoch": 0.69, + "grad_norm": 2.317821946191517, + "learning_rate": 2.3348714336732564e-06, + "loss": 0.3296, + "step": 4953 + }, + { + "epoch": 0.69, + "grad_norm": 2.1556395969739155, + "learning_rate": 2.3329668031456883e-06, + "loss": 0.2947, + "step": 4954 + }, + { + "epoch": 0.69, + "grad_norm": 1.901107572656685, + "learning_rate": 2.33106271336088e-06, + "loss": 0.3069, + "step": 4955 + }, + { + "epoch": 0.69, + "grad_norm": 2.3519751328307628, + "learning_rate": 2.3291591647048857e-06, + "loss": 0.313, + "step": 4956 + }, + { + "epoch": 0.69, + "grad_norm": 5.388115893542031, + "learning_rate": 2.3272561575636532e-06, + "loss": 0.3219, + "step": 4957 + }, + { + "epoch": 0.69, + "grad_norm": 2.8687077523655384, + "learning_rate": 2.3253536923230198e-06, + "loss": 0.3009, + "step": 4958 + }, + { + "epoch": 0.69, + "grad_norm": 2.1612272581319756, + "learning_rate": 2.3234517693687093e-06, + "loss": 0.2985, + "step": 4959 + }, + { + "epoch": 0.69, + "grad_norm": 2.168188483804853, + "learning_rate": 2.3215503890863367e-06, + "loss": 0.3249, + "step": 4960 + }, + { + "epoch": 0.69, + "grad_norm": 1.9841414363706427, + "learning_rate": 2.3196495518614122e-06, + "loss": 0.2954, + "step": 4961 + }, + { + "epoch": 0.69, + "grad_norm": 4.437787090725875, + "learning_rate": 2.3177492580793293e-06, + "loss": 0.3204, + "step": 4962 + }, + { + "epoch": 0.69, + "grad_norm": 3.5149612780221573, + "learning_rate": 2.315849508125374e-06, + "loss": 0.2986, + "step": 4963 + }, + { + "epoch": 0.69, + "grad_norm": 5.187583911370369, + "learning_rate": 2.313950302384718e-06, + "loss": 0.3013, + "step": 4964 + }, + { + "epoch": 0.69, + "grad_norm": 2.704805990786251, + "learning_rate": 2.3120516412424345e-06, + "loss": 0.3176, + "step": 4965 + }, + { + "epoch": 0.69, + "grad_norm": 1.9234957679084739, + "learning_rate": 2.310153525083473e-06, + "loss": 0.3243, + "step": 4966 + }, + { + "epoch": 0.69, + "grad_norm": 2.379957961442974, + "learning_rate": 2.3082559542926788e-06, + "loss": 0.3107, + "step": 4967 + }, + { + "epoch": 0.69, + "grad_norm": 2.0673605061574816, + "learning_rate": 2.3063589292547867e-06, + "loss": 0.323, + "step": 4968 + }, + { + "epoch": 0.69, + "grad_norm": 2.546856233171494, + "learning_rate": 2.3044624503544194e-06, + "loss": 0.3273, + "step": 4969 + }, + { + "epoch": 0.69, + "grad_norm": 2.4332978368681397, + "learning_rate": 2.3025665179760888e-06, + "loss": 0.292, + "step": 4970 + }, + { + "epoch": 0.69, + "grad_norm": 2.044300566164674, + "learning_rate": 2.300671132504194e-06, + "loss": 0.2998, + "step": 4971 + }, + { + "epoch": 0.69, + "grad_norm": 3.4863480252325325, + "learning_rate": 2.298776294323031e-06, + "loss": 0.3128, + "step": 4972 + }, + { + "epoch": 0.69, + "grad_norm": 2.4643202092421705, + "learning_rate": 2.296882003816774e-06, + "loss": 0.3145, + "step": 4973 + }, + { + "epoch": 0.69, + "grad_norm": 2.7592200463799905, + "learning_rate": 2.2949882613694945e-06, + "loss": 0.3149, + "step": 4974 + }, + { + "epoch": 0.69, + "grad_norm": 2.675386831634561, + "learning_rate": 2.2930950673651518e-06, + "loss": 0.3263, + "step": 4975 + }, + { + "epoch": 0.69, + "grad_norm": 2.5777072103212983, + "learning_rate": 2.29120242218759e-06, + "loss": 0.3124, + "step": 4976 + }, + { + "epoch": 0.69, + "grad_norm": 1.9257903415528523, + "learning_rate": 2.289310326220545e-06, + "loss": 0.3086, + "step": 4977 + }, + { + "epoch": 0.69, + "grad_norm": 2.3777722090820346, + "learning_rate": 2.2874187798476377e-06, + "loss": 0.3208, + "step": 4978 + }, + { + "epoch": 0.69, + "grad_norm": 2.568959045784916, + "learning_rate": 2.2855277834523855e-06, + "loss": 0.3039, + "step": 4979 + }, + { + "epoch": 0.69, + "grad_norm": 2.2689997695878272, + "learning_rate": 2.283637337418185e-06, + "loss": 0.2948, + "step": 4980 + }, + { + "epoch": 0.69, + "grad_norm": 2.2290751887619646, + "learning_rate": 2.2817474421283243e-06, + "loss": 0.3079, + "step": 4981 + }, + { + "epoch": 0.69, + "grad_norm": 2.2622678880539118, + "learning_rate": 2.2798580979659864e-06, + "loss": 0.3076, + "step": 4982 + }, + { + "epoch": 0.69, + "grad_norm": 2.458850549559735, + "learning_rate": 2.277969305314234e-06, + "loss": 0.3165, + "step": 4983 + }, + { + "epoch": 0.69, + "grad_norm": 2.3149051356111654, + "learning_rate": 2.2760810645560217e-06, + "loss": 0.2872, + "step": 4984 + }, + { + "epoch": 0.69, + "grad_norm": 2.124574014081766, + "learning_rate": 2.2741933760741885e-06, + "loss": 0.3216, + "step": 4985 + }, + { + "epoch": 0.69, + "grad_norm": 2.247446369335596, + "learning_rate": 2.2723062402514697e-06, + "loss": 0.327, + "step": 4986 + }, + { + "epoch": 0.69, + "grad_norm": 2.6905527092100194, + "learning_rate": 2.2704196574704797e-06, + "loss": 0.3048, + "step": 4987 + }, + { + "epoch": 0.69, + "grad_norm": 2.167109629144404, + "learning_rate": 2.2685336281137243e-06, + "loss": 0.3184, + "step": 4988 + }, + { + "epoch": 0.69, + "grad_norm": 3.2625807795114508, + "learning_rate": 2.2666481525636e-06, + "loss": 0.2908, + "step": 4989 + }, + { + "epoch": 0.69, + "grad_norm": 2.5762824812588763, + "learning_rate": 2.264763231202384e-06, + "loss": 0.3031, + "step": 4990 + }, + { + "epoch": 0.69, + "grad_norm": 2.4693033886696756, + "learning_rate": 2.262878864412251e-06, + "loss": 0.3026, + "step": 4991 + }, + { + "epoch": 0.69, + "grad_norm": 3.1770175542490433, + "learning_rate": 2.2609950525752517e-06, + "loss": 0.2873, + "step": 4992 + }, + { + "epoch": 0.69, + "grad_norm": 2.547993510787605, + "learning_rate": 2.259111796073335e-06, + "loss": 0.2844, + "step": 4993 + }, + { + "epoch": 0.69, + "grad_norm": 4.260451970104897, + "learning_rate": 2.257229095288331e-06, + "loss": 0.3064, + "step": 4994 + }, + { + "epoch": 0.69, + "grad_norm": 2.486014571569937, + "learning_rate": 2.2553469506019555e-06, + "loss": 0.2794, + "step": 4995 + }, + { + "epoch": 0.69, + "grad_norm": 2.830808478013778, + "learning_rate": 2.253465362395819e-06, + "loss": 0.3099, + "step": 4996 + }, + { + "epoch": 0.69, + "grad_norm": 2.1641117332417186, + "learning_rate": 2.251584331051413e-06, + "loss": 0.3313, + "step": 4997 + }, + { + "epoch": 0.69, + "grad_norm": 2.0457067727693063, + "learning_rate": 2.2497038569501163e-06, + "loss": 0.3226, + "step": 4998 + }, + { + "epoch": 0.69, + "grad_norm": 2.043542942214743, + "learning_rate": 2.247823940473197e-06, + "loss": 0.2848, + "step": 4999 + }, + { + "epoch": 0.7, + "grad_norm": 3.3793468165685807, + "learning_rate": 2.2459445820018123e-06, + "loss": 0.3041, + "step": 5000 + }, + { + "epoch": 0.7, + "grad_norm": 1.966267764307778, + "learning_rate": 2.244065781917001e-06, + "loss": 0.3045, + "step": 5001 + }, + { + "epoch": 0.7, + "grad_norm": 4.8483684527833955, + "learning_rate": 2.242187540599689e-06, + "loss": 0.2985, + "step": 5002 + }, + { + "epoch": 0.7, + "grad_norm": 2.3616495791831267, + "learning_rate": 2.240309858430696e-06, + "loss": 0.3096, + "step": 5003 + }, + { + "epoch": 0.7, + "grad_norm": 2.865197484597787, + "learning_rate": 2.23843273579072e-06, + "loss": 0.3086, + "step": 5004 + }, + { + "epoch": 0.7, + "grad_norm": 2.388427637803159, + "learning_rate": 2.2365561730603473e-06, + "loss": 0.3215, + "step": 5005 + }, + { + "epoch": 0.7, + "grad_norm": 2.2524925591404217, + "learning_rate": 2.2346801706200567e-06, + "loss": 0.2888, + "step": 5006 + }, + { + "epoch": 0.7, + "grad_norm": 2.175456185183331, + "learning_rate": 2.232804728850205e-06, + "loss": 0.3017, + "step": 5007 + }, + { + "epoch": 0.7, + "grad_norm": 4.73959228770871, + "learning_rate": 2.2309298481310415e-06, + "loss": 0.2976, + "step": 5008 + }, + { + "epoch": 0.7, + "grad_norm": 2.1058942646049625, + "learning_rate": 2.229055528842698e-06, + "loss": 0.3159, + "step": 5009 + }, + { + "epoch": 0.7, + "grad_norm": 7.069360769097597, + "learning_rate": 2.2271817713651966e-06, + "loss": 0.3144, + "step": 5010 + }, + { + "epoch": 0.7, + "grad_norm": 2.112334870904965, + "learning_rate": 2.225308576078441e-06, + "loss": 0.3292, + "step": 5011 + }, + { + "epoch": 0.7, + "grad_norm": 2.191269473707011, + "learning_rate": 2.2234359433622216e-06, + "loss": 0.2924, + "step": 5012 + }, + { + "epoch": 0.7, + "grad_norm": 2.1070208081662303, + "learning_rate": 2.2215638735962186e-06, + "loss": 0.3256, + "step": 5013 + }, + { + "epoch": 0.7, + "grad_norm": 3.0639145438499797, + "learning_rate": 2.219692367159994e-06, + "loss": 0.3136, + "step": 5014 + }, + { + "epoch": 0.7, + "grad_norm": 2.9076355918937713, + "learning_rate": 2.2178214244329956e-06, + "loss": 0.322, + "step": 5015 + }, + { + "epoch": 0.7, + "grad_norm": 1.9698812056267743, + "learning_rate": 2.2159510457945595e-06, + "loss": 0.3086, + "step": 5016 + }, + { + "epoch": 0.7, + "grad_norm": 1.7849462595654686, + "learning_rate": 2.214081231623908e-06, + "loss": 0.3023, + "step": 5017 + }, + { + "epoch": 0.7, + "grad_norm": 2.3772573209024768, + "learning_rate": 2.2122119823001464e-06, + "loss": 0.2905, + "step": 5018 + }, + { + "epoch": 0.7, + "grad_norm": 2.4721130806091423, + "learning_rate": 2.2103432982022634e-06, + "loss": 0.3092, + "step": 5019 + }, + { + "epoch": 0.7, + "grad_norm": 2.7498656572901, + "learning_rate": 2.2084751797091393e-06, + "loss": 0.3211, + "step": 5020 + }, + { + "epoch": 0.7, + "grad_norm": 2.232919140355247, + "learning_rate": 2.206607627199535e-06, + "loss": 0.2964, + "step": 5021 + }, + { + "epoch": 0.7, + "grad_norm": 1.876921708860168, + "learning_rate": 2.2047406410520974e-06, + "loss": 0.2996, + "step": 5022 + }, + { + "epoch": 0.7, + "grad_norm": 2.9962363742673266, + "learning_rate": 2.202874221645361e-06, + "loss": 0.3222, + "step": 5023 + }, + { + "epoch": 0.7, + "grad_norm": 2.5124640622797396, + "learning_rate": 2.201008369357742e-06, + "loss": 0.3064, + "step": 5024 + }, + { + "epoch": 0.7, + "grad_norm": 2.4294921112437193, + "learning_rate": 2.199143084567543e-06, + "loss": 0.3203, + "step": 5025 + }, + { + "epoch": 0.7, + "grad_norm": 2.3644841338795164, + "learning_rate": 2.197278367652951e-06, + "loss": 0.2796, + "step": 5026 + }, + { + "epoch": 0.7, + "grad_norm": 3.797484636056545, + "learning_rate": 2.195414218992043e-06, + "loss": 0.3115, + "step": 5027 + }, + { + "epoch": 0.7, + "grad_norm": 2.363524289820993, + "learning_rate": 2.193550638962773e-06, + "loss": 0.3158, + "step": 5028 + }, + { + "epoch": 0.7, + "grad_norm": 1.7484932015799832, + "learning_rate": 2.1916876279429817e-06, + "loss": 0.2794, + "step": 5029 + }, + { + "epoch": 0.7, + "grad_norm": 1.9792543318623885, + "learning_rate": 2.1898251863103993e-06, + "loss": 0.3358, + "step": 5030 + }, + { + "epoch": 0.7, + "grad_norm": 2.2645209605118444, + "learning_rate": 2.1879633144426355e-06, + "loss": 0.2933, + "step": 5031 + }, + { + "epoch": 0.7, + "grad_norm": 1.9834471652547312, + "learning_rate": 2.1861020127171843e-06, + "loss": 0.3098, + "step": 5032 + }, + { + "epoch": 0.7, + "grad_norm": 2.3980641415119015, + "learning_rate": 2.1842412815114274e-06, + "loss": 0.3109, + "step": 5033 + }, + { + "epoch": 0.7, + "grad_norm": 2.685739676443493, + "learning_rate": 2.1823811212026315e-06, + "loss": 0.351, + "step": 5034 + }, + { + "epoch": 0.7, + "grad_norm": 1.9543622416581743, + "learning_rate": 2.180521532167943e-06, + "loss": 0.3103, + "step": 5035 + }, + { + "epoch": 0.7, + "grad_norm": 2.6853306287038503, + "learning_rate": 2.1786625147843933e-06, + "loss": 0.3059, + "step": 5036 + }, + { + "epoch": 0.7, + "grad_norm": 2.3415197669640313, + "learning_rate": 2.1768040694289027e-06, + "loss": 0.3201, + "step": 5037 + }, + { + "epoch": 0.7, + "grad_norm": 2.423838934445162, + "learning_rate": 2.174946196478271e-06, + "loss": 0.3067, + "step": 5038 + }, + { + "epoch": 0.7, + "grad_norm": 2.732664955541631, + "learning_rate": 2.1730888963091826e-06, + "loss": 0.3291, + "step": 5039 + }, + { + "epoch": 0.7, + "grad_norm": 1.8841756240156333, + "learning_rate": 2.1712321692982053e-06, + "loss": 0.3044, + "step": 5040 + }, + { + "epoch": 0.7, + "grad_norm": 2.043154603807893, + "learning_rate": 2.1693760158217946e-06, + "loss": 0.3159, + "step": 5041 + }, + { + "epoch": 0.7, + "grad_norm": 2.167427858614579, + "learning_rate": 2.1675204362562842e-06, + "loss": 0.2984, + "step": 5042 + }, + { + "epoch": 0.7, + "grad_norm": 2.229735085808831, + "learning_rate": 2.1656654309778945e-06, + "loss": 0.3199, + "step": 5043 + }, + { + "epoch": 0.7, + "grad_norm": 4.326575756390337, + "learning_rate": 2.1638110003627317e-06, + "loss": 0.3018, + "step": 5044 + }, + { + "epoch": 0.7, + "grad_norm": 2.3070254263529404, + "learning_rate": 2.1619571447867816e-06, + "loss": 0.3118, + "step": 5045 + }, + { + "epoch": 0.7, + "grad_norm": 1.9730954118107207, + "learning_rate": 2.160103864625913e-06, + "loss": 0.3084, + "step": 5046 + }, + { + "epoch": 0.7, + "grad_norm": 2.5633338802626864, + "learning_rate": 2.1582511602558798e-06, + "loss": 0.3312, + "step": 5047 + }, + { + "epoch": 0.7, + "grad_norm": 2.056447670972618, + "learning_rate": 2.1563990320523213e-06, + "loss": 0.3191, + "step": 5048 + }, + { + "epoch": 0.7, + "grad_norm": 1.8958763837142498, + "learning_rate": 2.1545474803907564e-06, + "loss": 0.2987, + "step": 5049 + }, + { + "epoch": 0.7, + "grad_norm": 2.866069559642157, + "learning_rate": 2.152696505646585e-06, + "loss": 0.3058, + "step": 5050 + }, + { + "epoch": 0.7, + "grad_norm": 2.5056511243723936, + "learning_rate": 2.1508461081951003e-06, + "loss": 0.3024, + "step": 5051 + }, + { + "epoch": 0.7, + "grad_norm": 1.888133856152129, + "learning_rate": 2.1489962884114685e-06, + "loss": 0.3359, + "step": 5052 + }, + { + "epoch": 0.7, + "grad_norm": 2.5225525644343954, + "learning_rate": 2.147147046670741e-06, + "loss": 0.2732, + "step": 5053 + }, + { + "epoch": 0.7, + "grad_norm": 2.253072995451315, + "learning_rate": 2.145298383347852e-06, + "loss": 0.3021, + "step": 5054 + }, + { + "epoch": 0.7, + "grad_norm": 2.2206126969131024, + "learning_rate": 2.143450298817622e-06, + "loss": 0.2762, + "step": 5055 + }, + { + "epoch": 0.7, + "grad_norm": 2.5453120197003387, + "learning_rate": 2.1416027934547494e-06, + "loss": 0.3262, + "step": 5056 + }, + { + "epoch": 0.7, + "grad_norm": 3.1911699113048666, + "learning_rate": 2.1397558676338165e-06, + "loss": 0.2966, + "step": 5057 + }, + { + "epoch": 0.7, + "grad_norm": 2.004389419257761, + "learning_rate": 2.1379095217292916e-06, + "loss": 0.3151, + "step": 5058 + }, + { + "epoch": 0.7, + "grad_norm": 2.0601479036547774, + "learning_rate": 2.1360637561155188e-06, + "loss": 0.3113, + "step": 5059 + }, + { + "epoch": 0.7, + "grad_norm": 2.7734135710833097, + "learning_rate": 2.1342185711667325e-06, + "loss": 0.3085, + "step": 5060 + }, + { + "epoch": 0.7, + "grad_norm": 2.1501372035094932, + "learning_rate": 2.132373967257041e-06, + "loss": 0.3188, + "step": 5061 + }, + { + "epoch": 0.7, + "grad_norm": 2.607967745542745, + "learning_rate": 2.130529944760443e-06, + "loss": 0.2994, + "step": 5062 + }, + { + "epoch": 0.7, + "grad_norm": 2.0547656651929076, + "learning_rate": 2.1286865040508125e-06, + "loss": 0.2922, + "step": 5063 + }, + { + "epoch": 0.7, + "grad_norm": 4.002609417577497, + "learning_rate": 2.126843645501908e-06, + "loss": 0.3089, + "step": 5064 + }, + { + "epoch": 0.7, + "grad_norm": 2.202837238384428, + "learning_rate": 2.125001369487373e-06, + "loss": 0.3061, + "step": 5065 + }, + { + "epoch": 0.7, + "grad_norm": 2.1272701107215197, + "learning_rate": 2.1231596763807287e-06, + "loss": 0.2919, + "step": 5066 + }, + { + "epoch": 0.7, + "grad_norm": 2.6560827604542405, + "learning_rate": 2.121318566555378e-06, + "loss": 0.3191, + "step": 5067 + }, + { + "epoch": 0.7, + "grad_norm": 2.3147608801255983, + "learning_rate": 2.119478040384608e-06, + "loss": 0.3021, + "step": 5068 + }, + { + "epoch": 0.7, + "grad_norm": 4.0985315896487675, + "learning_rate": 2.11763809824159e-06, + "loss": 0.2953, + "step": 5069 + }, + { + "epoch": 0.7, + "grad_norm": 1.9951713868548242, + "learning_rate": 2.11579874049937e-06, + "loss": 0.3094, + "step": 5070 + }, + { + "epoch": 0.7, + "grad_norm": 2.4945420992488, + "learning_rate": 2.1139599675308785e-06, + "loss": 0.3173, + "step": 5071 + }, + { + "epoch": 0.71, + "grad_norm": 2.2230230684663215, + "learning_rate": 2.1121217797089306e-06, + "loss": 0.3028, + "step": 5072 + }, + { + "epoch": 0.71, + "grad_norm": 1.8835495980853594, + "learning_rate": 2.110284177406219e-06, + "loss": 0.3038, + "step": 5073 + }, + { + "epoch": 0.71, + "grad_norm": 5.178750107772932, + "learning_rate": 2.1084471609953166e-06, + "loss": 0.3085, + "step": 5074 + }, + { + "epoch": 0.71, + "grad_norm": 2.41125058608077, + "learning_rate": 2.1066107308486834e-06, + "loss": 0.3005, + "step": 5075 + }, + { + "epoch": 0.71, + "grad_norm": 2.8029455355693154, + "learning_rate": 2.104774887338653e-06, + "loss": 0.3159, + "step": 5076 + }, + { + "epoch": 0.71, + "grad_norm": 2.7981473527851417, + "learning_rate": 2.1029396308374485e-06, + "loss": 0.3238, + "step": 5077 + }, + { + "epoch": 0.71, + "grad_norm": 3.446270632463893, + "learning_rate": 2.1011049617171644e-06, + "loss": 0.3273, + "step": 5078 + }, + { + "epoch": 0.71, + "grad_norm": 2.440215129878047, + "learning_rate": 2.099270880349786e-06, + "loss": 0.3124, + "step": 5079 + }, + { + "epoch": 0.71, + "grad_norm": 2.431951626780361, + "learning_rate": 2.097437387107171e-06, + "loss": 0.3189, + "step": 5080 + }, + { + "epoch": 0.71, + "grad_norm": 3.3514233605976176, + "learning_rate": 2.095604482361061e-06, + "loss": 0.32, + "step": 5081 + }, + { + "epoch": 0.71, + "grad_norm": 2.39639926004607, + "learning_rate": 2.093772166483082e-06, + "loss": 0.3083, + "step": 5082 + }, + { + "epoch": 0.71, + "grad_norm": 1.9144997314121102, + "learning_rate": 2.091940439844736e-06, + "loss": 0.3219, + "step": 5083 + }, + { + "epoch": 0.71, + "grad_norm": 13.259548678355499, + "learning_rate": 2.0901093028174035e-06, + "loss": 0.3155, + "step": 5084 + }, + { + "epoch": 0.71, + "grad_norm": 2.962293802560424, + "learning_rate": 2.0882787557723514e-06, + "loss": 0.2879, + "step": 5085 + }, + { + "epoch": 0.71, + "grad_norm": 2.542232124325365, + "learning_rate": 2.086448799080727e-06, + "loss": 0.3212, + "step": 5086 + }, + { + "epoch": 0.71, + "grad_norm": 2.4210367700150237, + "learning_rate": 2.0846194331135524e-06, + "loss": 0.3204, + "step": 5087 + }, + { + "epoch": 0.71, + "grad_norm": 6.58904003140374, + "learning_rate": 2.08279065824173e-06, + "loss": 0.3208, + "step": 5088 + }, + { + "epoch": 0.71, + "grad_norm": 2.092856623037544, + "learning_rate": 2.0809624748360513e-06, + "loss": 0.3206, + "step": 5089 + }, + { + "epoch": 0.71, + "grad_norm": 2.6253136629098943, + "learning_rate": 2.0791348832671776e-06, + "loss": 0.3023, + "step": 5090 + }, + { + "epoch": 0.71, + "grad_norm": 2.573864105253433, + "learning_rate": 2.077307883905654e-06, + "loss": 0.3075, + "step": 5091 + }, + { + "epoch": 0.71, + "grad_norm": 2.624351155303346, + "learning_rate": 2.0754814771219083e-06, + "loss": 0.2875, + "step": 5092 + }, + { + "epoch": 0.71, + "grad_norm": 2.1450633987472645, + "learning_rate": 2.073655663286244e-06, + "loss": 0.2991, + "step": 5093 + }, + { + "epoch": 0.71, + "grad_norm": 2.0083088026534717, + "learning_rate": 2.0718304427688458e-06, + "loss": 0.2985, + "step": 5094 + }, + { + "epoch": 0.71, + "grad_norm": 2.027508890718246, + "learning_rate": 2.0700058159397785e-06, + "loss": 0.3178, + "step": 5095 + }, + { + "epoch": 0.71, + "grad_norm": 2.4880790482167856, + "learning_rate": 2.0681817831689886e-06, + "loss": 0.3217, + "step": 5096 + }, + { + "epoch": 0.71, + "grad_norm": 2.899037306857667, + "learning_rate": 2.066358344826299e-06, + "loss": 0.3055, + "step": 5097 + }, + { + "epoch": 0.71, + "grad_norm": 1.9717626170639813, + "learning_rate": 2.06453550128141e-06, + "loss": 0.3284, + "step": 5098 + }, + { + "epoch": 0.71, + "grad_norm": 5.384846193868419, + "learning_rate": 2.062713252903909e-06, + "loss": 0.3109, + "step": 5099 + }, + { + "epoch": 0.71, + "grad_norm": 3.469336555045967, + "learning_rate": 2.0608916000632556e-06, + "loss": 0.2873, + "step": 5100 + }, + { + "epoch": 0.71, + "grad_norm": 2.4124704523605804, + "learning_rate": 2.0590705431287895e-06, + "loss": 0.3209, + "step": 5101 + }, + { + "epoch": 0.71, + "grad_norm": 3.763845933775634, + "learning_rate": 2.057250082469733e-06, + "loss": 0.2737, + "step": 5102 + }, + { + "epoch": 0.71, + "grad_norm": 2.227788479868045, + "learning_rate": 2.055430218455188e-06, + "loss": 0.2983, + "step": 5103 + }, + { + "epoch": 0.71, + "grad_norm": 2.053454115515149, + "learning_rate": 2.05361095145413e-06, + "loss": 0.2952, + "step": 5104 + }, + { + "epoch": 0.71, + "grad_norm": 2.8102609445287285, + "learning_rate": 2.051792281835416e-06, + "loss": 0.2929, + "step": 5105 + }, + { + "epoch": 0.71, + "grad_norm": 4.2665461228517545, + "learning_rate": 2.0499742099677866e-06, + "loss": 0.3404, + "step": 5106 + }, + { + "epoch": 0.71, + "grad_norm": 2.2805503355048726, + "learning_rate": 2.0481567362198537e-06, + "loss": 0.2939, + "step": 5107 + }, + { + "epoch": 0.71, + "grad_norm": 2.016808251906013, + "learning_rate": 2.0463398609601122e-06, + "loss": 0.319, + "step": 5108 + }, + { + "epoch": 0.71, + "grad_norm": 2.0876821406927566, + "learning_rate": 2.0445235845569327e-06, + "loss": 0.3299, + "step": 5109 + }, + { + "epoch": 0.71, + "grad_norm": 2.6703777673936373, + "learning_rate": 2.0427079073785703e-06, + "loss": 0.3074, + "step": 5110 + }, + { + "epoch": 0.71, + "grad_norm": 2.4593196818319387, + "learning_rate": 2.0408928297931514e-06, + "loss": 0.3278, + "step": 5111 + }, + { + "epoch": 0.71, + "grad_norm": 2.9938797896941067, + "learning_rate": 2.0390783521686853e-06, + "loss": 0.3064, + "step": 5112 + }, + { + "epoch": 0.71, + "grad_norm": 2.6258416304516508, + "learning_rate": 2.0372644748730606e-06, + "loss": 0.32, + "step": 5113 + }, + { + "epoch": 0.71, + "grad_norm": 2.151878756299434, + "learning_rate": 2.035451198274041e-06, + "loss": 0.2922, + "step": 5114 + }, + { + "epoch": 0.71, + "grad_norm": 2.3494262513077846, + "learning_rate": 2.0336385227392684e-06, + "loss": 0.3094, + "step": 5115 + }, + { + "epoch": 0.71, + "grad_norm": 2.358044231725633, + "learning_rate": 2.031826448636263e-06, + "loss": 0.3379, + "step": 5116 + }, + { + "epoch": 0.71, + "grad_norm": 2.5273040967821996, + "learning_rate": 2.030014976332427e-06, + "loss": 0.3178, + "step": 5117 + }, + { + "epoch": 0.71, + "grad_norm": 5.078747747392151, + "learning_rate": 2.0282041061950367e-06, + "loss": 0.2919, + "step": 5118 + }, + { + "epoch": 0.71, + "grad_norm": 2.8146630015183756, + "learning_rate": 2.026393838591243e-06, + "loss": 0.3062, + "step": 5119 + }, + { + "epoch": 0.71, + "grad_norm": 2.2273719497938296, + "learning_rate": 2.0245841738880862e-06, + "loss": 0.2987, + "step": 5120 + }, + { + "epoch": 0.71, + "grad_norm": 2.2471229680692297, + "learning_rate": 2.0227751124524734e-06, + "loss": 0.2944, + "step": 5121 + }, + { + "epoch": 0.71, + "grad_norm": 7.15205693560755, + "learning_rate": 2.020966654651193e-06, + "loss": 0.3016, + "step": 5122 + }, + { + "epoch": 0.71, + "grad_norm": 4.3414653556463385, + "learning_rate": 2.0191588008509095e-06, + "loss": 0.2935, + "step": 5123 + }, + { + "epoch": 0.71, + "grad_norm": 2.4010468078158076, + "learning_rate": 2.0173515514181696e-06, + "loss": 0.3114, + "step": 5124 + }, + { + "epoch": 0.71, + "grad_norm": 2.401197735203011, + "learning_rate": 2.0155449067193928e-06, + "loss": 0.3099, + "step": 5125 + }, + { + "epoch": 0.71, + "grad_norm": 2.614381819058883, + "learning_rate": 2.0137388671208754e-06, + "loss": 0.3294, + "step": 5126 + }, + { + "epoch": 0.71, + "grad_norm": 2.609729296755374, + "learning_rate": 2.0119334329887974e-06, + "loss": 0.3026, + "step": 5127 + }, + { + "epoch": 0.71, + "grad_norm": 2.3986416764500635, + "learning_rate": 2.010128604689208e-06, + "loss": 0.317, + "step": 5128 + }, + { + "epoch": 0.71, + "grad_norm": 2.478411693823397, + "learning_rate": 2.00832438258804e-06, + "loss": 0.2949, + "step": 5129 + }, + { + "epoch": 0.71, + "grad_norm": 2.621776495300358, + "learning_rate": 2.006520767051098e-06, + "loss": 0.3053, + "step": 5130 + }, + { + "epoch": 0.71, + "grad_norm": 2.021221247924842, + "learning_rate": 2.0047177584440696e-06, + "loss": 0.3308, + "step": 5131 + }, + { + "epoch": 0.71, + "grad_norm": 2.0732120074962817, + "learning_rate": 2.0029153571325145e-06, + "loss": 0.2857, + "step": 5132 + }, + { + "epoch": 0.71, + "grad_norm": 2.832976878878806, + "learning_rate": 2.001113563481868e-06, + "loss": 0.2978, + "step": 5133 + }, + { + "epoch": 0.71, + "grad_norm": 2.289226380015739, + "learning_rate": 1.9993123778574497e-06, + "loss": 0.2893, + "step": 5134 + }, + { + "epoch": 0.71, + "grad_norm": 2.341421946875647, + "learning_rate": 1.9975118006244487e-06, + "loss": 0.3026, + "step": 5135 + }, + { + "epoch": 0.71, + "grad_norm": 2.020596791248385, + "learning_rate": 1.9957118321479326e-06, + "loss": 0.2961, + "step": 5136 + }, + { + "epoch": 0.71, + "grad_norm": 2.1258078831876133, + "learning_rate": 1.993912472792846e-06, + "loss": 0.3274, + "step": 5137 + }, + { + "epoch": 0.71, + "grad_norm": 2.5457397665135, + "learning_rate": 1.9921137229240134e-06, + "loss": 0.3183, + "step": 5138 + }, + { + "epoch": 0.71, + "grad_norm": 2.4191302745782575, + "learning_rate": 1.9903155829061307e-06, + "loss": 0.3109, + "step": 5139 + }, + { + "epoch": 0.71, + "grad_norm": 2.1360028526625374, + "learning_rate": 1.9885180531037694e-06, + "loss": 0.3015, + "step": 5140 + }, + { + "epoch": 0.71, + "grad_norm": 2.4393947920740517, + "learning_rate": 1.9867211338813842e-06, + "loss": 0.2904, + "step": 5141 + }, + { + "epoch": 0.71, + "grad_norm": 2.072079898592821, + "learning_rate": 1.9849248256033e-06, + "loss": 0.3091, + "step": 5142 + }, + { + "epoch": 0.71, + "grad_norm": 2.4218618180421068, + "learning_rate": 1.983129128633717e-06, + "loss": 0.3145, + "step": 5143 + }, + { + "epoch": 0.72, + "grad_norm": 2.0626107407123486, + "learning_rate": 1.981334043336718e-06, + "loss": 0.3131, + "step": 5144 + }, + { + "epoch": 0.72, + "grad_norm": 2.1549317628085687, + "learning_rate": 1.9795395700762537e-06, + "loss": 0.2863, + "step": 5145 + }, + { + "epoch": 0.72, + "grad_norm": 2.197578205701377, + "learning_rate": 1.977745709216159e-06, + "loss": 0.3147, + "step": 5146 + }, + { + "epoch": 0.72, + "grad_norm": 3.002858024392998, + "learning_rate": 1.9759524611201364e-06, + "loss": 0.2956, + "step": 5147 + }, + { + "epoch": 0.72, + "grad_norm": 1.9251087905333728, + "learning_rate": 1.974159826151772e-06, + "loss": 0.3081, + "step": 5148 + }, + { + "epoch": 0.72, + "grad_norm": 3.230591937911198, + "learning_rate": 1.9723678046745216e-06, + "loss": 0.3051, + "step": 5149 + }, + { + "epoch": 0.72, + "grad_norm": 1.906585324893317, + "learning_rate": 1.9705763970517174e-06, + "loss": 0.2954, + "step": 5150 + }, + { + "epoch": 0.72, + "grad_norm": 2.394444737406696, + "learning_rate": 1.9687856036465712e-06, + "loss": 0.2869, + "step": 5151 + }, + { + "epoch": 0.72, + "grad_norm": 1.9792973274858707, + "learning_rate": 1.966995424822167e-06, + "loss": 0.2973, + "step": 5152 + }, + { + "epoch": 0.72, + "grad_norm": 2.836577812787168, + "learning_rate": 1.9652058609414614e-06, + "loss": 0.3007, + "step": 5153 + }, + { + "epoch": 0.72, + "grad_norm": 2.6600050533328075, + "learning_rate": 1.9634169123672935e-06, + "loss": 0.3146, + "step": 5154 + }, + { + "epoch": 0.72, + "grad_norm": 2.013590635166472, + "learning_rate": 1.961628579462374e-06, + "loss": 0.3115, + "step": 5155 + }, + { + "epoch": 0.72, + "grad_norm": 1.7956563139568624, + "learning_rate": 1.9598408625892872e-06, + "loss": 0.3106, + "step": 5156 + }, + { + "epoch": 0.72, + "grad_norm": 2.1248513955475663, + "learning_rate": 1.958053762110492e-06, + "loss": 0.2913, + "step": 5157 + }, + { + "epoch": 0.72, + "grad_norm": 2.646402121763556, + "learning_rate": 1.9562672783883285e-06, + "loss": 0.2845, + "step": 5158 + }, + { + "epoch": 0.72, + "grad_norm": 2.124037584623824, + "learning_rate": 1.9544814117850053e-06, + "loss": 0.3184, + "step": 5159 + }, + { + "epoch": 0.72, + "grad_norm": 3.2523499823073867, + "learning_rate": 1.9526961626626057e-06, + "loss": 0.3016, + "step": 5160 + }, + { + "epoch": 0.72, + "grad_norm": 2.4864595774878224, + "learning_rate": 1.9509115313830946e-06, + "loss": 0.3033, + "step": 5161 + }, + { + "epoch": 0.72, + "grad_norm": 1.8919924324383057, + "learning_rate": 1.9491275183083024e-06, + "loss": 0.3098, + "step": 5162 + }, + { + "epoch": 0.72, + "grad_norm": 2.598677559742399, + "learning_rate": 1.9473441237999435e-06, + "loss": 0.3128, + "step": 5163 + }, + { + "epoch": 0.72, + "grad_norm": 5.51791002744955, + "learning_rate": 1.945561348219599e-06, + "loss": 0.3001, + "step": 5164 + }, + { + "epoch": 0.72, + "grad_norm": 2.4018247342223553, + "learning_rate": 1.943779191928731e-06, + "loss": 0.3007, + "step": 5165 + }, + { + "epoch": 0.72, + "grad_norm": 2.6944496918774252, + "learning_rate": 1.9419976552886704e-06, + "loss": 0.3081, + "step": 5166 + }, + { + "epoch": 0.72, + "grad_norm": 2.0746520989018538, + "learning_rate": 1.9402167386606243e-06, + "loss": 0.3088, + "step": 5167 + }, + { + "epoch": 0.72, + "grad_norm": 4.593627808109611, + "learning_rate": 1.9384364424056773e-06, + "loss": 0.3011, + "step": 5168 + }, + { + "epoch": 0.72, + "grad_norm": 2.0088255766375673, + "learning_rate": 1.9366567668847845e-06, + "loss": 0.2962, + "step": 5169 + }, + { + "epoch": 0.72, + "grad_norm": 2.816775756995665, + "learning_rate": 1.934877712458774e-06, + "loss": 0.3063, + "step": 5170 + }, + { + "epoch": 0.72, + "grad_norm": 2.809797442410455, + "learning_rate": 1.9330992794883537e-06, + "loss": 0.3057, + "step": 5171 + }, + { + "epoch": 0.72, + "grad_norm": 3.0439976957816217, + "learning_rate": 1.9313214683341015e-06, + "loss": 0.3159, + "step": 5172 + }, + { + "epoch": 0.72, + "grad_norm": 2.3947705371765315, + "learning_rate": 1.9295442793564693e-06, + "loss": 0.3213, + "step": 5173 + }, + { + "epoch": 0.72, + "grad_norm": 2.3952741602708265, + "learning_rate": 1.9277677129157813e-06, + "loss": 0.3134, + "step": 5174 + }, + { + "epoch": 0.72, + "grad_norm": 2.055251433939196, + "learning_rate": 1.9259917693722413e-06, + "loss": 0.3151, + "step": 5175 + }, + { + "epoch": 0.72, + "grad_norm": 2.0813922289138738, + "learning_rate": 1.924216449085921e-06, + "loss": 0.2987, + "step": 5176 + }, + { + "epoch": 0.72, + "grad_norm": 2.5316278365246307, + "learning_rate": 1.9224417524167667e-06, + "loss": 0.3307, + "step": 5177 + }, + { + "epoch": 0.72, + "grad_norm": 2.4920865804321832, + "learning_rate": 1.9206676797246014e-06, + "loss": 0.3099, + "step": 5178 + }, + { + "epoch": 0.72, + "grad_norm": 2.151652911975305, + "learning_rate": 1.9188942313691195e-06, + "loss": 0.3203, + "step": 5179 + }, + { + "epoch": 0.72, + "grad_norm": 2.178457299480979, + "learning_rate": 1.9171214077098853e-06, + "loss": 0.3038, + "step": 5180 + }, + { + "epoch": 0.72, + "grad_norm": 1.902812535886388, + "learning_rate": 1.9153492091063435e-06, + "loss": 0.29, + "step": 5181 + }, + { + "epoch": 0.72, + "grad_norm": 1.9958594271066215, + "learning_rate": 1.9135776359178094e-06, + "loss": 0.305, + "step": 5182 + }, + { + "epoch": 0.72, + "grad_norm": 1.9646920181388112, + "learning_rate": 1.9118066885034685e-06, + "loss": 0.2971, + "step": 5183 + }, + { + "epoch": 0.72, + "grad_norm": 2.2339471893320137, + "learning_rate": 1.910036367222382e-06, + "loss": 0.3098, + "step": 5184 + }, + { + "epoch": 0.72, + "grad_norm": 2.422912780193787, + "learning_rate": 1.9082666724334825e-06, + "loss": 0.296, + "step": 5185 + }, + { + "epoch": 0.72, + "grad_norm": 2.393315098395507, + "learning_rate": 1.9064976044955796e-06, + "loss": 0.3073, + "step": 5186 + }, + { + "epoch": 0.72, + "grad_norm": 2.8482553502505827, + "learning_rate": 1.904729163767351e-06, + "loss": 0.3047, + "step": 5187 + }, + { + "epoch": 0.72, + "grad_norm": 2.3328982860976413, + "learning_rate": 1.9029613506073468e-06, + "loss": 0.3057, + "step": 5188 + }, + { + "epoch": 0.72, + "grad_norm": 2.9156502733277985, + "learning_rate": 1.9011941653739985e-06, + "loss": 0.3275, + "step": 5189 + }, + { + "epoch": 0.72, + "grad_norm": 2.015785652283085, + "learning_rate": 1.8994276084256008e-06, + "loss": 0.2904, + "step": 5190 + }, + { + "epoch": 0.72, + "grad_norm": 2.0468149070790864, + "learning_rate": 1.897661680120324e-06, + "loss": 0.3155, + "step": 5191 + }, + { + "epoch": 0.72, + "grad_norm": 2.239508712558691, + "learning_rate": 1.89589638081621e-06, + "loss": 0.3297, + "step": 5192 + }, + { + "epoch": 0.72, + "grad_norm": 2.3131228749780317, + "learning_rate": 1.8941317108711777e-06, + "loss": 0.3223, + "step": 5193 + }, + { + "epoch": 0.72, + "grad_norm": 4.65614469151677, + "learning_rate": 1.8923676706430133e-06, + "loss": 0.3026, + "step": 5194 + }, + { + "epoch": 0.72, + "grad_norm": 2.276117473507176, + "learning_rate": 1.8906042604893753e-06, + "loss": 0.2968, + "step": 5195 + }, + { + "epoch": 0.72, + "grad_norm": 4.962351637946448, + "learning_rate": 1.8888414807678001e-06, + "loss": 0.2965, + "step": 5196 + }, + { + "epoch": 0.72, + "grad_norm": 2.2726794854845207, + "learning_rate": 1.8870793318356888e-06, + "loss": 0.3191, + "step": 5197 + }, + { + "epoch": 0.72, + "grad_norm": 2.4190730331956205, + "learning_rate": 1.8853178140503214e-06, + "loss": 0.3283, + "step": 5198 + }, + { + "epoch": 0.72, + "grad_norm": 2.3362204457795617, + "learning_rate": 1.8835569277688442e-06, + "loss": 0.2842, + "step": 5199 + }, + { + "epoch": 0.72, + "grad_norm": 2.081237543668682, + "learning_rate": 1.8817966733482807e-06, + "loss": 0.3234, + "step": 5200 + }, + { + "epoch": 0.72, + "grad_norm": 1.9767304648617348, + "learning_rate": 1.8800370511455218e-06, + "loss": 0.2798, + "step": 5201 + }, + { + "epoch": 0.72, + "grad_norm": 4.725139451602777, + "learning_rate": 1.8782780615173313e-06, + "loss": 0.2927, + "step": 5202 + }, + { + "epoch": 0.72, + "grad_norm": 2.9778137934216486, + "learning_rate": 1.876519704820348e-06, + "loss": 0.2967, + "step": 5203 + }, + { + "epoch": 0.72, + "grad_norm": 2.18151799992984, + "learning_rate": 1.8747619814110796e-06, + "loss": 0.309, + "step": 5204 + }, + { + "epoch": 0.72, + "grad_norm": 2.450708234225017, + "learning_rate": 1.873004891645902e-06, + "loss": 0.2973, + "step": 5205 + }, + { + "epoch": 0.72, + "grad_norm": 2.7795122558944465, + "learning_rate": 1.8712484358810695e-06, + "loss": 0.2918, + "step": 5206 + }, + { + "epoch": 0.72, + "grad_norm": 2.015315431505534, + "learning_rate": 1.8694926144727055e-06, + "loss": 0.3159, + "step": 5207 + }, + { + "epoch": 0.72, + "grad_norm": 6.848328517503497, + "learning_rate": 1.8677374277768034e-06, + "loss": 0.3004, + "step": 5208 + }, + { + "epoch": 0.72, + "grad_norm": 2.746606048503312, + "learning_rate": 1.8659828761492255e-06, + "loss": 0.2995, + "step": 5209 + }, + { + "epoch": 0.72, + "grad_norm": 2.3280361252538517, + "learning_rate": 1.8642289599457125e-06, + "loss": 0.3071, + "step": 5210 + }, + { + "epoch": 0.72, + "grad_norm": 2.501751164696882, + "learning_rate": 1.8624756795218696e-06, + "loss": 0.3006, + "step": 5211 + }, + { + "epoch": 0.72, + "grad_norm": 3.2534655615266437, + "learning_rate": 1.860723035233175e-06, + "loss": 0.2894, + "step": 5212 + }, + { + "epoch": 0.72, + "grad_norm": 2.7385949700419117, + "learning_rate": 1.858971027434981e-06, + "loss": 0.2959, + "step": 5213 + }, + { + "epoch": 0.72, + "grad_norm": 2.206372197032588, + "learning_rate": 1.8572196564825057e-06, + "loss": 0.3213, + "step": 5214 + }, + { + "epoch": 0.72, + "grad_norm": 4.49030179356795, + "learning_rate": 1.855468922730843e-06, + "loss": 0.3327, + "step": 5215 + }, + { + "epoch": 0.73, + "grad_norm": 2.9595510594802072, + "learning_rate": 1.8537188265349532e-06, + "loss": 0.3149, + "step": 5216 + }, + { + "epoch": 0.73, + "grad_norm": 2.1770034923559898, + "learning_rate": 1.851969368249672e-06, + "loss": 0.2978, + "step": 5217 + }, + { + "epoch": 0.73, + "grad_norm": 3.709841880629183, + "learning_rate": 1.850220548229702e-06, + "loss": 0.3483, + "step": 5218 + }, + { + "epoch": 0.73, + "grad_norm": 2.5077810030140206, + "learning_rate": 1.8484723668296157e-06, + "loss": 0.2921, + "step": 5219 + }, + { + "epoch": 0.73, + "grad_norm": 1.6621937300989984, + "learning_rate": 1.8467248244038616e-06, + "loss": 0.2937, + "step": 5220 + }, + { + "epoch": 0.73, + "grad_norm": 2.2296718001954585, + "learning_rate": 1.8449779213067532e-06, + "loss": 0.3335, + "step": 5221 + }, + { + "epoch": 0.73, + "grad_norm": 2.0924449464483006, + "learning_rate": 1.8432316578924747e-06, + "loss": 0.3006, + "step": 5222 + }, + { + "epoch": 0.73, + "grad_norm": 1.9945138665712623, + "learning_rate": 1.8414860345150837e-06, + "loss": 0.2743, + "step": 5223 + }, + { + "epoch": 0.73, + "grad_norm": 2.216645543340158, + "learning_rate": 1.8397410515285086e-06, + "loss": 0.3412, + "step": 5224 + }, + { + "epoch": 0.73, + "grad_norm": 2.13888909083997, + "learning_rate": 1.8379967092865437e-06, + "loss": 0.3148, + "step": 5225 + }, + { + "epoch": 0.73, + "grad_norm": 4.631725418871194, + "learning_rate": 1.8362530081428543e-06, + "loss": 0.3131, + "step": 5226 + }, + { + "epoch": 0.73, + "grad_norm": 2.123053030509626, + "learning_rate": 1.8345099484509799e-06, + "loss": 0.3016, + "step": 5227 + }, + { + "epoch": 0.73, + "grad_norm": 3.8251743367935704, + "learning_rate": 1.8327675305643255e-06, + "loss": 0.316, + "step": 5228 + }, + { + "epoch": 0.73, + "grad_norm": 2.3567840040505317, + "learning_rate": 1.8310257548361659e-06, + "loss": 0.307, + "step": 5229 + }, + { + "epoch": 0.73, + "grad_norm": 2.2816468961058245, + "learning_rate": 1.8292846216196508e-06, + "loss": 0.324, + "step": 5230 + }, + { + "epoch": 0.73, + "grad_norm": 2.1233810285527657, + "learning_rate": 1.8275441312677928e-06, + "loss": 0.3208, + "step": 5231 + }, + { + "epoch": 0.73, + "grad_norm": 2.4599065322306113, + "learning_rate": 1.82580428413348e-06, + "loss": 0.3245, + "step": 5232 + }, + { + "epoch": 0.73, + "grad_norm": 2.4282489008985015, + "learning_rate": 1.8240650805694653e-06, + "loss": 0.3104, + "step": 5233 + }, + { + "epoch": 0.73, + "grad_norm": 2.0288089519803427, + "learning_rate": 1.8223265209283758e-06, + "loss": 0.3039, + "step": 5234 + }, + { + "epoch": 0.73, + "grad_norm": 3.1956030141980123, + "learning_rate": 1.8205886055627048e-06, + "loss": 0.3208, + "step": 5235 + }, + { + "epoch": 0.73, + "grad_norm": 4.126198401355912, + "learning_rate": 1.8188513348248137e-06, + "loss": 0.3399, + "step": 5236 + }, + { + "epoch": 0.73, + "grad_norm": 2.135645117662053, + "learning_rate": 1.8171147090669383e-06, + "loss": 0.2784, + "step": 5237 + }, + { + "epoch": 0.73, + "grad_norm": 2.4240584249467334, + "learning_rate": 1.81537872864118e-06, + "loss": 0.3209, + "step": 5238 + }, + { + "epoch": 0.73, + "grad_norm": 2.510320455988758, + "learning_rate": 1.813643393899507e-06, + "loss": 0.2837, + "step": 5239 + }, + { + "epoch": 0.73, + "grad_norm": 2.07714160279079, + "learning_rate": 1.8119087051937623e-06, + "loss": 0.3016, + "step": 5240 + }, + { + "epoch": 0.73, + "grad_norm": 3.0509284952382227, + "learning_rate": 1.8101746628756561e-06, + "loss": 0.2759, + "step": 5241 + }, + { + "epoch": 0.73, + "grad_norm": 2.167816309303357, + "learning_rate": 1.8084412672967654e-06, + "loss": 0.3192, + "step": 5242 + }, + { + "epoch": 0.73, + "grad_norm": 4.690277173710603, + "learning_rate": 1.806708518808536e-06, + "loss": 0.3067, + "step": 5243 + }, + { + "epoch": 0.73, + "grad_norm": 2.461489596903332, + "learning_rate": 1.8049764177622864e-06, + "loss": 0.2853, + "step": 5244 + }, + { + "epoch": 0.73, + "grad_norm": 2.416866237680561, + "learning_rate": 1.8032449645092004e-06, + "loss": 0.3157, + "step": 5245 + }, + { + "epoch": 0.73, + "grad_norm": 2.8126878322403295, + "learning_rate": 1.8015141594003284e-06, + "loss": 0.3091, + "step": 5246 + }, + { + "epoch": 0.73, + "grad_norm": 2.076481629910843, + "learning_rate": 1.799784002786597e-06, + "loss": 0.3333, + "step": 5247 + }, + { + "epoch": 0.73, + "grad_norm": 2.961732179267052, + "learning_rate": 1.798054495018794e-06, + "loss": 0.3066, + "step": 5248 + }, + { + "epoch": 0.73, + "grad_norm": 2.303782273838917, + "learning_rate": 1.7963256364475768e-06, + "loss": 0.2954, + "step": 5249 + }, + { + "epoch": 0.73, + "grad_norm": 2.02802524188154, + "learning_rate": 1.7945974274234746e-06, + "loss": 0.3081, + "step": 5250 + }, + { + "epoch": 0.73, + "grad_norm": 2.0952386976096626, + "learning_rate": 1.7928698682968842e-06, + "loss": 0.3263, + "step": 5251 + }, + { + "epoch": 0.73, + "grad_norm": 7.57761710257232, + "learning_rate": 1.791142959418068e-06, + "loss": 0.3078, + "step": 5252 + }, + { + "epoch": 0.73, + "grad_norm": 2.6252605062618506, + "learning_rate": 1.7894167011371561e-06, + "loss": 0.3074, + "step": 5253 + }, + { + "epoch": 0.73, + "grad_norm": 3.361505566648093, + "learning_rate": 1.7876910938041513e-06, + "loss": 0.3064, + "step": 5254 + }, + { + "epoch": 0.73, + "grad_norm": 2.474111243877576, + "learning_rate": 1.7859661377689213e-06, + "loss": 0.2973, + "step": 5255 + }, + { + "epoch": 0.73, + "grad_norm": 2.033610625384094, + "learning_rate": 1.784241833381201e-06, + "loss": 0.3016, + "step": 5256 + }, + { + "epoch": 0.73, + "grad_norm": 2.850711541664, + "learning_rate": 1.7825181809905917e-06, + "loss": 0.315, + "step": 5257 + }, + { + "epoch": 0.73, + "grad_norm": 2.7260100232959106, + "learning_rate": 1.7807951809465712e-06, + "loss": 0.2867, + "step": 5258 + }, + { + "epoch": 0.73, + "grad_norm": 3.0375905442946993, + "learning_rate": 1.7790728335984759e-06, + "loss": 0.2926, + "step": 5259 + }, + { + "epoch": 0.73, + "grad_norm": 3.0959127569046174, + "learning_rate": 1.7773511392955122e-06, + "loss": 0.3082, + "step": 5260 + }, + { + "epoch": 0.73, + "grad_norm": 1.9699159742994141, + "learning_rate": 1.7756300983867537e-06, + "loss": 0.3139, + "step": 5261 + }, + { + "epoch": 0.73, + "grad_norm": 2.0138964292368353, + "learning_rate": 1.7739097112211457e-06, + "loss": 0.3155, + "step": 5262 + }, + { + "epoch": 0.73, + "grad_norm": 2.1960039696165037, + "learning_rate": 1.7721899781474966e-06, + "loss": 0.2843, + "step": 5263 + }, + { + "epoch": 0.73, + "grad_norm": 1.8738142499571757, + "learning_rate": 1.7704708995144815e-06, + "loss": 0.308, + "step": 5264 + }, + { + "epoch": 0.73, + "grad_norm": 7.12231587087768, + "learning_rate": 1.7687524756706476e-06, + "loss": 0.315, + "step": 5265 + }, + { + "epoch": 0.73, + "grad_norm": 2.2681045917384597, + "learning_rate": 1.7670347069644039e-06, + "loss": 0.3289, + "step": 5266 + }, + { + "epoch": 0.73, + "grad_norm": 2.2441017261123566, + "learning_rate": 1.765317593744032e-06, + "loss": 0.324, + "step": 5267 + }, + { + "epoch": 0.73, + "grad_norm": 2.296737723072994, + "learning_rate": 1.7636011363576743e-06, + "loss": 0.3146, + "step": 5268 + }, + { + "epoch": 0.73, + "grad_norm": 3.369648765812561, + "learning_rate": 1.7618853351533478e-06, + "loss": 0.2949, + "step": 5269 + }, + { + "epoch": 0.73, + "grad_norm": 1.9753461772673593, + "learning_rate": 1.7601701904789298e-06, + "loss": 0.3265, + "step": 5270 + }, + { + "epoch": 0.73, + "grad_norm": 2.635562835894985, + "learning_rate": 1.7584557026821658e-06, + "loss": 0.2969, + "step": 5271 + }, + { + "epoch": 0.73, + "grad_norm": 2.411006421881887, + "learning_rate": 1.7567418721106728e-06, + "loss": 0.2881, + "step": 5272 + }, + { + "epoch": 0.73, + "grad_norm": 1.7395583407913566, + "learning_rate": 1.7550286991119292e-06, + "loss": 0.318, + "step": 5273 + }, + { + "epoch": 0.73, + "grad_norm": 2.692037010961662, + "learning_rate": 1.7533161840332796e-06, + "loss": 0.2984, + "step": 5274 + }, + { + "epoch": 0.73, + "grad_norm": 2.377335557083588, + "learning_rate": 1.7516043272219406e-06, + "loss": 0.3042, + "step": 5275 + }, + { + "epoch": 0.73, + "grad_norm": 3.786467875462762, + "learning_rate": 1.7498931290249931e-06, + "loss": 0.2978, + "step": 5276 + }, + { + "epoch": 0.73, + "grad_norm": 1.8616633097717692, + "learning_rate": 1.748182589789383e-06, + "loss": 0.2837, + "step": 5277 + }, + { + "epoch": 0.73, + "grad_norm": 1.8625474422288129, + "learning_rate": 1.74647270986192e-06, + "loss": 0.3052, + "step": 5278 + }, + { + "epoch": 0.73, + "grad_norm": 2.67300463059191, + "learning_rate": 1.7447634895892878e-06, + "loss": 0.309, + "step": 5279 + }, + { + "epoch": 0.73, + "grad_norm": 2.0653070814700794, + "learning_rate": 1.74305492931803e-06, + "loss": 0.2967, + "step": 5280 + }, + { + "epoch": 0.73, + "grad_norm": 2.7678133361191612, + "learning_rate": 1.7413470293945566e-06, + "loss": 0.2947, + "step": 5281 + }, + { + "epoch": 0.73, + "grad_norm": 1.798467302100448, + "learning_rate": 1.7396397901651496e-06, + "loss": 0.3136, + "step": 5282 + }, + { + "epoch": 0.73, + "grad_norm": 2.649194434809387, + "learning_rate": 1.7379332119759484e-06, + "loss": 0.2803, + "step": 5283 + }, + { + "epoch": 0.73, + "grad_norm": 3.774143400457684, + "learning_rate": 1.7362272951729665e-06, + "loss": 0.2601, + "step": 5284 + }, + { + "epoch": 0.73, + "grad_norm": 2.7066171009153033, + "learning_rate": 1.7345220401020768e-06, + "loss": 0.3093, + "step": 5285 + }, + { + "epoch": 0.73, + "grad_norm": 3.7115419482909515, + "learning_rate": 1.7328174471090238e-06, + "loss": 0.3049, + "step": 5286 + }, + { + "epoch": 0.73, + "grad_norm": 2.1225615176646624, + "learning_rate": 1.7311135165394134e-06, + "loss": 0.3211, + "step": 5287 + }, + { + "epoch": 0.74, + "grad_norm": 2.8437901231128375, + "learning_rate": 1.7294102487387172e-06, + "loss": 0.2837, + "step": 5288 + }, + { + "epoch": 0.74, + "grad_norm": 2.0252762691990776, + "learning_rate": 1.7277076440522767e-06, + "loss": 0.3284, + "step": 5289 + }, + { + "epoch": 0.74, + "grad_norm": 1.8006053583724964, + "learning_rate": 1.7260057028252947e-06, + "loss": 0.3083, + "step": 5290 + }, + { + "epoch": 0.74, + "grad_norm": 2.4742260751309915, + "learning_rate": 1.7243044254028391e-06, + "loss": 0.2827, + "step": 5291 + }, + { + "epoch": 0.74, + "grad_norm": 1.9019759888757597, + "learning_rate": 1.722603812129847e-06, + "loss": 0.293, + "step": 5292 + }, + { + "epoch": 0.74, + "grad_norm": 2.2153004800685157, + "learning_rate": 1.7209038633511204e-06, + "loss": 0.3197, + "step": 5293 + }, + { + "epoch": 0.74, + "grad_norm": 2.0150831968900595, + "learning_rate": 1.7192045794113238e-06, + "loss": 0.2895, + "step": 5294 + }, + { + "epoch": 0.74, + "grad_norm": 3.1323414275566543, + "learning_rate": 1.7175059606549855e-06, + "loss": 0.2982, + "step": 5295 + }, + { + "epoch": 0.74, + "grad_norm": 2.462029226406001, + "learning_rate": 1.715808007426506e-06, + "loss": 0.3123, + "step": 5296 + }, + { + "epoch": 0.74, + "grad_norm": 2.3870252586386367, + "learning_rate": 1.714110720070144e-06, + "loss": 0.3031, + "step": 5297 + }, + { + "epoch": 0.74, + "grad_norm": 2.017125603058993, + "learning_rate": 1.7124140989300242e-06, + "loss": 0.2822, + "step": 5298 + }, + { + "epoch": 0.74, + "grad_norm": 2.575180124462474, + "learning_rate": 1.7107181443501413e-06, + "loss": 0.3121, + "step": 5299 + }, + { + "epoch": 0.74, + "grad_norm": 1.8696126921230374, + "learning_rate": 1.7090228566743477e-06, + "loss": 0.3135, + "step": 5300 + }, + { + "epoch": 0.74, + "grad_norm": 1.827757897758481, + "learning_rate": 1.7073282362463672e-06, + "loss": 0.3378, + "step": 5301 + }, + { + "epoch": 0.74, + "grad_norm": 2.1108302193993698, + "learning_rate": 1.7056342834097822e-06, + "loss": 0.3063, + "step": 5302 + }, + { + "epoch": 0.74, + "grad_norm": 8.789341706799487, + "learning_rate": 1.7039409985080463e-06, + "loss": 0.3076, + "step": 5303 + }, + { + "epoch": 0.74, + "grad_norm": 3.4567610242756373, + "learning_rate": 1.7022483818844714e-06, + "loss": 0.3087, + "step": 5304 + }, + { + "epoch": 0.74, + "grad_norm": 1.9307466411842444, + "learning_rate": 1.7005564338822366e-06, + "loss": 0.2898, + "step": 5305 + }, + { + "epoch": 0.74, + "grad_norm": 2.196262908083838, + "learning_rate": 1.6988651548443874e-06, + "loss": 0.2776, + "step": 5306 + }, + { + "epoch": 0.74, + "grad_norm": 2.0097103877362947, + "learning_rate": 1.697174545113831e-06, + "loss": 0.2899, + "step": 5307 + }, + { + "epoch": 0.74, + "grad_norm": 5.844913386248798, + "learning_rate": 1.6954846050333374e-06, + "loss": 0.3092, + "step": 5308 + }, + { + "epoch": 0.74, + "grad_norm": 2.565580178227051, + "learning_rate": 1.693795334945545e-06, + "loss": 0.3045, + "step": 5309 + }, + { + "epoch": 0.74, + "grad_norm": 3.586351389855022, + "learning_rate": 1.692106735192956e-06, + "loss": 0.2851, + "step": 5310 + }, + { + "epoch": 0.74, + "grad_norm": 2.4641718339545995, + "learning_rate": 1.6904188061179339e-06, + "loss": 0.2937, + "step": 5311 + }, + { + "epoch": 0.74, + "grad_norm": 2.0507270217396085, + "learning_rate": 1.6887315480627053e-06, + "loss": 0.3057, + "step": 5312 + }, + { + "epoch": 0.74, + "grad_norm": 2.340946044155803, + "learning_rate": 1.6870449613693663e-06, + "loss": 0.332, + "step": 5313 + }, + { + "epoch": 0.74, + "grad_norm": 3.206188629686213, + "learning_rate": 1.685359046379872e-06, + "loss": 0.3183, + "step": 5314 + }, + { + "epoch": 0.74, + "grad_norm": 2.1248816345459787, + "learning_rate": 1.6836738034360405e-06, + "loss": 0.29, + "step": 5315 + }, + { + "epoch": 0.74, + "grad_norm": 2.1232167314414307, + "learning_rate": 1.6819892328795601e-06, + "loss": 0.3098, + "step": 5316 + }, + { + "epoch": 0.74, + "grad_norm": 3.6767515773702506, + "learning_rate": 1.6803053350519771e-06, + "loss": 0.3057, + "step": 5317 + }, + { + "epoch": 0.74, + "grad_norm": 2.7151658293421335, + "learning_rate": 1.6786221102946998e-06, + "loss": 0.2948, + "step": 5318 + }, + { + "epoch": 0.74, + "grad_norm": 2.2771788924428136, + "learning_rate": 1.676939558949006e-06, + "loss": 0.3157, + "step": 5319 + }, + { + "epoch": 0.74, + "grad_norm": 2.571260445296058, + "learning_rate": 1.675257681356036e-06, + "loss": 0.2912, + "step": 5320 + }, + { + "epoch": 0.74, + "grad_norm": 2.1389940959370235, + "learning_rate": 1.6735764778567891e-06, + "loss": 0.3054, + "step": 5321 + }, + { + "epoch": 0.74, + "grad_norm": 2.4326886180515537, + "learning_rate": 1.6718959487921293e-06, + "loss": 0.3135, + "step": 5322 + }, + { + "epoch": 0.74, + "grad_norm": 1.7626432888860437, + "learning_rate": 1.6702160945027884e-06, + "loss": 0.2866, + "step": 5323 + }, + { + "epoch": 0.74, + "grad_norm": 2.555214747189403, + "learning_rate": 1.6685369153293556e-06, + "loss": 0.2887, + "step": 5324 + }, + { + "epoch": 0.74, + "grad_norm": 1.8819627725902601, + "learning_rate": 1.666858411612286e-06, + "loss": 0.3114, + "step": 5325 + }, + { + "epoch": 0.74, + "grad_norm": 1.9901715740572594, + "learning_rate": 1.6651805836918943e-06, + "loss": 0.3328, + "step": 5326 + }, + { + "epoch": 0.74, + "grad_norm": 2.2773692780879156, + "learning_rate": 1.6635034319083675e-06, + "loss": 0.2804, + "step": 5327 + }, + { + "epoch": 0.74, + "grad_norm": 2.157648429806104, + "learning_rate": 1.6618269566017465e-06, + "loss": 0.3333, + "step": 5328 + }, + { + "epoch": 0.74, + "grad_norm": 2.2412199401424884, + "learning_rate": 1.6601511581119374e-06, + "loss": 0.3218, + "step": 5329 + }, + { + "epoch": 0.74, + "grad_norm": 1.9720359244291823, + "learning_rate": 1.6584760367787072e-06, + "loss": 0.3087, + "step": 5330 + }, + { + "epoch": 0.74, + "grad_norm": 2.0420135907250314, + "learning_rate": 1.6568015929416919e-06, + "loss": 0.3106, + "step": 5331 + }, + { + "epoch": 0.74, + "grad_norm": 1.9411945135253215, + "learning_rate": 1.6551278269403842e-06, + "loss": 0.3121, + "step": 5332 + }, + { + "epoch": 0.74, + "grad_norm": 2.326106189687852, + "learning_rate": 1.6534547391141398e-06, + "loss": 0.2749, + "step": 5333 + }, + { + "epoch": 0.74, + "grad_norm": 2.1521394658785753, + "learning_rate": 1.6517823298021819e-06, + "loss": 0.3132, + "step": 5334 + }, + { + "epoch": 0.74, + "grad_norm": 1.9758551032344702, + "learning_rate": 1.650110599343589e-06, + "loss": 0.284, + "step": 5335 + }, + { + "epoch": 0.74, + "grad_norm": 3.126623842036868, + "learning_rate": 1.6484395480773085e-06, + "loss": 0.3018, + "step": 5336 + }, + { + "epoch": 0.74, + "grad_norm": 2.178546975338567, + "learning_rate": 1.6467691763421446e-06, + "loss": 0.2989, + "step": 5337 + }, + { + "epoch": 0.74, + "grad_norm": 2.266142034290527, + "learning_rate": 1.6450994844767688e-06, + "loss": 0.3137, + "step": 5338 + }, + { + "epoch": 0.74, + "grad_norm": 3.0004072630172716, + "learning_rate": 1.6434304728197109e-06, + "loss": 0.2809, + "step": 5339 + }, + { + "epoch": 0.74, + "grad_norm": 3.5825469851757252, + "learning_rate": 1.6417621417093633e-06, + "loss": 0.3121, + "step": 5340 + }, + { + "epoch": 0.74, + "grad_norm": 3.486472201632393, + "learning_rate": 1.6400944914839834e-06, + "loss": 0.289, + "step": 5341 + }, + { + "epoch": 0.74, + "grad_norm": 2.4916909954408855, + "learning_rate": 1.638427522481687e-06, + "loss": 0.3026, + "step": 5342 + }, + { + "epoch": 0.74, + "grad_norm": 2.0610897359006515, + "learning_rate": 1.6367612350404522e-06, + "loss": 0.3108, + "step": 5343 + }, + { + "epoch": 0.74, + "grad_norm": 2.540818440010136, + "learning_rate": 1.6350956294981201e-06, + "loss": 0.2869, + "step": 5344 + }, + { + "epoch": 0.74, + "grad_norm": 2.1850260493475373, + "learning_rate": 1.633430706192397e-06, + "loss": 0.3081, + "step": 5345 + }, + { + "epoch": 0.74, + "grad_norm": 3.234437903197098, + "learning_rate": 1.631766465460844e-06, + "loss": 0.3172, + "step": 5346 + }, + { + "epoch": 0.74, + "grad_norm": 2.4253101356658853, + "learning_rate": 1.6301029076408858e-06, + "loss": 0.2874, + "step": 5347 + }, + { + "epoch": 0.74, + "grad_norm": 2.390386385830457, + "learning_rate": 1.628440033069813e-06, + "loss": 0.326, + "step": 5348 + }, + { + "epoch": 0.74, + "grad_norm": 1.9846208890855035, + "learning_rate": 1.6267778420847735e-06, + "loss": 0.2876, + "step": 5349 + }, + { + "epoch": 0.74, + "grad_norm": 2.0163541258466933, + "learning_rate": 1.6251163350227756e-06, + "loss": 0.3059, + "step": 5350 + }, + { + "epoch": 0.74, + "grad_norm": 2.2593885064965, + "learning_rate": 1.623455512220694e-06, + "loss": 0.2898, + "step": 5351 + }, + { + "epoch": 0.74, + "grad_norm": 3.380667608700582, + "learning_rate": 1.6217953740152587e-06, + "loss": 0.3129, + "step": 5352 + }, + { + "epoch": 0.74, + "grad_norm": 3.1956481874695277, + "learning_rate": 1.6201359207430667e-06, + "loss": 0.2887, + "step": 5353 + }, + { + "epoch": 0.74, + "grad_norm": 2.090563177070576, + "learning_rate": 1.6184771527405706e-06, + "loss": 0.3152, + "step": 5354 + }, + { + "epoch": 0.74, + "grad_norm": 2.3339163450034057, + "learning_rate": 1.6168190703440894e-06, + "loss": 0.3046, + "step": 5355 + }, + { + "epoch": 0.74, + "grad_norm": 2.3676439015848003, + "learning_rate": 1.615161673889799e-06, + "loss": 0.3206, + "step": 5356 + }, + { + "epoch": 0.74, + "grad_norm": 2.206755125899246, + "learning_rate": 1.6135049637137362e-06, + "loss": 0.3011, + "step": 5357 + }, + { + "epoch": 0.74, + "grad_norm": 5.161337420935462, + "learning_rate": 1.6118489401518034e-06, + "loss": 0.331, + "step": 5358 + }, + { + "epoch": 0.74, + "grad_norm": 2.297995562532317, + "learning_rate": 1.6101936035397587e-06, + "loss": 0.3236, + "step": 5359 + }, + { + "epoch": 0.75, + "grad_norm": 2.2123836096727367, + "learning_rate": 1.6085389542132206e-06, + "loss": 0.2988, + "step": 5360 + }, + { + "epoch": 0.75, + "grad_norm": 2.7409044568678387, + "learning_rate": 1.6068849925076724e-06, + "loss": 0.3015, + "step": 5361 + }, + { + "epoch": 0.75, + "grad_norm": 1.8787191333935018, + "learning_rate": 1.6052317187584576e-06, + "loss": 0.2867, + "step": 5362 + }, + { + "epoch": 0.75, + "grad_norm": 1.9456056398029737, + "learning_rate": 1.6035791333007772e-06, + "loss": 0.2986, + "step": 5363 + }, + { + "epoch": 0.75, + "grad_norm": 6.7002532750342345, + "learning_rate": 1.601927236469692e-06, + "loss": 0.3132, + "step": 5364 + }, + { + "epoch": 0.75, + "grad_norm": 2.17082483092028, + "learning_rate": 1.6002760286001284e-06, + "loss": 0.2918, + "step": 5365 + }, + { + "epoch": 0.75, + "grad_norm": 2.5138874869103613, + "learning_rate": 1.5986255100268683e-06, + "loss": 0.3013, + "step": 5366 + }, + { + "epoch": 0.75, + "grad_norm": 1.945041250340327, + "learning_rate": 1.5969756810845543e-06, + "loss": 0.2977, + "step": 5367 + }, + { + "epoch": 0.75, + "grad_norm": 2.296678030845658, + "learning_rate": 1.595326542107693e-06, + "loss": 0.2995, + "step": 5368 + }, + { + "epoch": 0.75, + "grad_norm": 1.9886586059161182, + "learning_rate": 1.593678093430645e-06, + "loss": 0.2947, + "step": 5369 + }, + { + "epoch": 0.75, + "grad_norm": 1.9636007027238664, + "learning_rate": 1.5920303353876383e-06, + "loss": 0.3012, + "step": 5370 + }, + { + "epoch": 0.75, + "grad_norm": 2.539415870004738, + "learning_rate": 1.5903832683127535e-06, + "loss": 0.3307, + "step": 5371 + }, + { + "epoch": 0.75, + "grad_norm": 2.08705525961069, + "learning_rate": 1.588736892539937e-06, + "loss": 0.3128, + "step": 5372 + }, + { + "epoch": 0.75, + "grad_norm": 2.5970706971316373, + "learning_rate": 1.5870912084029921e-06, + "loss": 0.3094, + "step": 5373 + }, + { + "epoch": 0.75, + "grad_norm": 2.4468177785262264, + "learning_rate": 1.5854462162355804e-06, + "loss": 0.3221, + "step": 5374 + }, + { + "epoch": 0.75, + "grad_norm": 2.4397718453295596, + "learning_rate": 1.5838019163712282e-06, + "loss": 0.3081, + "step": 5375 + }, + { + "epoch": 0.75, + "grad_norm": 2.4991945579541914, + "learning_rate": 1.5821583091433168e-06, + "loss": 0.2857, + "step": 5376 + }, + { + "epoch": 0.75, + "grad_norm": 2.2880952914283235, + "learning_rate": 1.5805153948850866e-06, + "loss": 0.3145, + "step": 5377 + }, + { + "epoch": 0.75, + "grad_norm": 2.5320040520100053, + "learning_rate": 1.578873173929642e-06, + "loss": 0.3155, + "step": 5378 + }, + { + "epoch": 0.75, + "grad_norm": 2.3210469310861512, + "learning_rate": 1.5772316466099458e-06, + "loss": 0.3029, + "step": 5379 + }, + { + "epoch": 0.75, + "grad_norm": 1.9128195614458596, + "learning_rate": 1.575590813258817e-06, + "loss": 0.2819, + "step": 5380 + }, + { + "epoch": 0.75, + "grad_norm": 1.9150578940034684, + "learning_rate": 1.573950674208934e-06, + "loss": 0.3143, + "step": 5381 + }, + { + "epoch": 0.75, + "grad_norm": 2.0457370549410876, + "learning_rate": 1.5723112297928395e-06, + "loss": 0.3055, + "step": 5382 + }, + { + "epoch": 0.75, + "grad_norm": 2.261923817893563, + "learning_rate": 1.5706724803429297e-06, + "loss": 0.3156, + "step": 5383 + }, + { + "epoch": 0.75, + "grad_norm": 2.2077005714004314, + "learning_rate": 1.5690344261914615e-06, + "loss": 0.3031, + "step": 5384 + }, + { + "epoch": 0.75, + "grad_norm": 6.920054255922523, + "learning_rate": 1.5673970676705541e-06, + "loss": 0.3129, + "step": 5385 + }, + { + "epoch": 0.75, + "grad_norm": 3.228470187048207, + "learning_rate": 1.5657604051121816e-06, + "loss": 0.2937, + "step": 5386 + }, + { + "epoch": 0.75, + "grad_norm": 2.318408809483326, + "learning_rate": 1.564124438848177e-06, + "loss": 0.2884, + "step": 5387 + }, + { + "epoch": 0.75, + "grad_norm": 2.143828998009743, + "learning_rate": 1.5624891692102345e-06, + "loss": 0.3041, + "step": 5388 + }, + { + "epoch": 0.75, + "grad_norm": 2.3295979339958484, + "learning_rate": 1.5608545965299083e-06, + "loss": 0.2755, + "step": 5389 + }, + { + "epoch": 0.75, + "grad_norm": 1.903368439334569, + "learning_rate": 1.5592207211386073e-06, + "loss": 0.3177, + "step": 5390 + }, + { + "epoch": 0.75, + "grad_norm": 3.215966580745813, + "learning_rate": 1.5575875433675991e-06, + "loss": 0.315, + "step": 5391 + }, + { + "epoch": 0.75, + "grad_norm": 4.260668666645318, + "learning_rate": 1.5559550635480152e-06, + "loss": 0.3024, + "step": 5392 + }, + { + "epoch": 0.75, + "grad_norm": 2.0847447824704437, + "learning_rate": 1.55432328201084e-06, + "loss": 0.3081, + "step": 5393 + }, + { + "epoch": 0.75, + "grad_norm": 4.542280759782605, + "learning_rate": 1.5526921990869187e-06, + "loss": 0.2902, + "step": 5394 + }, + { + "epoch": 0.75, + "grad_norm": 2.1622698383343386, + "learning_rate": 1.5510618151069505e-06, + "loss": 0.2945, + "step": 5395 + }, + { + "epoch": 0.75, + "grad_norm": 2.1260707346084984, + "learning_rate": 1.5494321304015042e-06, + "loss": 0.3104, + "step": 5396 + }, + { + "epoch": 0.75, + "grad_norm": 2.3016525271694985, + "learning_rate": 1.5478031453009956e-06, + "loss": 0.3183, + "step": 5397 + }, + { + "epoch": 0.75, + "grad_norm": 2.2987115457317104, + "learning_rate": 1.5461748601357008e-06, + "loss": 0.3102, + "step": 5398 + }, + { + "epoch": 0.75, + "grad_norm": 2.079247750946993, + "learning_rate": 1.5445472752357594e-06, + "loss": 0.2947, + "step": 5399 + }, + { + "epoch": 0.75, + "grad_norm": 1.8581891676807543, + "learning_rate": 1.5429203909311636e-06, + "loss": 0.2754, + "step": 5400 + }, + { + "epoch": 0.75, + "grad_norm": 1.9845964607421032, + "learning_rate": 1.5412942075517651e-06, + "loss": 0.2838, + "step": 5401 + }, + { + "epoch": 0.75, + "grad_norm": 2.1956832472814733, + "learning_rate": 1.5396687254272724e-06, + "loss": 0.3214, + "step": 5402 + }, + { + "epoch": 0.75, + "grad_norm": 2.258731512663056, + "learning_rate": 1.5380439448872559e-06, + "loss": 0.3352, + "step": 5403 + }, + { + "epoch": 0.75, + "grad_norm": 2.518786568326642, + "learning_rate": 1.5364198662611374e-06, + "loss": 0.3438, + "step": 5404 + }, + { + "epoch": 0.75, + "grad_norm": 2.7610480291316866, + "learning_rate": 1.5347964898782042e-06, + "loss": 0.3174, + "step": 5405 + }, + { + "epoch": 0.75, + "grad_norm": 1.6252554057164228, + "learning_rate": 1.5331738160675929e-06, + "loss": 0.3051, + "step": 5406 + }, + { + "epoch": 0.75, + "grad_norm": 2.2514736975021994, + "learning_rate": 1.531551845158305e-06, + "loss": 0.3026, + "step": 5407 + }, + { + "epoch": 0.75, + "grad_norm": 2.7351717766572987, + "learning_rate": 1.5299305774791951e-06, + "loss": 0.3194, + "step": 5408 + }, + { + "epoch": 0.75, + "grad_norm": 2.7955195085536744, + "learning_rate": 1.5283100133589745e-06, + "loss": 0.297, + "step": 5409 + }, + { + "epoch": 0.75, + "grad_norm": 6.959946056668789, + "learning_rate": 1.5266901531262163e-06, + "loss": 0.3022, + "step": 5410 + }, + { + "epoch": 0.75, + "grad_norm": 1.967840215326755, + "learning_rate": 1.5250709971093468e-06, + "loss": 0.3022, + "step": 5411 + }, + { + "epoch": 0.75, + "grad_norm": 2.4730392791386855, + "learning_rate": 1.52345254563665e-06, + "loss": 0.3158, + "step": 5412 + }, + { + "epoch": 0.75, + "grad_norm": 2.199887266854136, + "learning_rate": 1.5218347990362692e-06, + "loss": 0.3119, + "step": 5413 + }, + { + "epoch": 0.75, + "grad_norm": 2.8538606786287226, + "learning_rate": 1.5202177576362054e-06, + "loss": 0.3173, + "step": 5414 + }, + { + "epoch": 0.75, + "grad_norm": 2.1082970988043863, + "learning_rate": 1.5186014217643125e-06, + "loss": 0.3105, + "step": 5415 + }, + { + "epoch": 0.75, + "grad_norm": 2.183888712259285, + "learning_rate": 1.5169857917483026e-06, + "loss": 0.3154, + "step": 5416 + }, + { + "epoch": 0.75, + "grad_norm": 2.80430809823187, + "learning_rate": 1.5153708679157498e-06, + "loss": 0.3084, + "step": 5417 + }, + { + "epoch": 0.75, + "grad_norm": 3.11388063568699, + "learning_rate": 1.513756650594078e-06, + "loss": 0.2856, + "step": 5418 + }, + { + "epoch": 0.75, + "grad_norm": 2.668338779670562, + "learning_rate": 1.5121431401105697e-06, + "loss": 0.2824, + "step": 5419 + }, + { + "epoch": 0.75, + "grad_norm": 2.017589913808478, + "learning_rate": 1.5105303367923685e-06, + "loss": 0.2955, + "step": 5420 + }, + { + "epoch": 0.75, + "grad_norm": 2.5864705778756196, + "learning_rate": 1.5089182409664688e-06, + "loss": 0.2792, + "step": 5421 + }, + { + "epoch": 0.75, + "grad_norm": 1.9951012756995068, + "learning_rate": 1.5073068529597263e-06, + "loss": 0.3235, + "step": 5422 + }, + { + "epoch": 0.75, + "grad_norm": 8.851614117154428, + "learning_rate": 1.505696173098848e-06, + "loss": 0.3026, + "step": 5423 + }, + { + "epoch": 0.75, + "grad_norm": 5.390968185306559, + "learning_rate": 1.5040862017104035e-06, + "loss": 0.3028, + "step": 5424 + }, + { + "epoch": 0.75, + "grad_norm": 2.5485300306965666, + "learning_rate": 1.502476939120815e-06, + "loss": 0.3196, + "step": 5425 + }, + { + "epoch": 0.75, + "grad_norm": 1.9396914649803465, + "learning_rate": 1.5008683856563583e-06, + "loss": 0.2954, + "step": 5426 + }, + { + "epoch": 0.75, + "grad_norm": 2.7785188158240723, + "learning_rate": 1.4992605416431732e-06, + "loss": 0.2888, + "step": 5427 + }, + { + "epoch": 0.75, + "grad_norm": 2.5574943090741513, + "learning_rate": 1.4976534074072485e-06, + "loss": 0.2976, + "step": 5428 + }, + { + "epoch": 0.75, + "grad_norm": 2.990949691971929, + "learning_rate": 1.4960469832744313e-06, + "loss": 0.2883, + "step": 5429 + }, + { + "epoch": 0.75, + "grad_norm": 2.5976316199649454, + "learning_rate": 1.4944412695704264e-06, + "loss": 0.3255, + "step": 5430 + }, + { + "epoch": 0.75, + "grad_norm": 3.270545437974554, + "learning_rate": 1.492836266620794e-06, + "loss": 0.2861, + "step": 5431 + }, + { + "epoch": 0.76, + "grad_norm": 2.006572045312078, + "learning_rate": 1.491231974750949e-06, + "loss": 0.3065, + "step": 5432 + }, + { + "epoch": 0.76, + "grad_norm": 2.595870931172905, + "learning_rate": 1.4896283942861605e-06, + "loss": 0.318, + "step": 5433 + }, + { + "epoch": 0.76, + "grad_norm": 2.0300873594111772, + "learning_rate": 1.4880255255515592e-06, + "loss": 0.2992, + "step": 5434 + }, + { + "epoch": 0.76, + "grad_norm": 1.9721679509611094, + "learning_rate": 1.4864233688721252e-06, + "loss": 0.2862, + "step": 5435 + }, + { + "epoch": 0.76, + "grad_norm": 2.1919653369267222, + "learning_rate": 1.4848219245726964e-06, + "loss": 0.3041, + "step": 5436 + }, + { + "epoch": 0.76, + "grad_norm": 2.3927608721872597, + "learning_rate": 1.4832211929779695e-06, + "loss": 0.3079, + "step": 5437 + }, + { + "epoch": 0.76, + "grad_norm": 1.991840867582317, + "learning_rate": 1.4816211744124903e-06, + "loss": 0.307, + "step": 5438 + }, + { + "epoch": 0.76, + "grad_norm": 3.472387542879791, + "learning_rate": 1.4800218692006674e-06, + "loss": 0.303, + "step": 5439 + }, + { + "epoch": 0.76, + "grad_norm": 2.3423540520526966, + "learning_rate": 1.4784232776667574e-06, + "loss": 0.3181, + "step": 5440 + }, + { + "epoch": 0.76, + "grad_norm": 1.9931948474318837, + "learning_rate": 1.4768254001348792e-06, + "loss": 0.3011, + "step": 5441 + }, + { + "epoch": 0.76, + "grad_norm": 2.197456934102365, + "learning_rate": 1.4752282369290016e-06, + "loss": 0.3108, + "step": 5442 + }, + { + "epoch": 0.76, + "grad_norm": 4.857007594352663, + "learning_rate": 1.4736317883729495e-06, + "loss": 0.3037, + "step": 5443 + }, + { + "epoch": 0.76, + "grad_norm": 2.1440240135598096, + "learning_rate": 1.4720360547904066e-06, + "loss": 0.3155, + "step": 5444 + }, + { + "epoch": 0.76, + "grad_norm": 3.15822652659379, + "learning_rate": 1.4704410365049076e-06, + "loss": 0.2807, + "step": 5445 + }, + { + "epoch": 0.76, + "grad_norm": 2.625663684804305, + "learning_rate": 1.468846733839842e-06, + "loss": 0.3268, + "step": 5446 + }, + { + "epoch": 0.76, + "grad_norm": 8.17736591014698, + "learning_rate": 1.467253147118457e-06, + "loss": 0.2985, + "step": 5447 + }, + { + "epoch": 0.76, + "grad_norm": 2.239522820174588, + "learning_rate": 1.4656602766638556e-06, + "loss": 0.2893, + "step": 5448 + }, + { + "epoch": 0.76, + "grad_norm": 2.0403821254318535, + "learning_rate": 1.4640681227989906e-06, + "loss": 0.2963, + "step": 5449 + }, + { + "epoch": 0.76, + "grad_norm": 3.4265020681493557, + "learning_rate": 1.4624766858466715e-06, + "loss": 0.3263, + "step": 5450 + }, + { + "epoch": 0.76, + "grad_norm": 3.6892190566084855, + "learning_rate": 1.4608859661295672e-06, + "loss": 0.2952, + "step": 5451 + }, + { + "epoch": 0.76, + "grad_norm": 2.4683037429322385, + "learning_rate": 1.4592959639701937e-06, + "loss": 0.2948, + "step": 5452 + }, + { + "epoch": 0.76, + "grad_norm": 2.2582004414666397, + "learning_rate": 1.4577066796909246e-06, + "loss": 0.3351, + "step": 5453 + }, + { + "epoch": 0.76, + "grad_norm": 2.2711556155643433, + "learning_rate": 1.4561181136139912e-06, + "loss": 0.3417, + "step": 5454 + }, + { + "epoch": 0.76, + "grad_norm": 2.853611447141902, + "learning_rate": 1.4545302660614752e-06, + "loss": 0.2946, + "step": 5455 + }, + { + "epoch": 0.76, + "grad_norm": 2.5777451278761085, + "learning_rate": 1.4529431373553115e-06, + "loss": 0.3022, + "step": 5456 + }, + { + "epoch": 0.76, + "grad_norm": 2.1611196143622564, + "learning_rate": 1.4513567278172934e-06, + "loss": 0.3208, + "step": 5457 + }, + { + "epoch": 0.76, + "grad_norm": 2.398898350032639, + "learning_rate": 1.4497710377690682e-06, + "loss": 0.2707, + "step": 5458 + }, + { + "epoch": 0.76, + "grad_norm": 3.0386414665858306, + "learning_rate": 1.4481860675321335e-06, + "loss": 0.3117, + "step": 5459 + }, + { + "epoch": 0.76, + "grad_norm": 2.3698607716550932, + "learning_rate": 1.4466018174278422e-06, + "loss": 0.3314, + "step": 5460 + }, + { + "epoch": 0.76, + "grad_norm": 2.780445229533878, + "learning_rate": 1.445018287777405e-06, + "loss": 0.3007, + "step": 5461 + }, + { + "epoch": 0.76, + "grad_norm": 3.596151441484516, + "learning_rate": 1.4434354789018817e-06, + "loss": 0.3123, + "step": 5462 + }, + { + "epoch": 0.76, + "grad_norm": 2.8673512880115246, + "learning_rate": 1.4418533911221882e-06, + "loss": 0.3167, + "step": 5463 + }, + { + "epoch": 0.76, + "grad_norm": 2.9516518598895343, + "learning_rate": 1.4402720247590908e-06, + "loss": 0.2755, + "step": 5464 + }, + { + "epoch": 0.76, + "grad_norm": 3.2029239754445182, + "learning_rate": 1.4386913801332185e-06, + "loss": 0.3143, + "step": 5465 + }, + { + "epoch": 0.76, + "grad_norm": 2.404457426378597, + "learning_rate": 1.4371114575650452e-06, + "loss": 0.2819, + "step": 5466 + }, + { + "epoch": 0.76, + "grad_norm": 2.6766864909014023, + "learning_rate": 1.4355322573748992e-06, + "loss": 0.3029, + "step": 5467 + }, + { + "epoch": 0.76, + "grad_norm": 2.7775841469534575, + "learning_rate": 1.4339537798829684e-06, + "loss": 0.3266, + "step": 5468 + }, + { + "epoch": 0.76, + "grad_norm": 2.7004585449456533, + "learning_rate": 1.4323760254092878e-06, + "loss": 0.3379, + "step": 5469 + }, + { + "epoch": 0.76, + "grad_norm": 2.2134114809385004, + "learning_rate": 1.4307989942737482e-06, + "loss": 0.2807, + "step": 5470 + }, + { + "epoch": 0.76, + "grad_norm": 2.2644700194507186, + "learning_rate": 1.429222686796093e-06, + "loss": 0.3, + "step": 5471 + }, + { + "epoch": 0.76, + "grad_norm": 2.5391194190506385, + "learning_rate": 1.4276471032959215e-06, + "loss": 0.2745, + "step": 5472 + }, + { + "epoch": 0.76, + "grad_norm": 2.2120004656276424, + "learning_rate": 1.4260722440926822e-06, + "loss": 0.3026, + "step": 5473 + }, + { + "epoch": 0.76, + "grad_norm": 2.2122760231088034, + "learning_rate": 1.4244981095056798e-06, + "loss": 0.326, + "step": 5474 + }, + { + "epoch": 0.76, + "grad_norm": 2.6223932613208887, + "learning_rate": 1.4229246998540736e-06, + "loss": 0.3138, + "step": 5475 + }, + { + "epoch": 0.76, + "grad_norm": 2.316074230005895, + "learning_rate": 1.4213520154568705e-06, + "loss": 0.3024, + "step": 5476 + }, + { + "epoch": 0.76, + "grad_norm": 2.1943010309838664, + "learning_rate": 1.419780056632934e-06, + "loss": 0.3118, + "step": 5477 + }, + { + "epoch": 0.76, + "grad_norm": 2.232536303221125, + "learning_rate": 1.4182088237009788e-06, + "loss": 0.2855, + "step": 5478 + }, + { + "epoch": 0.76, + "grad_norm": 2.0938493658745827, + "learning_rate": 1.4166383169795756e-06, + "loss": 0.2934, + "step": 5479 + }, + { + "epoch": 0.76, + "grad_norm": 2.1638352260102542, + "learning_rate": 1.4150685367871448e-06, + "loss": 0.3013, + "step": 5480 + }, + { + "epoch": 0.76, + "grad_norm": 2.0258290956116474, + "learning_rate": 1.4134994834419585e-06, + "loss": 0.2887, + "step": 5481 + }, + { + "epoch": 0.76, + "grad_norm": 3.7379061572423864, + "learning_rate": 1.4119311572621447e-06, + "loss": 0.2869, + "step": 5482 + }, + { + "epoch": 0.76, + "grad_norm": 4.113207977834115, + "learning_rate": 1.4103635585656844e-06, + "loss": 0.3121, + "step": 5483 + }, + { + "epoch": 0.76, + "grad_norm": 2.280065736244988, + "learning_rate": 1.4087966876704079e-06, + "loss": 0.2979, + "step": 5484 + }, + { + "epoch": 0.76, + "grad_norm": 1.8071460010035993, + "learning_rate": 1.4072305448939976e-06, + "loss": 0.306, + "step": 5485 + }, + { + "epoch": 0.76, + "grad_norm": 2.1215865018136713, + "learning_rate": 1.405665130553993e-06, + "loss": 0.3179, + "step": 5486 + }, + { + "epoch": 0.76, + "grad_norm": 2.214801017099937, + "learning_rate": 1.4041004449677815e-06, + "loss": 0.2994, + "step": 5487 + }, + { + "epoch": 0.76, + "grad_norm": 2.3185392826030937, + "learning_rate": 1.4025364884526022e-06, + "loss": 0.2916, + "step": 5488 + }, + { + "epoch": 0.76, + "grad_norm": 2.650598469286189, + "learning_rate": 1.4009732613255521e-06, + "loss": 0.2997, + "step": 5489 + }, + { + "epoch": 0.76, + "grad_norm": 2.9832927064622776, + "learning_rate": 1.3994107639035725e-06, + "loss": 0.3302, + "step": 5490 + }, + { + "epoch": 0.76, + "grad_norm": 2.611191705641627, + "learning_rate": 1.3978489965034652e-06, + "loss": 0.3196, + "step": 5491 + }, + { + "epoch": 0.76, + "grad_norm": 2.1070077343160714, + "learning_rate": 1.3962879594418754e-06, + "loss": 0.2961, + "step": 5492 + }, + { + "epoch": 0.76, + "grad_norm": 2.829031362220844, + "learning_rate": 1.3947276530353082e-06, + "loss": 0.3227, + "step": 5493 + }, + { + "epoch": 0.76, + "grad_norm": 2.861629702647296, + "learning_rate": 1.3931680776001144e-06, + "loss": 0.2954, + "step": 5494 + }, + { + "epoch": 0.76, + "grad_norm": 1.9659128349783894, + "learning_rate": 1.3916092334524984e-06, + "loss": 0.3089, + "step": 5495 + }, + { + "epoch": 0.76, + "grad_norm": 2.1222508426419333, + "learning_rate": 1.3900511209085188e-06, + "loss": 0.3193, + "step": 5496 + }, + { + "epoch": 0.76, + "grad_norm": 1.9368091289332523, + "learning_rate": 1.3884937402840837e-06, + "loss": 0.3007, + "step": 5497 + }, + { + "epoch": 0.76, + "grad_norm": 2.1127905705157133, + "learning_rate": 1.3869370918949504e-06, + "loss": 0.2789, + "step": 5498 + }, + { + "epoch": 0.76, + "grad_norm": 3.4913302188000657, + "learning_rate": 1.3853811760567326e-06, + "loss": 0.3176, + "step": 5499 + }, + { + "epoch": 0.76, + "grad_norm": 2.3751217047038997, + "learning_rate": 1.383825993084895e-06, + "loss": 0.3169, + "step": 5500 + }, + { + "epoch": 0.76, + "grad_norm": 2.3867181547244307, + "learning_rate": 1.3822715432947498e-06, + "loss": 0.3003, + "step": 5501 + }, + { + "epoch": 0.76, + "grad_norm": 2.4217147157818926, + "learning_rate": 1.3807178270014615e-06, + "loss": 0.261, + "step": 5502 + }, + { + "epoch": 0.76, + "grad_norm": 2.185503562498058, + "learning_rate": 1.3791648445200507e-06, + "loss": 0.3187, + "step": 5503 + }, + { + "epoch": 0.77, + "grad_norm": 2.4680416306299513, + "learning_rate": 1.377612596165383e-06, + "loss": 0.2937, + "step": 5504 + }, + { + "epoch": 0.77, + "grad_norm": 2.3240781822194103, + "learning_rate": 1.3760610822521769e-06, + "loss": 0.3014, + "step": 5505 + }, + { + "epoch": 0.77, + "grad_norm": 2.5640734641667806, + "learning_rate": 1.3745103030950057e-06, + "loss": 0.3107, + "step": 5506 + }, + { + "epoch": 0.77, + "grad_norm": 2.5317140643141145, + "learning_rate": 1.3729602590082886e-06, + "loss": 0.3107, + "step": 5507 + }, + { + "epoch": 0.77, + "grad_norm": 2.852403567696254, + "learning_rate": 1.3714109503063e-06, + "loss": 0.2811, + "step": 5508 + }, + { + "epoch": 0.77, + "grad_norm": 1.999987343054693, + "learning_rate": 1.3698623773031611e-06, + "loss": 0.2883, + "step": 5509 + }, + { + "epoch": 0.77, + "grad_norm": 2.234348865061137, + "learning_rate": 1.368314540312849e-06, + "loss": 0.2799, + "step": 5510 + }, + { + "epoch": 0.77, + "grad_norm": 1.8033139073363396, + "learning_rate": 1.366767439649187e-06, + "loss": 0.2957, + "step": 5511 + }, + { + "epoch": 0.77, + "grad_norm": 2.8967322501248596, + "learning_rate": 1.365221075625849e-06, + "loss": 0.2784, + "step": 5512 + }, + { + "epoch": 0.77, + "grad_norm": 3.596181250054731, + "learning_rate": 1.3636754485563647e-06, + "loss": 0.3116, + "step": 5513 + }, + { + "epoch": 0.77, + "grad_norm": 2.1829512228225356, + "learning_rate": 1.3621305587541096e-06, + "loss": 0.2958, + "step": 5514 + }, + { + "epoch": 0.77, + "grad_norm": 2.273022258532922, + "learning_rate": 1.3605864065323094e-06, + "loss": 0.3117, + "step": 5515 + }, + { + "epoch": 0.77, + "grad_norm": 5.816542183308082, + "learning_rate": 1.3590429922040438e-06, + "loss": 0.307, + "step": 5516 + }, + { + "epoch": 0.77, + "grad_norm": 2.112483881161411, + "learning_rate": 1.357500316082242e-06, + "loss": 0.3092, + "step": 5517 + }, + { + "epoch": 0.77, + "grad_norm": 3.3227695532662387, + "learning_rate": 1.355958378479682e-06, + "loss": 0.2992, + "step": 5518 + }, + { + "epoch": 0.77, + "grad_norm": 1.9714836404714209, + "learning_rate": 1.35441717970899e-06, + "loss": 0.2735, + "step": 5519 + }, + { + "epoch": 0.77, + "grad_norm": 2.119132484291261, + "learning_rate": 1.3528767200826493e-06, + "loss": 0.3056, + "step": 5520 + }, + { + "epoch": 0.77, + "grad_norm": 4.629184264803152, + "learning_rate": 1.3513369999129867e-06, + "loss": 0.303, + "step": 5521 + }, + { + "epoch": 0.77, + "grad_norm": 3.240379741347689, + "learning_rate": 1.3497980195121808e-06, + "loss": 0.3415, + "step": 5522 + }, + { + "epoch": 0.77, + "grad_norm": 5.680243323020423, + "learning_rate": 1.3482597791922631e-06, + "loss": 0.3025, + "step": 5523 + }, + { + "epoch": 0.77, + "grad_norm": 5.1502903983800765, + "learning_rate": 1.3467222792651113e-06, + "loss": 0.291, + "step": 5524 + }, + { + "epoch": 0.77, + "grad_norm": 2.933403960349093, + "learning_rate": 1.345185520042454e-06, + "loss": 0.3155, + "step": 5525 + }, + { + "epoch": 0.77, + "grad_norm": 2.1679329146540205, + "learning_rate": 1.3436495018358708e-06, + "loss": 0.3071, + "step": 5526 + }, + { + "epoch": 0.77, + "grad_norm": 2.117800053687113, + "learning_rate": 1.3421142249567925e-06, + "loss": 0.312, + "step": 5527 + }, + { + "epoch": 0.77, + "grad_norm": 2.6446074796970045, + "learning_rate": 1.3405796897164957e-06, + "loss": 0.3126, + "step": 5528 + }, + { + "epoch": 0.77, + "grad_norm": 2.8599056909987817, + "learning_rate": 1.3390458964261072e-06, + "loss": 0.2841, + "step": 5529 + }, + { + "epoch": 0.77, + "grad_norm": 3.0970001230954196, + "learning_rate": 1.3375128453966069e-06, + "loss": 0.2897, + "step": 5530 + }, + { + "epoch": 0.77, + "grad_norm": 4.803972534239417, + "learning_rate": 1.3359805369388212e-06, + "loss": 0.2938, + "step": 5531 + }, + { + "epoch": 0.77, + "grad_norm": 2.029342802284213, + "learning_rate": 1.3344489713634262e-06, + "loss": 0.3024, + "step": 5532 + }, + { + "epoch": 0.77, + "grad_norm": 2.16923757627728, + "learning_rate": 1.332918148980945e-06, + "loss": 0.2931, + "step": 5533 + }, + { + "epoch": 0.77, + "grad_norm": 3.3160300619581022, + "learning_rate": 1.331388070101759e-06, + "loss": 0.2823, + "step": 5534 + }, + { + "epoch": 0.77, + "grad_norm": 2.3815492391755853, + "learning_rate": 1.3298587350360892e-06, + "loss": 0.3176, + "step": 5535 + }, + { + "epoch": 0.77, + "grad_norm": 6.013134685134067, + "learning_rate": 1.328330144094009e-06, + "loss": 0.2903, + "step": 5536 + }, + { + "epoch": 0.77, + "grad_norm": 2.1193761339844954, + "learning_rate": 1.3268022975854428e-06, + "loss": 0.2793, + "step": 5537 + }, + { + "epoch": 0.77, + "grad_norm": 2.915604697046141, + "learning_rate": 1.3252751958201614e-06, + "loss": 0.301, + "step": 5538 + }, + { + "epoch": 0.77, + "grad_norm": 2.241111810818919, + "learning_rate": 1.3237488391077864e-06, + "loss": 0.3093, + "step": 5539 + }, + { + "epoch": 0.77, + "grad_norm": 2.2791622676371186, + "learning_rate": 1.3222232277577852e-06, + "loss": 0.3177, + "step": 5540 + }, + { + "epoch": 0.77, + "grad_norm": 2.4358183794925465, + "learning_rate": 1.3206983620794806e-06, + "loss": 0.3158, + "step": 5541 + }, + { + "epoch": 0.77, + "grad_norm": 1.9489801176306367, + "learning_rate": 1.3191742423820357e-06, + "loss": 0.3208, + "step": 5542 + }, + { + "epoch": 0.77, + "grad_norm": 2.0121232156781166, + "learning_rate": 1.3176508689744699e-06, + "loss": 0.2765, + "step": 5543 + }, + { + "epoch": 0.77, + "grad_norm": 3.8897739554115383, + "learning_rate": 1.3161282421656485e-06, + "loss": 0.2967, + "step": 5544 + }, + { + "epoch": 0.77, + "grad_norm": 2.178716324129663, + "learning_rate": 1.314606362264284e-06, + "loss": 0.2671, + "step": 5545 + }, + { + "epoch": 0.77, + "grad_norm": 2.6038090498726, + "learning_rate": 1.3130852295789386e-06, + "loss": 0.2932, + "step": 5546 + }, + { + "epoch": 0.77, + "grad_norm": 3.298081095356524, + "learning_rate": 1.3115648444180218e-06, + "loss": 0.3033, + "step": 5547 + }, + { + "epoch": 0.77, + "grad_norm": 2.1779233369420457, + "learning_rate": 1.3100452070897945e-06, + "loss": 0.2977, + "step": 5548 + }, + { + "epoch": 0.77, + "grad_norm": 4.877102590924938, + "learning_rate": 1.3085263179023638e-06, + "loss": 0.2703, + "step": 5549 + }, + { + "epoch": 0.77, + "grad_norm": 3.798489856517364, + "learning_rate": 1.3070081771636839e-06, + "loss": 0.3032, + "step": 5550 + }, + { + "epoch": 0.77, + "grad_norm": 2.2147346289036625, + "learning_rate": 1.3054907851815602e-06, + "loss": 0.3058, + "step": 5551 + }, + { + "epoch": 0.77, + "grad_norm": 1.9552981882337526, + "learning_rate": 1.3039741422636466e-06, + "loss": 0.2847, + "step": 5552 + }, + { + "epoch": 0.77, + "grad_norm": 2.370638769195423, + "learning_rate": 1.302458248717441e-06, + "loss": 0.2888, + "step": 5553 + }, + { + "epoch": 0.77, + "grad_norm": 2.414806211729241, + "learning_rate": 1.3009431048502913e-06, + "loss": 0.289, + "step": 5554 + }, + { + "epoch": 0.77, + "grad_norm": 1.9636633313137402, + "learning_rate": 1.299428710969397e-06, + "loss": 0.3041, + "step": 5555 + }, + { + "epoch": 0.77, + "grad_norm": 2.6470207868863, + "learning_rate": 1.2979150673818003e-06, + "loss": 0.3091, + "step": 5556 + }, + { + "epoch": 0.77, + "grad_norm": 2.3181068592825893, + "learning_rate": 1.2964021743943928e-06, + "loss": 0.3198, + "step": 5557 + }, + { + "epoch": 0.77, + "grad_norm": 2.5032064672364998, + "learning_rate": 1.294890032313917e-06, + "loss": 0.3173, + "step": 5558 + }, + { + "epoch": 0.77, + "grad_norm": 2.3009012574845458, + "learning_rate": 1.2933786414469574e-06, + "loss": 0.3163, + "step": 5559 + }, + { + "epoch": 0.77, + "grad_norm": 4.075801946995896, + "learning_rate": 1.291868002099953e-06, + "loss": 0.2938, + "step": 5560 + }, + { + "epoch": 0.77, + "grad_norm": 9.30473088774277, + "learning_rate": 1.2903581145791838e-06, + "loss": 0.3233, + "step": 5561 + }, + { + "epoch": 0.77, + "grad_norm": 2.300957190232922, + "learning_rate": 1.2888489791907837e-06, + "loss": 0.2952, + "step": 5562 + }, + { + "epoch": 0.77, + "grad_norm": 2.4568759140146468, + "learning_rate": 1.2873405962407293e-06, + "loss": 0.307, + "step": 5563 + }, + { + "epoch": 0.77, + "grad_norm": 2.6745036823836768, + "learning_rate": 1.2858329660348446e-06, + "loss": 0.3086, + "step": 5564 + }, + { + "epoch": 0.77, + "grad_norm": 1.9874096535448869, + "learning_rate": 1.2843260888788057e-06, + "loss": 0.309, + "step": 5565 + }, + { + "epoch": 0.77, + "grad_norm": 2.5664445650146943, + "learning_rate": 1.2828199650781315e-06, + "loss": 0.3074, + "step": 5566 + }, + { + "epoch": 0.77, + "grad_norm": 3.1670472720191785, + "learning_rate": 1.2813145949381884e-06, + "loss": 0.304, + "step": 5567 + }, + { + "epoch": 0.77, + "grad_norm": 2.3235713601238825, + "learning_rate": 1.2798099787641916e-06, + "loss": 0.2796, + "step": 5568 + }, + { + "epoch": 0.77, + "grad_norm": 2.408492871203998, + "learning_rate": 1.278306116861206e-06, + "loss": 0.3037, + "step": 5569 + }, + { + "epoch": 0.77, + "grad_norm": 3.0058109685510717, + "learning_rate": 1.2768030095341383e-06, + "loss": 0.3085, + "step": 5570 + }, + { + "epoch": 0.77, + "grad_norm": 3.8231389528116546, + "learning_rate": 1.275300657087743e-06, + "loss": 0.2953, + "step": 5571 + }, + { + "epoch": 0.77, + "grad_norm": 3.0136452184018756, + "learning_rate": 1.2737990598266259e-06, + "loss": 0.331, + "step": 5572 + }, + { + "epoch": 0.77, + "grad_norm": 2.0151138084283775, + "learning_rate": 1.2722982180552363e-06, + "loss": 0.2864, + "step": 5573 + }, + { + "epoch": 0.77, + "grad_norm": 2.093365242224289, + "learning_rate": 1.2707981320778683e-06, + "loss": 0.3344, + "step": 5574 + }, + { + "epoch": 0.78, + "grad_norm": 2.179001398925431, + "learning_rate": 1.2692988021986686e-06, + "loss": 0.2914, + "step": 5575 + }, + { + "epoch": 0.78, + "grad_norm": 2.240555687074227, + "learning_rate": 1.2678002287216246e-06, + "loss": 0.3158, + "step": 5576 + }, + { + "epoch": 0.78, + "grad_norm": 2.1257742738717864, + "learning_rate": 1.2663024119505762e-06, + "loss": 0.318, + "step": 5577 + }, + { + "epoch": 0.78, + "grad_norm": 2.1221557893579757, + "learning_rate": 1.2648053521892033e-06, + "loss": 0.2901, + "step": 5578 + }, + { + "epoch": 0.78, + "grad_norm": 2.3824088276007043, + "learning_rate": 1.2633090497410388e-06, + "loss": 0.3301, + "step": 5579 + }, + { + "epoch": 0.78, + "grad_norm": 2.479028620515614, + "learning_rate": 1.2618135049094582e-06, + "loss": 0.2965, + "step": 5580 + }, + { + "epoch": 0.78, + "grad_norm": 2.3199711691725557, + "learning_rate": 1.2603187179976816e-06, + "loss": 0.3086, + "step": 5581 + }, + { + "epoch": 0.78, + "grad_norm": 2.033427574825756, + "learning_rate": 1.2588246893087818e-06, + "loss": 0.29, + "step": 5582 + }, + { + "epoch": 0.78, + "grad_norm": 2.587790010992217, + "learning_rate": 1.2573314191456727e-06, + "loss": 0.299, + "step": 5583 + }, + { + "epoch": 0.78, + "grad_norm": 2.0612128745898115, + "learning_rate": 1.2558389078111138e-06, + "loss": 0.3168, + "step": 5584 + }, + { + "epoch": 0.78, + "grad_norm": 1.720237301537445, + "learning_rate": 1.2543471556077146e-06, + "loss": 0.2777, + "step": 5585 + }, + { + "epoch": 0.78, + "grad_norm": 2.9612972503057517, + "learning_rate": 1.252856162837931e-06, + "loss": 0.3041, + "step": 5586 + }, + { + "epoch": 0.78, + "grad_norm": 2.797253697725745, + "learning_rate": 1.2513659298040598e-06, + "loss": 0.2966, + "step": 5587 + }, + { + "epoch": 0.78, + "grad_norm": 2.832373021537092, + "learning_rate": 1.249876456808246e-06, + "loss": 0.3075, + "step": 5588 + }, + { + "epoch": 0.78, + "grad_norm": 2.0458464431334074, + "learning_rate": 1.2483877441524843e-06, + "loss": 0.2887, + "step": 5589 + }, + { + "epoch": 0.78, + "grad_norm": 2.134699563036615, + "learning_rate": 1.2468997921386112e-06, + "loss": 0.3088, + "step": 5590 + }, + { + "epoch": 0.78, + "grad_norm": 1.8454199911702935, + "learning_rate": 1.2454126010683071e-06, + "loss": 0.2977, + "step": 5591 + }, + { + "epoch": 0.78, + "grad_norm": 2.687464252569353, + "learning_rate": 1.2439261712431055e-06, + "loss": 0.2957, + "step": 5592 + }, + { + "epoch": 0.78, + "grad_norm": 2.0999807827866754, + "learning_rate": 1.2424405029643783e-06, + "loss": 0.2945, + "step": 5593 + }, + { + "epoch": 0.78, + "grad_norm": 2.4601917887788844, + "learning_rate": 1.2409555965333448e-06, + "loss": 0.3096, + "step": 5594 + }, + { + "epoch": 0.78, + "grad_norm": 2.9462407653072766, + "learning_rate": 1.2394714522510726e-06, + "loss": 0.3188, + "step": 5595 + }, + { + "epoch": 0.78, + "grad_norm": 1.9740319588150559, + "learning_rate": 1.2379880704184738e-06, + "loss": 0.2997, + "step": 5596 + }, + { + "epoch": 0.78, + "grad_norm": 2.6061648085867897, + "learning_rate": 1.236505451336304e-06, + "loss": 0.3157, + "step": 5597 + }, + { + "epoch": 0.78, + "grad_norm": 1.9816781277446625, + "learning_rate": 1.2350235953051637e-06, + "loss": 0.2824, + "step": 5598 + }, + { + "epoch": 0.78, + "grad_norm": 3.0916335629258302, + "learning_rate": 1.2335425026255026e-06, + "loss": 0.33, + "step": 5599 + }, + { + "epoch": 0.78, + "grad_norm": 2.3167647765723784, + "learning_rate": 1.2320621735976124e-06, + "loss": 0.2803, + "step": 5600 + }, + { + "epoch": 0.78, + "grad_norm": 1.9633669008932386, + "learning_rate": 1.2305826085216306e-06, + "loss": 0.3069, + "step": 5601 + }, + { + "epoch": 0.78, + "grad_norm": 1.9817152710447477, + "learning_rate": 1.229103807697537e-06, + "loss": 0.3018, + "step": 5602 + }, + { + "epoch": 0.78, + "grad_norm": 3.798595469404543, + "learning_rate": 1.2276257714251655e-06, + "loss": 0.2973, + "step": 5603 + }, + { + "epoch": 0.78, + "grad_norm": 2.1405663034947713, + "learning_rate": 1.2261485000041856e-06, + "loss": 0.2995, + "step": 5604 + }, + { + "epoch": 0.78, + "grad_norm": 2.38211795510811, + "learning_rate": 1.224671993734114e-06, + "loss": 0.3077, + "step": 5605 + }, + { + "epoch": 0.78, + "grad_norm": 2.8369394521626186, + "learning_rate": 1.223196252914316e-06, + "loss": 0.3013, + "step": 5606 + }, + { + "epoch": 0.78, + "grad_norm": 2.107906615221095, + "learning_rate": 1.221721277843998e-06, + "loss": 0.3069, + "step": 5607 + }, + { + "epoch": 0.78, + "grad_norm": 2.8249859935071635, + "learning_rate": 1.2202470688222113e-06, + "loss": 0.3141, + "step": 5608 + }, + { + "epoch": 0.78, + "grad_norm": 2.705274006123685, + "learning_rate": 1.2187736261478523e-06, + "loss": 0.307, + "step": 5609 + }, + { + "epoch": 0.78, + "grad_norm": 1.9909530731765908, + "learning_rate": 1.2173009501196652e-06, + "loss": 0.3221, + "step": 5610 + }, + { + "epoch": 0.78, + "grad_norm": 3.058702252798658, + "learning_rate": 1.2158290410362328e-06, + "loss": 0.3301, + "step": 5611 + }, + { + "epoch": 0.78, + "grad_norm": 7.282425597815352, + "learning_rate": 1.214357899195987e-06, + "loss": 0.2883, + "step": 5612 + }, + { + "epoch": 0.78, + "grad_norm": 2.332312701429817, + "learning_rate": 1.2128875248972043e-06, + "loss": 0.2862, + "step": 5613 + }, + { + "epoch": 0.78, + "grad_norm": 2.137256097657883, + "learning_rate": 1.2114179184380032e-06, + "loss": 0.3056, + "step": 5614 + }, + { + "epoch": 0.78, + "grad_norm": 2.211286984249388, + "learning_rate": 1.2099490801163472e-06, + "loss": 0.3268, + "step": 5615 + }, + { + "epoch": 0.78, + "grad_norm": 1.9114335882188414, + "learning_rate": 1.208481010230042e-06, + "loss": 0.3019, + "step": 5616 + }, + { + "epoch": 0.78, + "grad_norm": 2.055234826613935, + "learning_rate": 1.2070137090767436e-06, + "loss": 0.2956, + "step": 5617 + }, + { + "epoch": 0.78, + "grad_norm": 1.9036806600699978, + "learning_rate": 1.2055471769539463e-06, + "loss": 0.2852, + "step": 5618 + }, + { + "epoch": 0.78, + "grad_norm": 3.012641643379516, + "learning_rate": 1.204081414158989e-06, + "loss": 0.3093, + "step": 5619 + }, + { + "epoch": 0.78, + "grad_norm": 1.960581474043275, + "learning_rate": 1.2026164209890583e-06, + "loss": 0.3028, + "step": 5620 + }, + { + "epoch": 0.78, + "grad_norm": 3.5253722101579403, + "learning_rate": 1.2011521977411828e-06, + "loss": 0.3118, + "step": 5621 + }, + { + "epoch": 0.78, + "grad_norm": 2.826727621003026, + "learning_rate": 1.199688744712234e-06, + "loss": 0.2858, + "step": 5622 + }, + { + "epoch": 0.78, + "grad_norm": 2.359941747655638, + "learning_rate": 1.1982260621989267e-06, + "loss": 0.3035, + "step": 5623 + }, + { + "epoch": 0.78, + "grad_norm": 2.336393519382015, + "learning_rate": 1.1967641504978234e-06, + "loss": 0.2954, + "step": 5624 + }, + { + "epoch": 0.78, + "grad_norm": 2.2273862992908495, + "learning_rate": 1.1953030099053258e-06, + "loss": 0.319, + "step": 5625 + }, + { + "epoch": 0.78, + "grad_norm": 2.28230292808202, + "learning_rate": 1.1938426407176806e-06, + "loss": 0.3108, + "step": 5626 + }, + { + "epoch": 0.78, + "grad_norm": 3.733997845170595, + "learning_rate": 1.1923830432309808e-06, + "loss": 0.2907, + "step": 5627 + }, + { + "epoch": 0.78, + "grad_norm": 2.2948963038359493, + "learning_rate": 1.1909242177411578e-06, + "loss": 0.3105, + "step": 5628 + }, + { + "epoch": 0.78, + "grad_norm": 2.393119746218262, + "learning_rate": 1.189466164543993e-06, + "loss": 0.3122, + "step": 5629 + }, + { + "epoch": 0.78, + "grad_norm": 1.965315312606066, + "learning_rate": 1.1880088839351039e-06, + "loss": 0.2942, + "step": 5630 + }, + { + "epoch": 0.78, + "grad_norm": 2.751537273894358, + "learning_rate": 1.186552376209959e-06, + "loss": 0.2849, + "step": 5631 + }, + { + "epoch": 0.78, + "grad_norm": 2.3061731902113967, + "learning_rate": 1.1850966416638643e-06, + "loss": 0.3101, + "step": 5632 + }, + { + "epoch": 0.78, + "grad_norm": 3.495749888602386, + "learning_rate": 1.1836416805919687e-06, + "loss": 0.2972, + "step": 5633 + }, + { + "epoch": 0.78, + "grad_norm": 1.9941807560340965, + "learning_rate": 1.1821874932892708e-06, + "loss": 0.3096, + "step": 5634 + }, + { + "epoch": 0.78, + "grad_norm": 1.945905334248179, + "learning_rate": 1.1807340800506061e-06, + "loss": 0.3089, + "step": 5635 + }, + { + "epoch": 0.78, + "grad_norm": 2.7815478426840152, + "learning_rate": 1.1792814411706532e-06, + "loss": 0.2804, + "step": 5636 + }, + { + "epoch": 0.78, + "grad_norm": 3.2888805244489263, + "learning_rate": 1.1778295769439374e-06, + "loss": 0.3151, + "step": 5637 + }, + { + "epoch": 0.78, + "grad_norm": 2.105837519844848, + "learning_rate": 1.1763784876648271e-06, + "loss": 0.2877, + "step": 5638 + }, + { + "epoch": 0.78, + "grad_norm": 1.787941225194723, + "learning_rate": 1.1749281736275292e-06, + "loss": 0.3127, + "step": 5639 + }, + { + "epoch": 0.78, + "grad_norm": 2.1528549851464924, + "learning_rate": 1.1734786351260946e-06, + "loss": 0.2972, + "step": 5640 + }, + { + "epoch": 0.78, + "grad_norm": 1.85770368258291, + "learning_rate": 1.1720298724544216e-06, + "loss": 0.2913, + "step": 5641 + }, + { + "epoch": 0.78, + "grad_norm": 2.4260058467118717, + "learning_rate": 1.1705818859062462e-06, + "loss": 0.301, + "step": 5642 + }, + { + "epoch": 0.78, + "grad_norm": 1.9251686246612334, + "learning_rate": 1.169134675775147e-06, + "loss": 0.3016, + "step": 5643 + }, + { + "epoch": 0.78, + "grad_norm": 2.142860844805225, + "learning_rate": 1.1676882423545493e-06, + "loss": 0.3017, + "step": 5644 + }, + { + "epoch": 0.78, + "grad_norm": 3.416840570627123, + "learning_rate": 1.1662425859377169e-06, + "loss": 0.3015, + "step": 5645 + }, + { + "epoch": 0.78, + "grad_norm": 2.362663748547081, + "learning_rate": 1.164797706817759e-06, + "loss": 0.3217, + "step": 5646 + }, + { + "epoch": 0.79, + "grad_norm": 2.4477781566253003, + "learning_rate": 1.1633536052876237e-06, + "loss": 0.2801, + "step": 5647 + }, + { + "epoch": 0.79, + "grad_norm": 11.879188758022915, + "learning_rate": 1.1619102816401062e-06, + "loss": 0.3166, + "step": 5648 + }, + { + "epoch": 0.79, + "grad_norm": 2.232757060726138, + "learning_rate": 1.1604677361678402e-06, + "loss": 0.2954, + "step": 5649 + }, + { + "epoch": 0.79, + "grad_norm": 1.8558496600595802, + "learning_rate": 1.159025969163301e-06, + "loss": 0.3035, + "step": 5650 + }, + { + "epoch": 0.79, + "grad_norm": 2.5490088530206805, + "learning_rate": 1.1575849809188106e-06, + "loss": 0.2947, + "step": 5651 + }, + { + "epoch": 0.79, + "grad_norm": 2.111768381140235, + "learning_rate": 1.1561447717265295e-06, + "loss": 0.3222, + "step": 5652 + }, + { + "epoch": 0.79, + "grad_norm": 1.7926351496418547, + "learning_rate": 1.1547053418784586e-06, + "loss": 0.2933, + "step": 5653 + }, + { + "epoch": 0.79, + "grad_norm": 2.222695437262679, + "learning_rate": 1.153266691666446e-06, + "loss": 0.307, + "step": 5654 + }, + { + "epoch": 0.79, + "grad_norm": 2.3412355949270633, + "learning_rate": 1.1518288213821793e-06, + "loss": 0.275, + "step": 5655 + }, + { + "epoch": 0.79, + "grad_norm": 2.1665727333748923, + "learning_rate": 1.150391731317187e-06, + "loss": 0.3167, + "step": 5656 + }, + { + "epoch": 0.79, + "grad_norm": 4.172727047252683, + "learning_rate": 1.148955421762838e-06, + "loss": 0.3039, + "step": 5657 + }, + { + "epoch": 0.79, + "grad_norm": 2.253323249506165, + "learning_rate": 1.147519893010348e-06, + "loss": 0.2989, + "step": 5658 + }, + { + "epoch": 0.79, + "grad_norm": 2.188859870781199, + "learning_rate": 1.1460851453507705e-06, + "loss": 0.2913, + "step": 5659 + }, + { + "epoch": 0.79, + "grad_norm": 2.855802709407244, + "learning_rate": 1.1446511790749997e-06, + "loss": 0.3059, + "step": 5660 + }, + { + "epoch": 0.79, + "grad_norm": 1.9937912870634948, + "learning_rate": 1.1432179944737764e-06, + "loss": 0.2814, + "step": 5661 + }, + { + "epoch": 0.79, + "grad_norm": 2.4137909920923937, + "learning_rate": 1.141785591837678e-06, + "loss": 0.3328, + "step": 5662 + }, + { + "epoch": 0.79, + "grad_norm": 2.307043860172591, + "learning_rate": 1.1403539714571233e-06, + "loss": 0.3056, + "step": 5663 + }, + { + "epoch": 0.79, + "grad_norm": 2.2263145019541613, + "learning_rate": 1.1389231336223767e-06, + "loss": 0.2931, + "step": 5664 + }, + { + "epoch": 0.79, + "grad_norm": 2.106667546599523, + "learning_rate": 1.1374930786235416e-06, + "loss": 0.3075, + "step": 5665 + }, + { + "epoch": 0.79, + "grad_norm": 2.625575048124375, + "learning_rate": 1.1360638067505631e-06, + "loss": 0.3132, + "step": 5666 + }, + { + "epoch": 0.79, + "grad_norm": 2.8928361213731506, + "learning_rate": 1.1346353182932234e-06, + "loss": 0.3112, + "step": 5667 + }, + { + "epoch": 0.79, + "grad_norm": 2.214391130014123, + "learning_rate": 1.133207613541154e-06, + "loss": 0.2742, + "step": 5668 + }, + { + "epoch": 0.79, + "grad_norm": 2.290433476463271, + "learning_rate": 1.131780692783821e-06, + "loss": 0.285, + "step": 5669 + }, + { + "epoch": 0.79, + "grad_norm": 2.2583546441819315, + "learning_rate": 1.1303545563105334e-06, + "loss": 0.287, + "step": 5670 + }, + { + "epoch": 0.79, + "grad_norm": 3.3045473724795045, + "learning_rate": 1.1289292044104382e-06, + "loss": 0.3002, + "step": 5671 + }, + { + "epoch": 0.79, + "grad_norm": 3.8108365523438943, + "learning_rate": 1.127504637372533e-06, + "loss": 0.3011, + "step": 5672 + }, + { + "epoch": 0.79, + "grad_norm": 2.535307758690555, + "learning_rate": 1.1260808554856461e-06, + "loss": 0.2972, + "step": 5673 + }, + { + "epoch": 0.79, + "grad_norm": 2.255743418828071, + "learning_rate": 1.1246578590384493e-06, + "loss": 0.3035, + "step": 5674 + }, + { + "epoch": 0.79, + "grad_norm": 2.8209186532184947, + "learning_rate": 1.123235648319459e-06, + "loss": 0.3073, + "step": 5675 + }, + { + "epoch": 0.79, + "grad_norm": 2.0292650364343205, + "learning_rate": 1.1218142236170275e-06, + "loss": 0.2936, + "step": 5676 + }, + { + "epoch": 0.79, + "grad_norm": 3.1674389936397063, + "learning_rate": 1.1203935852193498e-06, + "loss": 0.3098, + "step": 5677 + }, + { + "epoch": 0.79, + "grad_norm": 2.3109318446083567, + "learning_rate": 1.1189737334144597e-06, + "loss": 0.2999, + "step": 5678 + }, + { + "epoch": 0.79, + "grad_norm": 2.8204036296357913, + "learning_rate": 1.117554668490236e-06, + "loss": 0.3109, + "step": 5679 + }, + { + "epoch": 0.79, + "grad_norm": 4.42260035442312, + "learning_rate": 1.1161363907343925e-06, + "loss": 0.3085, + "step": 5680 + }, + { + "epoch": 0.79, + "grad_norm": 2.1113786670936703, + "learning_rate": 1.114718900434487e-06, + "loss": 0.2983, + "step": 5681 + }, + { + "epoch": 0.79, + "grad_norm": 2.9225306969036144, + "learning_rate": 1.1133021978779185e-06, + "loss": 0.3171, + "step": 5682 + }, + { + "epoch": 0.79, + "grad_norm": 2.2054853392269145, + "learning_rate": 1.111886283351923e-06, + "loss": 0.3104, + "step": 5683 + }, + { + "epoch": 0.79, + "grad_norm": 2.0250861081188276, + "learning_rate": 1.1104711571435766e-06, + "loss": 0.2896, + "step": 5684 + }, + { + "epoch": 0.79, + "grad_norm": 2.3829365648677236, + "learning_rate": 1.1090568195397976e-06, + "loss": 0.3139, + "step": 5685 + }, + { + "epoch": 0.79, + "grad_norm": 3.152736167617577, + "learning_rate": 1.1076432708273467e-06, + "loss": 0.3201, + "step": 5686 + }, + { + "epoch": 0.79, + "grad_norm": 5.602430137042414, + "learning_rate": 1.1062305112928184e-06, + "loss": 0.2953, + "step": 5687 + }, + { + "epoch": 0.79, + "grad_norm": 6.362977860485569, + "learning_rate": 1.1048185412226515e-06, + "loss": 0.3209, + "step": 5688 + }, + { + "epoch": 0.79, + "grad_norm": 2.0099455017184846, + "learning_rate": 1.1034073609031242e-06, + "loss": 0.2994, + "step": 5689 + }, + { + "epoch": 0.79, + "grad_norm": 2.486192473185046, + "learning_rate": 1.1019969706203553e-06, + "loss": 0.3015, + "step": 5690 + }, + { + "epoch": 0.79, + "grad_norm": 2.770691809773771, + "learning_rate": 1.100587370660302e-06, + "loss": 0.3069, + "step": 5691 + }, + { + "epoch": 0.79, + "grad_norm": 2.049776144748334, + "learning_rate": 1.099178561308759e-06, + "loss": 0.2976, + "step": 5692 + }, + { + "epoch": 0.79, + "grad_norm": 2.807730309002486, + "learning_rate": 1.0977705428513674e-06, + "loss": 0.3016, + "step": 5693 + }, + { + "epoch": 0.79, + "grad_norm": 3.4183857390283467, + "learning_rate": 1.096363315573602e-06, + "loss": 0.3156, + "step": 5694 + }, + { + "epoch": 0.79, + "grad_norm": 7.1511353869301875, + "learning_rate": 1.0949568797607773e-06, + "loss": 0.299, + "step": 5695 + }, + { + "epoch": 0.79, + "grad_norm": 2.360876978706155, + "learning_rate": 1.0935512356980522e-06, + "loss": 0.3039, + "step": 5696 + }, + { + "epoch": 0.79, + "grad_norm": 6.120012368044757, + "learning_rate": 1.0921463836704193e-06, + "loss": 0.3245, + "step": 5697 + }, + { + "epoch": 0.79, + "grad_norm": 2.9630012688917637, + "learning_rate": 1.0907423239627157e-06, + "loss": 0.3267, + "step": 5698 + }, + { + "epoch": 0.79, + "grad_norm": 2.58204449735002, + "learning_rate": 1.089339056859613e-06, + "loss": 0.2879, + "step": 5699 + }, + { + "epoch": 0.79, + "grad_norm": 2.2507733598085227, + "learning_rate": 1.0879365826456273e-06, + "loss": 0.3027, + "step": 5700 + }, + { + "epoch": 0.79, + "grad_norm": 2.418473213237567, + "learning_rate": 1.0865349016051097e-06, + "loss": 0.2742, + "step": 5701 + }, + { + "epoch": 0.79, + "grad_norm": 1.7398361110590614, + "learning_rate": 1.0851340140222504e-06, + "loss": 0.3113, + "step": 5702 + }, + { + "epoch": 0.79, + "grad_norm": 4.64895698691993, + "learning_rate": 1.083733920181083e-06, + "loss": 0.3094, + "step": 5703 + }, + { + "epoch": 0.79, + "grad_norm": 4.1266810788754, + "learning_rate": 1.0823346203654771e-06, + "loss": 0.2857, + "step": 5704 + }, + { + "epoch": 0.79, + "grad_norm": 2.080071870242778, + "learning_rate": 1.080936114859139e-06, + "loss": 0.3017, + "step": 5705 + }, + { + "epoch": 0.79, + "grad_norm": 3.0623900280952308, + "learning_rate": 1.0795384039456186e-06, + "loss": 0.294, + "step": 5706 + }, + { + "epoch": 0.79, + "grad_norm": 3.6658272744713547, + "learning_rate": 1.078141487908304e-06, + "loss": 0.3011, + "step": 5707 + }, + { + "epoch": 0.79, + "grad_norm": 1.8292819144781172, + "learning_rate": 1.0767453670304196e-06, + "loss": 0.2819, + "step": 5708 + }, + { + "epoch": 0.79, + "grad_norm": 2.0199966528419195, + "learning_rate": 1.075350041595028e-06, + "loss": 0.3028, + "step": 5709 + }, + { + "epoch": 0.79, + "grad_norm": 1.9413821809480045, + "learning_rate": 1.0739555118850358e-06, + "loss": 0.317, + "step": 5710 + }, + { + "epoch": 0.79, + "grad_norm": 2.0262170087590685, + "learning_rate": 1.0725617781831827e-06, + "loss": 0.3188, + "step": 5711 + }, + { + "epoch": 0.79, + "grad_norm": 6.026113400183285, + "learning_rate": 1.0711688407720488e-06, + "loss": 0.3032, + "step": 5712 + }, + { + "epoch": 0.79, + "grad_norm": 2.1557489301236012, + "learning_rate": 1.0697766999340548e-06, + "loss": 0.29, + "step": 5713 + }, + { + "epoch": 0.79, + "grad_norm": 5.306643472548173, + "learning_rate": 1.0683853559514562e-06, + "loss": 0.2946, + "step": 5714 + }, + { + "epoch": 0.79, + "grad_norm": 3.1659301192013958, + "learning_rate": 1.066994809106351e-06, + "loss": 0.3359, + "step": 5715 + }, + { + "epoch": 0.79, + "grad_norm": 2.019166059702724, + "learning_rate": 1.065605059680671e-06, + "loss": 0.3064, + "step": 5716 + }, + { + "epoch": 0.79, + "grad_norm": 2.67133473645103, + "learning_rate": 1.064216107956192e-06, + "loss": 0.3118, + "step": 5717 + }, + { + "epoch": 0.79, + "grad_norm": 2.189967682686883, + "learning_rate": 1.0628279542145231e-06, + "loss": 0.2916, + "step": 5718 + }, + { + "epoch": 0.8, + "grad_norm": 2.168581924985359, + "learning_rate": 1.0614405987371118e-06, + "loss": 0.3133, + "step": 5719 + }, + { + "epoch": 0.8, + "grad_norm": 2.406599946498354, + "learning_rate": 1.0600540418052485e-06, + "loss": 0.288, + "step": 5720 + }, + { + "epoch": 0.8, + "grad_norm": 2.7332046739169154, + "learning_rate": 1.0586682837000572e-06, + "loss": 0.2819, + "step": 5721 + }, + { + "epoch": 0.8, + "grad_norm": 2.350478236630428, + "learning_rate": 1.0572833247025e-06, + "loss": 0.3214, + "step": 5722 + }, + { + "epoch": 0.8, + "grad_norm": 2.9809840278494915, + "learning_rate": 1.0558991650933792e-06, + "loss": 0.2945, + "step": 5723 + }, + { + "epoch": 0.8, + "grad_norm": 2.9752761020007887, + "learning_rate": 1.0545158051533355e-06, + "loss": 0.3202, + "step": 5724 + }, + { + "epoch": 0.8, + "grad_norm": 1.9796659083294565, + "learning_rate": 1.0531332451628452e-06, + "loss": 0.3125, + "step": 5725 + }, + { + "epoch": 0.8, + "grad_norm": 2.4357591925249116, + "learning_rate": 1.051751485402221e-06, + "loss": 0.2833, + "step": 5726 + }, + { + "epoch": 0.8, + "grad_norm": 2.6292086686802913, + "learning_rate": 1.0503705261516188e-06, + "loss": 0.3125, + "step": 5727 + }, + { + "epoch": 0.8, + "grad_norm": 1.9359426860757374, + "learning_rate": 1.0489903676910285e-06, + "loss": 0.3103, + "step": 5728 + }, + { + "epoch": 0.8, + "grad_norm": 2.4512501561022964, + "learning_rate": 1.0476110103002758e-06, + "loss": 0.3041, + "step": 5729 + }, + { + "epoch": 0.8, + "grad_norm": 2.163250473034427, + "learning_rate": 1.0462324542590286e-06, + "loss": 0.314, + "step": 5730 + }, + { + "epoch": 0.8, + "grad_norm": 1.8621431476746937, + "learning_rate": 1.0448546998467895e-06, + "loss": 0.2903, + "step": 5731 + }, + { + "epoch": 0.8, + "grad_norm": 2.3247862839790026, + "learning_rate": 1.043477747342898e-06, + "loss": 0.321, + "step": 5732 + }, + { + "epoch": 0.8, + "grad_norm": 2.776648577810426, + "learning_rate": 1.042101597026533e-06, + "loss": 0.3216, + "step": 5733 + }, + { + "epoch": 0.8, + "grad_norm": 2.087427328427302, + "learning_rate": 1.0407262491767117e-06, + "loss": 0.2875, + "step": 5734 + }, + { + "epoch": 0.8, + "grad_norm": 3.1466324353363118, + "learning_rate": 1.0393517040722844e-06, + "loss": 0.2964, + "step": 5735 + }, + { + "epoch": 0.8, + "grad_norm": 2.9072854154557892, + "learning_rate": 1.037977961991941e-06, + "loss": 0.3375, + "step": 5736 + }, + { + "epoch": 0.8, + "grad_norm": 1.9749170304769748, + "learning_rate": 1.0366050232142105e-06, + "loss": 0.307, + "step": 5737 + }, + { + "epoch": 0.8, + "grad_norm": 2.848486095749765, + "learning_rate": 1.0352328880174562e-06, + "loss": 0.274, + "step": 5738 + }, + { + "epoch": 0.8, + "grad_norm": 3.391266809207962, + "learning_rate": 1.0338615566798793e-06, + "loss": 0.3084, + "step": 5739 + }, + { + "epoch": 0.8, + "grad_norm": 2.1946525898512355, + "learning_rate": 1.032491029479515e-06, + "loss": 0.3002, + "step": 5740 + }, + { + "epoch": 0.8, + "grad_norm": 1.9282309004353613, + "learning_rate": 1.0311213066942445e-06, + "loss": 0.2861, + "step": 5741 + }, + { + "epoch": 0.8, + "grad_norm": 2.274883931217861, + "learning_rate": 1.0297523886017775e-06, + "loss": 0.2919, + "step": 5742 + }, + { + "epoch": 0.8, + "grad_norm": 2.8521093189181483, + "learning_rate": 1.0283842754796608e-06, + "loss": 0.3085, + "step": 5743 + }, + { + "epoch": 0.8, + "grad_norm": 2.0390644322681326, + "learning_rate": 1.0270169676052843e-06, + "loss": 0.2911, + "step": 5744 + }, + { + "epoch": 0.8, + "grad_norm": 2.3563661049768565, + "learning_rate": 1.025650465255868e-06, + "loss": 0.2672, + "step": 5745 + }, + { + "epoch": 0.8, + "grad_norm": 2.2296530991312533, + "learning_rate": 1.0242847687084718e-06, + "loss": 0.2984, + "step": 5746 + }, + { + "epoch": 0.8, + "grad_norm": 2.153720777400522, + "learning_rate": 1.0229198782399901e-06, + "loss": 0.3219, + "step": 5747 + }, + { + "epoch": 0.8, + "grad_norm": 4.1134961571019355, + "learning_rate": 1.0215557941271581e-06, + "loss": 0.2926, + "step": 5748 + }, + { + "epoch": 0.8, + "grad_norm": 2.278266176201173, + "learning_rate": 1.020192516646542e-06, + "loss": 0.2945, + "step": 5749 + }, + { + "epoch": 0.8, + "grad_norm": 2.524274420643571, + "learning_rate": 1.0188300460745487e-06, + "loss": 0.3074, + "step": 5750 + }, + { + "epoch": 0.8, + "grad_norm": 2.3861429776883596, + "learning_rate": 1.0174683826874209e-06, + "loss": 0.2996, + "step": 5751 + }, + { + "epoch": 0.8, + "grad_norm": 2.1504066944214966, + "learning_rate": 1.0161075267612358e-06, + "loss": 0.2671, + "step": 5752 + }, + { + "epoch": 0.8, + "grad_norm": 2.5041912224084646, + "learning_rate": 1.014747478571908e-06, + "loss": 0.3016, + "step": 5753 + }, + { + "epoch": 0.8, + "grad_norm": 2.328304686071762, + "learning_rate": 1.013388238395187e-06, + "loss": 0.3124, + "step": 5754 + }, + { + "epoch": 0.8, + "grad_norm": 5.841029884753378, + "learning_rate": 1.0120298065066614e-06, + "loss": 0.3135, + "step": 5755 + }, + { + "epoch": 0.8, + "grad_norm": 2.277157803021538, + "learning_rate": 1.0106721831817535e-06, + "loss": 0.3109, + "step": 5756 + }, + { + "epoch": 0.8, + "grad_norm": 2.5695078950161845, + "learning_rate": 1.0093153686957214e-06, + "loss": 0.2869, + "step": 5757 + }, + { + "epoch": 0.8, + "grad_norm": 4.39780636764067, + "learning_rate": 1.00795936332366e-06, + "loss": 0.3108, + "step": 5758 + }, + { + "epoch": 0.8, + "grad_norm": 3.099558040900706, + "learning_rate": 1.0066041673405036e-06, + "loss": 0.3161, + "step": 5759 + }, + { + "epoch": 0.8, + "grad_norm": 4.697125076003286, + "learning_rate": 1.005249781021016e-06, + "loss": 0.2835, + "step": 5760 + }, + { + "epoch": 0.8, + "grad_norm": 2.0757184157338244, + "learning_rate": 1.0038962046397998e-06, + "loss": 0.302, + "step": 5761 + }, + { + "epoch": 0.8, + "grad_norm": 3.8226989221891863, + "learning_rate": 1.0025434384712957e-06, + "loss": 0.3005, + "step": 5762 + }, + { + "epoch": 0.8, + "grad_norm": 2.3687994835422272, + "learning_rate": 1.0011914827897768e-06, + "loss": 0.2996, + "step": 5763 + }, + { + "epoch": 0.8, + "grad_norm": 2.210298629787843, + "learning_rate": 9.998403378693517e-07, + "loss": 0.2992, + "step": 5764 + }, + { + "epoch": 0.8, + "grad_norm": 1.9709085900958685, + "learning_rate": 9.984900039839678e-07, + "loss": 0.298, + "step": 5765 + }, + { + "epoch": 0.8, + "grad_norm": 2.003984028825483, + "learning_rate": 9.971404814074044e-07, + "loss": 0.3097, + "step": 5766 + }, + { + "epoch": 0.8, + "grad_norm": 2.4195718075522485, + "learning_rate": 9.957917704132808e-07, + "loss": 0.3153, + "step": 5767 + }, + { + "epoch": 0.8, + "grad_norm": 3.0096852152070435, + "learning_rate": 9.94443871275046e-07, + "loss": 0.3192, + "step": 5768 + }, + { + "epoch": 0.8, + "grad_norm": 2.3721571653469873, + "learning_rate": 9.9309678426599e-07, + "loss": 0.3103, + "step": 5769 + }, + { + "epoch": 0.8, + "grad_norm": 2.8823733729653607, + "learning_rate": 9.91750509659235e-07, + "loss": 0.2872, + "step": 5770 + }, + { + "epoch": 0.8, + "grad_norm": 2.485998992985023, + "learning_rate": 9.904050477277367e-07, + "loss": 0.3168, + "step": 5771 + }, + { + "epoch": 0.8, + "grad_norm": 1.9461249652517347, + "learning_rate": 9.89060398744292e-07, + "loss": 0.3076, + "step": 5772 + }, + { + "epoch": 0.8, + "grad_norm": 8.606841883666835, + "learning_rate": 9.877165629815272e-07, + "loss": 0.3128, + "step": 5773 + }, + { + "epoch": 0.8, + "grad_norm": 2.323393615237803, + "learning_rate": 9.863735407119057e-07, + "loss": 0.3008, + "step": 5774 + }, + { + "epoch": 0.8, + "grad_norm": 2.9978294343572403, + "learning_rate": 9.850313322077266e-07, + "loss": 0.3158, + "step": 5775 + }, + { + "epoch": 0.8, + "grad_norm": 2.256819894617559, + "learning_rate": 9.836899377411246e-07, + "loss": 0.3125, + "step": 5776 + }, + { + "epoch": 0.8, + "grad_norm": 2.5526897627545275, + "learning_rate": 9.82349357584068e-07, + "loss": 0.2934, + "step": 5777 + }, + { + "epoch": 0.8, + "grad_norm": 2.1500422963364794, + "learning_rate": 9.810095920083584e-07, + "loss": 0.2751, + "step": 5778 + }, + { + "epoch": 0.8, + "grad_norm": 2.5048708578162295, + "learning_rate": 9.796706412856371e-07, + "loss": 0.3213, + "step": 5779 + }, + { + "epoch": 0.8, + "grad_norm": 2.0806529701006653, + "learning_rate": 9.783325056873756e-07, + "loss": 0.3037, + "step": 5780 + }, + { + "epoch": 0.8, + "grad_norm": 2.25639824283049, + "learning_rate": 9.769951854848803e-07, + "loss": 0.2979, + "step": 5781 + }, + { + "epoch": 0.8, + "grad_norm": 2.3514668396374203, + "learning_rate": 9.756586809492968e-07, + "loss": 0.3054, + "step": 5782 + }, + { + "epoch": 0.8, + "grad_norm": 3.3658058087942315, + "learning_rate": 9.743229923516e-07, + "loss": 0.3016, + "step": 5783 + }, + { + "epoch": 0.8, + "grad_norm": 3.5207059852303186, + "learning_rate": 9.729881199626034e-07, + "loss": 0.2953, + "step": 5784 + }, + { + "epoch": 0.8, + "grad_norm": 3.152666582966602, + "learning_rate": 9.716540640529514e-07, + "loss": 0.293, + "step": 5785 + }, + { + "epoch": 0.8, + "grad_norm": 2.3085948992500915, + "learning_rate": 9.703208248931272e-07, + "loss": 0.3371, + "step": 5786 + }, + { + "epoch": 0.8, + "grad_norm": 2.2693613729004602, + "learning_rate": 9.68988402753444e-07, + "loss": 0.3126, + "step": 5787 + }, + { + "epoch": 0.8, + "grad_norm": 3.518298831161134, + "learning_rate": 9.676567979040507e-07, + "loss": 0.3064, + "step": 5788 + }, + { + "epoch": 0.8, + "grad_norm": 1.9713090182018465, + "learning_rate": 9.66326010614933e-07, + "loss": 0.3145, + "step": 5789 + }, + { + "epoch": 0.8, + "grad_norm": 3.1818323684340184, + "learning_rate": 9.649960411559085e-07, + "loss": 0.2995, + "step": 5790 + }, + { + "epoch": 0.81, + "grad_norm": 2.8740287406254366, + "learning_rate": 9.636668897966278e-07, + "loss": 0.2999, + "step": 5791 + }, + { + "epoch": 0.81, + "grad_norm": 4.56905157280196, + "learning_rate": 9.62338556806578e-07, + "loss": 0.3109, + "step": 5792 + }, + { + "epoch": 0.81, + "grad_norm": 2.057375027010398, + "learning_rate": 9.610110424550812e-07, + "loss": 0.2787, + "step": 5793 + }, + { + "epoch": 0.81, + "grad_norm": 2.330564402959801, + "learning_rate": 9.596843470112905e-07, + "loss": 0.3177, + "step": 5794 + }, + { + "epoch": 0.81, + "grad_norm": 2.643224076372307, + "learning_rate": 9.58358470744193e-07, + "loss": 0.3049, + "step": 5795 + }, + { + "epoch": 0.81, + "grad_norm": 3.398798852388914, + "learning_rate": 9.57033413922613e-07, + "loss": 0.2905, + "step": 5796 + }, + { + "epoch": 0.81, + "grad_norm": 2.091115636897083, + "learning_rate": 9.557091768152066e-07, + "loss": 0.305, + "step": 5797 + }, + { + "epoch": 0.81, + "grad_norm": 1.8889956317286591, + "learning_rate": 9.54385759690461e-07, + "loss": 0.3173, + "step": 5798 + }, + { + "epoch": 0.81, + "grad_norm": 1.9206654617448848, + "learning_rate": 9.530631628167037e-07, + "loss": 0.2914, + "step": 5799 + }, + { + "epoch": 0.81, + "grad_norm": 2.1109132118147578, + "learning_rate": 9.517413864620906e-07, + "loss": 0.2907, + "step": 5800 + }, + { + "epoch": 0.81, + "grad_norm": 2.294315477379491, + "learning_rate": 9.504204308946108e-07, + "loss": 0.2898, + "step": 5801 + }, + { + "epoch": 0.81, + "grad_norm": 27.658340647439573, + "learning_rate": 9.491002963820911e-07, + "loss": 0.291, + "step": 5802 + }, + { + "epoch": 0.81, + "grad_norm": 2.5237309893127535, + "learning_rate": 9.477809831921903e-07, + "loss": 0.2839, + "step": 5803 + }, + { + "epoch": 0.81, + "grad_norm": 2.66510419581668, + "learning_rate": 9.464624915923987e-07, + "loss": 0.2969, + "step": 5804 + }, + { + "epoch": 0.81, + "grad_norm": 2.922773336717644, + "learning_rate": 9.451448218500403e-07, + "loss": 0.2771, + "step": 5805 + }, + { + "epoch": 0.81, + "grad_norm": 1.987529824042319, + "learning_rate": 9.438279742322759e-07, + "loss": 0.2836, + "step": 5806 + }, + { + "epoch": 0.81, + "grad_norm": 2.519561065779652, + "learning_rate": 9.425119490060963e-07, + "loss": 0.3151, + "step": 5807 + }, + { + "epoch": 0.81, + "grad_norm": 3.3489607997632973, + "learning_rate": 9.411967464383254e-07, + "loss": 0.287, + "step": 5808 + }, + { + "epoch": 0.81, + "grad_norm": 2.8134770454779625, + "learning_rate": 9.398823667956219e-07, + "loss": 0.2783, + "step": 5809 + }, + { + "epoch": 0.81, + "grad_norm": 4.455778373466491, + "learning_rate": 9.385688103444784e-07, + "loss": 0.2934, + "step": 5810 + }, + { + "epoch": 0.81, + "grad_norm": 1.9565361334139337, + "learning_rate": 9.372560773512191e-07, + "loss": 0.3093, + "step": 5811 + }, + { + "epoch": 0.81, + "grad_norm": 2.7820451506362174, + "learning_rate": 9.359441680819997e-07, + "loss": 0.3129, + "step": 5812 + }, + { + "epoch": 0.81, + "grad_norm": 3.0204303014871132, + "learning_rate": 9.346330828028127e-07, + "loss": 0.3308, + "step": 5813 + }, + { + "epoch": 0.81, + "grad_norm": 2.1282835327079566, + "learning_rate": 9.33322821779481e-07, + "loss": 0.3022, + "step": 5814 + }, + { + "epoch": 0.81, + "grad_norm": 2.417634756153475, + "learning_rate": 9.320133852776602e-07, + "loss": 0.3152, + "step": 5815 + }, + { + "epoch": 0.81, + "grad_norm": 2.7636895607775265, + "learning_rate": 9.307047735628389e-07, + "loss": 0.3047, + "step": 5816 + }, + { + "epoch": 0.81, + "grad_norm": 2.065317302903407, + "learning_rate": 9.293969869003411e-07, + "loss": 0.3112, + "step": 5817 + }, + { + "epoch": 0.81, + "grad_norm": 1.9207474414855448, + "learning_rate": 9.280900255553193e-07, + "loss": 0.2981, + "step": 5818 + }, + { + "epoch": 0.81, + "grad_norm": 2.0441219014519545, + "learning_rate": 9.267838897927611e-07, + "loss": 0.33, + "step": 5819 + }, + { + "epoch": 0.81, + "grad_norm": 3.1264188247539417, + "learning_rate": 9.254785798774884e-07, + "loss": 0.2999, + "step": 5820 + }, + { + "epoch": 0.81, + "grad_norm": 2.836047151970783, + "learning_rate": 9.241740960741525e-07, + "loss": 0.3038, + "step": 5821 + }, + { + "epoch": 0.81, + "grad_norm": 2.7324364976468405, + "learning_rate": 9.228704386472381e-07, + "loss": 0.3121, + "step": 5822 + }, + { + "epoch": 0.81, + "grad_norm": 2.1993693508421317, + "learning_rate": 9.215676078610608e-07, + "loss": 0.2936, + "step": 5823 + }, + { + "epoch": 0.81, + "grad_norm": 3.202881993101825, + "learning_rate": 9.202656039797742e-07, + "loss": 0.321, + "step": 5824 + }, + { + "epoch": 0.81, + "grad_norm": 2.302508639575075, + "learning_rate": 9.189644272673582e-07, + "loss": 0.2605, + "step": 5825 + }, + { + "epoch": 0.81, + "grad_norm": 3.495909501452235, + "learning_rate": 9.176640779876262e-07, + "loss": 0.3173, + "step": 5826 + }, + { + "epoch": 0.81, + "grad_norm": 2.1033732423829137, + "learning_rate": 9.163645564042268e-07, + "loss": 0.2977, + "step": 5827 + }, + { + "epoch": 0.81, + "grad_norm": 2.196719983320135, + "learning_rate": 9.150658627806386e-07, + "loss": 0.2741, + "step": 5828 + }, + { + "epoch": 0.81, + "grad_norm": 3.532842705466112, + "learning_rate": 9.137679973801727e-07, + "loss": 0.2787, + "step": 5829 + }, + { + "epoch": 0.81, + "grad_norm": 4.714589036205647, + "learning_rate": 9.12470960465971e-07, + "loss": 0.307, + "step": 5830 + }, + { + "epoch": 0.81, + "grad_norm": 3.5488513034907854, + "learning_rate": 9.1117475230101e-07, + "loss": 0.3249, + "step": 5831 + }, + { + "epoch": 0.81, + "grad_norm": 2.157964107410892, + "learning_rate": 9.098793731480965e-07, + "loss": 0.2828, + "step": 5832 + }, + { + "epoch": 0.81, + "grad_norm": 2.87703992040802, + "learning_rate": 9.08584823269868e-07, + "loss": 0.2831, + "step": 5833 + }, + { + "epoch": 0.81, + "grad_norm": 2.1719149519369267, + "learning_rate": 9.072911029287972e-07, + "loss": 0.2968, + "step": 5834 + }, + { + "epoch": 0.81, + "grad_norm": 2.615167718912739, + "learning_rate": 9.059982123871852e-07, + "loss": 0.2946, + "step": 5835 + }, + { + "epoch": 0.81, + "grad_norm": 2.1908613470333655, + "learning_rate": 9.047061519071681e-07, + "loss": 0.2974, + "step": 5836 + }, + { + "epoch": 0.81, + "grad_norm": 2.6198030104474843, + "learning_rate": 9.034149217507104e-07, + "loss": 0.3025, + "step": 5837 + }, + { + "epoch": 0.81, + "grad_norm": 2.256341617083383, + "learning_rate": 9.021245221796121e-07, + "loss": 0.2949, + "step": 5838 + }, + { + "epoch": 0.81, + "grad_norm": 2.480531259491731, + "learning_rate": 9.008349534555005e-07, + "loss": 0.3043, + "step": 5839 + }, + { + "epoch": 0.81, + "grad_norm": 1.9378225333338657, + "learning_rate": 8.99546215839836e-07, + "loss": 0.321, + "step": 5840 + }, + { + "epoch": 0.81, + "grad_norm": 2.358447935372379, + "learning_rate": 8.98258309593914e-07, + "loss": 0.297, + "step": 5841 + }, + { + "epoch": 0.81, + "grad_norm": 3.1815104204871143, + "learning_rate": 8.969712349788561e-07, + "loss": 0.2832, + "step": 5842 + }, + { + "epoch": 0.81, + "grad_norm": 2.2043284578367666, + "learning_rate": 8.956849922556166e-07, + "loss": 0.2994, + "step": 5843 + }, + { + "epoch": 0.81, + "grad_norm": 2.150052861775342, + "learning_rate": 8.943995816849837e-07, + "loss": 0.2914, + "step": 5844 + }, + { + "epoch": 0.81, + "grad_norm": 2.477845835797498, + "learning_rate": 8.931150035275759e-07, + "loss": 0.2895, + "step": 5845 + }, + { + "epoch": 0.81, + "grad_norm": 1.9253585237076443, + "learning_rate": 8.918312580438415e-07, + "loss": 0.2869, + "step": 5846 + }, + { + "epoch": 0.81, + "grad_norm": 2.203068918330735, + "learning_rate": 8.905483454940594e-07, + "loss": 0.3075, + "step": 5847 + }, + { + "epoch": 0.81, + "grad_norm": 2.29454571491427, + "learning_rate": 8.892662661383433e-07, + "loss": 0.334, + "step": 5848 + }, + { + "epoch": 0.81, + "grad_norm": 4.039870940328333, + "learning_rate": 8.879850202366347e-07, + "loss": 0.3201, + "step": 5849 + }, + { + "epoch": 0.81, + "grad_norm": 2.9487876228715986, + "learning_rate": 8.867046080487058e-07, + "loss": 0.3061, + "step": 5850 + }, + { + "epoch": 0.81, + "grad_norm": 2.0329190122078993, + "learning_rate": 8.854250298341632e-07, + "loss": 0.3076, + "step": 5851 + }, + { + "epoch": 0.81, + "grad_norm": 2.1487219018635497, + "learning_rate": 8.841462858524397e-07, + "loss": 0.2933, + "step": 5852 + }, + { + "epoch": 0.81, + "grad_norm": 2.775982616057377, + "learning_rate": 8.828683763628049e-07, + "loss": 0.3181, + "step": 5853 + }, + { + "epoch": 0.81, + "grad_norm": 3.1957304350478046, + "learning_rate": 8.815913016243515e-07, + "loss": 0.3068, + "step": 5854 + }, + { + "epoch": 0.81, + "grad_norm": 6.711940286543404, + "learning_rate": 8.803150618960115e-07, + "loss": 0.2881, + "step": 5855 + }, + { + "epoch": 0.81, + "grad_norm": 2.147108360292467, + "learning_rate": 8.790396574365418e-07, + "loss": 0.2841, + "step": 5856 + }, + { + "epoch": 0.81, + "grad_norm": 2.8624517665198366, + "learning_rate": 8.777650885045291e-07, + "loss": 0.2805, + "step": 5857 + }, + { + "epoch": 0.81, + "grad_norm": 2.2585691147814506, + "learning_rate": 8.76491355358397e-07, + "loss": 0.2775, + "step": 5858 + }, + { + "epoch": 0.81, + "grad_norm": 2.254777197558909, + "learning_rate": 8.752184582563938e-07, + "loss": 0.2845, + "step": 5859 + }, + { + "epoch": 0.81, + "grad_norm": 1.9774783419967514, + "learning_rate": 8.739463974565992e-07, + "loss": 0.2984, + "step": 5860 + }, + { + "epoch": 0.81, + "grad_norm": 2.109919570848368, + "learning_rate": 8.726751732169253e-07, + "loss": 0.3166, + "step": 5861 + }, + { + "epoch": 0.81, + "grad_norm": 2.1768065761417756, + "learning_rate": 8.714047857951153e-07, + "loss": 0.2919, + "step": 5862 + }, + { + "epoch": 0.82, + "grad_norm": 2.591286885894731, + "learning_rate": 8.701352354487392e-07, + "loss": 0.282, + "step": 5863 + }, + { + "epoch": 0.82, + "grad_norm": 2.27508037644575, + "learning_rate": 8.688665224351989e-07, + "loss": 0.2885, + "step": 5864 + }, + { + "epoch": 0.82, + "grad_norm": 2.121877279437965, + "learning_rate": 8.675986470117287e-07, + "loss": 0.307, + "step": 5865 + }, + { + "epoch": 0.82, + "grad_norm": 2.793381122350201, + "learning_rate": 8.663316094353901e-07, + "loss": 0.2835, + "step": 5866 + }, + { + "epoch": 0.82, + "grad_norm": 2.2577400182470044, + "learning_rate": 8.650654099630745e-07, + "loss": 0.2936, + "step": 5867 + }, + { + "epoch": 0.82, + "grad_norm": 2.3322296767830766, + "learning_rate": 8.638000488515075e-07, + "loss": 0.3096, + "step": 5868 + }, + { + "epoch": 0.82, + "grad_norm": 2.5173573517218832, + "learning_rate": 8.625355263572399e-07, + "loss": 0.2606, + "step": 5869 + }, + { + "epoch": 0.82, + "grad_norm": 11.318425642097617, + "learning_rate": 8.61271842736654e-07, + "loss": 0.3299, + "step": 5870 + }, + { + "epoch": 0.82, + "grad_norm": 4.215564293726661, + "learning_rate": 8.600089982459636e-07, + "loss": 0.3146, + "step": 5871 + }, + { + "epoch": 0.82, + "grad_norm": 2.2385796633540243, + "learning_rate": 8.587469931412118e-07, + "loss": 0.3317, + "step": 5872 + }, + { + "epoch": 0.82, + "grad_norm": 2.6188572037129445, + "learning_rate": 8.57485827678271e-07, + "loss": 0.2986, + "step": 5873 + }, + { + "epoch": 0.82, + "grad_norm": 2.3410399657583323, + "learning_rate": 8.562255021128407e-07, + "loss": 0.3028, + "step": 5874 + }, + { + "epoch": 0.82, + "grad_norm": 2.6117445842290903, + "learning_rate": 8.549660167004565e-07, + "loss": 0.2902, + "step": 5875 + }, + { + "epoch": 0.82, + "grad_norm": 2.187200726155899, + "learning_rate": 8.537073716964777e-07, + "loss": 0.2781, + "step": 5876 + }, + { + "epoch": 0.82, + "grad_norm": 2.860670574367722, + "learning_rate": 8.524495673560946e-07, + "loss": 0.2932, + "step": 5877 + }, + { + "epoch": 0.82, + "grad_norm": 2.0278897169951393, + "learning_rate": 8.511926039343294e-07, + "loss": 0.2843, + "step": 5878 + }, + { + "epoch": 0.82, + "grad_norm": 2.1338508634933726, + "learning_rate": 8.499364816860328e-07, + "loss": 0.3082, + "step": 5879 + }, + { + "epoch": 0.82, + "grad_norm": 1.8807897058037018, + "learning_rate": 8.48681200865884e-07, + "loss": 0.2836, + "step": 5880 + }, + { + "epoch": 0.82, + "grad_norm": 2.2080240173171024, + "learning_rate": 8.474267617283899e-07, + "loss": 0.3037, + "step": 5881 + }, + { + "epoch": 0.82, + "grad_norm": 1.9528704263674677, + "learning_rate": 8.46173164527892e-07, + "loss": 0.3065, + "step": 5882 + }, + { + "epoch": 0.82, + "grad_norm": 2.0268753576192124, + "learning_rate": 8.449204095185565e-07, + "loss": 0.287, + "step": 5883 + }, + { + "epoch": 0.82, + "grad_norm": 2.194123803754626, + "learning_rate": 8.436684969543801e-07, + "loss": 0.2938, + "step": 5884 + }, + { + "epoch": 0.82, + "grad_norm": 2.0896043201817447, + "learning_rate": 8.424174270891883e-07, + "loss": 0.3054, + "step": 5885 + }, + { + "epoch": 0.82, + "grad_norm": 1.983363060069243, + "learning_rate": 8.411672001766386e-07, + "loss": 0.3084, + "step": 5886 + }, + { + "epoch": 0.82, + "grad_norm": 2.550191218617372, + "learning_rate": 8.399178164702121e-07, + "loss": 0.3276, + "step": 5887 + }, + { + "epoch": 0.82, + "grad_norm": 2.209577154779753, + "learning_rate": 8.386692762232246e-07, + "loss": 0.32, + "step": 5888 + }, + { + "epoch": 0.82, + "grad_norm": 2.041001757729898, + "learning_rate": 8.374215796888186e-07, + "loss": 0.3102, + "step": 5889 + }, + { + "epoch": 0.82, + "grad_norm": 2.0823488500298746, + "learning_rate": 8.361747271199649e-07, + "loss": 0.2916, + "step": 5890 + }, + { + "epoch": 0.82, + "grad_norm": 2.658204647350575, + "learning_rate": 8.349287187694638e-07, + "loss": 0.3172, + "step": 5891 + }, + { + "epoch": 0.82, + "grad_norm": 3.0319304173031045, + "learning_rate": 8.336835548899425e-07, + "loss": 0.2893, + "step": 5892 + }, + { + "epoch": 0.82, + "grad_norm": 2.040814836415692, + "learning_rate": 8.324392357338618e-07, + "loss": 0.2896, + "step": 5893 + }, + { + "epoch": 0.82, + "grad_norm": 2.152441973967516, + "learning_rate": 8.311957615535066e-07, + "loss": 0.3031, + "step": 5894 + }, + { + "epoch": 0.82, + "grad_norm": 3.0060078817563287, + "learning_rate": 8.299531326009908e-07, + "loss": 0.3064, + "step": 5895 + }, + { + "epoch": 0.82, + "grad_norm": 2.5145676709832094, + "learning_rate": 8.2871134912826e-07, + "loss": 0.3063, + "step": 5896 + }, + { + "epoch": 0.82, + "grad_norm": 2.141985797819801, + "learning_rate": 8.274704113870874e-07, + "loss": 0.3223, + "step": 5897 + }, + { + "epoch": 0.82, + "grad_norm": 4.642042228209124, + "learning_rate": 8.262303196290733e-07, + "loss": 0.2995, + "step": 5898 + }, + { + "epoch": 0.82, + "grad_norm": 2.088430321085508, + "learning_rate": 8.24991074105645e-07, + "loss": 0.2991, + "step": 5899 + }, + { + "epoch": 0.82, + "grad_norm": 2.353199558989279, + "learning_rate": 8.237526750680636e-07, + "loss": 0.3223, + "step": 5900 + }, + { + "epoch": 0.82, + "grad_norm": 2.51672250843316, + "learning_rate": 8.225151227674138e-07, + "loss": 0.3075, + "step": 5901 + }, + { + "epoch": 0.82, + "grad_norm": 1.8797649886374987, + "learning_rate": 8.21278417454609e-07, + "loss": 0.3019, + "step": 5902 + }, + { + "epoch": 0.82, + "grad_norm": 1.8221569760026237, + "learning_rate": 8.200425593803946e-07, + "loss": 0.2959, + "step": 5903 + }, + { + "epoch": 0.82, + "grad_norm": 2.2192862584077813, + "learning_rate": 8.18807548795339e-07, + "loss": 0.3124, + "step": 5904 + }, + { + "epoch": 0.82, + "grad_norm": 2.828306520864469, + "learning_rate": 8.175733859498436e-07, + "loss": 0.2983, + "step": 5905 + }, + { + "epoch": 0.82, + "grad_norm": 2.210075378240779, + "learning_rate": 8.163400710941339e-07, + "loss": 0.2918, + "step": 5906 + }, + { + "epoch": 0.82, + "grad_norm": 2.545818551789337, + "learning_rate": 8.151076044782669e-07, + "loss": 0.3059, + "step": 5907 + }, + { + "epoch": 0.82, + "grad_norm": 3.8054265403245244, + "learning_rate": 8.138759863521262e-07, + "loss": 0.3073, + "step": 5908 + }, + { + "epoch": 0.82, + "grad_norm": 2.1608803987748133, + "learning_rate": 8.126452169654203e-07, + "loss": 0.3094, + "step": 5909 + }, + { + "epoch": 0.82, + "grad_norm": 2.3114841945933917, + "learning_rate": 8.114152965676925e-07, + "loss": 0.2789, + "step": 5910 + }, + { + "epoch": 0.82, + "grad_norm": 1.9460344750633012, + "learning_rate": 8.101862254083076e-07, + "loss": 0.3139, + "step": 5911 + }, + { + "epoch": 0.82, + "grad_norm": 2.1720204027459813, + "learning_rate": 8.089580037364602e-07, + "loss": 0.3077, + "step": 5912 + }, + { + "epoch": 0.82, + "grad_norm": 2.9202630353747208, + "learning_rate": 8.077306318011735e-07, + "loss": 0.3068, + "step": 5913 + }, + { + "epoch": 0.82, + "grad_norm": 2.074521410313853, + "learning_rate": 8.065041098512999e-07, + "loss": 0.3094, + "step": 5914 + }, + { + "epoch": 0.82, + "grad_norm": 2.8441071609156268, + "learning_rate": 8.052784381355161e-07, + "loss": 0.3006, + "step": 5915 + }, + { + "epoch": 0.82, + "grad_norm": 3.6006967475789575, + "learning_rate": 8.04053616902326e-07, + "loss": 0.3188, + "step": 5916 + }, + { + "epoch": 0.82, + "grad_norm": 2.0580594300137265, + "learning_rate": 8.028296464000668e-07, + "loss": 0.2818, + "step": 5917 + }, + { + "epoch": 0.82, + "grad_norm": 2.12407114870154, + "learning_rate": 8.016065268768969e-07, + "loss": 0.2867, + "step": 5918 + }, + { + "epoch": 0.82, + "grad_norm": 2.6174899575144295, + "learning_rate": 8.003842585808036e-07, + "loss": 0.3153, + "step": 5919 + }, + { + "epoch": 0.82, + "grad_norm": 3.480829615860544, + "learning_rate": 7.991628417596047e-07, + "loss": 0.3052, + "step": 5920 + }, + { + "epoch": 0.82, + "grad_norm": 3.219203273467932, + "learning_rate": 7.979422766609424e-07, + "loss": 0.3112, + "step": 5921 + }, + { + "epoch": 0.82, + "grad_norm": 2.0295204499695028, + "learning_rate": 7.967225635322878e-07, + "loss": 0.2829, + "step": 5922 + }, + { + "epoch": 0.82, + "grad_norm": 3.005820857691128, + "learning_rate": 7.955037026209367e-07, + "loss": 0.3152, + "step": 5923 + }, + { + "epoch": 0.82, + "grad_norm": 1.8021500284399146, + "learning_rate": 7.942856941740168e-07, + "loss": 0.3018, + "step": 5924 + }, + { + "epoch": 0.82, + "grad_norm": 2.435966402711488, + "learning_rate": 7.930685384384784e-07, + "loss": 0.3056, + "step": 5925 + }, + { + "epoch": 0.82, + "grad_norm": 2.2845223244649104, + "learning_rate": 7.918522356610997e-07, + "loss": 0.3125, + "step": 5926 + }, + { + "epoch": 0.82, + "grad_norm": 2.5121964048685035, + "learning_rate": 7.906367860884889e-07, + "loss": 0.2728, + "step": 5927 + }, + { + "epoch": 0.82, + "grad_norm": 2.110831079556406, + "learning_rate": 7.894221899670784e-07, + "loss": 0.2826, + "step": 5928 + }, + { + "epoch": 0.82, + "grad_norm": 4.088092211776802, + "learning_rate": 7.882084475431273e-07, + "loss": 0.306, + "step": 5929 + }, + { + "epoch": 0.82, + "grad_norm": 1.8863411586411354, + "learning_rate": 7.869955590627238e-07, + "loss": 0.3121, + "step": 5930 + }, + { + "epoch": 0.82, + "grad_norm": 2.467587885748708, + "learning_rate": 7.857835247717827e-07, + "loss": 0.2869, + "step": 5931 + }, + { + "epoch": 0.82, + "grad_norm": 2.1000264918430402, + "learning_rate": 7.845723449160436e-07, + "loss": 0.2973, + "step": 5932 + }, + { + "epoch": 0.82, + "grad_norm": 2.448681692563208, + "learning_rate": 7.833620197410735e-07, + "loss": 0.3021, + "step": 5933 + }, + { + "epoch": 0.82, + "grad_norm": 3.74361311879113, + "learning_rate": 7.821525494922683e-07, + "loss": 0.2814, + "step": 5934 + }, + { + "epoch": 0.83, + "grad_norm": 2.2492323282645668, + "learning_rate": 7.809439344148479e-07, + "loss": 0.3006, + "step": 5935 + }, + { + "epoch": 0.83, + "grad_norm": 2.0360228946390833, + "learning_rate": 7.797361747538584e-07, + "loss": 0.2973, + "step": 5936 + }, + { + "epoch": 0.83, + "grad_norm": 2.061169093531497, + "learning_rate": 7.78529270754177e-07, + "loss": 0.2825, + "step": 5937 + }, + { + "epoch": 0.83, + "grad_norm": 2.0288499607995387, + "learning_rate": 7.773232226605027e-07, + "loss": 0.3156, + "step": 5938 + }, + { + "epoch": 0.83, + "grad_norm": 2.1790663285275307, + "learning_rate": 7.761180307173605e-07, + "loss": 0.2868, + "step": 5939 + }, + { + "epoch": 0.83, + "grad_norm": 2.547395595901256, + "learning_rate": 7.749136951691066e-07, + "loss": 0.2958, + "step": 5940 + }, + { + "epoch": 0.83, + "grad_norm": 3.827341464199626, + "learning_rate": 7.737102162599208e-07, + "loss": 0.3039, + "step": 5941 + }, + { + "epoch": 0.83, + "grad_norm": 2.038337140403736, + "learning_rate": 7.72507594233809e-07, + "loss": 0.3176, + "step": 5942 + }, + { + "epoch": 0.83, + "grad_norm": 4.118165426554545, + "learning_rate": 7.71305829334602e-07, + "loss": 0.3003, + "step": 5943 + }, + { + "epoch": 0.83, + "grad_norm": 2.376653797451755, + "learning_rate": 7.701049218059603e-07, + "loss": 0.3405, + "step": 5944 + }, + { + "epoch": 0.83, + "grad_norm": 3.612365849668618, + "learning_rate": 7.689048718913683e-07, + "loss": 0.2878, + "step": 5945 + }, + { + "epoch": 0.83, + "grad_norm": 2.2887178235703245, + "learning_rate": 7.677056798341353e-07, + "loss": 0.3018, + "step": 5946 + }, + { + "epoch": 0.83, + "grad_norm": 2.504531563961863, + "learning_rate": 7.665073458774003e-07, + "loss": 0.3082, + "step": 5947 + }, + { + "epoch": 0.83, + "grad_norm": 2.143026621015412, + "learning_rate": 7.65309870264126e-07, + "loss": 0.3108, + "step": 5948 + }, + { + "epoch": 0.83, + "grad_norm": 2.1161459780546266, + "learning_rate": 7.64113253237101e-07, + "loss": 0.3178, + "step": 5949 + }, + { + "epoch": 0.83, + "grad_norm": 3.8085346949992065, + "learning_rate": 7.629174950389396e-07, + "loss": 0.3269, + "step": 5950 + }, + { + "epoch": 0.83, + "grad_norm": 2.440664735586667, + "learning_rate": 7.617225959120839e-07, + "loss": 0.3016, + "step": 5951 + }, + { + "epoch": 0.83, + "grad_norm": 2.854084534653671, + "learning_rate": 7.605285560987996e-07, + "loss": 0.302, + "step": 5952 + }, + { + "epoch": 0.83, + "grad_norm": 2.09048469819597, + "learning_rate": 7.593353758411787e-07, + "loss": 0.3103, + "step": 5953 + }, + { + "epoch": 0.83, + "grad_norm": 2.666724648374528, + "learning_rate": 7.581430553811403e-07, + "loss": 0.297, + "step": 5954 + }, + { + "epoch": 0.83, + "grad_norm": 2.7417069094329625, + "learning_rate": 7.569515949604284e-07, + "loss": 0.2864, + "step": 5955 + }, + { + "epoch": 0.83, + "grad_norm": 2.0673603063271955, + "learning_rate": 7.557609948206102e-07, + "loss": 0.2915, + "step": 5956 + }, + { + "epoch": 0.83, + "grad_norm": 2.2459480160741214, + "learning_rate": 7.54571255203082e-07, + "loss": 0.2957, + "step": 5957 + }, + { + "epoch": 0.83, + "grad_norm": 2.744318721208998, + "learning_rate": 7.533823763490655e-07, + "loss": 0.3027, + "step": 5958 + }, + { + "epoch": 0.83, + "grad_norm": 3.874224985755747, + "learning_rate": 7.521943584996061e-07, + "loss": 0.2856, + "step": 5959 + }, + { + "epoch": 0.83, + "grad_norm": 2.2606204149149076, + "learning_rate": 7.510072018955745e-07, + "loss": 0.2911, + "step": 5960 + }, + { + "epoch": 0.83, + "grad_norm": 2.4145206883679844, + "learning_rate": 7.498209067776668e-07, + "loss": 0.2895, + "step": 5961 + }, + { + "epoch": 0.83, + "grad_norm": 1.8369923057068032, + "learning_rate": 7.486354733864071e-07, + "loss": 0.3143, + "step": 5962 + }, + { + "epoch": 0.83, + "grad_norm": 5.467771207816722, + "learning_rate": 7.474509019621418e-07, + "loss": 0.3033, + "step": 5963 + }, + { + "epoch": 0.83, + "grad_norm": 2.874256948489365, + "learning_rate": 7.462671927450421e-07, + "loss": 0.2811, + "step": 5964 + }, + { + "epoch": 0.83, + "grad_norm": 2.152007757118412, + "learning_rate": 7.450843459751079e-07, + "loss": 0.2954, + "step": 5965 + }, + { + "epoch": 0.83, + "grad_norm": 1.7884510831244398, + "learning_rate": 7.43902361892162e-07, + "loss": 0.2944, + "step": 5966 + }, + { + "epoch": 0.83, + "grad_norm": 3.2631080913090416, + "learning_rate": 7.427212407358525e-07, + "loss": 0.3086, + "step": 5967 + }, + { + "epoch": 0.83, + "grad_norm": 2.6688014745502713, + "learning_rate": 7.415409827456504e-07, + "loss": 0.2788, + "step": 5968 + }, + { + "epoch": 0.83, + "grad_norm": 2.494520773122793, + "learning_rate": 7.403615881608572e-07, + "loss": 0.2981, + "step": 5969 + }, + { + "epoch": 0.83, + "grad_norm": 2.001969056858894, + "learning_rate": 7.391830572205939e-07, + "loss": 0.2855, + "step": 5970 + }, + { + "epoch": 0.83, + "grad_norm": 3.063815343501887, + "learning_rate": 7.380053901638073e-07, + "loss": 0.3105, + "step": 5971 + }, + { + "epoch": 0.83, + "grad_norm": 3.019349525883373, + "learning_rate": 7.36828587229273e-07, + "loss": 0.3002, + "step": 5972 + }, + { + "epoch": 0.83, + "grad_norm": 2.324975146609703, + "learning_rate": 7.356526486555854e-07, + "loss": 0.2672, + "step": 5973 + }, + { + "epoch": 0.83, + "grad_norm": 2.9774821233202027, + "learning_rate": 7.344775746811705e-07, + "loss": 0.302, + "step": 5974 + }, + { + "epoch": 0.83, + "grad_norm": 2.332487876323499, + "learning_rate": 7.333033655442718e-07, + "loss": 0.3186, + "step": 5975 + }, + { + "epoch": 0.83, + "grad_norm": 2.1978112488015134, + "learning_rate": 7.321300214829641e-07, + "loss": 0.3069, + "step": 5976 + }, + { + "epoch": 0.83, + "grad_norm": 2.4140866630619344, + "learning_rate": 7.309575427351417e-07, + "loss": 0.3163, + "step": 5977 + }, + { + "epoch": 0.83, + "grad_norm": 2.4008472633527385, + "learning_rate": 7.297859295385251e-07, + "loss": 0.3226, + "step": 5978 + }, + { + "epoch": 0.83, + "grad_norm": 1.8354415843682188, + "learning_rate": 7.286151821306614e-07, + "loss": 0.2988, + "step": 5979 + }, + { + "epoch": 0.83, + "grad_norm": 2.0954618822589364, + "learning_rate": 7.274453007489197e-07, + "loss": 0.3147, + "step": 5980 + }, + { + "epoch": 0.83, + "grad_norm": 3.513324630976773, + "learning_rate": 7.262762856304928e-07, + "loss": 0.3064, + "step": 5981 + }, + { + "epoch": 0.83, + "grad_norm": 2.419919097854537, + "learning_rate": 7.251081370124003e-07, + "loss": 0.2818, + "step": 5982 + }, + { + "epoch": 0.83, + "grad_norm": 3.560435601874338, + "learning_rate": 7.23940855131487e-07, + "loss": 0.3229, + "step": 5983 + }, + { + "epoch": 0.83, + "grad_norm": 2.3204201354868577, + "learning_rate": 7.22774440224418e-07, + "loss": 0.3213, + "step": 5984 + }, + { + "epoch": 0.83, + "grad_norm": 2.5617759435087155, + "learning_rate": 7.216088925276843e-07, + "loss": 0.2875, + "step": 5985 + }, + { + "epoch": 0.83, + "grad_norm": 1.8777930884651148, + "learning_rate": 7.204442122776029e-07, + "loss": 0.2715, + "step": 5986 + }, + { + "epoch": 0.83, + "grad_norm": 4.09445423793148, + "learning_rate": 7.192803997103138e-07, + "loss": 0.2991, + "step": 5987 + }, + { + "epoch": 0.83, + "grad_norm": 2.303932246836585, + "learning_rate": 7.181174550617781e-07, + "loss": 0.2948, + "step": 5988 + }, + { + "epoch": 0.83, + "grad_norm": 2.3997942307630464, + "learning_rate": 7.169553785677863e-07, + "loss": 0.3041, + "step": 5989 + }, + { + "epoch": 0.83, + "grad_norm": 3.040972642257623, + "learning_rate": 7.157941704639481e-07, + "loss": 0.3271, + "step": 5990 + }, + { + "epoch": 0.83, + "grad_norm": 2.180337869531993, + "learning_rate": 7.146338309857021e-07, + "loss": 0.303, + "step": 5991 + }, + { + "epoch": 0.83, + "grad_norm": 3.8693391586124566, + "learning_rate": 7.134743603683047e-07, + "loss": 0.2963, + "step": 5992 + }, + { + "epoch": 0.83, + "grad_norm": 2.8066749174849215, + "learning_rate": 7.123157588468421e-07, + "loss": 0.2931, + "step": 5993 + }, + { + "epoch": 0.83, + "grad_norm": 1.8530516266821266, + "learning_rate": 7.111580266562201e-07, + "loss": 0.295, + "step": 5994 + }, + { + "epoch": 0.83, + "grad_norm": 2.8520744262725324, + "learning_rate": 7.100011640311688e-07, + "loss": 0.274, + "step": 5995 + }, + { + "epoch": 0.83, + "grad_norm": 2.280499343181778, + "learning_rate": 7.08845171206245e-07, + "loss": 0.3095, + "step": 5996 + }, + { + "epoch": 0.83, + "grad_norm": 2.1993669647061225, + "learning_rate": 7.076900484158261e-07, + "loss": 0.2925, + "step": 5997 + }, + { + "epoch": 0.83, + "grad_norm": 2.061942748392188, + "learning_rate": 7.065357958941122e-07, + "loss": 0.3129, + "step": 5998 + }, + { + "epoch": 0.83, + "grad_norm": 1.939085425342766, + "learning_rate": 7.053824138751308e-07, + "loss": 0.2924, + "step": 5999 + }, + { + "epoch": 0.83, + "grad_norm": 2.9361603082496046, + "learning_rate": 7.042299025927312e-07, + "loss": 0.3005, + "step": 6000 + }, + { + "epoch": 0.83, + "grad_norm": 2.1447213157957004, + "learning_rate": 7.030782622805854e-07, + "loss": 0.2935, + "step": 6001 + }, + { + "epoch": 0.83, + "grad_norm": 2.291063573829537, + "learning_rate": 7.019274931721881e-07, + "loss": 0.2668, + "step": 6002 + }, + { + "epoch": 0.83, + "grad_norm": 2.7290724329667015, + "learning_rate": 7.0077759550086e-07, + "loss": 0.2956, + "step": 6003 + }, + { + "epoch": 0.83, + "grad_norm": 2.3581843665919116, + "learning_rate": 6.996285694997429e-07, + "loss": 0.2969, + "step": 6004 + }, + { + "epoch": 0.83, + "grad_norm": 2.3598608404683734, + "learning_rate": 6.984804154018021e-07, + "loss": 0.2846, + "step": 6005 + }, + { + "epoch": 0.83, + "grad_norm": 2.231036046475555, + "learning_rate": 6.973331334398276e-07, + "loss": 0.2935, + "step": 6006 + }, + { + "epoch": 0.84, + "grad_norm": 2.603750283992693, + "learning_rate": 6.961867238464315e-07, + "loss": 0.3326, + "step": 6007 + }, + { + "epoch": 0.84, + "grad_norm": 2.4899107011463144, + "learning_rate": 6.950411868540475e-07, + "loss": 0.2939, + "step": 6008 + }, + { + "epoch": 0.84, + "grad_norm": 2.211745457952753, + "learning_rate": 6.938965226949357e-07, + "loss": 0.3062, + "step": 6009 + }, + { + "epoch": 0.84, + "grad_norm": 2.667196726849978, + "learning_rate": 6.927527316011778e-07, + "loss": 0.2926, + "step": 6010 + }, + { + "epoch": 0.84, + "grad_norm": 2.9947990882580324, + "learning_rate": 6.916098138046773e-07, + "loss": 0.3049, + "step": 6011 + }, + { + "epoch": 0.84, + "grad_norm": 3.2279735569086414, + "learning_rate": 6.904677695371608e-07, + "loss": 0.3074, + "step": 6012 + }, + { + "epoch": 0.84, + "grad_norm": 1.9912871582023697, + "learning_rate": 6.893265990301806e-07, + "loss": 0.2837, + "step": 6013 + }, + { + "epoch": 0.84, + "grad_norm": 4.886091654651776, + "learning_rate": 6.881863025151087e-07, + "loss": 0.3109, + "step": 6014 + }, + { + "epoch": 0.84, + "grad_norm": 2.7267883540092446, + "learning_rate": 6.8704688022314e-07, + "loss": 0.2913, + "step": 6015 + }, + { + "epoch": 0.84, + "grad_norm": 2.1832121610231168, + "learning_rate": 6.859083323852933e-07, + "loss": 0.3035, + "step": 6016 + }, + { + "epoch": 0.84, + "grad_norm": 1.9921939818642334, + "learning_rate": 6.847706592324127e-07, + "loss": 0.2878, + "step": 6017 + }, + { + "epoch": 0.84, + "grad_norm": 2.6390771838140092, + "learning_rate": 6.836338609951599e-07, + "loss": 0.3193, + "step": 6018 + }, + { + "epoch": 0.84, + "grad_norm": 3.1374542191611523, + "learning_rate": 6.824979379040203e-07, + "loss": 0.3276, + "step": 6019 + }, + { + "epoch": 0.84, + "grad_norm": 2.717616531298631, + "learning_rate": 6.81362890189306e-07, + "loss": 0.3014, + "step": 6020 + }, + { + "epoch": 0.84, + "grad_norm": 3.714387257614362, + "learning_rate": 6.802287180811468e-07, + "loss": 0.2946, + "step": 6021 + }, + { + "epoch": 0.84, + "grad_norm": 1.869354972128528, + "learning_rate": 6.790954218094964e-07, + "loss": 0.2825, + "step": 6022 + }, + { + "epoch": 0.84, + "grad_norm": 1.9614044817295333, + "learning_rate": 6.779630016041339e-07, + "loss": 0.3112, + "step": 6023 + }, + { + "epoch": 0.84, + "grad_norm": 9.398283630225086, + "learning_rate": 6.768314576946561e-07, + "loss": 0.2907, + "step": 6024 + }, + { + "epoch": 0.84, + "grad_norm": 2.038656679152053, + "learning_rate": 6.757007903104834e-07, + "loss": 0.314, + "step": 6025 + }, + { + "epoch": 0.84, + "grad_norm": 2.2815236169986473, + "learning_rate": 6.745709996808609e-07, + "loss": 0.2857, + "step": 6026 + }, + { + "epoch": 0.84, + "grad_norm": 2.6222056854199374, + "learning_rate": 6.734420860348556e-07, + "loss": 0.332, + "step": 6027 + }, + { + "epoch": 0.84, + "grad_norm": 1.9996208172377847, + "learning_rate": 6.723140496013542e-07, + "loss": 0.291, + "step": 6028 + }, + { + "epoch": 0.84, + "grad_norm": 1.768759502471853, + "learning_rate": 6.711868906090657e-07, + "loss": 0.278, + "step": 6029 + }, + { + "epoch": 0.84, + "grad_norm": 2.184907554525499, + "learning_rate": 6.700606092865241e-07, + "loss": 0.3087, + "step": 6030 + }, + { + "epoch": 0.84, + "grad_norm": 2.231525078812747, + "learning_rate": 6.689352058620835e-07, + "loss": 0.2986, + "step": 6031 + }, + { + "epoch": 0.84, + "grad_norm": 2.0405475353733533, + "learning_rate": 6.678106805639195e-07, + "loss": 0.315, + "step": 6032 + }, + { + "epoch": 0.84, + "grad_norm": 2.1339021176675192, + "learning_rate": 6.666870336200293e-07, + "loss": 0.3098, + "step": 6033 + }, + { + "epoch": 0.84, + "grad_norm": 2.254232185826937, + "learning_rate": 6.655642652582345e-07, + "loss": 0.276, + "step": 6034 + }, + { + "epoch": 0.84, + "grad_norm": 2.070350785402926, + "learning_rate": 6.64442375706178e-07, + "loss": 0.3117, + "step": 6035 + }, + { + "epoch": 0.84, + "grad_norm": 2.1001222897156566, + "learning_rate": 6.633213651913229e-07, + "loss": 0.3061, + "step": 6036 + }, + { + "epoch": 0.84, + "grad_norm": 2.5974115418766153, + "learning_rate": 6.622012339409528e-07, + "loss": 0.2864, + "step": 6037 + }, + { + "epoch": 0.84, + "grad_norm": 2.2026784538558846, + "learning_rate": 6.610819821821784e-07, + "loss": 0.2797, + "step": 6038 + }, + { + "epoch": 0.84, + "grad_norm": 1.8970245349524084, + "learning_rate": 6.599636101419271e-07, + "loss": 0.2954, + "step": 6039 + }, + { + "epoch": 0.84, + "grad_norm": 2.5850855091246117, + "learning_rate": 6.588461180469485e-07, + "loss": 0.2982, + "step": 6040 + }, + { + "epoch": 0.84, + "grad_norm": 2.1948260471065524, + "learning_rate": 6.57729506123817e-07, + "loss": 0.3031, + "step": 6041 + }, + { + "epoch": 0.84, + "grad_norm": 2.229394753969474, + "learning_rate": 6.566137745989248e-07, + "loss": 0.3117, + "step": 6042 + }, + { + "epoch": 0.84, + "grad_norm": 2.7902737947293503, + "learning_rate": 6.554989236984883e-07, + "loss": 0.283, + "step": 6043 + }, + { + "epoch": 0.84, + "grad_norm": 1.9971260298557896, + "learning_rate": 6.54384953648543e-07, + "loss": 0.2781, + "step": 6044 + }, + { + "epoch": 0.84, + "grad_norm": 2.460858858860733, + "learning_rate": 6.532718646749492e-07, + "loss": 0.2905, + "step": 6045 + }, + { + "epoch": 0.84, + "grad_norm": 2.2548198011145546, + "learning_rate": 6.521596570033845e-07, + "loss": 0.2939, + "step": 6046 + }, + { + "epoch": 0.84, + "grad_norm": 2.2729471035438236, + "learning_rate": 6.510483308593496e-07, + "loss": 0.2831, + "step": 6047 + }, + { + "epoch": 0.84, + "grad_norm": 2.223803200425216, + "learning_rate": 6.499378864681683e-07, + "loss": 0.2905, + "step": 6048 + }, + { + "epoch": 0.84, + "grad_norm": 2.2315438068899742, + "learning_rate": 6.488283240549831e-07, + "loss": 0.3158, + "step": 6049 + }, + { + "epoch": 0.84, + "grad_norm": 1.8131256502594963, + "learning_rate": 6.477196438447575e-07, + "loss": 0.3155, + "step": 6050 + }, + { + "epoch": 0.84, + "grad_norm": 2.5602440375854987, + "learning_rate": 6.466118460622777e-07, + "loss": 0.3158, + "step": 6051 + }, + { + "epoch": 0.84, + "grad_norm": 2.6790137139802113, + "learning_rate": 6.45504930932152e-07, + "loss": 0.3071, + "step": 6052 + }, + { + "epoch": 0.84, + "grad_norm": 2.415552010741699, + "learning_rate": 6.443988986788069e-07, + "loss": 0.3284, + "step": 6053 + }, + { + "epoch": 0.84, + "grad_norm": 2.2331759689337387, + "learning_rate": 6.432937495264902e-07, + "loss": 0.3409, + "step": 6054 + }, + { + "epoch": 0.84, + "grad_norm": 1.9407974828550434, + "learning_rate": 6.421894836992742e-07, + "loss": 0.295, + "step": 6055 + }, + { + "epoch": 0.84, + "grad_norm": 3.165499756698066, + "learning_rate": 6.410861014210473e-07, + "loss": 0.3033, + "step": 6056 + }, + { + "epoch": 0.84, + "grad_norm": 2.349202364138289, + "learning_rate": 6.399836029155215e-07, + "loss": 0.2901, + "step": 6057 + }, + { + "epoch": 0.84, + "grad_norm": 2.799663482757012, + "learning_rate": 6.388819884062297e-07, + "loss": 0.3143, + "step": 6058 + }, + { + "epoch": 0.84, + "grad_norm": 2.122068479045297, + "learning_rate": 6.377812581165238e-07, + "loss": 0.2918, + "step": 6059 + }, + { + "epoch": 0.84, + "grad_norm": 2.049283455505625, + "learning_rate": 6.366814122695797e-07, + "loss": 0.3109, + "step": 6060 + }, + { + "epoch": 0.84, + "grad_norm": 2.1339032053270364, + "learning_rate": 6.35582451088389e-07, + "loss": 0.3077, + "step": 6061 + }, + { + "epoch": 0.84, + "grad_norm": 2.573620902614056, + "learning_rate": 6.344843747957697e-07, + "loss": 0.2843, + "step": 6062 + }, + { + "epoch": 0.84, + "grad_norm": 2.4294234158590826, + "learning_rate": 6.333871836143562e-07, + "loss": 0.2974, + "step": 6063 + }, + { + "epoch": 0.84, + "grad_norm": 1.9998414770689272, + "learning_rate": 6.322908777666037e-07, + "loss": 0.3066, + "step": 6064 + }, + { + "epoch": 0.84, + "grad_norm": 2.1769535517781096, + "learning_rate": 6.311954574747908e-07, + "loss": 0.3148, + "step": 6065 + }, + { + "epoch": 0.84, + "grad_norm": 4.3957668265826255, + "learning_rate": 6.301009229610144e-07, + "loss": 0.3158, + "step": 6066 + }, + { + "epoch": 0.84, + "grad_norm": 2.4257061727088174, + "learning_rate": 6.290072744471904e-07, + "loss": 0.2956, + "step": 6067 + }, + { + "epoch": 0.84, + "grad_norm": 2.375096017894144, + "learning_rate": 6.279145121550578e-07, + "loss": 0.2865, + "step": 6068 + }, + { + "epoch": 0.84, + "grad_norm": 2.634267670143357, + "learning_rate": 6.268226363061758e-07, + "loss": 0.3024, + "step": 6069 + }, + { + "epoch": 0.84, + "grad_norm": 1.899918260618669, + "learning_rate": 6.257316471219227e-07, + "loss": 0.2947, + "step": 6070 + }, + { + "epoch": 0.84, + "grad_norm": 2.2638591112978, + "learning_rate": 6.246415448234954e-07, + "loss": 0.3047, + "step": 6071 + }, + { + "epoch": 0.84, + "grad_norm": 1.9317502445257748, + "learning_rate": 6.235523296319152e-07, + "loss": 0.2912, + "step": 6072 + }, + { + "epoch": 0.84, + "grad_norm": 2.123541596066536, + "learning_rate": 6.224640017680206e-07, + "loss": 0.3024, + "step": 6073 + }, + { + "epoch": 0.84, + "grad_norm": 2.0313892330127135, + "learning_rate": 6.213765614524692e-07, + "loss": 0.3028, + "step": 6074 + }, + { + "epoch": 0.84, + "grad_norm": 1.9265659589241313, + "learning_rate": 6.202900089057423e-07, + "loss": 0.2816, + "step": 6075 + }, + { + "epoch": 0.84, + "grad_norm": 1.9249783265649485, + "learning_rate": 6.192043443481388e-07, + "loss": 0.2794, + "step": 6076 + }, + { + "epoch": 0.84, + "grad_norm": 3.783229825406677, + "learning_rate": 6.181195679997759e-07, + "loss": 0.3174, + "step": 6077 + }, + { + "epoch": 0.84, + "grad_norm": 2.5856091507503316, + "learning_rate": 6.170356800805949e-07, + "loss": 0.2752, + "step": 6078 + }, + { + "epoch": 0.85, + "grad_norm": 4.017705606837016, + "learning_rate": 6.159526808103549e-07, + "loss": 0.2841, + "step": 6079 + }, + { + "epoch": 0.85, + "grad_norm": 2.5177096177635363, + "learning_rate": 6.148705704086339e-07, + "loss": 0.3397, + "step": 6080 + }, + { + "epoch": 0.85, + "grad_norm": 2.329206167387738, + "learning_rate": 6.137893490948299e-07, + "loss": 0.2798, + "step": 6081 + }, + { + "epoch": 0.85, + "grad_norm": 4.552198136374419, + "learning_rate": 6.127090170881628e-07, + "loss": 0.2815, + "step": 6082 + }, + { + "epoch": 0.85, + "grad_norm": 3.6900935732812585, + "learning_rate": 6.1162957460767e-07, + "loss": 0.2827, + "step": 6083 + }, + { + "epoch": 0.85, + "grad_norm": 1.9478348785578976, + "learning_rate": 6.105510218722077e-07, + "loss": 0.3047, + "step": 6084 + }, + { + "epoch": 0.85, + "grad_norm": 1.9428902070586525, + "learning_rate": 6.094733591004548e-07, + "loss": 0.3053, + "step": 6085 + }, + { + "epoch": 0.85, + "grad_norm": 6.459893551480095, + "learning_rate": 6.083965865109088e-07, + "loss": 0.304, + "step": 6086 + }, + { + "epoch": 0.85, + "grad_norm": 1.7928814102768011, + "learning_rate": 6.073207043218848e-07, + "loss": 0.3016, + "step": 6087 + }, + { + "epoch": 0.85, + "grad_norm": 5.735544375770854, + "learning_rate": 6.062457127515175e-07, + "loss": 0.2983, + "step": 6088 + }, + { + "epoch": 0.85, + "grad_norm": 2.306619691094795, + "learning_rate": 6.051716120177653e-07, + "loss": 0.2884, + "step": 6089 + }, + { + "epoch": 0.85, + "grad_norm": 4.041683250843848, + "learning_rate": 6.040984023384011e-07, + "loss": 0.3046, + "step": 6090 + }, + { + "epoch": 0.85, + "grad_norm": 2.1758232874113026, + "learning_rate": 6.030260839310176e-07, + "loss": 0.2951, + "step": 6091 + }, + { + "epoch": 0.85, + "grad_norm": 2.4095058533422544, + "learning_rate": 6.019546570130302e-07, + "loss": 0.2828, + "step": 6092 + }, + { + "epoch": 0.85, + "grad_norm": 2.565689950690757, + "learning_rate": 6.008841218016709e-07, + "loss": 0.3002, + "step": 6093 + }, + { + "epoch": 0.85, + "grad_norm": 2.145104439016104, + "learning_rate": 5.998144785139898e-07, + "loss": 0.299, + "step": 6094 + }, + { + "epoch": 0.85, + "grad_norm": 1.973516156572293, + "learning_rate": 5.987457273668595e-07, + "loss": 0.2817, + "step": 6095 + }, + { + "epoch": 0.85, + "grad_norm": 4.875402573364802, + "learning_rate": 5.976778685769702e-07, + "loss": 0.2934, + "step": 6096 + }, + { + "epoch": 0.85, + "grad_norm": 2.137554921559245, + "learning_rate": 5.966109023608307e-07, + "loss": 0.3259, + "step": 6097 + }, + { + "epoch": 0.85, + "grad_norm": 2.3306978011598614, + "learning_rate": 5.955448289347676e-07, + "loss": 0.3015, + "step": 6098 + }, + { + "epoch": 0.85, + "grad_norm": 2.0911030603741474, + "learning_rate": 5.944796485149307e-07, + "loss": 0.3009, + "step": 6099 + }, + { + "epoch": 0.85, + "grad_norm": 3.0404189992813473, + "learning_rate": 5.934153613172839e-07, + "loss": 0.3079, + "step": 6100 + }, + { + "epoch": 0.85, + "grad_norm": 1.8306393161506052, + "learning_rate": 5.923519675576134e-07, + "loss": 0.2883, + "step": 6101 + }, + { + "epoch": 0.85, + "grad_norm": 4.409800489518733, + "learning_rate": 5.912894674515207e-07, + "loss": 0.2936, + "step": 6102 + }, + { + "epoch": 0.85, + "grad_norm": 9.323002106814542, + "learning_rate": 5.902278612144302e-07, + "loss": 0.2972, + "step": 6103 + }, + { + "epoch": 0.85, + "grad_norm": 2.3935511584643203, + "learning_rate": 5.891671490615847e-07, + "loss": 0.2584, + "step": 6104 + }, + { + "epoch": 0.85, + "grad_norm": 1.7500327955459232, + "learning_rate": 5.88107331208042e-07, + "loss": 0.2876, + "step": 6105 + }, + { + "epoch": 0.85, + "grad_norm": 2.069723526646527, + "learning_rate": 5.870484078686811e-07, + "loss": 0.3021, + "step": 6106 + }, + { + "epoch": 0.85, + "grad_norm": 2.645318908360607, + "learning_rate": 5.859903792582011e-07, + "loss": 0.3154, + "step": 6107 + }, + { + "epoch": 0.85, + "grad_norm": 3.1207244111659462, + "learning_rate": 5.849332455911167e-07, + "loss": 0.2697, + "step": 6108 + }, + { + "epoch": 0.85, + "grad_norm": 1.9804223990636598, + "learning_rate": 5.838770070817612e-07, + "loss": 0.2827, + "step": 6109 + }, + { + "epoch": 0.85, + "grad_norm": 3.588802451748933, + "learning_rate": 5.828216639442907e-07, + "loss": 0.3063, + "step": 6110 + }, + { + "epoch": 0.85, + "grad_norm": 2.5558902804031716, + "learning_rate": 5.817672163926735e-07, + "loss": 0.3056, + "step": 6111 + }, + { + "epoch": 0.85, + "grad_norm": 2.8648247118627896, + "learning_rate": 5.807136646407024e-07, + "loss": 0.28, + "step": 6112 + }, + { + "epoch": 0.85, + "grad_norm": 2.0665213613304596, + "learning_rate": 5.796610089019833e-07, + "loss": 0.3102, + "step": 6113 + }, + { + "epoch": 0.85, + "grad_norm": 3.226097066956584, + "learning_rate": 5.786092493899453e-07, + "loss": 0.312, + "step": 6114 + }, + { + "epoch": 0.85, + "grad_norm": 2.102388663045705, + "learning_rate": 5.775583863178319e-07, + "loss": 0.2959, + "step": 6115 + }, + { + "epoch": 0.85, + "grad_norm": 1.8938815046626518, + "learning_rate": 5.765084198987053e-07, + "loss": 0.3027, + "step": 6116 + }, + { + "epoch": 0.85, + "grad_norm": 3.096271941548048, + "learning_rate": 5.754593503454498e-07, + "loss": 0.2946, + "step": 6117 + }, + { + "epoch": 0.85, + "grad_norm": 2.253629393822221, + "learning_rate": 5.744111778707623e-07, + "loss": 0.3048, + "step": 6118 + }, + { + "epoch": 0.85, + "grad_norm": 2.4505281550567553, + "learning_rate": 5.733639026871607e-07, + "loss": 0.2746, + "step": 6119 + }, + { + "epoch": 0.85, + "grad_norm": 2.353694215704874, + "learning_rate": 5.723175250069812e-07, + "loss": 0.3104, + "step": 6120 + }, + { + "epoch": 0.85, + "grad_norm": 2.2380703500161574, + "learning_rate": 5.712720450423792e-07, + "loss": 0.2843, + "step": 6121 + }, + { + "epoch": 0.85, + "grad_norm": 2.449634477208329, + "learning_rate": 5.702274630053245e-07, + "loss": 0.3091, + "step": 6122 + }, + { + "epoch": 0.85, + "grad_norm": 4.291562886180386, + "learning_rate": 5.691837791076066e-07, + "loss": 0.3036, + "step": 6123 + }, + { + "epoch": 0.85, + "grad_norm": 2.461013737566301, + "learning_rate": 5.681409935608351e-07, + "loss": 0.3154, + "step": 6124 + }, + { + "epoch": 0.85, + "grad_norm": 2.04692788427433, + "learning_rate": 5.670991065764336e-07, + "loss": 0.2672, + "step": 6125 + }, + { + "epoch": 0.85, + "grad_norm": 2.377623853659998, + "learning_rate": 5.660581183656455e-07, + "loss": 0.3136, + "step": 6126 + }, + { + "epoch": 0.85, + "grad_norm": 2.043398931405381, + "learning_rate": 5.65018029139533e-07, + "loss": 0.3054, + "step": 6127 + }, + { + "epoch": 0.85, + "grad_norm": 3.2336473049953853, + "learning_rate": 5.639788391089735e-07, + "loss": 0.2803, + "step": 6128 + }, + { + "epoch": 0.85, + "grad_norm": 3.489884366832778, + "learning_rate": 5.62940548484665e-07, + "loss": 0.2981, + "step": 6129 + }, + { + "epoch": 0.85, + "grad_norm": 3.33616747414503, + "learning_rate": 5.6190315747712e-07, + "loss": 0.3004, + "step": 6130 + }, + { + "epoch": 0.85, + "grad_norm": 2.2286321002863376, + "learning_rate": 5.608666662966716e-07, + "loss": 0.3179, + "step": 6131 + }, + { + "epoch": 0.85, + "grad_norm": 2.385267257745008, + "learning_rate": 5.598310751534691e-07, + "loss": 0.3161, + "step": 6132 + }, + { + "epoch": 0.85, + "grad_norm": 2.1881458857867147, + "learning_rate": 5.587963842574773e-07, + "loss": 0.3136, + "step": 6133 + }, + { + "epoch": 0.85, + "grad_norm": 2.193557472655874, + "learning_rate": 5.577625938184833e-07, + "loss": 0.2764, + "step": 6134 + }, + { + "epoch": 0.85, + "grad_norm": 2.264045557538153, + "learning_rate": 5.567297040460867e-07, + "loss": 0.3081, + "step": 6135 + }, + { + "epoch": 0.85, + "grad_norm": 1.9550478697477234, + "learning_rate": 5.556977151497067e-07, + "loss": 0.3064, + "step": 6136 + }, + { + "epoch": 0.85, + "grad_norm": 1.9599725059468036, + "learning_rate": 5.546666273385804e-07, + "loss": 0.2849, + "step": 6137 + }, + { + "epoch": 0.85, + "grad_norm": 2.5495270916231836, + "learning_rate": 5.536364408217626e-07, + "loss": 0.2915, + "step": 6138 + }, + { + "epoch": 0.85, + "grad_norm": 2.0601720298308064, + "learning_rate": 5.52607155808123e-07, + "loss": 0.2914, + "step": 6139 + }, + { + "epoch": 0.85, + "grad_norm": 2.6881607842550124, + "learning_rate": 5.515787725063487e-07, + "loss": 0.304, + "step": 6140 + }, + { + "epoch": 0.85, + "grad_norm": 1.8513502278326137, + "learning_rate": 5.505512911249477e-07, + "loss": 0.2905, + "step": 6141 + }, + { + "epoch": 0.85, + "grad_norm": 2.315319710249675, + "learning_rate": 5.495247118722408e-07, + "loss": 0.3039, + "step": 6142 + }, + { + "epoch": 0.85, + "grad_norm": 6.3839115489395235, + "learning_rate": 5.484990349563673e-07, + "loss": 0.3263, + "step": 6143 + }, + { + "epoch": 0.85, + "grad_norm": 5.6302320735314755, + "learning_rate": 5.474742605852856e-07, + "loss": 0.3109, + "step": 6144 + }, + { + "epoch": 0.85, + "grad_norm": 3.0930572910786487, + "learning_rate": 5.464503889667688e-07, + "loss": 0.2806, + "step": 6145 + }, + { + "epoch": 0.85, + "grad_norm": 3.2798588934079667, + "learning_rate": 5.454274203084054e-07, + "loss": 0.3036, + "step": 6146 + }, + { + "epoch": 0.85, + "grad_norm": 1.7512542222150898, + "learning_rate": 5.444053548176054e-07, + "loss": 0.3117, + "step": 6147 + }, + { + "epoch": 0.85, + "grad_norm": 2.6150976403533273, + "learning_rate": 5.433841927015932e-07, + "loss": 0.3045, + "step": 6148 + }, + { + "epoch": 0.85, + "grad_norm": 5.244513348035178, + "learning_rate": 5.423639341674097e-07, + "loss": 0.2925, + "step": 6149 + }, + { + "epoch": 0.85, + "grad_norm": 2.7570447599983945, + "learning_rate": 5.413445794219119e-07, + "loss": 0.31, + "step": 6150 + }, + { + "epoch": 0.86, + "grad_norm": 2.039728227692426, + "learning_rate": 5.403261286717765e-07, + "loss": 0.2878, + "step": 6151 + }, + { + "epoch": 0.86, + "grad_norm": 3.2250877109591216, + "learning_rate": 5.393085821234939e-07, + "loss": 0.3003, + "step": 6152 + }, + { + "epoch": 0.86, + "grad_norm": 2.541446424045787, + "learning_rate": 5.382919399833713e-07, + "loss": 0.3196, + "step": 6153 + }, + { + "epoch": 0.86, + "grad_norm": 2.0433433751394743, + "learning_rate": 5.372762024575356e-07, + "loss": 0.279, + "step": 6154 + }, + { + "epoch": 0.86, + "grad_norm": 2.2480867843353014, + "learning_rate": 5.362613697519281e-07, + "loss": 0.3056, + "step": 6155 + }, + { + "epoch": 0.86, + "grad_norm": 2.7288307742581246, + "learning_rate": 5.352474420723058e-07, + "loss": 0.2846, + "step": 6156 + }, + { + "epoch": 0.86, + "grad_norm": 3.221996420986124, + "learning_rate": 5.342344196242433e-07, + "loss": 0.2928, + "step": 6157 + }, + { + "epoch": 0.86, + "grad_norm": 2.294195638493793, + "learning_rate": 5.332223026131328e-07, + "loss": 0.2808, + "step": 6158 + }, + { + "epoch": 0.86, + "grad_norm": 2.6543154386618615, + "learning_rate": 5.322110912441803e-07, + "loss": 0.3215, + "step": 6159 + }, + { + "epoch": 0.86, + "grad_norm": 2.832822311545668, + "learning_rate": 5.312007857224094e-07, + "loss": 0.2896, + "step": 6160 + }, + { + "epoch": 0.86, + "grad_norm": 1.9955711465078587, + "learning_rate": 5.30191386252662e-07, + "loss": 0.2942, + "step": 6161 + }, + { + "epoch": 0.86, + "grad_norm": 2.047398370535908, + "learning_rate": 5.291828930395931e-07, + "loss": 0.2906, + "step": 6162 + }, + { + "epoch": 0.86, + "grad_norm": 7.889863012132134, + "learning_rate": 5.281753062876749e-07, + "loss": 0.2796, + "step": 6163 + }, + { + "epoch": 0.86, + "grad_norm": 2.1743869707640004, + "learning_rate": 5.271686262011971e-07, + "loss": 0.2967, + "step": 6164 + }, + { + "epoch": 0.86, + "grad_norm": 2.3819671838383694, + "learning_rate": 5.26162852984266e-07, + "loss": 0.2973, + "step": 6165 + }, + { + "epoch": 0.86, + "grad_norm": 3.923876122889801, + "learning_rate": 5.251579868408008e-07, + "loss": 0.3105, + "step": 6166 + }, + { + "epoch": 0.86, + "grad_norm": 2.9550404863927136, + "learning_rate": 5.241540279745388e-07, + "loss": 0.3038, + "step": 6167 + }, + { + "epoch": 0.86, + "grad_norm": 2.746721314115344, + "learning_rate": 5.231509765890353e-07, + "loss": 0.3073, + "step": 6168 + }, + { + "epoch": 0.86, + "grad_norm": 1.934587841484992, + "learning_rate": 5.221488328876584e-07, + "loss": 0.2966, + "step": 6169 + }, + { + "epoch": 0.86, + "grad_norm": 3.297013704914575, + "learning_rate": 5.211475970735929e-07, + "loss": 0.2973, + "step": 6170 + }, + { + "epoch": 0.86, + "grad_norm": 2.8521159500176445, + "learning_rate": 5.201472693498399e-07, + "loss": 0.2882, + "step": 6171 + }, + { + "epoch": 0.86, + "grad_norm": 2.761222221492505, + "learning_rate": 5.191478499192171e-07, + "loss": 0.2904, + "step": 6172 + }, + { + "epoch": 0.86, + "grad_norm": 2.9468296469172217, + "learning_rate": 5.181493389843584e-07, + "loss": 0.2915, + "step": 6173 + }, + { + "epoch": 0.86, + "grad_norm": 2.848904895266979, + "learning_rate": 5.171517367477102e-07, + "loss": 0.3017, + "step": 6174 + }, + { + "epoch": 0.86, + "grad_norm": 2.290688731779601, + "learning_rate": 5.161550434115403e-07, + "loss": 0.2958, + "step": 6175 + }, + { + "epoch": 0.86, + "grad_norm": 2.43416743200975, + "learning_rate": 5.151592591779264e-07, + "loss": 0.3143, + "step": 6176 + }, + { + "epoch": 0.86, + "grad_norm": 4.636346234297209, + "learning_rate": 5.141643842487654e-07, + "loss": 0.2977, + "step": 6177 + }, + { + "epoch": 0.86, + "grad_norm": 2.098063437127936, + "learning_rate": 5.131704188257674e-07, + "loss": 0.3071, + "step": 6178 + }, + { + "epoch": 0.86, + "grad_norm": 4.696198649934945, + "learning_rate": 5.121773631104621e-07, + "loss": 0.2904, + "step": 6179 + }, + { + "epoch": 0.86, + "grad_norm": 2.49747914426161, + "learning_rate": 5.111852173041893e-07, + "loss": 0.3111, + "step": 6180 + }, + { + "epoch": 0.86, + "grad_norm": 3.1451134847058038, + "learning_rate": 5.101939816081097e-07, + "loss": 0.2832, + "step": 6181 + }, + { + "epoch": 0.86, + "grad_norm": 2.988207651878736, + "learning_rate": 5.092036562231956e-07, + "loss": 0.2959, + "step": 6182 + }, + { + "epoch": 0.86, + "grad_norm": 2.3551473186746112, + "learning_rate": 5.082142413502373e-07, + "loss": 0.3238, + "step": 6183 + }, + { + "epoch": 0.86, + "grad_norm": 2.9160016022542234, + "learning_rate": 5.072257371898387e-07, + "loss": 0.3117, + "step": 6184 + }, + { + "epoch": 0.86, + "grad_norm": 2.13706164014224, + "learning_rate": 5.06238143942418e-07, + "loss": 0.3212, + "step": 6185 + }, + { + "epoch": 0.86, + "grad_norm": 2.105031513634257, + "learning_rate": 5.052514618082138e-07, + "loss": 0.2715, + "step": 6186 + }, + { + "epoch": 0.86, + "grad_norm": 2.648208982132296, + "learning_rate": 5.04265690987274e-07, + "loss": 0.2967, + "step": 6187 + }, + { + "epoch": 0.86, + "grad_norm": 3.6752026310467487, + "learning_rate": 5.032808316794647e-07, + "loss": 0.2977, + "step": 6188 + }, + { + "epoch": 0.86, + "grad_norm": 2.2032127552262852, + "learning_rate": 5.022968840844666e-07, + "loss": 0.303, + "step": 6189 + }, + { + "epoch": 0.86, + "grad_norm": 2.742869959358281, + "learning_rate": 5.013138484017777e-07, + "loss": 0.2933, + "step": 6190 + }, + { + "epoch": 0.86, + "grad_norm": 3.2150911821356307, + "learning_rate": 5.003317248307071e-07, + "loss": 0.2972, + "step": 6191 + }, + { + "epoch": 0.86, + "grad_norm": 3.3808425859620144, + "learning_rate": 4.993505135703808e-07, + "loss": 0.2934, + "step": 6192 + }, + { + "epoch": 0.86, + "grad_norm": 2.6188862302501548, + "learning_rate": 4.983702148197422e-07, + "loss": 0.3258, + "step": 6193 + }, + { + "epoch": 0.86, + "grad_norm": 2.0341398250592992, + "learning_rate": 4.973908287775458e-07, + "loss": 0.2913, + "step": 6194 + }, + { + "epoch": 0.86, + "grad_norm": 2.407206161215797, + "learning_rate": 4.964123556423622e-07, + "loss": 0.2961, + "step": 6195 + }, + { + "epoch": 0.86, + "grad_norm": 1.8637019453372157, + "learning_rate": 4.954347956125794e-07, + "loss": 0.2831, + "step": 6196 + }, + { + "epoch": 0.86, + "grad_norm": 2.079281676206457, + "learning_rate": 4.944581488863965e-07, + "loss": 0.2721, + "step": 6197 + }, + { + "epoch": 0.86, + "grad_norm": 2.3324997557942595, + "learning_rate": 4.934824156618306e-07, + "loss": 0.3113, + "step": 6198 + }, + { + "epoch": 0.86, + "grad_norm": 2.2288317967616864, + "learning_rate": 4.925075961367115e-07, + "loss": 0.3021, + "step": 6199 + }, + { + "epoch": 0.86, + "grad_norm": 2.183238676659212, + "learning_rate": 4.915336905086854e-07, + "loss": 0.2952, + "step": 6200 + }, + { + "epoch": 0.86, + "grad_norm": 1.9318384358776122, + "learning_rate": 4.905606989752115e-07, + "loss": 0.2974, + "step": 6201 + }, + { + "epoch": 0.86, + "grad_norm": 2.2192488472109537, + "learning_rate": 4.895886217335644e-07, + "loss": 0.3053, + "step": 6202 + }, + { + "epoch": 0.86, + "grad_norm": 2.122320364316513, + "learning_rate": 4.886174589808341e-07, + "loss": 0.286, + "step": 6203 + }, + { + "epoch": 0.86, + "grad_norm": 3.0114936682233373, + "learning_rate": 4.87647210913924e-07, + "loss": 0.3104, + "step": 6204 + }, + { + "epoch": 0.86, + "grad_norm": 2.518305121485446, + "learning_rate": 4.866778777295522e-07, + "loss": 0.2926, + "step": 6205 + }, + { + "epoch": 0.86, + "grad_norm": 2.5328847213030556, + "learning_rate": 4.857094596242517e-07, + "loss": 0.2805, + "step": 6206 + }, + { + "epoch": 0.86, + "grad_norm": 2.689507674271696, + "learning_rate": 4.847419567943718e-07, + "loss": 0.2857, + "step": 6207 + }, + { + "epoch": 0.86, + "grad_norm": 2.4306347278904226, + "learning_rate": 4.837753694360725e-07, + "loss": 0.3275, + "step": 6208 + }, + { + "epoch": 0.86, + "grad_norm": 2.4201970842545353, + "learning_rate": 4.828096977453295e-07, + "loss": 0.3016, + "step": 6209 + }, + { + "epoch": 0.86, + "grad_norm": 2.2004218812992575, + "learning_rate": 4.81844941917935e-07, + "loss": 0.2947, + "step": 6210 + }, + { + "epoch": 0.86, + "grad_norm": 2.1819764026337407, + "learning_rate": 4.80881102149493e-07, + "loss": 0.2863, + "step": 6211 + }, + { + "epoch": 0.86, + "grad_norm": 7.732019114526687, + "learning_rate": 4.799181786354223e-07, + "loss": 0.3189, + "step": 6212 + }, + { + "epoch": 0.86, + "grad_norm": 1.6420100517502803, + "learning_rate": 4.789561715709578e-07, + "loss": 0.3036, + "step": 6213 + }, + { + "epoch": 0.86, + "grad_norm": 1.8594717370033174, + "learning_rate": 4.779950811511458e-07, + "loss": 0.2966, + "step": 6214 + }, + { + "epoch": 0.86, + "grad_norm": 2.400580941191009, + "learning_rate": 4.770349075708475e-07, + "loss": 0.2893, + "step": 6215 + }, + { + "epoch": 0.86, + "grad_norm": 2.7186799590529334, + "learning_rate": 4.760756510247394e-07, + "loss": 0.2961, + "step": 6216 + }, + { + "epoch": 0.86, + "grad_norm": 1.6905557137332579, + "learning_rate": 4.7511731170731266e-07, + "loss": 0.263, + "step": 6217 + }, + { + "epoch": 0.86, + "grad_norm": 2.361263519146902, + "learning_rate": 4.7415988981286965e-07, + "loss": 0.2979, + "step": 6218 + }, + { + "epoch": 0.86, + "grad_norm": 2.3898105754703813, + "learning_rate": 4.7320338553552835e-07, + "loss": 0.2801, + "step": 6219 + }, + { + "epoch": 0.86, + "grad_norm": 2.46946426995346, + "learning_rate": 4.7224779906922146e-07, + "loss": 0.3183, + "step": 6220 + }, + { + "epoch": 0.86, + "grad_norm": 2.0252429109228607, + "learning_rate": 4.712931306076951e-07, + "loss": 0.308, + "step": 6221 + }, + { + "epoch": 0.86, + "grad_norm": 2.3848788774827816, + "learning_rate": 4.7033938034450666e-07, + "loss": 0.2909, + "step": 6222 + }, + { + "epoch": 0.87, + "grad_norm": 1.9973685974622528, + "learning_rate": 4.6938654847303086e-07, + "loss": 0.2914, + "step": 6223 + }, + { + "epoch": 0.87, + "grad_norm": 2.4417547499353045, + "learning_rate": 4.6843463518645715e-07, + "loss": 0.3015, + "step": 6224 + }, + { + "epoch": 0.87, + "grad_norm": 2.4581172910319724, + "learning_rate": 4.674836406777838e-07, + "loss": 0.2885, + "step": 6225 + }, + { + "epoch": 0.87, + "grad_norm": 2.239622852061465, + "learning_rate": 4.665335651398262e-07, + "loss": 0.2675, + "step": 6226 + }, + { + "epoch": 0.87, + "grad_norm": 2.498322619769942, + "learning_rate": 4.655844087652139e-07, + "loss": 0.2952, + "step": 6227 + }, + { + "epoch": 0.87, + "grad_norm": 2.6106820485651223, + "learning_rate": 4.646361717463882e-07, + "loss": 0.291, + "step": 6228 + }, + { + "epoch": 0.87, + "grad_norm": 2.3560075248842787, + "learning_rate": 4.6368885427560405e-07, + "loss": 0.3246, + "step": 6229 + }, + { + "epoch": 0.87, + "grad_norm": 2.597123324559832, + "learning_rate": 4.6274245654493175e-07, + "loss": 0.3082, + "step": 6230 + }, + { + "epoch": 0.87, + "grad_norm": 12.340115780624085, + "learning_rate": 4.617969787462545e-07, + "loss": 0.3274, + "step": 6231 + }, + { + "epoch": 0.87, + "grad_norm": 1.9869156303303843, + "learning_rate": 4.6085242107126617e-07, + "loss": 0.3192, + "step": 6232 + }, + { + "epoch": 0.87, + "grad_norm": 3.042630242274935, + "learning_rate": 4.5990878371147853e-07, + "loss": 0.2758, + "step": 6233 + }, + { + "epoch": 0.87, + "grad_norm": 2.9347119289411396, + "learning_rate": 4.589660668582152e-07, + "loss": 0.3227, + "step": 6234 + }, + { + "epoch": 0.87, + "grad_norm": 2.967272692835142, + "learning_rate": 4.5802427070261166e-07, + "loss": 0.3002, + "step": 6235 + }, + { + "epoch": 0.87, + "grad_norm": 2.5979809404612335, + "learning_rate": 4.5708339543561686e-07, + "loss": 0.2855, + "step": 6236 + }, + { + "epoch": 0.87, + "grad_norm": 2.0949416850336746, + "learning_rate": 4.5614344124799594e-07, + "loss": 0.2712, + "step": 6237 + }, + { + "epoch": 0.87, + "grad_norm": 2.181227703346149, + "learning_rate": 4.5520440833032363e-07, + "loss": 0.3044, + "step": 6238 + }, + { + "epoch": 0.87, + "grad_norm": 2.3559451741598263, + "learning_rate": 4.5426629687298997e-07, + "loss": 0.3076, + "step": 6239 + }, + { + "epoch": 0.87, + "grad_norm": 2.498909134648591, + "learning_rate": 4.5332910706619715e-07, + "loss": 0.2859, + "step": 6240 + }, + { + "epoch": 0.87, + "grad_norm": 2.1348714564942832, + "learning_rate": 4.52392839099961e-07, + "loss": 0.2865, + "step": 6241 + }, + { + "epoch": 0.87, + "grad_norm": 3.515048571331362, + "learning_rate": 4.514574931641125e-07, + "loss": 0.3163, + "step": 6242 + }, + { + "epoch": 0.87, + "grad_norm": 2.4316555454199014, + "learning_rate": 4.505230694482909e-07, + "loss": 0.2733, + "step": 6243 + }, + { + "epoch": 0.87, + "grad_norm": 2.7686612280747567, + "learning_rate": 4.495895681419532e-07, + "loss": 0.3148, + "step": 6244 + }, + { + "epoch": 0.87, + "grad_norm": 1.866378559515508, + "learning_rate": 4.486569894343673e-07, + "loss": 0.3125, + "step": 6245 + }, + { + "epoch": 0.87, + "grad_norm": 2.158903362386379, + "learning_rate": 4.4772533351461313e-07, + "loss": 0.3012, + "step": 6246 + }, + { + "epoch": 0.87, + "grad_norm": 3.042306877937786, + "learning_rate": 4.46794600571584e-07, + "loss": 0.292, + "step": 6247 + }, + { + "epoch": 0.87, + "grad_norm": 2.4473313663201752, + "learning_rate": 4.4586479079398895e-07, + "loss": 0.3001, + "step": 6248 + }, + { + "epoch": 0.87, + "grad_norm": 3.484547591036173, + "learning_rate": 4.4493590437034496e-07, + "loss": 0.3054, + "step": 6249 + }, + { + "epoch": 0.87, + "grad_norm": 1.8842429749195742, + "learning_rate": 4.4400794148898583e-07, + "loss": 0.2849, + "step": 6250 + }, + { + "epoch": 0.87, + "grad_norm": 3.0524878388070804, + "learning_rate": 4.430809023380572e-07, + "loss": 0.2687, + "step": 6251 + }, + { + "epoch": 0.87, + "grad_norm": 2.2550398696053606, + "learning_rate": 4.42154787105516e-07, + "loss": 0.3017, + "step": 6252 + }, + { + "epoch": 0.87, + "grad_norm": 2.361717160834891, + "learning_rate": 4.41229595979133e-07, + "loss": 0.2749, + "step": 6253 + }, + { + "epoch": 0.87, + "grad_norm": 1.7592849332890248, + "learning_rate": 4.403053291464904e-07, + "loss": 0.2958, + "step": 6254 + }, + { + "epoch": 0.87, + "grad_norm": 2.183227791305827, + "learning_rate": 4.39381986794985e-07, + "loss": 0.2771, + "step": 6255 + }, + { + "epoch": 0.87, + "grad_norm": 3.23471703186508, + "learning_rate": 4.384595691118254e-07, + "loss": 0.3053, + "step": 6256 + }, + { + "epoch": 0.87, + "grad_norm": 2.3620992146203865, + "learning_rate": 4.375380762840303e-07, + "loss": 0.2903, + "step": 6257 + }, + { + "epoch": 0.87, + "grad_norm": 2.0391577391187, + "learning_rate": 4.3661750849843475e-07, + "loss": 0.3029, + "step": 6258 + }, + { + "epoch": 0.87, + "grad_norm": 1.820697183795479, + "learning_rate": 4.3569786594168496e-07, + "loss": 0.3119, + "step": 6259 + }, + { + "epoch": 0.87, + "grad_norm": 9.68824607965409, + "learning_rate": 4.347791488002384e-07, + "loss": 0.2898, + "step": 6260 + }, + { + "epoch": 0.87, + "grad_norm": 1.8854266748272126, + "learning_rate": 4.338613572603645e-07, + "loss": 0.2857, + "step": 6261 + }, + { + "epoch": 0.87, + "grad_norm": 2.4763316181341386, + "learning_rate": 4.329444915081482e-07, + "loss": 0.292, + "step": 6262 + }, + { + "epoch": 0.87, + "grad_norm": 2.464574879124222, + "learning_rate": 4.320285517294831e-07, + "loss": 0.3114, + "step": 6263 + }, + { + "epoch": 0.87, + "grad_norm": 2.270229129806179, + "learning_rate": 4.311135381100762e-07, + "loss": 0.2869, + "step": 6264 + }, + { + "epoch": 0.87, + "grad_norm": 2.1496005370054223, + "learning_rate": 4.30199450835449e-07, + "loss": 0.2946, + "step": 6265 + }, + { + "epoch": 0.87, + "grad_norm": 2.5443025144071356, + "learning_rate": 4.2928629009093115e-07, + "loss": 0.2946, + "step": 6266 + }, + { + "epoch": 0.87, + "grad_norm": 2.3903121941248404, + "learning_rate": 4.2837405606166835e-07, + "loss": 0.2828, + "step": 6267 + }, + { + "epoch": 0.87, + "grad_norm": 5.1010485815849895, + "learning_rate": 4.274627489326155e-07, + "loss": 0.3071, + "step": 6268 + }, + { + "epoch": 0.87, + "grad_norm": 2.3909195551136264, + "learning_rate": 4.26552368888542e-07, + "loss": 0.3024, + "step": 6269 + }, + { + "epoch": 0.87, + "grad_norm": 2.768406590892085, + "learning_rate": 4.256429161140274e-07, + "loss": 0.2913, + "step": 6270 + }, + { + "epoch": 0.87, + "grad_norm": 2.1759617866100154, + "learning_rate": 4.2473439079346257e-07, + "loss": 0.2881, + "step": 6271 + }, + { + "epoch": 0.87, + "grad_norm": 2.2325663207880186, + "learning_rate": 4.2382679311105345e-07, + "loss": 0.3069, + "step": 6272 + }, + { + "epoch": 0.87, + "grad_norm": 6.144760636285787, + "learning_rate": 4.2292012325081567e-07, + "loss": 0.3041, + "step": 6273 + }, + { + "epoch": 0.87, + "grad_norm": 2.2788892182973086, + "learning_rate": 4.22014381396576e-07, + "loss": 0.2746, + "step": 6274 + }, + { + "epoch": 0.87, + "grad_norm": 2.313805752147569, + "learning_rate": 4.2110956773197486e-07, + "loss": 0.2954, + "step": 6275 + }, + { + "epoch": 0.87, + "grad_norm": 2.1959166981883813, + "learning_rate": 4.202056824404649e-07, + "loss": 0.3016, + "step": 6276 + }, + { + "epoch": 0.87, + "grad_norm": 2.9805635004003865, + "learning_rate": 4.193027257053089e-07, + "loss": 0.2821, + "step": 6277 + }, + { + "epoch": 0.87, + "grad_norm": 2.8320193534032456, + "learning_rate": 4.184006977095806e-07, + "loss": 0.2922, + "step": 6278 + }, + { + "epoch": 0.87, + "grad_norm": 2.666271095784794, + "learning_rate": 4.174995986361685e-07, + "loss": 0.3083, + "step": 6279 + }, + { + "epoch": 0.87, + "grad_norm": 2.1834307829006936, + "learning_rate": 4.165994286677705e-07, + "loss": 0.3275, + "step": 6280 + }, + { + "epoch": 0.87, + "grad_norm": 3.324760631912092, + "learning_rate": 4.157001879868955e-07, + "loss": 0.3071, + "step": 6281 + }, + { + "epoch": 0.87, + "grad_norm": 1.8643292455196308, + "learning_rate": 4.1480187677586716e-07, + "loss": 0.2834, + "step": 6282 + }, + { + "epoch": 0.87, + "grad_norm": 2.068939987909746, + "learning_rate": 4.1390449521681753e-07, + "loss": 0.2914, + "step": 6283 + }, + { + "epoch": 0.87, + "grad_norm": 2.1426004163122903, + "learning_rate": 4.1300804349169057e-07, + "loss": 0.3044, + "step": 6284 + }, + { + "epoch": 0.87, + "grad_norm": 5.111866379699838, + "learning_rate": 4.121125217822436e-07, + "loss": 0.2889, + "step": 6285 + }, + { + "epoch": 0.87, + "grad_norm": 2.508224235149877, + "learning_rate": 4.1121793027004475e-07, + "loss": 0.3181, + "step": 6286 + }, + { + "epoch": 0.87, + "grad_norm": 2.1119093129875512, + "learning_rate": 4.1032426913647174e-07, + "loss": 0.3189, + "step": 6287 + }, + { + "epoch": 0.87, + "grad_norm": 2.83439501736679, + "learning_rate": 4.0943153856271565e-07, + "loss": 0.3012, + "step": 6288 + }, + { + "epoch": 0.87, + "grad_norm": 2.240316855720349, + "learning_rate": 4.0853973872977794e-07, + "loss": 0.2948, + "step": 6289 + }, + { + "epoch": 0.87, + "grad_norm": 1.9416548802123939, + "learning_rate": 4.0764886981847274e-07, + "loss": 0.288, + "step": 6290 + }, + { + "epoch": 0.87, + "grad_norm": 2.2191269252441903, + "learning_rate": 4.067589320094217e-07, + "loss": 0.3036, + "step": 6291 + }, + { + "epoch": 0.87, + "grad_norm": 2.0459608282796693, + "learning_rate": 4.0586992548306227e-07, + "loss": 0.2794, + "step": 6292 + }, + { + "epoch": 0.87, + "grad_norm": 18.633745640600907, + "learning_rate": 4.049818504196412e-07, + "loss": 0.3011, + "step": 6293 + }, + { + "epoch": 0.87, + "grad_norm": 1.8959789521732366, + "learning_rate": 4.0409470699921627e-07, + "loss": 0.2927, + "step": 6294 + }, + { + "epoch": 0.88, + "grad_norm": 2.1268875469615676, + "learning_rate": 4.032084954016552e-07, + "loss": 0.2959, + "step": 6295 + }, + { + "epoch": 0.88, + "grad_norm": 3.811886090406985, + "learning_rate": 4.023232158066398e-07, + "loss": 0.2887, + "step": 6296 + }, + { + "epoch": 0.88, + "grad_norm": 2.6399000378701016, + "learning_rate": 4.014388683936599e-07, + "loss": 0.2866, + "step": 6297 + }, + { + "epoch": 0.88, + "grad_norm": 4.337380890421858, + "learning_rate": 4.005554533420164e-07, + "loss": 0.2857, + "step": 6298 + }, + { + "epoch": 0.88, + "grad_norm": 2.0235451470597288, + "learning_rate": 3.9967297083082503e-07, + "loss": 0.3141, + "step": 6299 + }, + { + "epoch": 0.88, + "grad_norm": 2.0551659360647005, + "learning_rate": 3.9879142103900815e-07, + "loss": 0.317, + "step": 6300 + }, + { + "epoch": 0.88, + "grad_norm": 2.4010789388369265, + "learning_rate": 3.979108041452995e-07, + "loss": 0.2923, + "step": 6301 + }, + { + "epoch": 0.88, + "grad_norm": 2.3909125654448427, + "learning_rate": 3.970311203282462e-07, + "loss": 0.2807, + "step": 6302 + }, + { + "epoch": 0.88, + "grad_norm": 2.2192781546457847, + "learning_rate": 3.9615236976620565e-07, + "loss": 0.3151, + "step": 6303 + }, + { + "epoch": 0.88, + "grad_norm": 2.001557714883169, + "learning_rate": 3.952745526373436e-07, + "loss": 0.3034, + "step": 6304 + }, + { + "epoch": 0.88, + "grad_norm": 1.818195797047972, + "learning_rate": 3.9439766911963716e-07, + "loss": 0.3082, + "step": 6305 + }, + { + "epoch": 0.88, + "grad_norm": 2.4774735330229736, + "learning_rate": 3.93521719390878e-07, + "loss": 0.288, + "step": 6306 + }, + { + "epoch": 0.88, + "grad_norm": 2.096321476668874, + "learning_rate": 3.9264670362866295e-07, + "loss": 0.2702, + "step": 6307 + }, + { + "epoch": 0.88, + "grad_norm": 2.4896426971656522, + "learning_rate": 3.9177262201040347e-07, + "loss": 0.305, + "step": 6308 + }, + { + "epoch": 0.88, + "grad_norm": 2.1974403061491707, + "learning_rate": 3.908994747133188e-07, + "loss": 0.3092, + "step": 6309 + }, + { + "epoch": 0.88, + "grad_norm": 3.798698272799998, + "learning_rate": 3.9002726191444075e-07, + "loss": 0.3176, + "step": 6310 + }, + { + "epoch": 0.88, + "grad_norm": 2.6179386619311993, + "learning_rate": 3.891559837906117e-07, + "loss": 0.3073, + "step": 6311 + }, + { + "epoch": 0.88, + "grad_norm": 2.1713809334041008, + "learning_rate": 3.8828564051848317e-07, + "loss": 0.2787, + "step": 6312 + }, + { + "epoch": 0.88, + "grad_norm": 2.6830103025411334, + "learning_rate": 3.874162322745184e-07, + "loss": 0.2929, + "step": 6313 + }, + { + "epoch": 0.88, + "grad_norm": 2.723583605894665, + "learning_rate": 3.8654775923499087e-07, + "loss": 0.2927, + "step": 6314 + }, + { + "epoch": 0.88, + "grad_norm": 2.279559751663549, + "learning_rate": 3.856802215759825e-07, + "loss": 0.2937, + "step": 6315 + }, + { + "epoch": 0.88, + "grad_norm": 2.429341734024937, + "learning_rate": 3.8481361947338756e-07, + "loss": 0.3264, + "step": 6316 + }, + { + "epoch": 0.88, + "grad_norm": 3.271647765955777, + "learning_rate": 3.839479531029117e-07, + "loss": 0.2902, + "step": 6317 + }, + { + "epoch": 0.88, + "grad_norm": 2.404355641431419, + "learning_rate": 3.830832226400671e-07, + "loss": 0.2853, + "step": 6318 + }, + { + "epoch": 0.88, + "grad_norm": 2.0613135916697014, + "learning_rate": 3.8221942826017986e-07, + "loss": 0.3198, + "step": 6319 + }, + { + "epoch": 0.88, + "grad_norm": 2.9999111676335373, + "learning_rate": 3.813565701383859e-07, + "loss": 0.3005, + "step": 6320 + }, + { + "epoch": 0.88, + "grad_norm": 2.464012041459089, + "learning_rate": 3.8049464844962855e-07, + "loss": 0.3144, + "step": 6321 + }, + { + "epoch": 0.88, + "grad_norm": 2.76848941728378, + "learning_rate": 3.7963366336866317e-07, + "loss": 0.2836, + "step": 6322 + }, + { + "epoch": 0.88, + "grad_norm": 2.818860601466171, + "learning_rate": 3.7877361507005495e-07, + "loss": 0.3049, + "step": 6323 + }, + { + "epoch": 0.88, + "grad_norm": 4.804686606359718, + "learning_rate": 3.779145037281806e-07, + "loss": 0.2947, + "step": 6324 + }, + { + "epoch": 0.88, + "grad_norm": 2.474534997159345, + "learning_rate": 3.7705632951722414e-07, + "loss": 0.2981, + "step": 6325 + }, + { + "epoch": 0.88, + "grad_norm": 4.075570461319048, + "learning_rate": 3.7619909261118137e-07, + "loss": 0.3014, + "step": 6326 + }, + { + "epoch": 0.88, + "grad_norm": 2.5488066008447654, + "learning_rate": 3.753427931838571e-07, + "loss": 0.2895, + "step": 6327 + }, + { + "epoch": 0.88, + "grad_norm": 2.1383286963744017, + "learning_rate": 3.744874314088681e-07, + "loss": 0.3222, + "step": 6328 + }, + { + "epoch": 0.88, + "grad_norm": 2.395347566046638, + "learning_rate": 3.736330074596384e-07, + "loss": 0.2932, + "step": 6329 + }, + { + "epoch": 0.88, + "grad_norm": 2.152118004056397, + "learning_rate": 3.7277952150940264e-07, + "loss": 0.3074, + "step": 6330 + }, + { + "epoch": 0.88, + "grad_norm": 2.098801753258982, + "learning_rate": 3.7192697373120754e-07, + "loss": 0.3174, + "step": 6331 + }, + { + "epoch": 0.88, + "grad_norm": 2.2811763175278847, + "learning_rate": 3.710753642979059e-07, + "loss": 0.2909, + "step": 6332 + }, + { + "epoch": 0.88, + "grad_norm": 2.778577329948304, + "learning_rate": 3.702246933821624e-07, + "loss": 0.2994, + "step": 6333 + }, + { + "epoch": 0.88, + "grad_norm": 2.5639577223630914, + "learning_rate": 3.693749611564518e-07, + "loss": 0.2828, + "step": 6334 + }, + { + "epoch": 0.88, + "grad_norm": 1.9451241644337673, + "learning_rate": 3.6852616779305683e-07, + "loss": 0.2942, + "step": 6335 + }, + { + "epoch": 0.88, + "grad_norm": 2.378678768657625, + "learning_rate": 3.6767831346407267e-07, + "loss": 0.2794, + "step": 6336 + }, + { + "epoch": 0.88, + "grad_norm": 2.1037672945850305, + "learning_rate": 3.668313983414001e-07, + "loss": 0.2966, + "step": 6337 + }, + { + "epoch": 0.88, + "grad_norm": 3.3593620012570105, + "learning_rate": 3.6598542259675353e-07, + "loss": 0.3075, + "step": 6338 + }, + { + "epoch": 0.88, + "grad_norm": 2.0035716211407455, + "learning_rate": 3.6514038640165505e-07, + "loss": 0.2831, + "step": 6339 + }, + { + "epoch": 0.88, + "grad_norm": 2.0902856439213404, + "learning_rate": 3.6429628992743436e-07, + "loss": 0.2887, + "step": 6340 + }, + { + "epoch": 0.88, + "grad_norm": 2.5907323472078967, + "learning_rate": 3.6345313334523514e-07, + "loss": 0.3235, + "step": 6341 + }, + { + "epoch": 0.88, + "grad_norm": 1.856447307016341, + "learning_rate": 3.626109168260067e-07, + "loss": 0.2978, + "step": 6342 + }, + { + "epoch": 0.88, + "grad_norm": 3.199136152328295, + "learning_rate": 3.61769640540508e-07, + "loss": 0.3028, + "step": 6343 + }, + { + "epoch": 0.88, + "grad_norm": 2.081030948028401, + "learning_rate": 3.6092930465930933e-07, + "loss": 0.2917, + "step": 6344 + }, + { + "epoch": 0.88, + "grad_norm": 1.8994404411976307, + "learning_rate": 3.600899093527904e-07, + "loss": 0.2705, + "step": 6345 + }, + { + "epoch": 0.88, + "grad_norm": 2.208440857815209, + "learning_rate": 3.5925145479113854e-07, + "loss": 0.2882, + "step": 6346 + }, + { + "epoch": 0.88, + "grad_norm": 4.25453750356661, + "learning_rate": 3.584139411443499e-07, + "loss": 0.2969, + "step": 6347 + }, + { + "epoch": 0.88, + "grad_norm": 3.659928643459064, + "learning_rate": 3.57577368582232e-07, + "loss": 0.2779, + "step": 6348 + }, + { + "epoch": 0.88, + "grad_norm": 2.1997443483504786, + "learning_rate": 3.567417372744003e-07, + "loss": 0.302, + "step": 6349 + }, + { + "epoch": 0.88, + "grad_norm": 1.8388802216271962, + "learning_rate": 3.559070473902787e-07, + "loss": 0.2915, + "step": 6350 + }, + { + "epoch": 0.88, + "grad_norm": 2.331769007291177, + "learning_rate": 3.550732990991029e-07, + "loss": 0.2874, + "step": 6351 + }, + { + "epoch": 0.88, + "grad_norm": 2.005031341441581, + "learning_rate": 3.54240492569915e-07, + "loss": 0.3034, + "step": 6352 + }, + { + "epoch": 0.88, + "grad_norm": 2.5236015173903055, + "learning_rate": 3.5340862797156594e-07, + "loss": 0.287, + "step": 6353 + }, + { + "epoch": 0.88, + "grad_norm": 2.1883292705197204, + "learning_rate": 3.5257770547271753e-07, + "loss": 0.2856, + "step": 6354 + }, + { + "epoch": 0.88, + "grad_norm": 5.290050031499952, + "learning_rate": 3.5174772524184163e-07, + "loss": 0.3, + "step": 6355 + }, + { + "epoch": 0.88, + "grad_norm": 1.8505032446327758, + "learning_rate": 3.509186874472159e-07, + "loss": 0.3048, + "step": 6356 + }, + { + "epoch": 0.88, + "grad_norm": 2.030083016777442, + "learning_rate": 3.5009059225692697e-07, + "loss": 0.3232, + "step": 6357 + }, + { + "epoch": 0.88, + "grad_norm": 4.747446678534353, + "learning_rate": 3.492634398388739e-07, + "loss": 0.3178, + "step": 6358 + }, + { + "epoch": 0.88, + "grad_norm": 2.3132017890489296, + "learning_rate": 3.4843723036076195e-07, + "loss": 0.2764, + "step": 6359 + }, + { + "epoch": 0.88, + "grad_norm": 2.0557977674029053, + "learning_rate": 3.4761196399010375e-07, + "loss": 0.2934, + "step": 6360 + }, + { + "epoch": 0.88, + "grad_norm": 2.0055743570053948, + "learning_rate": 3.4678764089422445e-07, + "loss": 0.2775, + "step": 6361 + }, + { + "epoch": 0.88, + "grad_norm": 2.754694492316087, + "learning_rate": 3.459642612402564e-07, + "loss": 0.2973, + "step": 6362 + }, + { + "epoch": 0.88, + "grad_norm": 1.7693804614177788, + "learning_rate": 3.4514182519513996e-07, + "loss": 0.2753, + "step": 6363 + }, + { + "epoch": 0.88, + "grad_norm": 1.9813619409816032, + "learning_rate": 3.443203329256234e-07, + "loss": 0.251, + "step": 6364 + }, + { + "epoch": 0.88, + "grad_norm": 2.146622399128637, + "learning_rate": 3.434997845982663e-07, + "loss": 0.3073, + "step": 6365 + }, + { + "epoch": 0.88, + "grad_norm": 2.1802567647195294, + "learning_rate": 3.4268018037943493e-07, + "loss": 0.3091, + "step": 6366 + }, + { + "epoch": 0.89, + "grad_norm": 2.350925265645299, + "learning_rate": 3.4186152043530376e-07, + "loss": 0.2963, + "step": 6367 + }, + { + "epoch": 0.89, + "grad_norm": 5.192164723386829, + "learning_rate": 3.410438049318576e-07, + "loss": 0.2965, + "step": 6368 + }, + { + "epoch": 0.89, + "grad_norm": 2.2159364812451416, + "learning_rate": 3.4022703403488956e-07, + "loss": 0.294, + "step": 6369 + }, + { + "epoch": 0.89, + "grad_norm": 3.5693485513059873, + "learning_rate": 3.394112079099976e-07, + "loss": 0.3324, + "step": 6370 + }, + { + "epoch": 0.89, + "grad_norm": 2.6674408467220685, + "learning_rate": 3.385963267225939e-07, + "loss": 0.2831, + "step": 6371 + }, + { + "epoch": 0.89, + "grad_norm": 2.0011039075312858, + "learning_rate": 3.377823906378952e-07, + "loss": 0.2873, + "step": 6372 + }, + { + "epoch": 0.89, + "grad_norm": 2.178216078027947, + "learning_rate": 3.369693998209278e-07, + "loss": 0.2972, + "step": 6373 + }, + { + "epoch": 0.89, + "grad_norm": 3.4906036520732306, + "learning_rate": 3.361573544365254e-07, + "loss": 0.3209, + "step": 6374 + }, + { + "epoch": 0.89, + "grad_norm": 2.3153353557292586, + "learning_rate": 3.353462546493319e-07, + "loss": 0.297, + "step": 6375 + }, + { + "epoch": 0.89, + "grad_norm": 1.8446227449912236, + "learning_rate": 3.345361006237974e-07, + "loss": 0.2847, + "step": 6376 + }, + { + "epoch": 0.89, + "grad_norm": 2.1054469903771964, + "learning_rate": 3.3372689252418153e-07, + "loss": 0.3002, + "step": 6377 + }, + { + "epoch": 0.89, + "grad_norm": 2.2185128387496023, + "learning_rate": 3.3291863051455097e-07, + "loss": 0.291, + "step": 6378 + }, + { + "epoch": 0.89, + "grad_norm": 2.43460575904419, + "learning_rate": 3.321113147587818e-07, + "loss": 0.2812, + "step": 6379 + }, + { + "epoch": 0.89, + "grad_norm": 2.1027879156063736, + "learning_rate": 3.313049454205591e-07, + "loss": 0.3099, + "step": 6380 + }, + { + "epoch": 0.89, + "grad_norm": 2.4646039223154146, + "learning_rate": 3.304995226633728e-07, + "loss": 0.3209, + "step": 6381 + }, + { + "epoch": 0.89, + "grad_norm": 4.113141311129472, + "learning_rate": 3.29695046650525e-07, + "loss": 0.2784, + "step": 6382 + }, + { + "epoch": 0.89, + "grad_norm": 1.8649059429784358, + "learning_rate": 3.2889151754512237e-07, + "loss": 0.2798, + "step": 6383 + }, + { + "epoch": 0.89, + "grad_norm": 2.9422577680152444, + "learning_rate": 3.2808893551008136e-07, + "loss": 0.2944, + "step": 6384 + }, + { + "epoch": 0.89, + "grad_norm": 3.7092148627063524, + "learning_rate": 3.2728730070812506e-07, + "loss": 0.3, + "step": 6385 + }, + { + "epoch": 0.89, + "grad_norm": 2.4001766455824085, + "learning_rate": 3.264866133017874e-07, + "loss": 0.3152, + "step": 6386 + }, + { + "epoch": 0.89, + "grad_norm": 3.347802306423061, + "learning_rate": 3.2568687345340576e-07, + "loss": 0.3102, + "step": 6387 + }, + { + "epoch": 0.89, + "grad_norm": 2.7485269318345424, + "learning_rate": 3.2488808132512986e-07, + "loss": 0.2838, + "step": 6388 + }, + { + "epoch": 0.89, + "grad_norm": 1.9365212469822761, + "learning_rate": 3.240902370789156e-07, + "loss": 0.3083, + "step": 6389 + }, + { + "epoch": 0.89, + "grad_norm": 2.083038025554839, + "learning_rate": 3.2329334087652544e-07, + "loss": 0.3275, + "step": 6390 + }, + { + "epoch": 0.89, + "grad_norm": 1.956370323652657, + "learning_rate": 3.2249739287953106e-07, + "loss": 0.294, + "step": 6391 + }, + { + "epoch": 0.89, + "grad_norm": 2.3707892145688207, + "learning_rate": 3.217023932493102e-07, + "loss": 0.3312, + "step": 6392 + }, + { + "epoch": 0.89, + "grad_norm": 3.6199717276663033, + "learning_rate": 3.209083421470516e-07, + "loss": 0.3269, + "step": 6393 + }, + { + "epoch": 0.89, + "grad_norm": 2.1682014501999722, + "learning_rate": 3.201152397337487e-07, + "loss": 0.2905, + "step": 6394 + }, + { + "epoch": 0.89, + "grad_norm": 2.1991219056780107, + "learning_rate": 3.193230861702024e-07, + "loss": 0.2903, + "step": 6395 + }, + { + "epoch": 0.89, + "grad_norm": 2.075754380474622, + "learning_rate": 3.185318816170235e-07, + "loss": 0.3016, + "step": 6396 + }, + { + "epoch": 0.89, + "grad_norm": 33.16082912171405, + "learning_rate": 3.1774162623463e-07, + "loss": 0.2947, + "step": 6397 + }, + { + "epoch": 0.89, + "grad_norm": 2.166819254771771, + "learning_rate": 3.169523201832458e-07, + "loss": 0.3274, + "step": 6398 + }, + { + "epoch": 0.89, + "grad_norm": 1.8630886594691658, + "learning_rate": 3.16163963622903e-07, + "loss": 0.2969, + "step": 6399 + }, + { + "epoch": 0.89, + "grad_norm": 2.460250033344366, + "learning_rate": 3.1537655671344204e-07, + "loss": 0.2945, + "step": 6400 + }, + { + "epoch": 0.89, + "grad_norm": 2.5788301538054457, + "learning_rate": 3.145900996145101e-07, + "loss": 0.2845, + "step": 6401 + }, + { + "epoch": 0.89, + "grad_norm": 2.1092757480809885, + "learning_rate": 3.1380459248556095e-07, + "loss": 0.281, + "step": 6402 + }, + { + "epoch": 0.89, + "grad_norm": 2.294507812500146, + "learning_rate": 3.130200354858587e-07, + "loss": 0.2972, + "step": 6403 + }, + { + "epoch": 0.89, + "grad_norm": 1.829896144454502, + "learning_rate": 3.1223642877447e-07, + "loss": 0.2982, + "step": 6404 + }, + { + "epoch": 0.89, + "grad_norm": 2.226347284674858, + "learning_rate": 3.1145377251027395e-07, + "loss": 0.2809, + "step": 6405 + }, + { + "epoch": 0.89, + "grad_norm": 2.065419684977257, + "learning_rate": 3.106720668519536e-07, + "loss": 0.2898, + "step": 6406 + }, + { + "epoch": 0.89, + "grad_norm": 2.1983631166522946, + "learning_rate": 3.0989131195800106e-07, + "loss": 0.2928, + "step": 6407 + }, + { + "epoch": 0.89, + "grad_norm": 2.857031211380462, + "learning_rate": 3.091115079867141e-07, + "loss": 0.327, + "step": 6408 + }, + { + "epoch": 0.89, + "grad_norm": 2.0737047648066826, + "learning_rate": 3.083326550961985e-07, + "loss": 0.2987, + "step": 6409 + }, + { + "epoch": 0.89, + "grad_norm": 2.3289194451451722, + "learning_rate": 3.075547534443679e-07, + "loss": 0.301, + "step": 6410 + }, + { + "epoch": 0.89, + "grad_norm": 4.000850453294434, + "learning_rate": 3.067778031889418e-07, + "loss": 0.3201, + "step": 6411 + }, + { + "epoch": 0.89, + "grad_norm": 2.833252859337546, + "learning_rate": 3.0600180448744744e-07, + "loss": 0.2779, + "step": 6412 + }, + { + "epoch": 0.89, + "grad_norm": 5.186963984858741, + "learning_rate": 3.0522675749721843e-07, + "loss": 0.3037, + "step": 6413 + }, + { + "epoch": 0.89, + "grad_norm": 1.9503008106391064, + "learning_rate": 3.0445266237539793e-07, + "loss": 0.2879, + "step": 6414 + }, + { + "epoch": 0.89, + "grad_norm": 2.660916847741793, + "learning_rate": 3.036795192789332e-07, + "loss": 0.3001, + "step": 6415 + }, + { + "epoch": 0.89, + "grad_norm": 2.3537470711167887, + "learning_rate": 3.029073283645789e-07, + "loss": 0.3133, + "step": 6416 + }, + { + "epoch": 0.89, + "grad_norm": 2.656983979026003, + "learning_rate": 3.021360897888981e-07, + "loss": 0.3064, + "step": 6417 + }, + { + "epoch": 0.89, + "grad_norm": 3.275410485919855, + "learning_rate": 3.0136580370825954e-07, + "loss": 0.2908, + "step": 6418 + }, + { + "epoch": 0.89, + "grad_norm": 4.220363384495607, + "learning_rate": 3.005964702788389e-07, + "loss": 0.3052, + "step": 6419 + }, + { + "epoch": 0.89, + "grad_norm": 2.7141296569936624, + "learning_rate": 2.9982808965662027e-07, + "loss": 0.3306, + "step": 6420 + }, + { + "epoch": 0.89, + "grad_norm": 2.7935477626344336, + "learning_rate": 2.9906066199739235e-07, + "loss": 0.282, + "step": 6421 + }, + { + "epoch": 0.89, + "grad_norm": 1.7104673907249677, + "learning_rate": 2.982941874567513e-07, + "loss": 0.2678, + "step": 6422 + }, + { + "epoch": 0.89, + "grad_norm": 2.731139710317933, + "learning_rate": 2.975286661901006e-07, + "loss": 0.3143, + "step": 6423 + }, + { + "epoch": 0.89, + "grad_norm": 2.8467375576865943, + "learning_rate": 2.9676409835265095e-07, + "loss": 0.2996, + "step": 6424 + }, + { + "epoch": 0.89, + "grad_norm": 2.8862262974787836, + "learning_rate": 2.960004840994191e-07, + "loss": 0.2754, + "step": 6425 + }, + { + "epoch": 0.89, + "grad_norm": 3.6058157078362667, + "learning_rate": 2.952378235852266e-07, + "loss": 0.3073, + "step": 6426 + }, + { + "epoch": 0.89, + "grad_norm": 3.5924338782996035, + "learning_rate": 2.944761169647053e-07, + "loss": 0.2861, + "step": 6427 + }, + { + "epoch": 0.89, + "grad_norm": 1.9704791006238758, + "learning_rate": 2.9371536439229074e-07, + "loss": 0.2945, + "step": 6428 + }, + { + "epoch": 0.89, + "grad_norm": 2.7866241254022595, + "learning_rate": 2.929555660222261e-07, + "loss": 0.3141, + "step": 6429 + }, + { + "epoch": 0.89, + "grad_norm": 2.8465488001648294, + "learning_rate": 2.9219672200856033e-07, + "loss": 0.2932, + "step": 6430 + }, + { + "epoch": 0.89, + "grad_norm": 2.5467432045024903, + "learning_rate": 2.9143883250515146e-07, + "loss": 0.2888, + "step": 6431 + }, + { + "epoch": 0.89, + "grad_norm": 2.291666422008314, + "learning_rate": 2.906818976656611e-07, + "loss": 0.2846, + "step": 6432 + }, + { + "epoch": 0.89, + "grad_norm": 2.071540681251761, + "learning_rate": 2.8992591764355707e-07, + "loss": 0.3252, + "step": 6433 + }, + { + "epoch": 0.89, + "grad_norm": 2.1861028814345174, + "learning_rate": 2.8917089259211716e-07, + "loss": 0.3043, + "step": 6434 + }, + { + "epoch": 0.89, + "grad_norm": 4.466660003858824, + "learning_rate": 2.8841682266442127e-07, + "loss": 0.3108, + "step": 6435 + }, + { + "epoch": 0.89, + "grad_norm": 2.943137454950847, + "learning_rate": 2.876637080133582e-07, + "loss": 0.2934, + "step": 6436 + }, + { + "epoch": 0.89, + "grad_norm": 2.0144797741571057, + "learning_rate": 2.86911548791623e-07, + "loss": 0.3021, + "step": 6437 + }, + { + "epoch": 0.89, + "grad_norm": 3.145487491239878, + "learning_rate": 2.861603451517159e-07, + "loss": 0.2905, + "step": 6438 + }, + { + "epoch": 0.9, + "grad_norm": 5.831941847453555, + "learning_rate": 2.8541009724594337e-07, + "loss": 0.3127, + "step": 6439 + }, + { + "epoch": 0.9, + "grad_norm": 1.947343010619186, + "learning_rate": 2.846608052264194e-07, + "loss": 0.295, + "step": 6440 + }, + { + "epoch": 0.9, + "grad_norm": 2.2502825591222697, + "learning_rate": 2.83912469245064e-07, + "loss": 0.2922, + "step": 6441 + }, + { + "epoch": 0.9, + "grad_norm": 2.0391037336248186, + "learning_rate": 2.83165089453602e-07, + "loss": 0.2903, + "step": 6442 + }, + { + "epoch": 0.9, + "grad_norm": 2.3038283925088456, + "learning_rate": 2.8241866600356494e-07, + "loss": 0.3141, + "step": 6443 + }, + { + "epoch": 0.9, + "grad_norm": 3.6089847127684833, + "learning_rate": 2.8167319904629174e-07, + "loss": 0.2646, + "step": 6444 + }, + { + "epoch": 0.9, + "grad_norm": 2.4950209293019285, + "learning_rate": 2.8092868873292545e-07, + "loss": 0.3395, + "step": 6445 + }, + { + "epoch": 0.9, + "grad_norm": 2.5962906593343935, + "learning_rate": 2.801851352144164e-07, + "loss": 0.2807, + "step": 6446 + }, + { + "epoch": 0.9, + "grad_norm": 3.030331999954378, + "learning_rate": 2.7944253864151895e-07, + "loss": 0.3116, + "step": 6447 + }, + { + "epoch": 0.9, + "grad_norm": 2.346011019210976, + "learning_rate": 2.7870089916479715e-07, + "loss": 0.3156, + "step": 6448 + }, + { + "epoch": 0.9, + "grad_norm": 2.6461711538802763, + "learning_rate": 2.7796021693461904e-07, + "loss": 0.2649, + "step": 6449 + }, + { + "epoch": 0.9, + "grad_norm": 2.752450431786507, + "learning_rate": 2.772204921011562e-07, + "loss": 0.3031, + "step": 6450 + }, + { + "epoch": 0.9, + "grad_norm": 5.946298126117611, + "learning_rate": 2.7648172481439137e-07, + "loss": 0.301, + "step": 6451 + }, + { + "epoch": 0.9, + "grad_norm": 2.0862403997681462, + "learning_rate": 2.757439152241076e-07, + "loss": 0.2796, + "step": 6452 + }, + { + "epoch": 0.9, + "grad_norm": 2.469761078788315, + "learning_rate": 2.750070634798979e-07, + "loss": 0.3191, + "step": 6453 + }, + { + "epoch": 0.9, + "grad_norm": 1.7802100134995773, + "learning_rate": 2.742711697311573e-07, + "loss": 0.2676, + "step": 6454 + }, + { + "epoch": 0.9, + "grad_norm": 2.5057257913824413, + "learning_rate": 2.735362341270914e-07, + "loss": 0.2946, + "step": 6455 + }, + { + "epoch": 0.9, + "grad_norm": 2.1690392066496162, + "learning_rate": 2.7280225681670714e-07, + "loss": 0.3051, + "step": 6456 + }, + { + "epoch": 0.9, + "grad_norm": 2.4657307222743703, + "learning_rate": 2.720692379488188e-07, + "loss": 0.2942, + "step": 6457 + }, + { + "epoch": 0.9, + "grad_norm": 2.769110747687511, + "learning_rate": 2.7133717767204803e-07, + "loss": 0.2749, + "step": 6458 + }, + { + "epoch": 0.9, + "grad_norm": 2.179612210063726, + "learning_rate": 2.706060761348195e-07, + "loss": 0.293, + "step": 6459 + }, + { + "epoch": 0.9, + "grad_norm": 4.250469545019834, + "learning_rate": 2.698759334853651e-07, + "loss": 0.2817, + "step": 6460 + }, + { + "epoch": 0.9, + "grad_norm": 2.383170015761046, + "learning_rate": 2.6914674987171984e-07, + "loss": 0.3182, + "step": 6461 + }, + { + "epoch": 0.9, + "grad_norm": 3.8503620044991624, + "learning_rate": 2.6841852544172875e-07, + "loss": 0.2733, + "step": 6462 + }, + { + "epoch": 0.9, + "grad_norm": 2.5270972222659625, + "learning_rate": 2.676912603430387e-07, + "loss": 0.287, + "step": 6463 + }, + { + "epoch": 0.9, + "grad_norm": 1.8994179869611967, + "learning_rate": 2.669649547231018e-07, + "loss": 0.2608, + "step": 6464 + }, + { + "epoch": 0.9, + "grad_norm": 3.780459393285751, + "learning_rate": 2.662396087291791e-07, + "loss": 0.3045, + "step": 6465 + }, + { + "epoch": 0.9, + "grad_norm": 2.029295942109929, + "learning_rate": 2.6551522250833464e-07, + "loss": 0.3146, + "step": 6466 + }, + { + "epoch": 0.9, + "grad_norm": 2.300584592952039, + "learning_rate": 2.6479179620743756e-07, + "loss": 0.2967, + "step": 6467 + }, + { + "epoch": 0.9, + "grad_norm": 2.4380118228989085, + "learning_rate": 2.6406932997316283e-07, + "loss": 0.3177, + "step": 6468 + }, + { + "epoch": 0.9, + "grad_norm": 2.5192317286255976, + "learning_rate": 2.633478239519921e-07, + "loss": 0.2947, + "step": 6469 + }, + { + "epoch": 0.9, + "grad_norm": 2.296074892673541, + "learning_rate": 2.6262727829021005e-07, + "loss": 0.2814, + "step": 6470 + }, + { + "epoch": 0.9, + "grad_norm": 2.5001057264970115, + "learning_rate": 2.6190769313390817e-07, + "loss": 0.2941, + "step": 6471 + }, + { + "epoch": 0.9, + "grad_norm": 2.3083195385700352, + "learning_rate": 2.611890686289831e-07, + "loss": 0.2957, + "step": 6472 + }, + { + "epoch": 0.9, + "grad_norm": 1.8991235255836245, + "learning_rate": 2.60471404921136e-07, + "loss": 0.2915, + "step": 6473 + }, + { + "epoch": 0.9, + "grad_norm": 2.3963137681622726, + "learning_rate": 2.597547021558744e-07, + "loss": 0.3032, + "step": 6474 + }, + { + "epoch": 0.9, + "grad_norm": 2.06726677698708, + "learning_rate": 2.590389604785087e-07, + "loss": 0.2991, + "step": 6475 + }, + { + "epoch": 0.9, + "grad_norm": 3.2513338989267218, + "learning_rate": 2.5832418003415795e-07, + "loss": 0.3071, + "step": 6476 + }, + { + "epoch": 0.9, + "grad_norm": 5.044989229258927, + "learning_rate": 2.5761036096774327e-07, + "loss": 0.3025, + "step": 6477 + }, + { + "epoch": 0.9, + "grad_norm": 2.3457285375556483, + "learning_rate": 2.568975034239918e-07, + "loss": 0.2942, + "step": 6478 + }, + { + "epoch": 0.9, + "grad_norm": 2.2641846751972645, + "learning_rate": 2.5618560754743725e-07, + "loss": 0.2975, + "step": 6479 + }, + { + "epoch": 0.9, + "grad_norm": 2.7665175128411272, + "learning_rate": 2.554746734824154e-07, + "loss": 0.3086, + "step": 6480 + }, + { + "epoch": 0.9, + "grad_norm": 2.410515414290349, + "learning_rate": 2.547647013730686e-07, + "loss": 0.3031, + "step": 6481 + }, + { + "epoch": 0.9, + "grad_norm": 2.580303336903884, + "learning_rate": 2.540556913633446e-07, + "loss": 0.3137, + "step": 6482 + }, + { + "epoch": 0.9, + "grad_norm": 3.1798906316282385, + "learning_rate": 2.533476435969967e-07, + "loss": 0.2971, + "step": 6483 + }, + { + "epoch": 0.9, + "grad_norm": 3.9201238735273716, + "learning_rate": 2.5264055821758114e-07, + "loss": 0.2903, + "step": 6484 + }, + { + "epoch": 0.9, + "grad_norm": 2.053711394920938, + "learning_rate": 2.519344353684594e-07, + "loss": 0.2915, + "step": 6485 + }, + { + "epoch": 0.9, + "grad_norm": 1.935043520656106, + "learning_rate": 2.512292751927992e-07, + "loss": 0.3058, + "step": 6486 + }, + { + "epoch": 0.9, + "grad_norm": 2.5779038546907693, + "learning_rate": 2.5052507783357215e-07, + "loss": 0.2969, + "step": 6487 + }, + { + "epoch": 0.9, + "grad_norm": 2.6481897521663527, + "learning_rate": 2.498218434335542e-07, + "loss": 0.2857, + "step": 6488 + }, + { + "epoch": 0.9, + "grad_norm": 1.886301361261367, + "learning_rate": 2.4911957213532736e-07, + "loss": 0.27, + "step": 6489 + }, + { + "epoch": 0.9, + "grad_norm": 3.231556653166008, + "learning_rate": 2.484182640812771e-07, + "loss": 0.284, + "step": 6490 + }, + { + "epoch": 0.9, + "grad_norm": 1.7445302024770426, + "learning_rate": 2.4771791941359313e-07, + "loss": 0.2883, + "step": 6491 + }, + { + "epoch": 0.9, + "grad_norm": 3.434430792812523, + "learning_rate": 2.470185382742724e-07, + "loss": 0.2599, + "step": 6492 + }, + { + "epoch": 0.9, + "grad_norm": 3.245248790709296, + "learning_rate": 2.4632012080511526e-07, + "loss": 0.3076, + "step": 6493 + }, + { + "epoch": 0.9, + "grad_norm": 5.127317854809318, + "learning_rate": 2.4562266714772465e-07, + "loss": 0.3061, + "step": 6494 + }, + { + "epoch": 0.9, + "grad_norm": 2.0355402572550716, + "learning_rate": 2.449261774435102e-07, + "loss": 0.2879, + "step": 6495 + }, + { + "epoch": 0.9, + "grad_norm": 2.015321027168965, + "learning_rate": 2.4423065183368677e-07, + "loss": 0.3114, + "step": 6496 + }, + { + "epoch": 0.9, + "grad_norm": 2.7251106377778958, + "learning_rate": 2.435360904592721e-07, + "loss": 0.3084, + "step": 6497 + }, + { + "epoch": 0.9, + "grad_norm": 3.0690256771380917, + "learning_rate": 2.428424934610873e-07, + "loss": 0.2987, + "step": 6498 + }, + { + "epoch": 0.9, + "grad_norm": 2.2231932200327877, + "learning_rate": 2.421498609797618e-07, + "loss": 0.2813, + "step": 6499 + }, + { + "epoch": 0.9, + "grad_norm": 2.0676089301864358, + "learning_rate": 2.4145819315572696e-07, + "loss": 0.292, + "step": 6500 + }, + { + "epoch": 0.9, + "grad_norm": 2.140907176347114, + "learning_rate": 2.4076749012921843e-07, + "loss": 0.2808, + "step": 6501 + }, + { + "epoch": 0.9, + "grad_norm": 2.814677903365687, + "learning_rate": 2.4007775204027585e-07, + "loss": 0.2883, + "step": 6502 + }, + { + "epoch": 0.9, + "grad_norm": 3.203880672998539, + "learning_rate": 2.393889790287457e-07, + "loss": 0.2874, + "step": 6503 + }, + { + "epoch": 0.9, + "grad_norm": 8.91772792466935, + "learning_rate": 2.387011712342768e-07, + "loss": 0.2764, + "step": 6504 + }, + { + "epoch": 0.9, + "grad_norm": 5.573345689608759, + "learning_rate": 2.380143287963216e-07, + "loss": 0.301, + "step": 6505 + }, + { + "epoch": 0.9, + "grad_norm": 2.9181036718761777, + "learning_rate": 2.3732845185413912e-07, + "loss": 0.3185, + "step": 6506 + }, + { + "epoch": 0.9, + "grad_norm": 2.4772789627299576, + "learning_rate": 2.366435405467904e-07, + "loss": 0.3097, + "step": 6507 + }, + { + "epoch": 0.9, + "grad_norm": 2.071445939450012, + "learning_rate": 2.3595959501314213e-07, + "loss": 0.2855, + "step": 6508 + }, + { + "epoch": 0.9, + "grad_norm": 5.81415545255427, + "learning_rate": 2.3527661539186397e-07, + "loss": 0.3047, + "step": 6509 + }, + { + "epoch": 0.9, + "grad_norm": 2.304565644365921, + "learning_rate": 2.3459460182143235e-07, + "loss": 0.3011, + "step": 6510 + }, + { + "epoch": 0.91, + "grad_norm": 2.3106514430621083, + "learning_rate": 2.3391355444012442e-07, + "loss": 0.3173, + "step": 6511 + }, + { + "epoch": 0.91, + "grad_norm": 2.077095017322027, + "learning_rate": 2.332334733860231e-07, + "loss": 0.2887, + "step": 6512 + }, + { + "epoch": 0.91, + "grad_norm": 2.372907296468559, + "learning_rate": 2.3255435879701637e-07, + "loss": 0.2887, + "step": 6513 + }, + { + "epoch": 0.91, + "grad_norm": 2.0195019231338187, + "learning_rate": 2.3187621081079415e-07, + "loss": 0.3154, + "step": 6514 + }, + { + "epoch": 0.91, + "grad_norm": 2.1593073222080656, + "learning_rate": 2.3119902956485142e-07, + "loss": 0.295, + "step": 6515 + }, + { + "epoch": 0.91, + "grad_norm": 2.6473932121416524, + "learning_rate": 2.3052281519648612e-07, + "loss": 0.2881, + "step": 6516 + }, + { + "epoch": 0.91, + "grad_norm": 4.090463474434713, + "learning_rate": 2.298475678428036e-07, + "loss": 0.2752, + "step": 6517 + }, + { + "epoch": 0.91, + "grad_norm": 2.044340613985299, + "learning_rate": 2.2917328764070935e-07, + "loss": 0.2835, + "step": 6518 + }, + { + "epoch": 0.91, + "grad_norm": 2.4953281556262863, + "learning_rate": 2.28499974726914e-07, + "loss": 0.275, + "step": 6519 + }, + { + "epoch": 0.91, + "grad_norm": 2.3400441193218677, + "learning_rate": 2.2782762923793277e-07, + "loss": 0.3042, + "step": 6520 + }, + { + "epoch": 0.91, + "grad_norm": 2.3601057256857114, + "learning_rate": 2.271562513100839e-07, + "loss": 0.2837, + "step": 6521 + }, + { + "epoch": 0.91, + "grad_norm": 2.2790562797096676, + "learning_rate": 2.2648584107948956e-07, + "loss": 0.291, + "step": 6522 + }, + { + "epoch": 0.91, + "grad_norm": 2.569265749302196, + "learning_rate": 2.258163986820755e-07, + "loss": 0.2804, + "step": 6523 + }, + { + "epoch": 0.91, + "grad_norm": 1.8448827591339048, + "learning_rate": 2.2514792425357256e-07, + "loss": 0.3107, + "step": 6524 + }, + { + "epoch": 0.91, + "grad_norm": 2.390981553030664, + "learning_rate": 2.2448041792951348e-07, + "loss": 0.2879, + "step": 6525 + }, + { + "epoch": 0.91, + "grad_norm": 1.82793092498999, + "learning_rate": 2.238138798452355e-07, + "loss": 0.2596, + "step": 6526 + }, + { + "epoch": 0.91, + "grad_norm": 2.8606967476288814, + "learning_rate": 2.2314831013588112e-07, + "loss": 0.2912, + "step": 6527 + }, + { + "epoch": 0.91, + "grad_norm": 2.1237027473608534, + "learning_rate": 2.224837089363946e-07, + "loss": 0.2981, + "step": 6528 + }, + { + "epoch": 0.91, + "grad_norm": 2.1138409502413547, + "learning_rate": 2.2182007638152314e-07, + "loss": 0.2886, + "step": 6529 + }, + { + "epoch": 0.91, + "grad_norm": 3.143740130218538, + "learning_rate": 2.2115741260581857e-07, + "loss": 0.2684, + "step": 6530 + }, + { + "epoch": 0.91, + "grad_norm": 2.886956877433389, + "learning_rate": 2.2049571774363844e-07, + "loss": 0.3077, + "step": 6531 + }, + { + "epoch": 0.91, + "grad_norm": 2.1505611960672573, + "learning_rate": 2.1983499192913982e-07, + "loss": 0.2958, + "step": 6532 + }, + { + "epoch": 0.91, + "grad_norm": 2.282006133611076, + "learning_rate": 2.1917523529628615e-07, + "loss": 0.294, + "step": 6533 + }, + { + "epoch": 0.91, + "grad_norm": 3.1538353678757454, + "learning_rate": 2.1851644797884263e-07, + "loss": 0.2964, + "step": 6534 + }, + { + "epoch": 0.91, + "grad_norm": 2.608124281556718, + "learning_rate": 2.1785863011038078e-07, + "loss": 0.291, + "step": 6535 + }, + { + "epoch": 0.91, + "grad_norm": 6.010429562661825, + "learning_rate": 2.1720178182427276e-07, + "loss": 0.3073, + "step": 6536 + }, + { + "epoch": 0.91, + "grad_norm": 2.3203277967617355, + "learning_rate": 2.1654590325369317e-07, + "loss": 0.321, + "step": 6537 + }, + { + "epoch": 0.91, + "grad_norm": 3.7865783732428775, + "learning_rate": 2.1589099453162455e-07, + "loss": 0.2926, + "step": 6538 + }, + { + "epoch": 0.91, + "grad_norm": 3.239800561502043, + "learning_rate": 2.1523705579084896e-07, + "loss": 0.2954, + "step": 6539 + }, + { + "epoch": 0.91, + "grad_norm": 2.060991397254291, + "learning_rate": 2.1458408716395152e-07, + "loss": 0.3095, + "step": 6540 + }, + { + "epoch": 0.91, + "grad_norm": 2.0746061491290897, + "learning_rate": 2.1393208878332462e-07, + "loss": 0.2868, + "step": 6541 + }, + { + "epoch": 0.91, + "grad_norm": 2.240252433522403, + "learning_rate": 2.132810607811592e-07, + "loss": 0.2991, + "step": 6542 + }, + { + "epoch": 0.91, + "grad_norm": 2.029443197481481, + "learning_rate": 2.1263100328945353e-07, + "loss": 0.3094, + "step": 6543 + }, + { + "epoch": 0.91, + "grad_norm": 2.089421946226848, + "learning_rate": 2.11981916440005e-07, + "loss": 0.2907, + "step": 6544 + }, + { + "epoch": 0.91, + "grad_norm": 2.293327111521, + "learning_rate": 2.1133380036441887e-07, + "loss": 0.2883, + "step": 6545 + }, + { + "epoch": 0.91, + "grad_norm": 2.71741023569152, + "learning_rate": 2.1068665519409948e-07, + "loss": 0.2823, + "step": 6546 + }, + { + "epoch": 0.91, + "grad_norm": 2.463678648166353, + "learning_rate": 2.1004048106025577e-07, + "loss": 0.293, + "step": 6547 + }, + { + "epoch": 0.91, + "grad_norm": 2.135895390845283, + "learning_rate": 2.093952780939007e-07, + "loss": 0.2985, + "step": 6548 + }, + { + "epoch": 0.91, + "grad_norm": 2.1569199207743917, + "learning_rate": 2.087510464258502e-07, + "loss": 0.2786, + "step": 6549 + }, + { + "epoch": 0.91, + "grad_norm": 2.0734870328688007, + "learning_rate": 2.081077861867209e-07, + "loss": 0.3076, + "step": 6550 + }, + { + "epoch": 0.91, + "grad_norm": 2.905442500780436, + "learning_rate": 2.0746549750693513e-07, + "loss": 0.312, + "step": 6551 + }, + { + "epoch": 0.91, + "grad_norm": 2.6719013519165995, + "learning_rate": 2.0682418051671815e-07, + "loss": 0.3066, + "step": 6552 + }, + { + "epoch": 0.91, + "grad_norm": 2.172930023246484, + "learning_rate": 2.0618383534609655e-07, + "loss": 0.3079, + "step": 6553 + }, + { + "epoch": 0.91, + "grad_norm": 2.0500147234683723, + "learning_rate": 2.0554446212490031e-07, + "loss": 0.2931, + "step": 6554 + }, + { + "epoch": 0.91, + "grad_norm": 2.118207680895123, + "learning_rate": 2.049060609827641e-07, + "loss": 0.2883, + "step": 6555 + }, + { + "epoch": 0.91, + "grad_norm": 2.117786313617464, + "learning_rate": 2.0426863204912328e-07, + "loss": 0.2933, + "step": 6556 + }, + { + "epoch": 0.91, + "grad_norm": 2.669551959326981, + "learning_rate": 2.0363217545321668e-07, + "loss": 0.3006, + "step": 6557 + }, + { + "epoch": 0.91, + "grad_norm": 3.312044576024641, + "learning_rate": 2.0299669132408716e-07, + "loss": 0.2769, + "step": 6558 + }, + { + "epoch": 0.91, + "grad_norm": 2.760082516725813, + "learning_rate": 2.0236217979057948e-07, + "loss": 0.3197, + "step": 6559 + }, + { + "epoch": 0.91, + "grad_norm": 2.443743184926829, + "learning_rate": 2.0172864098134014e-07, + "loss": 0.28, + "step": 6560 + }, + { + "epoch": 0.91, + "grad_norm": 2.215484781232322, + "learning_rate": 2.010960750248203e-07, + "loss": 0.2925, + "step": 6561 + }, + { + "epoch": 0.91, + "grad_norm": 2.6439755898500956, + "learning_rate": 2.0046448204927404e-07, + "loss": 0.2719, + "step": 6562 + }, + { + "epoch": 0.91, + "grad_norm": 2.7487240283092795, + "learning_rate": 1.9983386218275614e-07, + "loss": 0.3003, + "step": 6563 + }, + { + "epoch": 0.91, + "grad_norm": 3.279077410405548, + "learning_rate": 1.992042155531254e-07, + "loss": 0.2928, + "step": 6564 + }, + { + "epoch": 0.91, + "grad_norm": 2.150284268602392, + "learning_rate": 1.985755422880431e-07, + "loss": 0.2818, + "step": 6565 + }, + { + "epoch": 0.91, + "grad_norm": 2.4962602689258424, + "learning_rate": 1.979478425149739e-07, + "loss": 0.3018, + "step": 6566 + }, + { + "epoch": 0.91, + "grad_norm": 2.0004338522476584, + "learning_rate": 1.9732111636118268e-07, + "loss": 0.291, + "step": 6567 + }, + { + "epoch": 0.91, + "grad_norm": 2.683734171178302, + "learning_rate": 1.9669536395373945e-07, + "loss": 0.2888, + "step": 6568 + }, + { + "epoch": 0.91, + "grad_norm": 1.8723216064626511, + "learning_rate": 1.9607058541951717e-07, + "loss": 0.2903, + "step": 6569 + }, + { + "epoch": 0.91, + "grad_norm": 2.638328923745587, + "learning_rate": 1.9544678088518843e-07, + "loss": 0.2923, + "step": 6570 + }, + { + "epoch": 0.91, + "grad_norm": 2.9041316172747087, + "learning_rate": 1.9482395047723036e-07, + "loss": 0.3127, + "step": 6571 + }, + { + "epoch": 0.91, + "grad_norm": 3.7411959617056545, + "learning_rate": 1.9420209432192306e-07, + "loss": 0.3269, + "step": 6572 + }, + { + "epoch": 0.91, + "grad_norm": 2.4644630418401907, + "learning_rate": 1.935812125453479e-07, + "loss": 0.289, + "step": 6573 + }, + { + "epoch": 0.91, + "grad_norm": 2.4529657235979596, + "learning_rate": 1.9296130527338753e-07, + "loss": 0.302, + "step": 6574 + }, + { + "epoch": 0.91, + "grad_norm": 2.2978316567607453, + "learning_rate": 1.923423726317314e-07, + "loss": 0.2784, + "step": 6575 + }, + { + "epoch": 0.91, + "grad_norm": 2.788020449580225, + "learning_rate": 1.9172441474586635e-07, + "loss": 0.3142, + "step": 6576 + }, + { + "epoch": 0.91, + "grad_norm": 2.0384630493718765, + "learning_rate": 1.911074317410838e-07, + "loss": 0.2849, + "step": 6577 + }, + { + "epoch": 0.91, + "grad_norm": 2.317508818805263, + "learning_rate": 1.9049142374247875e-07, + "loss": 0.2927, + "step": 6578 + }, + { + "epoch": 0.91, + "grad_norm": 2.922567053263571, + "learning_rate": 1.898763908749468e-07, + "loss": 0.2942, + "step": 6579 + }, + { + "epoch": 0.91, + "grad_norm": 3.410375343858465, + "learning_rate": 1.8926233326318598e-07, + "loss": 0.3039, + "step": 6580 + }, + { + "epoch": 0.91, + "grad_norm": 2.352934067791198, + "learning_rate": 1.8864925103169618e-07, + "loss": 0.284, + "step": 6581 + }, + { + "epoch": 0.91, + "grad_norm": 2.5773615406158, + "learning_rate": 1.8803714430478183e-07, + "loss": 0.3042, + "step": 6582 + }, + { + "epoch": 0.92, + "grad_norm": 2.3575991384542307, + "learning_rate": 1.8742601320654697e-07, + "loss": 0.2786, + "step": 6583 + }, + { + "epoch": 0.92, + "grad_norm": 2.79229977206754, + "learning_rate": 1.8681585786089862e-07, + "loss": 0.3078, + "step": 6584 + }, + { + "epoch": 0.92, + "grad_norm": 3.9786924980833014, + "learning_rate": 1.8620667839154616e-07, + "loss": 0.3013, + "step": 6585 + }, + { + "epoch": 0.92, + "grad_norm": 3.044929373371704, + "learning_rate": 1.8559847492200245e-07, + "loss": 0.3169, + "step": 6586 + }, + { + "epoch": 0.92, + "grad_norm": 2.8557685124689707, + "learning_rate": 1.8499124757557995e-07, + "loss": 0.322, + "step": 6587 + }, + { + "epoch": 0.92, + "grad_norm": 2.3113055657605406, + "learning_rate": 1.8438499647539465e-07, + "loss": 0.303, + "step": 6588 + }, + { + "epoch": 0.92, + "grad_norm": 2.736184738592496, + "learning_rate": 1.8377972174436487e-07, + "loss": 0.2947, + "step": 6589 + }, + { + "epoch": 0.92, + "grad_norm": 1.942981428552251, + "learning_rate": 1.8317542350520967e-07, + "loss": 0.2916, + "step": 6590 + }, + { + "epoch": 0.92, + "grad_norm": 2.3041171539887797, + "learning_rate": 1.825721018804516e-07, + "loss": 0.2973, + "step": 6591 + }, + { + "epoch": 0.92, + "grad_norm": 2.053918198505069, + "learning_rate": 1.8196975699241336e-07, + "loss": 0.3268, + "step": 6592 + }, + { + "epoch": 0.92, + "grad_norm": 2.5531262730209803, + "learning_rate": 1.8136838896322227e-07, + "loss": 0.3064, + "step": 6593 + }, + { + "epoch": 0.92, + "grad_norm": 3.0454595593454385, + "learning_rate": 1.807679979148047e-07, + "loss": 0.272, + "step": 6594 + }, + { + "epoch": 0.92, + "grad_norm": 1.901763843756674, + "learning_rate": 1.8016858396889158e-07, + "loss": 0.2709, + "step": 6595 + }, + { + "epoch": 0.92, + "grad_norm": 3.443734815731879, + "learning_rate": 1.7957014724701404e-07, + "loss": 0.296, + "step": 6596 + }, + { + "epoch": 0.92, + "grad_norm": 1.8975321708874124, + "learning_rate": 1.7897268787050558e-07, + "loss": 0.2943, + "step": 6597 + }, + { + "epoch": 0.92, + "grad_norm": 2.118337655040421, + "learning_rate": 1.7837620596050154e-07, + "loss": 0.2911, + "step": 6598 + }, + { + "epoch": 0.92, + "grad_norm": 2.0263347800995475, + "learning_rate": 1.777807016379385e-07, + "loss": 0.3095, + "step": 6599 + }, + { + "epoch": 0.92, + "grad_norm": 3.842397364134635, + "learning_rate": 1.7718617502355594e-07, + "loss": 0.2748, + "step": 6600 + }, + { + "epoch": 0.92, + "grad_norm": 1.9261667913748226, + "learning_rate": 1.765926262378942e-07, + "loss": 0.2924, + "step": 6601 + }, + { + "epoch": 0.92, + "grad_norm": 3.15669869105199, + "learning_rate": 1.7600005540129528e-07, + "loss": 0.2696, + "step": 6602 + }, + { + "epoch": 0.92, + "grad_norm": 2.5469726838507643, + "learning_rate": 1.754084626339042e-07, + "loss": 0.306, + "step": 6603 + }, + { + "epoch": 0.92, + "grad_norm": 2.6604801500226296, + "learning_rate": 1.7481784805566672e-07, + "loss": 0.2639, + "step": 6604 + }, + { + "epoch": 0.92, + "grad_norm": 6.454812884853911, + "learning_rate": 1.7422821178632976e-07, + "loss": 0.2948, + "step": 6605 + }, + { + "epoch": 0.92, + "grad_norm": 3.4262374945024145, + "learning_rate": 1.736395539454422e-07, + "loss": 0.298, + "step": 6606 + }, + { + "epoch": 0.92, + "grad_norm": 2.583891584789574, + "learning_rate": 1.730518746523563e-07, + "loss": 0.3053, + "step": 6607 + }, + { + "epoch": 0.92, + "grad_norm": 3.8289954946675353, + "learning_rate": 1.7246517402622342e-07, + "loss": 0.3056, + "step": 6608 + }, + { + "epoch": 0.92, + "grad_norm": 3.555512759667357, + "learning_rate": 1.718794521859968e-07, + "loss": 0.3029, + "step": 6609 + }, + { + "epoch": 0.92, + "grad_norm": 2.0815189823175064, + "learning_rate": 1.7129470925043358e-07, + "loss": 0.293, + "step": 6610 + }, + { + "epoch": 0.92, + "grad_norm": 2.3459784395244645, + "learning_rate": 1.7071094533808895e-07, + "loss": 0.3141, + "step": 6611 + }, + { + "epoch": 0.92, + "grad_norm": 1.8465790248006062, + "learning_rate": 1.7012816056732273e-07, + "loss": 0.2994, + "step": 6612 + }, + { + "epoch": 0.92, + "grad_norm": 5.833800161764868, + "learning_rate": 1.695463550562948e-07, + "loss": 0.2962, + "step": 6613 + }, + { + "epoch": 0.92, + "grad_norm": 3.4545372811288533, + "learning_rate": 1.6896552892296637e-07, + "loss": 0.2839, + "step": 6614 + }, + { + "epoch": 0.92, + "grad_norm": 5.848605383756321, + "learning_rate": 1.6838568228510043e-07, + "loss": 0.3057, + "step": 6615 + }, + { + "epoch": 0.92, + "grad_norm": 2.6804220331987962, + "learning_rate": 1.6780681526026078e-07, + "loss": 0.3028, + "step": 6616 + }, + { + "epoch": 0.92, + "grad_norm": 3.133554751238899, + "learning_rate": 1.6722892796581401e-07, + "loss": 0.3039, + "step": 6617 + }, + { + "epoch": 0.92, + "grad_norm": 2.0993454430563436, + "learning_rate": 1.6665202051892647e-07, + "loss": 0.2896, + "step": 6618 + }, + { + "epoch": 0.92, + "grad_norm": 2.360523098246917, + "learning_rate": 1.6607609303656623e-07, + "loss": 0.2963, + "step": 6619 + }, + { + "epoch": 0.92, + "grad_norm": 2.3988234800736796, + "learning_rate": 1.655011456355038e-07, + "loss": 0.3086, + "step": 6620 + }, + { + "epoch": 0.92, + "grad_norm": 1.9327469879426178, + "learning_rate": 1.6492717843230978e-07, + "loss": 0.2929, + "step": 6621 + }, + { + "epoch": 0.92, + "grad_norm": 2.326094524296026, + "learning_rate": 1.6435419154335662e-07, + "loss": 0.2966, + "step": 6622 + }, + { + "epoch": 0.92, + "grad_norm": 2.856435158186086, + "learning_rate": 1.6378218508481702e-07, + "loss": 0.3083, + "step": 6623 + }, + { + "epoch": 0.92, + "grad_norm": 2.3071425593973287, + "learning_rate": 1.6321115917266704e-07, + "loss": 0.3182, + "step": 6624 + }, + { + "epoch": 0.92, + "grad_norm": 2.303634378798891, + "learning_rate": 1.626411139226819e-07, + "loss": 0.2833, + "step": 6625 + }, + { + "epoch": 0.92, + "grad_norm": 1.912493733907217, + "learning_rate": 1.6207204945043742e-07, + "loss": 0.3035, + "step": 6626 + }, + { + "epoch": 0.92, + "grad_norm": 2.125199602170809, + "learning_rate": 1.6150396587131412e-07, + "loss": 0.3217, + "step": 6627 + }, + { + "epoch": 0.92, + "grad_norm": 2.7646745429565103, + "learning_rate": 1.609368633004893e-07, + "loss": 0.3123, + "step": 6628 + }, + { + "epoch": 0.92, + "grad_norm": 2.2659873979936287, + "learning_rate": 1.6037074185294432e-07, + "loss": 0.2843, + "step": 6629 + }, + { + "epoch": 0.92, + "grad_norm": 2.0025231391775917, + "learning_rate": 1.5980560164346015e-07, + "loss": 0.3204, + "step": 6630 + }, + { + "epoch": 0.92, + "grad_norm": 2.6897994744008242, + "learning_rate": 1.5924144278662068e-07, + "loss": 0.2946, + "step": 6631 + }, + { + "epoch": 0.92, + "grad_norm": 2.0943818204074303, + "learning_rate": 1.586782653968083e-07, + "loss": 0.291, + "step": 6632 + }, + { + "epoch": 0.92, + "grad_norm": 3.8296035438782816, + "learning_rate": 1.5811606958820726e-07, + "loss": 0.323, + "step": 6633 + }, + { + "epoch": 0.92, + "grad_norm": 8.417707796058986, + "learning_rate": 1.575548554748041e-07, + "loss": 0.3152, + "step": 6634 + }, + { + "epoch": 0.92, + "grad_norm": 2.6037337179224203, + "learning_rate": 1.5699462317038562e-07, + "loss": 0.2979, + "step": 6635 + }, + { + "epoch": 0.92, + "grad_norm": 2.3597109219783454, + "learning_rate": 1.564353727885376e-07, + "loss": 0.2705, + "step": 6636 + }, + { + "epoch": 0.92, + "grad_norm": 2.0827816423970313, + "learning_rate": 1.558771044426499e-07, + "loss": 0.3011, + "step": 6637 + }, + { + "epoch": 0.92, + "grad_norm": 2.2066358889345645, + "learning_rate": 1.5531981824591147e-07, + "loss": 0.3093, + "step": 6638 + }, + { + "epoch": 0.92, + "grad_norm": 2.165601419125921, + "learning_rate": 1.54763514311313e-07, + "loss": 0.2846, + "step": 6639 + }, + { + "epoch": 0.92, + "grad_norm": 2.4188024049962404, + "learning_rate": 1.5420819275164368e-07, + "loss": 0.2969, + "step": 6640 + }, + { + "epoch": 0.92, + "grad_norm": 2.353070144181737, + "learning_rate": 1.5365385367949738e-07, + "loss": 0.318, + "step": 6641 + }, + { + "epoch": 0.92, + "grad_norm": 2.129349941640151, + "learning_rate": 1.5310049720726583e-07, + "loss": 0.3199, + "step": 6642 + }, + { + "epoch": 0.92, + "grad_norm": 2.1156819151872552, + "learning_rate": 1.525481234471421e-07, + "loss": 0.2928, + "step": 6643 + }, + { + "epoch": 0.92, + "grad_norm": 2.6534262134368696, + "learning_rate": 1.51996732511121e-07, + "loss": 0.3046, + "step": 6644 + }, + { + "epoch": 0.92, + "grad_norm": 2.214881936241278, + "learning_rate": 1.5144632451099706e-07, + "loss": 0.3014, + "step": 6645 + }, + { + "epoch": 0.92, + "grad_norm": 2.6994371407319235, + "learning_rate": 1.508968995583654e-07, + "loss": 0.3087, + "step": 6646 + }, + { + "epoch": 0.92, + "grad_norm": 2.1567002871374497, + "learning_rate": 1.5034845776462303e-07, + "loss": 0.2767, + "step": 6647 + }, + { + "epoch": 0.92, + "grad_norm": 2.215866223864004, + "learning_rate": 1.4980099924096714e-07, + "loss": 0.3122, + "step": 6648 + }, + { + "epoch": 0.92, + "grad_norm": 2.2944589110071765, + "learning_rate": 1.4925452409839446e-07, + "loss": 0.2856, + "step": 6649 + }, + { + "epoch": 0.92, + "grad_norm": 2.1104317627413103, + "learning_rate": 1.4870903244770307e-07, + "loss": 0.2825, + "step": 6650 + }, + { + "epoch": 0.92, + "grad_norm": 2.0796908288803695, + "learning_rate": 1.4816452439949335e-07, + "loss": 0.2613, + "step": 6651 + }, + { + "epoch": 0.92, + "grad_norm": 2.324513317595476, + "learning_rate": 1.4762100006416314e-07, + "loss": 0.3188, + "step": 6652 + }, + { + "epoch": 0.92, + "grad_norm": 2.0111273591643184, + "learning_rate": 1.4707845955191202e-07, + "loss": 0.3055, + "step": 6653 + }, + { + "epoch": 0.93, + "grad_norm": 2.8203568397744725, + "learning_rate": 1.4653690297274203e-07, + "loss": 0.3148, + "step": 6654 + }, + { + "epoch": 0.93, + "grad_norm": 1.9662540131089774, + "learning_rate": 1.4599633043645311e-07, + "loss": 0.28, + "step": 6655 + }, + { + "epoch": 0.93, + "grad_norm": 2.8438913167099926, + "learning_rate": 1.4545674205264704e-07, + "loss": 0.3245, + "step": 6656 + }, + { + "epoch": 0.93, + "grad_norm": 2.297574344614063, + "learning_rate": 1.4491813793072517e-07, + "loss": 0.2868, + "step": 6657 + }, + { + "epoch": 0.93, + "grad_norm": 5.683572241477095, + "learning_rate": 1.443805181798913e-07, + "loss": 0.2882, + "step": 6658 + }, + { + "epoch": 0.93, + "grad_norm": 1.9405308376344368, + "learning_rate": 1.4384388290914653e-07, + "loss": 0.2899, + "step": 6659 + }, + { + "epoch": 0.93, + "grad_norm": 2.499550946788578, + "learning_rate": 1.4330823222729496e-07, + "loss": 0.2941, + "step": 6660 + }, + { + "epoch": 0.93, + "grad_norm": 1.875257615546078, + "learning_rate": 1.4277356624293915e-07, + "loss": 0.2807, + "step": 6661 + }, + { + "epoch": 0.93, + "grad_norm": 2.461209490391293, + "learning_rate": 1.4223988506448405e-07, + "loss": 0.292, + "step": 6662 + }, + { + "epoch": 0.93, + "grad_norm": 1.9842336322357437, + "learning_rate": 1.417071888001331e-07, + "loss": 0.2836, + "step": 6663 + }, + { + "epoch": 0.93, + "grad_norm": 3.2869097304475785, + "learning_rate": 1.4117547755789108e-07, + "loss": 0.2835, + "step": 6664 + }, + { + "epoch": 0.93, + "grad_norm": 2.6448775382567224, + "learning_rate": 1.406447514455639e-07, + "loss": 0.2911, + "step": 6665 + }, + { + "epoch": 0.93, + "grad_norm": 2.1439641907067455, + "learning_rate": 1.4011501057075493e-07, + "loss": 0.303, + "step": 6666 + }, + { + "epoch": 0.93, + "grad_norm": 2.323935641584942, + "learning_rate": 1.3958625504087055e-07, + "loss": 0.3035, + "step": 6667 + }, + { + "epoch": 0.93, + "grad_norm": 2.2906765034881236, + "learning_rate": 1.3905848496311546e-07, + "loss": 0.2674, + "step": 6668 + }, + { + "epoch": 0.93, + "grad_norm": 2.27796528236284, + "learning_rate": 1.3853170044449638e-07, + "loss": 0.2874, + "step": 6669 + }, + { + "epoch": 0.93, + "grad_norm": 2.3836561954428714, + "learning_rate": 1.380059015918189e-07, + "loss": 0.3199, + "step": 6670 + }, + { + "epoch": 0.93, + "grad_norm": 1.9535405642952657, + "learning_rate": 1.374810885116884e-07, + "loss": 0.2946, + "step": 6671 + }, + { + "epoch": 0.93, + "grad_norm": 2.261258878219232, + "learning_rate": 1.3695726131051134e-07, + "loss": 0.2813, + "step": 6672 + }, + { + "epoch": 0.93, + "grad_norm": 2.055334139188353, + "learning_rate": 1.3643442009449503e-07, + "loss": 0.2923, + "step": 6673 + }, + { + "epoch": 0.93, + "grad_norm": 2.8279735816975684, + "learning_rate": 1.3591256496964522e-07, + "loss": 0.3083, + "step": 6674 + }, + { + "epoch": 0.93, + "grad_norm": 2.104382138996217, + "learning_rate": 1.3539169604176728e-07, + "loss": 0.2963, + "step": 6675 + }, + { + "epoch": 0.93, + "grad_norm": 2.9095253865410515, + "learning_rate": 1.3487181341647003e-07, + "loss": 0.2856, + "step": 6676 + }, + { + "epoch": 0.93, + "grad_norm": 2.634906019788433, + "learning_rate": 1.343529171991581e-07, + "loss": 0.2996, + "step": 6677 + }, + { + "epoch": 0.93, + "grad_norm": 3.02635338151754, + "learning_rate": 1.338350074950384e-07, + "loss": 0.3028, + "step": 6678 + }, + { + "epoch": 0.93, + "grad_norm": 2.5830915632261506, + "learning_rate": 1.3331808440911865e-07, + "loss": 0.2917, + "step": 6679 + }, + { + "epoch": 0.93, + "grad_norm": 2.2266509557852894, + "learning_rate": 1.3280214804620327e-07, + "loss": 0.2821, + "step": 6680 + }, + { + "epoch": 0.93, + "grad_norm": 2.2464253135662333, + "learning_rate": 1.322871985109009e-07, + "loss": 0.2974, + "step": 6681 + }, + { + "epoch": 0.93, + "grad_norm": 2.2937717544746508, + "learning_rate": 1.3177323590761627e-07, + "loss": 0.3187, + "step": 6682 + }, + { + "epoch": 0.93, + "grad_norm": 2.2567616755146367, + "learning_rate": 1.312602603405566e-07, + "loss": 0.288, + "step": 6683 + }, + { + "epoch": 0.93, + "grad_norm": 2.387281556065795, + "learning_rate": 1.307482719137282e-07, + "loss": 0.301, + "step": 6684 + }, + { + "epoch": 0.93, + "grad_norm": 2.2340086959610943, + "learning_rate": 1.3023727073093573e-07, + "loss": 0.2729, + "step": 6685 + }, + { + "epoch": 0.93, + "grad_norm": 17.373797855228748, + "learning_rate": 1.297272568957858e-07, + "loss": 0.3052, + "step": 6686 + }, + { + "epoch": 0.93, + "grad_norm": 2.6641760554720264, + "learning_rate": 1.2921823051168514e-07, + "loss": 0.3053, + "step": 6687 + }, + { + "epoch": 0.93, + "grad_norm": 2.0499964941072553, + "learning_rate": 1.2871019168183674e-07, + "loss": 0.2687, + "step": 6688 + }, + { + "epoch": 0.93, + "grad_norm": 2.38850517918126, + "learning_rate": 1.2820314050924765e-07, + "loss": 0.2942, + "step": 6689 + }, + { + "epoch": 0.93, + "grad_norm": 2.348254702571419, + "learning_rate": 1.2769707709672285e-07, + "loss": 0.2886, + "step": 6690 + }, + { + "epoch": 0.93, + "grad_norm": 3.8353898949353153, + "learning_rate": 1.2719200154686694e-07, + "loss": 0.3107, + "step": 6691 + }, + { + "epoch": 0.93, + "grad_norm": 2.1029515181390757, + "learning_rate": 1.2668791396208303e-07, + "loss": 0.3246, + "step": 6692 + }, + { + "epoch": 0.93, + "grad_norm": 2.089034323928231, + "learning_rate": 1.2618481444457708e-07, + "loss": 0.3096, + "step": 6693 + }, + { + "epoch": 0.93, + "grad_norm": 1.9187672221657879, + "learning_rate": 1.2568270309635144e-07, + "loss": 0.3027, + "step": 6694 + }, + { + "epoch": 0.93, + "grad_norm": 2.093660747246413, + "learning_rate": 1.2518158001920965e-07, + "loss": 0.3079, + "step": 6695 + }, + { + "epoch": 0.93, + "grad_norm": 2.3269619874192125, + "learning_rate": 1.2468144531475602e-07, + "loss": 0.3008, + "step": 6696 + }, + { + "epoch": 0.93, + "grad_norm": 2.8785729998987573, + "learning_rate": 1.241822990843916e-07, + "loss": 0.3206, + "step": 6697 + }, + { + "epoch": 0.93, + "grad_norm": 2.109772095333126, + "learning_rate": 1.2368414142931938e-07, + "loss": 0.315, + "step": 6698 + }, + { + "epoch": 0.93, + "grad_norm": 1.9737834182640874, + "learning_rate": 1.2318697245054078e-07, + "loss": 0.3154, + "step": 6699 + }, + { + "epoch": 0.93, + "grad_norm": 2.209929721964194, + "learning_rate": 1.226907922488585e-07, + "loss": 0.2961, + "step": 6700 + }, + { + "epoch": 0.93, + "grad_norm": 2.2382208674332156, + "learning_rate": 1.2219560092487149e-07, + "loss": 0.2789, + "step": 6701 + }, + { + "epoch": 0.93, + "grad_norm": 1.8759624146752834, + "learning_rate": 1.217013985789811e-07, + "loss": 0.2579, + "step": 6702 + }, + { + "epoch": 0.93, + "grad_norm": 2.1645593799306684, + "learning_rate": 1.2120818531138722e-07, + "loss": 0.2945, + "step": 6703 + }, + { + "epoch": 0.93, + "grad_norm": 2.321333362025475, + "learning_rate": 1.2071596122208928e-07, + "loss": 0.306, + "step": 6704 + }, + { + "epoch": 0.93, + "grad_norm": 3.919491277429195, + "learning_rate": 1.2022472641088522e-07, + "loss": 0.305, + "step": 6705 + }, + { + "epoch": 0.93, + "grad_norm": 2.5687268747084455, + "learning_rate": 1.1973448097737427e-07, + "loss": 0.3053, + "step": 6706 + }, + { + "epoch": 0.93, + "grad_norm": 2.266663376157991, + "learning_rate": 1.1924522502095359e-07, + "loss": 0.2936, + "step": 6707 + }, + { + "epoch": 0.93, + "grad_norm": 2.199666455480945, + "learning_rate": 1.1875695864081993e-07, + "loss": 0.2835, + "step": 6708 + }, + { + "epoch": 0.93, + "grad_norm": 2.11511796428247, + "learning_rate": 1.1826968193596911e-07, + "loss": 0.2962, + "step": 6709 + }, + { + "epoch": 0.93, + "grad_norm": 2.937888858942829, + "learning_rate": 1.1778339500519875e-07, + "loss": 0.3134, + "step": 6710 + }, + { + "epoch": 0.93, + "grad_norm": 1.787835549300808, + "learning_rate": 1.1729809794710223e-07, + "loss": 0.3121, + "step": 6711 + }, + { + "epoch": 0.93, + "grad_norm": 5.974628688664136, + "learning_rate": 1.1681379086007361e-07, + "loss": 0.3167, + "step": 6712 + }, + { + "epoch": 0.93, + "grad_norm": 2.2750469880217246, + "learning_rate": 1.1633047384230767e-07, + "loss": 0.2807, + "step": 6713 + }, + { + "epoch": 0.93, + "grad_norm": 1.9719498417629115, + "learning_rate": 1.1584814699179659e-07, + "loss": 0.312, + "step": 6714 + }, + { + "epoch": 0.93, + "grad_norm": 2.2714067915653895, + "learning_rate": 1.1536681040633214e-07, + "loss": 0.2934, + "step": 6715 + }, + { + "epoch": 0.93, + "grad_norm": 2.609635388327927, + "learning_rate": 1.1488646418350568e-07, + "loss": 0.322, + "step": 6716 + }, + { + "epoch": 0.93, + "grad_norm": 2.2786725156522243, + "learning_rate": 1.1440710842070935e-07, + "loss": 0.3097, + "step": 6717 + }, + { + "epoch": 0.93, + "grad_norm": 7.045816900720868, + "learning_rate": 1.1392874321513092e-07, + "loss": 0.2878, + "step": 6718 + }, + { + "epoch": 0.93, + "grad_norm": 2.1042762018082657, + "learning_rate": 1.134513686637595e-07, + "loss": 0.3207, + "step": 6719 + }, + { + "epoch": 0.93, + "grad_norm": 2.848954723478304, + "learning_rate": 1.1297498486338432e-07, + "loss": 0.2874, + "step": 6720 + }, + { + "epoch": 0.93, + "grad_norm": 2.0707537903487685, + "learning_rate": 1.1249959191059146e-07, + "loss": 0.3115, + "step": 6721 + }, + { + "epoch": 0.93, + "grad_norm": 2.6547124590725577, + "learning_rate": 1.1202518990176714e-07, + "loss": 0.3224, + "step": 6722 + }, + { + "epoch": 0.93, + "grad_norm": 2.403450922488013, + "learning_rate": 1.1155177893309665e-07, + "loss": 0.3092, + "step": 6723 + }, + { + "epoch": 0.93, + "grad_norm": 2.5972123666152416, + "learning_rate": 1.1107935910056545e-07, + "loss": 0.2594, + "step": 6724 + }, + { + "epoch": 0.93, + "grad_norm": 2.2021306002743053, + "learning_rate": 1.1060793049995578e-07, + "loss": 0.2807, + "step": 6725 + }, + { + "epoch": 0.94, + "grad_norm": 2.0028063007691075, + "learning_rate": 1.1013749322685063e-07, + "loss": 0.2803, + "step": 6726 + }, + { + "epoch": 0.94, + "grad_norm": 2.307741725711019, + "learning_rate": 1.096680473766315e-07, + "loss": 0.2913, + "step": 6727 + }, + { + "epoch": 0.94, + "grad_norm": 2.273162471224583, + "learning_rate": 1.0919959304447891e-07, + "loss": 0.3075, + "step": 6728 + }, + { + "epoch": 0.94, + "grad_norm": 2.306306208768607, + "learning_rate": 1.0873213032537133e-07, + "loss": 0.299, + "step": 6729 + }, + { + "epoch": 0.94, + "grad_norm": 1.9097255390906251, + "learning_rate": 1.082656593140885e-07, + "loss": 0.2914, + "step": 6730 + }, + { + "epoch": 0.94, + "grad_norm": 2.2012054581329963, + "learning_rate": 1.0780018010520754e-07, + "loss": 0.292, + "step": 6731 + }, + { + "epoch": 0.94, + "grad_norm": 2.0528505645206088, + "learning_rate": 1.0733569279310297e-07, + "loss": 0.296, + "step": 6732 + }, + { + "epoch": 0.94, + "grad_norm": 2.1573588369224184, + "learning_rate": 1.0687219747195165e-07, + "loss": 0.3123, + "step": 6733 + }, + { + "epoch": 0.94, + "grad_norm": 3.0000541027088254, + "learning_rate": 1.0640969423572789e-07, + "loss": 0.3106, + "step": 6734 + }, + { + "epoch": 0.94, + "grad_norm": 2.4687253903362563, + "learning_rate": 1.0594818317820332e-07, + "loss": 0.2957, + "step": 6735 + }, + { + "epoch": 0.94, + "grad_norm": 1.9314399470261374, + "learning_rate": 1.0548766439295033e-07, + "loss": 0.296, + "step": 6736 + }, + { + "epoch": 0.94, + "grad_norm": 2.579004782728943, + "learning_rate": 1.0502813797333866e-07, + "loss": 0.3211, + "step": 6737 + }, + { + "epoch": 0.94, + "grad_norm": 4.429944046370937, + "learning_rate": 1.0456960401253824e-07, + "loss": 0.3051, + "step": 6738 + }, + { + "epoch": 0.94, + "grad_norm": 2.602229966221889, + "learning_rate": 1.0411206260351747e-07, + "loss": 0.2957, + "step": 6739 + }, + { + "epoch": 0.94, + "grad_norm": 2.222959405547427, + "learning_rate": 1.036555138390416e-07, + "loss": 0.3037, + "step": 6740 + }, + { + "epoch": 0.94, + "grad_norm": 3.301268159993142, + "learning_rate": 1.0319995781167769e-07, + "loss": 0.3105, + "step": 6741 + }, + { + "epoch": 0.94, + "grad_norm": 1.9243671212805158, + "learning_rate": 1.0274539461378962e-07, + "loss": 0.3179, + "step": 6742 + }, + { + "epoch": 0.94, + "grad_norm": 2.0377601996460974, + "learning_rate": 1.0229182433754092e-07, + "loss": 0.2979, + "step": 6743 + }, + { + "epoch": 0.94, + "grad_norm": 2.119928760296577, + "learning_rate": 1.018392470748919e-07, + "loss": 0.308, + "step": 6744 + }, + { + "epoch": 0.94, + "grad_norm": 1.9344421045988756, + "learning_rate": 1.0138766291760416e-07, + "loss": 0.2944, + "step": 6745 + }, + { + "epoch": 0.94, + "grad_norm": 2.003345770980031, + "learning_rate": 1.0093707195723668e-07, + "loss": 0.2968, + "step": 6746 + }, + { + "epoch": 0.94, + "grad_norm": 2.739997775095551, + "learning_rate": 1.0048747428514582e-07, + "loss": 0.3024, + "step": 6747 + }, + { + "epoch": 0.94, + "grad_norm": 2.4053149080676652, + "learning_rate": 1.0003886999248924e-07, + "loss": 0.3224, + "step": 6748 + }, + { + "epoch": 0.94, + "grad_norm": 2.197657844364964, + "learning_rate": 9.959125917022028e-08, + "loss": 0.2811, + "step": 6749 + }, + { + "epoch": 0.94, + "grad_norm": 3.346853336442677, + "learning_rate": 9.914464190909412e-08, + "loss": 0.3091, + "step": 6750 + }, + { + "epoch": 0.94, + "grad_norm": 4.1620141530271715, + "learning_rate": 9.869901829966111e-08, + "loss": 0.2974, + "step": 6751 + }, + { + "epoch": 0.94, + "grad_norm": 2.3900128575240545, + "learning_rate": 9.825438843227287e-08, + "loss": 0.2748, + "step": 6752 + }, + { + "epoch": 0.94, + "grad_norm": 2.994184425319869, + "learning_rate": 9.781075239707838e-08, + "loss": 0.3066, + "step": 6753 + }, + { + "epoch": 0.94, + "grad_norm": 3.0149444210012453, + "learning_rate": 9.736811028402349e-08, + "loss": 0.2769, + "step": 6754 + }, + { + "epoch": 0.94, + "grad_norm": 2.0233051602818444, + "learning_rate": 9.69264621828564e-08, + "loss": 0.3267, + "step": 6755 + }, + { + "epoch": 0.94, + "grad_norm": 1.997872400476646, + "learning_rate": 9.648580818312048e-08, + "loss": 0.3042, + "step": 6756 + }, + { + "epoch": 0.94, + "grad_norm": 4.492861053480205, + "learning_rate": 9.604614837415815e-08, + "loss": 0.313, + "step": 6757 + }, + { + "epoch": 0.94, + "grad_norm": 1.8361232103668084, + "learning_rate": 9.560748284511146e-08, + "loss": 0.2882, + "step": 6758 + }, + { + "epoch": 0.94, + "grad_norm": 2.528299221419997, + "learning_rate": 9.516981168492034e-08, + "loss": 0.3077, + "step": 6759 + }, + { + "epoch": 0.94, + "grad_norm": 2.916224913550913, + "learning_rate": 9.473313498232218e-08, + "loss": 0.3132, + "step": 6760 + }, + { + "epoch": 0.94, + "grad_norm": 2.4822033182729215, + "learning_rate": 9.429745282585389e-08, + "loss": 0.3163, + "step": 6761 + }, + { + "epoch": 0.94, + "grad_norm": 2.0243094659060166, + "learning_rate": 9.386276530385042e-08, + "loss": 0.3123, + "step": 6762 + }, + { + "epoch": 0.94, + "grad_norm": 2.428912713897224, + "learning_rate": 9.34290725044451e-08, + "loss": 0.3005, + "step": 6763 + }, + { + "epoch": 0.94, + "grad_norm": 2.6265977775802316, + "learning_rate": 9.29963745155682e-08, + "loss": 0.2933, + "step": 6764 + }, + { + "epoch": 0.94, + "grad_norm": 1.9681028555571702, + "learning_rate": 9.256467142495118e-08, + "loss": 0.2835, + "step": 6765 + }, + { + "epoch": 0.94, + "grad_norm": 2.2341642315524854, + "learning_rate": 9.213396332012126e-08, + "loss": 0.303, + "step": 6766 + }, + { + "epoch": 0.94, + "grad_norm": 1.945209495764932, + "learning_rate": 9.170425028840468e-08, + "loss": 0.3309, + "step": 6767 + }, + { + "epoch": 0.94, + "grad_norm": 4.575985677035671, + "learning_rate": 9.127553241692621e-08, + "loss": 0.3239, + "step": 6768 + }, + { + "epoch": 0.94, + "grad_norm": 2.0068983426176854, + "learning_rate": 9.08478097926091e-08, + "loss": 0.3061, + "step": 6769 + }, + { + "epoch": 0.94, + "grad_norm": 2.024785932521108, + "learning_rate": 9.042108250217396e-08, + "loss": 0.2817, + "step": 6770 + }, + { + "epoch": 0.94, + "grad_norm": 2.2163368919225426, + "learning_rate": 8.999535063213994e-08, + "loss": 0.2765, + "step": 6771 + }, + { + "epoch": 0.94, + "grad_norm": 1.8958156722606223, + "learning_rate": 8.957061426882518e-08, + "loss": 0.3085, + "step": 6772 + }, + { + "epoch": 0.94, + "grad_norm": 2.5144805951484925, + "learning_rate": 8.914687349834473e-08, + "loss": 0.301, + "step": 6773 + }, + { + "epoch": 0.94, + "grad_norm": 5.0312851188720344, + "learning_rate": 8.872412840661149e-08, + "loss": 0.289, + "step": 6774 + }, + { + "epoch": 0.94, + "grad_norm": 2.417468828278365, + "learning_rate": 8.830237907933859e-08, + "loss": 0.3038, + "step": 6775 + }, + { + "epoch": 0.94, + "grad_norm": 1.950530596571687, + "learning_rate": 8.788162560203594e-08, + "loss": 0.2955, + "step": 6776 + }, + { + "epoch": 0.94, + "grad_norm": 2.12187814753126, + "learning_rate": 8.746186806001088e-08, + "loss": 0.3298, + "step": 6777 + }, + { + "epoch": 0.94, + "grad_norm": 4.003535088878299, + "learning_rate": 8.704310653836978e-08, + "loss": 0.2924, + "step": 6778 + }, + { + "epoch": 0.94, + "grad_norm": 2.4624674366052153, + "learning_rate": 8.662534112201748e-08, + "loss": 0.285, + "step": 6779 + }, + { + "epoch": 0.94, + "grad_norm": 2.5709516163138417, + "learning_rate": 8.620857189565512e-08, + "loss": 0.2825, + "step": 6780 + }, + { + "epoch": 0.94, + "grad_norm": 4.167346530371781, + "learning_rate": 8.579279894378289e-08, + "loss": 0.3099, + "step": 6781 + }, + { + "epoch": 0.94, + "grad_norm": 2.5739099565023604, + "learning_rate": 8.537802235070058e-08, + "loss": 0.2843, + "step": 6782 + }, + { + "epoch": 0.94, + "grad_norm": 5.248253309426239, + "learning_rate": 8.49642422005026e-08, + "loss": 0.2912, + "step": 6783 + }, + { + "epoch": 0.94, + "grad_norm": 2.571745599779672, + "learning_rate": 8.455145857708402e-08, + "loss": 0.3021, + "step": 6784 + }, + { + "epoch": 0.94, + "grad_norm": 7.853159646172959, + "learning_rate": 8.413967156413738e-08, + "loss": 0.312, + "step": 6785 + }, + { + "epoch": 0.94, + "grad_norm": 2.9148535026720817, + "learning_rate": 8.3728881245152e-08, + "loss": 0.2642, + "step": 6786 + }, + { + "epoch": 0.94, + "grad_norm": 1.8104617612330733, + "learning_rate": 8.331908770341624e-08, + "loss": 0.2792, + "step": 6787 + }, + { + "epoch": 0.94, + "grad_norm": 3.8256827995259273, + "learning_rate": 8.291029102201587e-08, + "loss": 0.3083, + "step": 6788 + }, + { + "epoch": 0.94, + "grad_norm": 4.615013569627178, + "learning_rate": 8.250249128383514e-08, + "loss": 0.2758, + "step": 6789 + }, + { + "epoch": 0.94, + "grad_norm": 2.346566408564908, + "learning_rate": 8.209568857155569e-08, + "loss": 0.2905, + "step": 6790 + }, + { + "epoch": 0.94, + "grad_norm": 2.0785798561662934, + "learning_rate": 8.1689882967656e-08, + "loss": 0.3093, + "step": 6791 + }, + { + "epoch": 0.94, + "grad_norm": 2.073528814338229, + "learning_rate": 8.128507455441415e-08, + "loss": 0.3112, + "step": 6792 + }, + { + "epoch": 0.94, + "grad_norm": 1.9138873068747364, + "learning_rate": 8.088126341390557e-08, + "loss": 0.2787, + "step": 6793 + }, + { + "epoch": 0.94, + "grad_norm": 1.7430003540637267, + "learning_rate": 8.04784496280031e-08, + "loss": 0.2717, + "step": 6794 + }, + { + "epoch": 0.94, + "grad_norm": 2.445667521418728, + "learning_rate": 8.007663327837755e-08, + "loss": 0.2663, + "step": 6795 + }, + { + "epoch": 0.94, + "grad_norm": 2.942137175344497, + "learning_rate": 7.967581444649708e-08, + "loss": 0.2746, + "step": 6796 + }, + { + "epoch": 0.94, + "grad_norm": 6.536976101197265, + "learning_rate": 7.927599321362833e-08, + "loss": 0.281, + "step": 6797 + }, + { + "epoch": 0.95, + "grad_norm": 2.3285198279643304, + "learning_rate": 7.88771696608348e-08, + "loss": 0.3143, + "step": 6798 + }, + { + "epoch": 0.95, + "grad_norm": 3.0108432761153687, + "learning_rate": 7.847934386897904e-08, + "loss": 0.2827, + "step": 6799 + }, + { + "epoch": 0.95, + "grad_norm": 2.729789999222763, + "learning_rate": 7.80825159187204e-08, + "loss": 0.2888, + "step": 6800 + }, + { + "epoch": 0.95, + "grad_norm": 2.7025838400268927, + "learning_rate": 7.768668589051453e-08, + "loss": 0.2806, + "step": 6801 + }, + { + "epoch": 0.95, + "grad_norm": 2.749697312148453, + "learning_rate": 7.729185386461835e-08, + "loss": 0.266, + "step": 6802 + }, + { + "epoch": 0.95, + "grad_norm": 2.0175336762888314, + "learning_rate": 7.689801992108281e-08, + "loss": 0.315, + "step": 6803 + }, + { + "epoch": 0.95, + "grad_norm": 2.6232587734186694, + "learning_rate": 7.650518413975905e-08, + "loss": 0.3188, + "step": 6804 + }, + { + "epoch": 0.95, + "grad_norm": 13.789275971787571, + "learning_rate": 7.61133466002939e-08, + "loss": 0.2864, + "step": 6805 + }, + { + "epoch": 0.95, + "grad_norm": 2.196466039622626, + "learning_rate": 7.572250738213327e-08, + "loss": 0.2906, + "step": 6806 + }, + { + "epoch": 0.95, + "grad_norm": 5.393221923794878, + "learning_rate": 7.533266656452043e-08, + "loss": 0.3013, + "step": 6807 + }, + { + "epoch": 0.95, + "grad_norm": 2.1877480150786517, + "learning_rate": 7.494382422649494e-08, + "loss": 0.3185, + "step": 6808 + }, + { + "epoch": 0.95, + "grad_norm": 2.3285323121864114, + "learning_rate": 7.455598044689482e-08, + "loss": 0.2822, + "step": 6809 + }, + { + "epoch": 0.95, + "grad_norm": 2.5696815411168563, + "learning_rate": 7.416913530435554e-08, + "loss": 0.3001, + "step": 6810 + }, + { + "epoch": 0.95, + "grad_norm": 2.014515443475921, + "learning_rate": 7.378328887731156e-08, + "loss": 0.295, + "step": 6811 + }, + { + "epoch": 0.95, + "grad_norm": 1.9547424532817295, + "learning_rate": 7.339844124399309e-08, + "loss": 0.2949, + "step": 6812 + }, + { + "epoch": 0.95, + "grad_norm": 2.50090747618428, + "learning_rate": 7.30145924824266e-08, + "loss": 0.2817, + "step": 6813 + }, + { + "epoch": 0.95, + "grad_norm": 3.3203809504662978, + "learning_rate": 7.263174267043981e-08, + "loss": 0.2853, + "step": 6814 + }, + { + "epoch": 0.95, + "grad_norm": 2.6358194636191468, + "learning_rate": 7.224989188565513e-08, + "loss": 0.3182, + "step": 6815 + }, + { + "epoch": 0.95, + "grad_norm": 3.1037618680787262, + "learning_rate": 7.186904020549168e-08, + "loss": 0.2885, + "step": 6816 + }, + { + "epoch": 0.95, + "grad_norm": 2.182438890524169, + "learning_rate": 7.14891877071694e-08, + "loss": 0.272, + "step": 6817 + }, + { + "epoch": 0.95, + "grad_norm": 2.222155668062196, + "learning_rate": 7.111033446770221e-08, + "loss": 0.3139, + "step": 6818 + }, + { + "epoch": 0.95, + "grad_norm": 2.0961081629719636, + "learning_rate": 7.07324805639037e-08, + "loss": 0.2832, + "step": 6819 + }, + { + "epoch": 0.95, + "grad_norm": 2.093091381176708, + "learning_rate": 7.035562607238311e-08, + "loss": 0.2829, + "step": 6820 + }, + { + "epoch": 0.95, + "grad_norm": 2.5001622638314434, + "learning_rate": 6.997977106954934e-08, + "loss": 0.3116, + "step": 6821 + }, + { + "epoch": 0.95, + "grad_norm": 3.126719147601388, + "learning_rate": 6.960491563160643e-08, + "loss": 0.2841, + "step": 6822 + }, + { + "epoch": 0.95, + "grad_norm": 1.9950374632089825, + "learning_rate": 6.92310598345558e-08, + "loss": 0.269, + "step": 6823 + }, + { + "epoch": 0.95, + "grad_norm": 1.9180826637119264, + "learning_rate": 6.88582037541985e-08, + "loss": 0.2974, + "step": 6824 + }, + { + "epoch": 0.95, + "grad_norm": 2.3109008871748595, + "learning_rate": 6.848634746613014e-08, + "loss": 0.2845, + "step": 6825 + }, + { + "epoch": 0.95, + "grad_norm": 2.3097965015647715, + "learning_rate": 6.811549104574544e-08, + "loss": 0.3124, + "step": 6826 + }, + { + "epoch": 0.95, + "grad_norm": 3.0382762304020163, + "learning_rate": 6.774563456823535e-08, + "loss": 0.3027, + "step": 6827 + }, + { + "epoch": 0.95, + "grad_norm": 2.6331306005832262, + "learning_rate": 6.737677810858879e-08, + "loss": 0.304, + "step": 6828 + }, + { + "epoch": 0.95, + "grad_norm": 3.2061474266897467, + "learning_rate": 6.700892174159202e-08, + "loss": 0.3034, + "step": 6829 + }, + { + "epoch": 0.95, + "grad_norm": 2.671662850381845, + "learning_rate": 6.664206554182762e-08, + "loss": 0.3082, + "step": 6830 + }, + { + "epoch": 0.95, + "grad_norm": 3.003696429237943, + "learning_rate": 6.627620958367609e-08, + "loss": 0.3053, + "step": 6831 + }, + { + "epoch": 0.95, + "grad_norm": 2.365821613601773, + "learning_rate": 6.591135394131477e-08, + "loss": 0.3022, + "step": 6832 + }, + { + "epoch": 0.95, + "grad_norm": 2.165289958611475, + "learning_rate": 6.554749868871891e-08, + "loss": 0.2852, + "step": 6833 + }, + { + "epoch": 0.95, + "grad_norm": 2.120203334372289, + "learning_rate": 6.518464389966006e-08, + "loss": 0.3192, + "step": 6834 + }, + { + "epoch": 0.95, + "grad_norm": 4.072090853866627, + "learning_rate": 6.482278964770717e-08, + "loss": 0.2888, + "step": 6835 + }, + { + "epoch": 0.95, + "grad_norm": 2.3542147326745346, + "learning_rate": 6.446193600622652e-08, + "loss": 0.3123, + "step": 6836 + }, + { + "epoch": 0.95, + "grad_norm": 3.5959557139226312, + "learning_rate": 6.410208304838128e-08, + "loss": 0.2786, + "step": 6837 + }, + { + "epoch": 0.95, + "grad_norm": 5.653286408648208, + "learning_rate": 6.374323084713252e-08, + "loss": 0.3102, + "step": 6838 + }, + { + "epoch": 0.95, + "grad_norm": 1.9895158870812844, + "learning_rate": 6.338537947523704e-08, + "loss": 0.3045, + "step": 6839 + }, + { + "epoch": 0.95, + "grad_norm": 8.064763795527913, + "learning_rate": 6.302852900524958e-08, + "loss": 0.31, + "step": 6840 + }, + { + "epoch": 0.95, + "grad_norm": 2.6960076196167546, + "learning_rate": 6.267267950952228e-08, + "loss": 0.3039, + "step": 6841 + }, + { + "epoch": 0.95, + "grad_norm": 2.403719077279177, + "learning_rate": 6.23178310602035e-08, + "loss": 0.3001, + "step": 6842 + }, + { + "epoch": 0.95, + "grad_norm": 2.1885570719731486, + "learning_rate": 6.196398372923851e-08, + "loss": 0.2949, + "step": 6843 + }, + { + "epoch": 0.95, + "grad_norm": 2.3588232351991065, + "learning_rate": 6.1611137588371e-08, + "loss": 0.2878, + "step": 6844 + }, + { + "epoch": 0.95, + "grad_norm": 3.126719688384761, + "learning_rate": 6.125929270914044e-08, + "loss": 0.2962, + "step": 6845 + }, + { + "epoch": 0.95, + "grad_norm": 2.618808058148791, + "learning_rate": 6.090844916288363e-08, + "loss": 0.2948, + "step": 6846 + }, + { + "epoch": 0.95, + "grad_norm": 2.3258038272419754, + "learning_rate": 6.055860702073368e-08, + "loss": 0.2962, + "step": 6847 + }, + { + "epoch": 0.95, + "grad_norm": 5.135320610079998, + "learning_rate": 6.020976635362219e-08, + "loss": 0.3104, + "step": 6848 + }, + { + "epoch": 0.95, + "grad_norm": 2.2877364747139235, + "learning_rate": 5.986192723227701e-08, + "loss": 0.3002, + "step": 6849 + }, + { + "epoch": 0.95, + "grad_norm": 2.411221501632654, + "learning_rate": 5.9515089727221195e-08, + "loss": 0.3023, + "step": 6850 + }, + { + "epoch": 0.95, + "grad_norm": 2.6105929141447026, + "learning_rate": 5.9169253908778476e-08, + "loss": 0.3061, + "step": 6851 + }, + { + "epoch": 0.95, + "grad_norm": 2.2552934019455604, + "learning_rate": 5.882441984706555e-08, + "loss": 0.2836, + "step": 6852 + }, + { + "epoch": 0.95, + "grad_norm": 3.710475322602936, + "learning_rate": 5.8480587611998705e-08, + "loss": 0.2864, + "step": 6853 + }, + { + "epoch": 0.95, + "grad_norm": 1.894599576358644, + "learning_rate": 5.813775727328941e-08, + "loss": 0.3147, + "step": 6854 + }, + { + "epoch": 0.95, + "grad_norm": 3.1454281807308346, + "learning_rate": 5.7795928900447606e-08, + "loss": 0.3165, + "step": 6855 + }, + { + "epoch": 0.95, + "grad_norm": 2.4718596636609886, + "learning_rate": 5.745510256277842e-08, + "loss": 0.3075, + "step": 6856 + }, + { + "epoch": 0.95, + "grad_norm": 2.5565197208985806, + "learning_rate": 5.71152783293849e-08, + "loss": 0.2974, + "step": 6857 + }, + { + "epoch": 0.95, + "grad_norm": 2.493806825822216, + "learning_rate": 5.67764562691675e-08, + "loss": 0.3078, + "step": 6858 + }, + { + "epoch": 0.95, + "grad_norm": 1.7897424710305139, + "learning_rate": 5.643863645082126e-08, + "loss": 0.2984, + "step": 6859 + }, + { + "epoch": 0.95, + "grad_norm": 3.8227078677078654, + "learning_rate": 5.610181894283972e-08, + "loss": 0.3167, + "step": 6860 + }, + { + "epoch": 0.95, + "grad_norm": 2.1475643731447813, + "learning_rate": 5.576600381351327e-08, + "loss": 0.3053, + "step": 6861 + }, + { + "epoch": 0.95, + "grad_norm": 5.027432572112507, + "learning_rate": 5.543119113092854e-08, + "loss": 0.2978, + "step": 6862 + }, + { + "epoch": 0.95, + "grad_norm": 2.374415455030059, + "learning_rate": 5.509738096296957e-08, + "loss": 0.276, + "step": 6863 + }, + { + "epoch": 0.95, + "grad_norm": 3.059697020448554, + "learning_rate": 5.4764573377315e-08, + "loss": 0.2777, + "step": 6864 + }, + { + "epoch": 0.95, + "grad_norm": 2.376379566526783, + "learning_rate": 5.443276844144363e-08, + "loss": 0.3068, + "step": 6865 + }, + { + "epoch": 0.95, + "grad_norm": 2.1782554574927175, + "learning_rate": 5.4101966222627753e-08, + "loss": 0.2922, + "step": 6866 + }, + { + "epoch": 0.95, + "grad_norm": 2.2679631154228965, + "learning_rate": 5.377216678793817e-08, + "loss": 0.3098, + "step": 6867 + }, + { + "epoch": 0.95, + "grad_norm": 2.3025419664044686, + "learning_rate": 5.3443370204241954e-08, + "loss": 0.3134, + "step": 6868 + }, + { + "epoch": 0.95, + "grad_norm": 4.674573810499435, + "learning_rate": 5.311557653820354e-08, + "loss": 0.3073, + "step": 6869 + }, + { + "epoch": 0.96, + "grad_norm": 2.437258076113971, + "learning_rate": 5.2788785856281996e-08, + "loss": 0.3186, + "step": 6870 + }, + { + "epoch": 0.96, + "grad_norm": 2.2853007265178813, + "learning_rate": 5.2462998224734885e-08, + "loss": 0.3004, + "step": 6871 + }, + { + "epoch": 0.96, + "grad_norm": 2.067788155596286, + "learning_rate": 5.213821370961658e-08, + "loss": 0.321, + "step": 6872 + }, + { + "epoch": 0.96, + "grad_norm": 2.270138420241632, + "learning_rate": 5.1814432376776635e-08, + "loss": 0.3025, + "step": 6873 + }, + { + "epoch": 0.96, + "grad_norm": 3.382768160487235, + "learning_rate": 5.149165429186198e-08, + "loss": 0.2945, + "step": 6874 + }, + { + "epoch": 0.96, + "grad_norm": 3.2270401907185784, + "learning_rate": 5.116987952031638e-08, + "loss": 0.2964, + "step": 6875 + }, + { + "epoch": 0.96, + "grad_norm": 2.726679205708226, + "learning_rate": 5.084910812737987e-08, + "loss": 0.3093, + "step": 6876 + }, + { + "epoch": 0.96, + "grad_norm": 2.8728147063728056, + "learning_rate": 5.052934017808875e-08, + "loss": 0.2906, + "step": 6877 + }, + { + "epoch": 0.96, + "grad_norm": 2.506968724010001, + "learning_rate": 5.021057573727617e-08, + "loss": 0.3179, + "step": 6878 + }, + { + "epoch": 0.96, + "grad_norm": 3.0392870845936457, + "learning_rate": 4.989281486957154e-08, + "loss": 0.2771, + "step": 6879 + }, + { + "epoch": 0.96, + "grad_norm": 2.9287697728888644, + "learning_rate": 4.957605763940221e-08, + "loss": 0.2787, + "step": 6880 + }, + { + "epoch": 0.96, + "grad_norm": 2.465435053380733, + "learning_rate": 4.9260304110990145e-08, + "loss": 0.2918, + "step": 6881 + }, + { + "epoch": 0.96, + "grad_norm": 1.9187393422621923, + "learning_rate": 4.8945554348354685e-08, + "loss": 0.2917, + "step": 6882 + }, + { + "epoch": 0.96, + "grad_norm": 2.3693295213019634, + "learning_rate": 4.863180841531146e-08, + "loss": 0.2856, + "step": 6883 + }, + { + "epoch": 0.96, + "grad_norm": 1.9454298878341028, + "learning_rate": 4.83190663754729e-08, + "loss": 0.288, + "step": 6884 + }, + { + "epoch": 0.96, + "grad_norm": 2.009032932412054, + "learning_rate": 4.800732829224719e-08, + "loss": 0.3099, + "step": 6885 + }, + { + "epoch": 0.96, + "grad_norm": 2.427927756934816, + "learning_rate": 4.769659422883988e-08, + "loss": 0.2952, + "step": 6886 + }, + { + "epoch": 0.96, + "grad_norm": 2.8747360995704505, + "learning_rate": 4.738686424825223e-08, + "loss": 0.3004, + "step": 6887 + }, + { + "epoch": 0.96, + "grad_norm": 2.577898175325833, + "learning_rate": 4.7078138413282904e-08, + "loss": 0.2957, + "step": 6888 + }, + { + "epoch": 0.96, + "grad_norm": 1.967098742748558, + "learning_rate": 4.6770416786525163e-08, + "loss": 0.295, + "step": 6889 + }, + { + "epoch": 0.96, + "grad_norm": 3.3537925779290823, + "learning_rate": 4.646369943037077e-08, + "loss": 0.3217, + "step": 6890 + }, + { + "epoch": 0.96, + "grad_norm": 2.526000934604789, + "learning_rate": 4.6157986407006085e-08, + "loss": 0.3062, + "step": 6891 + }, + { + "epoch": 0.96, + "grad_norm": 2.037229853356466, + "learning_rate": 4.5853277778415416e-08, + "loss": 0.3034, + "step": 6892 + }, + { + "epoch": 0.96, + "grad_norm": 1.9319414922444598, + "learning_rate": 4.5549573606377686e-08, + "loss": 0.2994, + "step": 6893 + }, + { + "epoch": 0.96, + "grad_norm": 2.0679092756197632, + "learning_rate": 4.5246873952470296e-08, + "loss": 0.2772, + "step": 6894 + }, + { + "epoch": 0.96, + "grad_norm": 2.2887948746047853, + "learning_rate": 4.494517887806471e-08, + "loss": 0.3038, + "step": 6895 + }, + { + "epoch": 0.96, + "grad_norm": 3.820553387201725, + "learning_rate": 4.464448844433034e-08, + "loss": 0.3029, + "step": 6896 + }, + { + "epoch": 0.96, + "grad_norm": 4.218633199507957, + "learning_rate": 4.434480271223229e-08, + "loss": 0.2821, + "step": 6897 + }, + { + "epoch": 0.96, + "grad_norm": 2.2549100250223204, + "learning_rate": 4.4046121742531976e-08, + "loss": 0.2991, + "step": 6898 + }, + { + "epoch": 0.96, + "grad_norm": 1.6982872969742953, + "learning_rate": 4.374844559578706e-08, + "loss": 0.316, + "step": 6899 + }, + { + "epoch": 0.96, + "grad_norm": 2.233339147497143, + "learning_rate": 4.3451774332352036e-08, + "loss": 0.3015, + "step": 6900 + }, + { + "epoch": 0.96, + "grad_norm": 2.045617087515471, + "learning_rate": 4.315610801237713e-08, + "loss": 0.2791, + "step": 6901 + }, + { + "epoch": 0.96, + "grad_norm": 2.2630524919619766, + "learning_rate": 4.2861446695808276e-08, + "loss": 0.3156, + "step": 6902 + }, + { + "epoch": 0.96, + "grad_norm": 1.9675918043877716, + "learning_rate": 4.256779044238879e-08, + "loss": 0.3017, + "step": 6903 + }, + { + "epoch": 0.96, + "grad_norm": 2.2006904223007315, + "learning_rate": 4.2275139311657165e-08, + "loss": 0.2882, + "step": 6904 + }, + { + "epoch": 0.96, + "grad_norm": 2.3145159763482974, + "learning_rate": 4.1983493362949265e-08, + "loss": 0.3276, + "step": 6905 + }, + { + "epoch": 0.96, + "grad_norm": 2.2430437187897367, + "learning_rate": 4.169285265539613e-08, + "loss": 0.3108, + "step": 6906 + }, + { + "epoch": 0.96, + "grad_norm": 2.4782694863393604, + "learning_rate": 4.140321724792562e-08, + "loss": 0.2757, + "step": 6907 + }, + { + "epoch": 0.96, + "grad_norm": 3.5780968995248887, + "learning_rate": 4.1114587199261316e-08, + "loss": 0.302, + "step": 6908 + }, + { + "epoch": 0.96, + "grad_norm": 1.9681794448235554, + "learning_rate": 4.0826962567922515e-08, + "loss": 0.2985, + "step": 6909 + }, + { + "epoch": 0.96, + "grad_norm": 2.881914108364124, + "learning_rate": 4.0540343412226455e-08, + "loss": 0.2784, + "step": 6910 + }, + { + "epoch": 0.96, + "grad_norm": 2.1110306753678616, + "learning_rate": 4.0254729790285e-08, + "loss": 0.2847, + "step": 6911 + }, + { + "epoch": 0.96, + "grad_norm": 2.4555341719677153, + "learning_rate": 3.997012176000625e-08, + "loss": 0.3078, + "step": 6912 + }, + { + "epoch": 0.96, + "grad_norm": 3.146837208586492, + "learning_rate": 3.9686519379094624e-08, + "loss": 0.2839, + "step": 6913 + }, + { + "epoch": 0.96, + "grad_norm": 4.157660058295381, + "learning_rate": 3.940392270505078e-08, + "loss": 0.3115, + "step": 6914 + }, + { + "epoch": 0.96, + "grad_norm": 3.001449620862552, + "learning_rate": 3.912233179517166e-08, + "loss": 0.2881, + "step": 6915 + }, + { + "epoch": 0.96, + "grad_norm": 3.6276643341338204, + "learning_rate": 3.884174670654994e-08, + "loss": 0.2628, + "step": 6916 + }, + { + "epoch": 0.96, + "grad_norm": 2.8363896567793545, + "learning_rate": 3.8562167496074e-08, + "loss": 0.2788, + "step": 6917 + }, + { + "epoch": 0.96, + "grad_norm": 2.2759043584501146, + "learning_rate": 3.8283594220429045e-08, + "loss": 0.2965, + "step": 6918 + }, + { + "epoch": 0.96, + "grad_norm": 2.1495921909388205, + "learning_rate": 3.8006026936096005e-08, + "loss": 0.3005, + "step": 6919 + }, + { + "epoch": 0.96, + "grad_norm": 3.714872478237356, + "learning_rate": 3.772946569935154e-08, + "loss": 0.3039, + "step": 6920 + }, + { + "epoch": 0.96, + "grad_norm": 3.3275981838822917, + "learning_rate": 3.7453910566269123e-08, + "loss": 0.2964, + "step": 6921 + }, + { + "epoch": 0.96, + "grad_norm": 2.935206828374694, + "learning_rate": 3.71793615927174e-08, + "loss": 0.2789, + "step": 6922 + }, + { + "epoch": 0.96, + "grad_norm": 2.7581304703832847, + "learning_rate": 3.690581883436128e-08, + "loss": 0.2933, + "step": 6923 + }, + { + "epoch": 0.96, + "grad_norm": 2.2089520682714125, + "learning_rate": 3.6633282346662526e-08, + "loss": 0.3045, + "step": 6924 + }, + { + "epoch": 0.96, + "grad_norm": 2.459226140837491, + "learning_rate": 3.636175218487692e-08, + "loss": 0.3239, + "step": 6925 + }, + { + "epoch": 0.96, + "grad_norm": 2.221379924928662, + "learning_rate": 3.6091228404057654e-08, + "loss": 0.3147, + "step": 6926 + }, + { + "epoch": 0.96, + "grad_norm": 2.900599165182176, + "learning_rate": 3.5821711059054745e-08, + "loss": 0.2729, + "step": 6927 + }, + { + "epoch": 0.96, + "grad_norm": 2.538275999486867, + "learning_rate": 3.5553200204511696e-08, + "loss": 0.3044, + "step": 6928 + }, + { + "epoch": 0.96, + "grad_norm": 3.4409544785695085, + "learning_rate": 3.528569589486941e-08, + "loss": 0.2885, + "step": 6929 + }, + { + "epoch": 0.96, + "grad_norm": 2.1209726859348907, + "learning_rate": 3.5019198184365053e-08, + "loss": 0.285, + "step": 6930 + }, + { + "epoch": 0.96, + "grad_norm": 2.384947304630262, + "learning_rate": 3.4753707127030964e-08, + "loss": 0.3046, + "step": 6931 + }, + { + "epoch": 0.96, + "grad_norm": 2.0805094208138466, + "learning_rate": 3.448922277669631e-08, + "loss": 0.3, + "step": 6932 + }, + { + "epoch": 0.96, + "grad_norm": 2.132401459170794, + "learning_rate": 3.4225745186984296e-08, + "loss": 0.302, + "step": 6933 + }, + { + "epoch": 0.96, + "grad_norm": 2.2954289430653914, + "learning_rate": 3.3963274411315526e-08, + "loss": 0.3205, + "step": 6934 + }, + { + "epoch": 0.96, + "grad_norm": 2.2777461644843613, + "learning_rate": 3.3701810502906885e-08, + "loss": 0.2953, + "step": 6935 + }, + { + "epoch": 0.96, + "grad_norm": 1.9764090387590203, + "learning_rate": 3.344135351476929e-08, + "loss": 0.2986, + "step": 6936 + }, + { + "epoch": 0.96, + "grad_norm": 2.424632732844378, + "learning_rate": 3.318190349971162e-08, + "loss": 0.3003, + "step": 6937 + }, + { + "epoch": 0.96, + "grad_norm": 2.847431200276354, + "learning_rate": 3.292346051033679e-08, + "loss": 0.2764, + "step": 6938 + }, + { + "epoch": 0.96, + "grad_norm": 3.25103301099074, + "learning_rate": 3.2666024599044554e-08, + "loss": 0.2749, + "step": 6939 + }, + { + "epoch": 0.96, + "grad_norm": 7.371186825238795, + "learning_rate": 3.240959581802983e-08, + "loss": 0.3085, + "step": 6940 + }, + { + "epoch": 0.96, + "grad_norm": 1.8774313744530322, + "learning_rate": 3.215417421928435e-08, + "loss": 0.3013, + "step": 6941 + }, + { + "epoch": 0.97, + "grad_norm": 2.683601586789307, + "learning_rate": 3.189975985459503e-08, + "loss": 0.3126, + "step": 6942 + }, + { + "epoch": 0.97, + "grad_norm": 2.7362949404558403, + "learning_rate": 3.16463527755434e-08, + "loss": 0.3092, + "step": 6943 + }, + { + "epoch": 0.97, + "grad_norm": 2.2673354457759425, + "learning_rate": 3.139395303350945e-08, + "loss": 0.3191, + "step": 6944 + }, + { + "epoch": 0.97, + "grad_norm": 2.1623137355506765, + "learning_rate": 3.11425606796667e-08, + "loss": 0.2814, + "step": 6945 + }, + { + "epoch": 0.97, + "grad_norm": 2.9415812734536506, + "learning_rate": 3.089217576498493e-08, + "loss": 0.2892, + "step": 6946 + }, + { + "epoch": 0.97, + "grad_norm": 2.1284849460488746, + "learning_rate": 3.0642798340230206e-08, + "loss": 0.2702, + "step": 6947 + }, + { + "epoch": 0.97, + "grad_norm": 1.9758709424866105, + "learning_rate": 3.039442845596319e-08, + "loss": 0.2978, + "step": 6948 + }, + { + "epoch": 0.97, + "grad_norm": 1.9842162385996505, + "learning_rate": 3.014706616254248e-08, + "loss": 0.2979, + "step": 6949 + }, + { + "epoch": 0.97, + "grad_norm": 1.9374969145106165, + "learning_rate": 2.990071151011964e-08, + "loss": 0.2981, + "step": 6950 + }, + { + "epoch": 0.97, + "grad_norm": 2.502848982418737, + "learning_rate": 2.9655364548643595e-08, + "loss": 0.2924, + "step": 6951 + }, + { + "epoch": 0.97, + "grad_norm": 3.3897869001863614, + "learning_rate": 2.9411025327859e-08, + "loss": 0.3113, + "step": 6952 + }, + { + "epoch": 0.97, + "grad_norm": 2.317520964924901, + "learning_rate": 2.916769389730567e-08, + "loss": 0.3046, + "step": 6953 + }, + { + "epoch": 0.97, + "grad_norm": 2.2637027238611673, + "learning_rate": 2.892537030631859e-08, + "loss": 0.3145, + "step": 6954 + }, + { + "epoch": 0.97, + "grad_norm": 2.098242024477537, + "learning_rate": 2.868405460403012e-08, + "loss": 0.2958, + "step": 6955 + }, + { + "epoch": 0.97, + "grad_norm": 2.1488551816263017, + "learning_rate": 2.8443746839366126e-08, + "loss": 0.2935, + "step": 6956 + }, + { + "epoch": 0.97, + "grad_norm": 3.002277539346751, + "learning_rate": 2.8204447061049856e-08, + "loss": 0.2924, + "step": 6957 + }, + { + "epoch": 0.97, + "grad_norm": 2.9676638326827245, + "learning_rate": 2.7966155317599163e-08, + "loss": 0.2951, + "step": 6958 + }, + { + "epoch": 0.97, + "grad_norm": 4.740904563332658, + "learning_rate": 2.772887165732763e-08, + "loss": 0.2938, + "step": 6959 + }, + { + "epoch": 0.97, + "grad_norm": 1.9703127764731727, + "learning_rate": 2.749259612834565e-08, + "loss": 0.282, + "step": 6960 + }, + { + "epoch": 0.97, + "grad_norm": 2.070446855249786, + "learning_rate": 2.7257328778557135e-08, + "loss": 0.2888, + "step": 6961 + }, + { + "epoch": 0.97, + "grad_norm": 2.3434902024971107, + "learning_rate": 2.702306965566337e-08, + "loss": 0.3022, + "step": 6962 + }, + { + "epoch": 0.97, + "grad_norm": 1.8863125896353785, + "learning_rate": 2.678981880716025e-08, + "loss": 0.304, + "step": 6963 + }, + { + "epoch": 0.97, + "grad_norm": 2.1374568221792347, + "learning_rate": 2.6557576280339393e-08, + "loss": 0.2966, + "step": 6964 + }, + { + "epoch": 0.97, + "grad_norm": 2.262949636281106, + "learning_rate": 2.6326342122288685e-08, + "loss": 0.2957, + "step": 6965 + }, + { + "epoch": 0.97, + "grad_norm": 2.8426373373182683, + "learning_rate": 2.6096116379890622e-08, + "loss": 0.3068, + "step": 6966 + }, + { + "epoch": 0.97, + "grad_norm": 2.113277638655162, + "learning_rate": 2.586689909982343e-08, + "loss": 0.2902, + "step": 6967 + }, + { + "epoch": 0.97, + "grad_norm": 2.685525652412725, + "learning_rate": 2.5638690328561588e-08, + "loss": 0.2877, + "step": 6968 + }, + { + "epoch": 0.97, + "grad_norm": 2.1386889776615416, + "learning_rate": 2.5411490112374203e-08, + "loss": 0.2955, + "step": 6969 + }, + { + "epoch": 0.97, + "grad_norm": 3.1985027597518063, + "learning_rate": 2.518529849732665e-08, + "loss": 0.3059, + "step": 6970 + }, + { + "epoch": 0.97, + "grad_norm": 2.261597892620359, + "learning_rate": 2.4960115529278907e-08, + "loss": 0.2648, + "step": 6971 + }, + { + "epoch": 0.97, + "grad_norm": 1.9610937766936098, + "learning_rate": 2.473594125388723e-08, + "loss": 0.297, + "step": 6972 + }, + { + "epoch": 0.97, + "grad_norm": 2.881861746123438, + "learning_rate": 2.4512775716603042e-08, + "loss": 0.3062, + "step": 6973 + }, + { + "epoch": 0.97, + "grad_norm": 4.77036731583757, + "learning_rate": 2.429061896267293e-08, + "loss": 0.2981, + "step": 6974 + }, + { + "epoch": 0.97, + "grad_norm": 2.3951490091788545, + "learning_rate": 2.40694710371403e-08, + "loss": 0.2844, + "step": 6975 + }, + { + "epoch": 0.97, + "grad_norm": 4.182544982935222, + "learning_rate": 2.3849331984842627e-08, + "loss": 0.2981, + "step": 6976 + }, + { + "epoch": 0.97, + "grad_norm": 2.1583929401390907, + "learning_rate": 2.3630201850412538e-08, + "loss": 0.3109, + "step": 6977 + }, + { + "epoch": 0.97, + "grad_norm": 3.4346741974589294, + "learning_rate": 2.3412080678280046e-08, + "loss": 0.3121, + "step": 6978 + }, + { + "epoch": 0.97, + "grad_norm": 2.0098417125981434, + "learning_rate": 2.3194968512668114e-08, + "loss": 0.3228, + "step": 6979 + }, + { + "epoch": 0.97, + "grad_norm": 2.052114697095985, + "learning_rate": 2.297886539759764e-08, + "loss": 0.2668, + "step": 6980 + }, + { + "epoch": 0.97, + "grad_norm": 2.5835469713800516, + "learning_rate": 2.276377137688246e-08, + "loss": 0.3045, + "step": 6981 + }, + { + "epoch": 0.97, + "grad_norm": 3.2916762139773206, + "learning_rate": 2.254968649413325e-08, + "loss": 0.2993, + "step": 6982 + }, + { + "epoch": 0.97, + "grad_norm": 2.0765647214274727, + "learning_rate": 2.2336610792756954e-08, + "loss": 0.3193, + "step": 6983 + }, + { + "epoch": 0.97, + "grad_norm": 2.4071952347616548, + "learning_rate": 2.2124544315953455e-08, + "loss": 0.2835, + "step": 6984 + }, + { + "epoch": 0.97, + "grad_norm": 2.229089369128523, + "learning_rate": 2.1913487106720588e-08, + "loss": 0.3132, + "step": 6985 + }, + { + "epoch": 0.97, + "grad_norm": 2.799435696360796, + "learning_rate": 2.170343920784912e-08, + "loss": 0.2769, + "step": 6986 + }, + { + "epoch": 0.97, + "grad_norm": 4.855126664266868, + "learning_rate": 2.1494400661927207e-08, + "loss": 0.2972, + "step": 6987 + }, + { + "epoch": 0.97, + "grad_norm": 2.749951948932612, + "learning_rate": 2.1286371511337613e-08, + "loss": 0.2949, + "step": 6988 + }, + { + "epoch": 0.97, + "grad_norm": 2.1686176492979623, + "learning_rate": 2.107935179825771e-08, + "loss": 0.2903, + "step": 6989 + }, + { + "epoch": 0.97, + "grad_norm": 2.07840234745873, + "learning_rate": 2.0873341564661697e-08, + "loss": 0.2957, + "step": 6990 + }, + { + "epoch": 0.97, + "grad_norm": 2.5605850619586508, + "learning_rate": 2.0668340852317837e-08, + "loss": 0.2936, + "step": 6991 + }, + { + "epoch": 0.97, + "grad_norm": 3.4362116818639317, + "learning_rate": 2.0464349702789542e-08, + "loss": 0.3005, + "step": 6992 + }, + { + "epoch": 0.97, + "grad_norm": 2.679965748038844, + "learning_rate": 2.0261368157437624e-08, + "loss": 0.2898, + "step": 6993 + }, + { + "epoch": 0.97, + "grad_norm": 18.19071074471285, + "learning_rate": 2.005939625741582e-08, + "loss": 0.2762, + "step": 6994 + }, + { + "epoch": 0.97, + "grad_norm": 2.155435202470057, + "learning_rate": 1.9858434043673602e-08, + "loss": 0.2737, + "step": 6995 + }, + { + "epoch": 0.97, + "grad_norm": 4.871516088551677, + "learning_rate": 1.9658481556957264e-08, + "loss": 0.2912, + "step": 6996 + }, + { + "epoch": 0.97, + "grad_norm": 2.380063882618967, + "learning_rate": 1.9459538837806603e-08, + "loss": 0.2846, + "step": 6997 + }, + { + "epoch": 0.97, + "grad_norm": 2.1777014819865506, + "learning_rate": 1.926160592655768e-08, + "loss": 0.3203, + "step": 6998 + }, + { + "epoch": 0.97, + "grad_norm": 2.28469449847175, + "learning_rate": 1.9064682863341177e-08, + "loss": 0.316, + "step": 6999 + }, + { + "epoch": 0.97, + "grad_norm": 2.33241024714059, + "learning_rate": 1.8868769688084043e-08, + "loss": 0.2956, + "step": 7000 + }, + { + "epoch": 0.97, + "grad_norm": 3.2850029288156493, + "learning_rate": 1.8673866440507282e-08, + "loss": 0.2935, + "step": 7001 + }, + { + "epoch": 0.97, + "grad_norm": 2.8033914378101183, + "learning_rate": 1.8479973160127617e-08, + "loss": 0.2809, + "step": 7002 + }, + { + "epoch": 0.97, + "grad_norm": 2.2373936345409686, + "learning_rate": 1.8287089886257493e-08, + "loss": 0.31, + "step": 7003 + }, + { + "epoch": 0.97, + "grad_norm": 2.7934607757692507, + "learning_rate": 1.8095216658003956e-08, + "loss": 0.2767, + "step": 7004 + }, + { + "epoch": 0.97, + "grad_norm": 1.963263488320258, + "learning_rate": 1.7904353514268668e-08, + "loss": 0.2864, + "step": 7005 + }, + { + "epoch": 0.97, + "grad_norm": 2.1096309225648127, + "learning_rate": 1.771450049375012e-08, + "loss": 0.2777, + "step": 7006 + }, + { + "epoch": 0.97, + "grad_norm": 6.8857191515937854, + "learning_rate": 1.7525657634941407e-08, + "loss": 0.2824, + "step": 7007 + }, + { + "epoch": 0.97, + "grad_norm": 2.0072081029947024, + "learning_rate": 1.7337824976129126e-08, + "loss": 0.2848, + "step": 7008 + }, + { + "epoch": 0.97, + "grad_norm": 3.1419645379192773, + "learning_rate": 1.7151002555397813e-08, + "loss": 0.2963, + "step": 7009 + }, + { + "epoch": 0.97, + "grad_norm": 2.0466731123957302, + "learning_rate": 1.6965190410625497e-08, + "loss": 0.3007, + "step": 7010 + }, + { + "epoch": 0.97, + "grad_norm": 3.24172104404421, + "learning_rate": 1.678038857948594e-08, + "loss": 0.3112, + "step": 7011 + }, + { + "epoch": 0.97, + "grad_norm": 2.1232732810169352, + "learning_rate": 1.6596597099446942e-08, + "loss": 0.2722, + "step": 7012 + }, + { + "epoch": 0.97, + "grad_norm": 2.1833541554779963, + "learning_rate": 1.6413816007773696e-08, + "loss": 0.3053, + "step": 7013 + }, + { + "epoch": 0.98, + "grad_norm": 2.693475274607186, + "learning_rate": 1.6232045341523784e-08, + "loss": 0.3097, + "step": 7014 + }, + { + "epoch": 0.98, + "grad_norm": 2.127391376196053, + "learning_rate": 1.6051285137552163e-08, + "loss": 0.2995, + "step": 7015 + }, + { + "epoch": 0.98, + "grad_norm": 2.5080808407113966, + "learning_rate": 1.5871535432507856e-08, + "loss": 0.2968, + "step": 7016 + }, + { + "epoch": 0.98, + "grad_norm": 2.4299470958472584, + "learning_rate": 1.5692796262835043e-08, + "loss": 0.3015, + "step": 7017 + }, + { + "epoch": 0.98, + "grad_norm": 2.1783834676419525, + "learning_rate": 1.5515067664774176e-08, + "loss": 0.3, + "step": 7018 + }, + { + "epoch": 0.98, + "grad_norm": 2.1460707260227396, + "learning_rate": 1.5338349674358098e-08, + "loss": 0.2993, + "step": 7019 + }, + { + "epoch": 0.98, + "grad_norm": 2.921793591827244, + "learning_rate": 1.5162642327418154e-08, + "loss": 0.3233, + "step": 7020 + }, + { + "epoch": 0.98, + "grad_norm": 1.8949644441286997, + "learning_rate": 1.4987945659578617e-08, + "loss": 0.3109, + "step": 7021 + }, + { + "epoch": 0.98, + "grad_norm": 2.447504381357835, + "learning_rate": 1.4814259706258938e-08, + "loss": 0.2872, + "step": 7022 + }, + { + "epoch": 0.98, + "grad_norm": 2.6648300412063235, + "learning_rate": 1.4641584502674832e-08, + "loss": 0.3061, + "step": 7023 + }, + { + "epoch": 0.98, + "grad_norm": 2.189216544284509, + "learning_rate": 1.4469920083834964e-08, + "loss": 0.3069, + "step": 7024 + }, + { + "epoch": 0.98, + "grad_norm": 2.9601633689872706, + "learning_rate": 1.4299266484545937e-08, + "loss": 0.3147, + "step": 7025 + }, + { + "epoch": 0.98, + "grad_norm": 2.2438138865177706, + "learning_rate": 1.4129623739407294e-08, + "loss": 0.2936, + "step": 7026 + }, + { + "epoch": 0.98, + "grad_norm": 2.060565240943316, + "learning_rate": 1.3960991882813745e-08, + "loss": 0.2834, + "step": 7027 + }, + { + "epoch": 0.98, + "grad_norm": 2.817650764851987, + "learning_rate": 1.379337094895572e-08, + "loss": 0.3028, + "step": 7028 + }, + { + "epoch": 0.98, + "grad_norm": 1.8069345889967823, + "learning_rate": 1.3626760971819365e-08, + "loss": 0.3122, + "step": 7029 + }, + { + "epoch": 0.98, + "grad_norm": 2.172252893251784, + "learning_rate": 1.3461161985183768e-08, + "loss": 0.2699, + "step": 7030 + }, + { + "epoch": 0.98, + "grad_norm": 2.7009506483888805, + "learning_rate": 1.3296574022624854e-08, + "loss": 0.2978, + "step": 7031 + }, + { + "epoch": 0.98, + "grad_norm": 2.0512312887627564, + "learning_rate": 1.3132997117512591e-08, + "loss": 0.2762, + "step": 7032 + }, + { + "epoch": 0.98, + "grad_norm": 2.246241395543875, + "learning_rate": 1.2970431303012676e-08, + "loss": 0.3008, + "step": 7033 + }, + { + "epoch": 0.98, + "grad_norm": 4.51074377146337, + "learning_rate": 1.2808876612085408e-08, + "loss": 0.2759, + "step": 7034 + }, + { + "epoch": 0.98, + "grad_norm": 2.082083553927425, + "learning_rate": 1.2648333077486252e-08, + "loss": 0.2776, + "step": 7035 + }, + { + "epoch": 0.98, + "grad_norm": 1.9063546510478555, + "learning_rate": 1.2488800731764728e-08, + "loss": 0.3017, + "step": 7036 + }, + { + "epoch": 0.98, + "grad_norm": 2.2556740156278483, + "learning_rate": 1.2330279607266626e-08, + "loss": 0.3031, + "step": 7037 + }, + { + "epoch": 0.98, + "grad_norm": 3.8742658296205317, + "learning_rate": 1.2172769736132905e-08, + "loss": 0.3005, + "step": 7038 + }, + { + "epoch": 0.98, + "grad_norm": 2.1982577082944204, + "learning_rate": 1.2016271150297465e-08, + "loss": 0.2908, + "step": 7039 + }, + { + "epoch": 0.98, + "grad_norm": 2.2295913923874546, + "learning_rate": 1.1860783881491589e-08, + "loss": 0.3087, + "step": 7040 + }, + { + "epoch": 0.98, + "grad_norm": 2.360644231285657, + "learning_rate": 1.1706307961239504e-08, + "loss": 0.3051, + "step": 7041 + }, + { + "epoch": 0.98, + "grad_norm": 2.8999875964273163, + "learning_rate": 1.155284342086227e-08, + "loss": 0.2975, + "step": 7042 + }, + { + "epoch": 0.98, + "grad_norm": 10.733915384097726, + "learning_rate": 1.140039029147444e-08, + "loss": 0.3091, + "step": 7043 + }, + { + "epoch": 0.98, + "grad_norm": 2.218452004644328, + "learning_rate": 1.1248948603985732e-08, + "loss": 0.3007, + "step": 7044 + }, + { + "epoch": 0.98, + "grad_norm": 1.9880272447490754, + "learning_rate": 1.1098518389101587e-08, + "loss": 0.2902, + "step": 7045 + }, + { + "epoch": 0.98, + "grad_norm": 2.092352250673357, + "learning_rate": 1.0949099677321495e-08, + "loss": 0.3242, + "step": 7046 + }, + { + "epoch": 0.98, + "grad_norm": 2.060160034858198, + "learning_rate": 1.0800692498940113e-08, + "loss": 0.2808, + "step": 7047 + }, + { + "epoch": 0.98, + "grad_norm": 2.9777948176279443, + "learning_rate": 1.065329688404726e-08, + "loss": 0.3117, + "step": 7048 + }, + { + "epoch": 0.98, + "grad_norm": 2.5429608809635305, + "learning_rate": 1.0506912862527363e-08, + "loss": 0.2917, + "step": 7049 + }, + { + "epoch": 0.98, + "grad_norm": 2.5557744395491637, + "learning_rate": 1.0361540464060016e-08, + "loss": 0.2912, + "step": 7050 + }, + { + "epoch": 0.98, + "grad_norm": 2.1038156079169057, + "learning_rate": 1.0217179718119974e-08, + "loss": 0.2904, + "step": 7051 + }, + { + "epoch": 0.98, + "grad_norm": 4.3995445492095895, + "learning_rate": 1.007383065397549e-08, + "loss": 0.3048, + "step": 7052 + }, + { + "epoch": 0.98, + "grad_norm": 2.8085272940216006, + "learning_rate": 9.931493300691652e-09, + "loss": 0.2968, + "step": 7053 + }, + { + "epoch": 0.98, + "grad_norm": 2.0892072201077054, + "learning_rate": 9.790167687127039e-09, + "loss": 0.3245, + "step": 7054 + }, + { + "epoch": 0.98, + "grad_norm": 2.7809611953500637, + "learning_rate": 9.649853841935397e-09, + "loss": 0.2783, + "step": 7055 + }, + { + "epoch": 0.98, + "grad_norm": 1.9153664996555688, + "learning_rate": 9.510551793565636e-09, + "loss": 0.2896, + "step": 7056 + }, + { + "epoch": 0.98, + "grad_norm": 2.9326293168360724, + "learning_rate": 9.372261570261276e-09, + "loss": 0.2852, + "step": 7057 + }, + { + "epoch": 0.98, + "grad_norm": 2.114381591790378, + "learning_rate": 9.234983200060999e-09, + "loss": 0.2937, + "step": 7058 + }, + { + "epoch": 0.98, + "grad_norm": 2.2580661322330826, + "learning_rate": 9.09871671079754e-09, + "loss": 0.3269, + "step": 7059 + }, + { + "epoch": 0.98, + "grad_norm": 2.517778094181239, + "learning_rate": 8.963462130099354e-09, + "loss": 0.2846, + "step": 7060 + }, + { + "epoch": 0.98, + "grad_norm": 2.2089379688905155, + "learning_rate": 8.829219485389507e-09, + "loss": 0.2965, + "step": 7061 + }, + { + "epoch": 0.98, + "grad_norm": 2.4798850244741217, + "learning_rate": 8.695988803885668e-09, + "loss": 0.2947, + "step": 7062 + }, + { + "epoch": 0.98, + "grad_norm": 4.343096336733268, + "learning_rate": 8.563770112600122e-09, + "loss": 0.2714, + "step": 7063 + }, + { + "epoch": 0.98, + "grad_norm": 2.1551531136811333, + "learning_rate": 8.432563438340869e-09, + "loss": 0.2913, + "step": 7064 + }, + { + "epoch": 0.98, + "grad_norm": 2.317898429816688, + "learning_rate": 8.30236880770996e-09, + "loss": 0.3117, + "step": 7065 + }, + { + "epoch": 0.98, + "grad_norm": 7.4509448679504064, + "learning_rate": 8.17318624710406e-09, + "loss": 0.2918, + "step": 7066 + }, + { + "epoch": 0.98, + "grad_norm": 2.345878756780549, + "learning_rate": 8.045015782715549e-09, + "loss": 0.2789, + "step": 7067 + }, + { + "epoch": 0.98, + "grad_norm": 1.9961713873373848, + "learning_rate": 7.917857440530863e-09, + "loss": 0.282, + "step": 7068 + }, + { + "epoch": 0.98, + "grad_norm": 2.6582983550620796, + "learning_rate": 7.791711246331046e-09, + "loss": 0.3013, + "step": 7069 + }, + { + "epoch": 0.98, + "grad_norm": 1.7958493176037231, + "learning_rate": 7.666577225693417e-09, + "loss": 0.2637, + "step": 7070 + }, + { + "epoch": 0.98, + "grad_norm": 2.2507753136066064, + "learning_rate": 7.54245540398768e-09, + "loss": 0.3135, + "step": 7071 + }, + { + "epoch": 0.98, + "grad_norm": 2.392170847529988, + "learning_rate": 7.419345806380374e-09, + "loss": 0.3056, + "step": 7072 + }, + { + "epoch": 0.98, + "grad_norm": 2.06985931836313, + "learning_rate": 7.297248457832085e-09, + "loss": 0.3016, + "step": 7073 + }, + { + "epoch": 0.98, + "grad_norm": 3.9730408756912903, + "learning_rate": 7.176163383097457e-09, + "loss": 0.2846, + "step": 7074 + }, + { + "epoch": 0.98, + "grad_norm": 2.6333266130005026, + "learning_rate": 7.056090606727406e-09, + "loss": 0.2853, + "step": 7075 + }, + { + "epoch": 0.98, + "grad_norm": 1.824328002002521, + "learning_rate": 6.937030153066349e-09, + "loss": 0.2959, + "step": 7076 + }, + { + "epoch": 0.98, + "grad_norm": 1.8457419183677488, + "learning_rate": 6.818982046253863e-09, + "loss": 0.3045, + "step": 7077 + }, + { + "epoch": 0.98, + "grad_norm": 1.8561327224778976, + "learning_rate": 6.701946310224694e-09, + "loss": 0.2896, + "step": 7078 + }, + { + "epoch": 0.98, + "grad_norm": 2.573075845873155, + "learning_rate": 6.585922968707081e-09, + "loss": 0.3026, + "step": 7079 + }, + { + "epoch": 0.98, + "grad_norm": 2.2179628308627026, + "learning_rate": 6.470912045225541e-09, + "loss": 0.2986, + "step": 7080 + }, + { + "epoch": 0.98, + "grad_norm": 2.292312143109516, + "learning_rate": 6.356913563098643e-09, + "loss": 0.2947, + "step": 7081 + }, + { + "epoch": 0.98, + "grad_norm": 2.4617506275525676, + "learning_rate": 6.24392754543901e-09, + "loss": 0.2969, + "step": 7082 + }, + { + "epoch": 0.98, + "grad_norm": 2.22869503336811, + "learning_rate": 6.131954015154984e-09, + "loss": 0.2852, + "step": 7083 + }, + { + "epoch": 0.98, + "grad_norm": 2.421837990739564, + "learning_rate": 6.020992994949515e-09, + "loss": 0.3107, + "step": 7084 + }, + { + "epoch": 0.98, + "grad_norm": 2.1065276029908273, + "learning_rate": 5.911044507320163e-09, + "loss": 0.2867, + "step": 7085 + }, + { + "epoch": 0.99, + "grad_norm": 2.564671436573622, + "learning_rate": 5.802108574557985e-09, + "loss": 0.3103, + "step": 7086 + }, + { + "epoch": 0.99, + "grad_norm": 3.221548038908074, + "learning_rate": 5.694185218750869e-09, + "loss": 0.3062, + "step": 7087 + }, + { + "epoch": 0.99, + "grad_norm": 3.4229810315865814, + "learning_rate": 5.587274461780201e-09, + "loss": 0.2994, + "step": 7088 + }, + { + "epoch": 0.99, + "grad_norm": 3.684904940530182, + "learning_rate": 5.481376325321974e-09, + "loss": 0.3009, + "step": 7089 + }, + { + "epoch": 0.99, + "grad_norm": 2.621342430966824, + "learning_rate": 5.376490830847347e-09, + "loss": 0.2946, + "step": 7090 + }, + { + "epoch": 0.99, + "grad_norm": 3.1265644487344, + "learning_rate": 5.272617999621532e-09, + "loss": 0.286, + "step": 7091 + }, + { + "epoch": 0.99, + "grad_norm": 2.667083297805384, + "learning_rate": 5.1697578527049045e-09, + "loss": 0.3023, + "step": 7092 + }, + { + "epoch": 0.99, + "grad_norm": 4.544090381919779, + "learning_rate": 5.067910410953003e-09, + "loss": 0.2761, + "step": 7093 + }, + { + "epoch": 0.99, + "grad_norm": 1.8555015122869907, + "learning_rate": 4.967075695014312e-09, + "loss": 0.307, + "step": 7094 + }, + { + "epoch": 0.99, + "grad_norm": 2.975480958613091, + "learning_rate": 4.867253725334142e-09, + "loss": 0.3066, + "step": 7095 + }, + { + "epoch": 0.99, + "grad_norm": 2.1754830018100817, + "learning_rate": 4.768444522151306e-09, + "loss": 0.2924, + "step": 7096 + }, + { + "epoch": 0.99, + "grad_norm": 2.237995069982689, + "learning_rate": 4.67064810549922e-09, + "loss": 0.2965, + "step": 7097 + }, + { + "epoch": 0.99, + "grad_norm": 2.3236457590297217, + "learning_rate": 4.57386449520647e-09, + "loss": 0.2867, + "step": 7098 + }, + { + "epoch": 0.99, + "grad_norm": 2.3358756859753442, + "learning_rate": 4.478093710895692e-09, + "loss": 0.2921, + "step": 7099 + }, + { + "epoch": 0.99, + "grad_norm": 3.925342249088067, + "learning_rate": 4.383335771984132e-09, + "loss": 0.3019, + "step": 7100 + }, + { + "epoch": 0.99, + "grad_norm": 1.8275210449572183, + "learning_rate": 4.289590697684754e-09, + "loss": 0.2882, + "step": 7101 + }, + { + "epoch": 0.99, + "grad_norm": 2.4304297199613782, + "learning_rate": 4.19685850700402e-09, + "loss": 0.2921, + "step": 7102 + }, + { + "epoch": 0.99, + "grad_norm": 2.0055711373553504, + "learning_rate": 4.1051392187441144e-09, + "loss": 0.3032, + "step": 7103 + }, + { + "epoch": 0.99, + "grad_norm": 2.3727722939272877, + "learning_rate": 4.014432851500161e-09, + "loss": 0.2968, + "step": 7104 + }, + { + "epoch": 0.99, + "grad_norm": 2.0879950445046576, + "learning_rate": 3.9247394236630045e-09, + "loss": 0.2534, + "step": 7105 + }, + { + "epoch": 0.99, + "grad_norm": 1.9670854258990922, + "learning_rate": 3.836058953419208e-09, + "loss": 0.2964, + "step": 7106 + }, + { + "epoch": 0.99, + "grad_norm": 8.995299582879094, + "learning_rate": 3.748391458747724e-09, + "loss": 0.2848, + "step": 7107 + }, + { + "epoch": 0.99, + "grad_norm": 2.3466449830986953, + "learning_rate": 3.661736957423223e-09, + "loss": 0.303, + "step": 7108 + }, + { + "epoch": 0.99, + "grad_norm": 1.783768140791652, + "learning_rate": 3.576095467015539e-09, + "loss": 0.2842, + "step": 7109 + }, + { + "epoch": 0.99, + "grad_norm": 2.176681884763865, + "learning_rate": 3.491467004888005e-09, + "loss": 0.2865, + "step": 7110 + }, + { + "epoch": 0.99, + "grad_norm": 2.070185296890889, + "learning_rate": 3.407851588199673e-09, + "loss": 0.3139, + "step": 7111 + }, + { + "epoch": 0.99, + "grad_norm": 2.149970309290007, + "learning_rate": 3.325249233903094e-09, + "loss": 0.2929, + "step": 7112 + }, + { + "epoch": 0.99, + "grad_norm": 2.855731006465313, + "learning_rate": 3.2436599587459815e-09, + "loss": 0.3116, + "step": 7113 + }, + { + "epoch": 0.99, + "grad_norm": 3.3943341862271055, + "learning_rate": 3.1630837792712143e-09, + "loss": 0.2889, + "step": 7114 + }, + { + "epoch": 0.99, + "grad_norm": 2.6664633727638067, + "learning_rate": 3.0835207118146136e-09, + "loss": 0.2969, + "step": 7115 + }, + { + "epoch": 0.99, + "grad_norm": 2.2786640548004735, + "learning_rate": 3.0049707725088307e-09, + "loss": 0.2736, + "step": 7116 + }, + { + "epoch": 0.99, + "grad_norm": 1.9536071276573457, + "learning_rate": 2.927433977278904e-09, + "loss": 0.2983, + "step": 7117 + }, + { + "epoch": 0.99, + "grad_norm": 2.135298052528821, + "learning_rate": 2.850910341846147e-09, + "loss": 0.314, + "step": 7118 + }, + { + "epoch": 0.99, + "grad_norm": 3.0328945884250027, + "learning_rate": 2.7753998817259264e-09, + "loss": 0.2906, + "step": 7119 + }, + { + "epoch": 0.99, + "grad_norm": 2.8384174335015326, + "learning_rate": 2.7009026122271076e-09, + "loss": 0.3159, + "step": 7120 + }, + { + "epoch": 0.99, + "grad_norm": 3.6853858660484304, + "learning_rate": 2.62741854845483e-09, + "loss": 0.2919, + "step": 7121 + }, + { + "epoch": 0.99, + "grad_norm": 3.795494617433961, + "learning_rate": 2.554947705307731e-09, + "loss": 0.2878, + "step": 7122 + }, + { + "epoch": 0.99, + "grad_norm": 3.5586481224848536, + "learning_rate": 2.4834900974796126e-09, + "loss": 0.3331, + "step": 7123 + }, + { + "epoch": 0.99, + "grad_norm": 3.74733941195272, + "learning_rate": 2.4130457394577754e-09, + "loss": 0.3018, + "step": 7124 + }, + { + "epoch": 0.99, + "grad_norm": 5.856290110157313, + "learning_rate": 2.3436146455263485e-09, + "loss": 0.3049, + "step": 7125 + }, + { + "epoch": 0.99, + "grad_norm": 3.71864342233616, + "learning_rate": 2.2751968297607396e-09, + "loss": 0.2895, + "step": 7126 + }, + { + "epoch": 0.99, + "grad_norm": 2.178632280571307, + "learning_rate": 2.2077923060342956e-09, + "loss": 0.319, + "step": 7127 + }, + { + "epoch": 0.99, + "grad_norm": 2.4176901008984477, + "learning_rate": 2.1414010880121962e-09, + "loss": 0.269, + "step": 7128 + }, + { + "epoch": 0.99, + "grad_norm": 2.796652122186013, + "learning_rate": 2.0760231891558957e-09, + "loss": 0.2928, + "step": 7129 + }, + { + "epoch": 0.99, + "grad_norm": 2.845881133824049, + "learning_rate": 2.0116586227203473e-09, + "loss": 0.3077, + "step": 7130 + }, + { + "epoch": 0.99, + "grad_norm": 2.3297372220317345, + "learning_rate": 1.948307401756222e-09, + "loss": 0.2939, + "step": 7131 + }, + { + "epoch": 0.99, + "grad_norm": 2.5408908590569674, + "learning_rate": 1.8859695391071353e-09, + "loss": 0.2882, + "step": 7132 + }, + { + "epoch": 0.99, + "grad_norm": 6.14966316129949, + "learning_rate": 1.8246450474129763e-09, + "loss": 0.2984, + "step": 7133 + }, + { + "epoch": 0.99, + "grad_norm": 2.2727369400813964, + "learning_rate": 1.7643339391065773e-09, + "loss": 0.3008, + "step": 7134 + }, + { + "epoch": 0.99, + "grad_norm": 2.589756091529717, + "learning_rate": 1.705036226417045e-09, + "loss": 0.2938, + "step": 7135 + }, + { + "epoch": 0.99, + "grad_norm": 2.389354343534786, + "learning_rate": 1.6467519213658745e-09, + "loss": 0.3134, + "step": 7136 + }, + { + "epoch": 0.99, + "grad_norm": 1.9269739859711894, + "learning_rate": 1.5894810357708347e-09, + "loss": 0.2892, + "step": 7137 + }, + { + "epoch": 0.99, + "grad_norm": 5.827562462690043, + "learning_rate": 1.533223581243748e-09, + "loss": 0.3082, + "step": 7138 + }, + { + "epoch": 0.99, + "grad_norm": 2.115317637600868, + "learning_rate": 1.4779795691904907e-09, + "loss": 0.2867, + "step": 7139 + }, + { + "epoch": 0.99, + "grad_norm": 2.7794671182814508, + "learning_rate": 1.4237490108121033e-09, + "loss": 0.3209, + "step": 7140 + }, + { + "epoch": 0.99, + "grad_norm": 3.575497737319559, + "learning_rate": 1.3705319171042341e-09, + "loss": 0.3104, + "step": 7141 + }, + { + "epoch": 0.99, + "grad_norm": 2.3038198950573605, + "learning_rate": 1.3183282988554757e-09, + "loss": 0.2745, + "step": 7142 + }, + { + "epoch": 0.99, + "grad_norm": 3.1838969396289856, + "learning_rate": 1.2671381666512494e-09, + "loss": 0.304, + "step": 7143 + }, + { + "epoch": 0.99, + "grad_norm": 2.308318774004277, + "learning_rate": 1.2169615308704751e-09, + "loss": 0.2914, + "step": 7144 + }, + { + "epoch": 0.99, + "grad_norm": 1.980357997220453, + "learning_rate": 1.1677984016855715e-09, + "loss": 0.296, + "step": 7145 + }, + { + "epoch": 0.99, + "grad_norm": 2.520367937033757, + "learning_rate": 1.1196487890652308e-09, + "loss": 0.2945, + "step": 7146 + }, + { + "epoch": 0.99, + "grad_norm": 1.9155403388705128, + "learning_rate": 1.072512702771089e-09, + "loss": 0.2881, + "step": 7147 + }, + { + "epoch": 0.99, + "grad_norm": 1.902012742057838, + "learning_rate": 1.0263901523605014e-09, + "loss": 0.2799, + "step": 7148 + }, + { + "epoch": 0.99, + "grad_norm": 2.136327857063892, + "learning_rate": 9.81281147184876e-10, + "loss": 0.2928, + "step": 7149 + }, + { + "epoch": 0.99, + "grad_norm": 2.005594581893038, + "learning_rate": 9.371856963896753e-10, + "loss": 0.298, + "step": 7150 + }, + { + "epoch": 0.99, + "grad_norm": 4.966319426875389, + "learning_rate": 8.941038089160803e-10, + "loss": 0.2887, + "step": 7151 + }, + { + "epoch": 0.99, + "grad_norm": 2.0732962107957404, + "learning_rate": 8.520354934982155e-10, + "loss": 0.2934, + "step": 7152 + }, + { + "epoch": 0.99, + "grad_norm": 1.8209284886782144, + "learning_rate": 8.109807586653695e-10, + "loss": 0.2921, + "step": 7153 + }, + { + "epoch": 0.99, + "grad_norm": 2.834278643686234, + "learning_rate": 7.709396127425494e-10, + "loss": 0.3201, + "step": 7154 + }, + { + "epoch": 0.99, + "grad_norm": 2.148770154761996, + "learning_rate": 7.31912063846596e-10, + "loss": 0.3097, + "step": 7155 + }, + { + "epoch": 0.99, + "grad_norm": 2.1787133763777673, + "learning_rate": 6.938981198917338e-10, + "loss": 0.3016, + "step": 7156 + }, + { + "epoch": 0.99, + "grad_norm": 2.2673973210384033, + "learning_rate": 6.568977885840211e-10, + "loss": 0.292, + "step": 7157 + }, + { + "epoch": 1.0, + "grad_norm": 2.95462352452137, + "learning_rate": 6.209110774263449e-10, + "loss": 0.307, + "step": 7158 + }, + { + "epoch": 1.0, + "grad_norm": 2.3256311974084176, + "learning_rate": 5.85937993714536e-10, + "loss": 0.326, + "step": 7159 + }, + { + "epoch": 1.0, + "grad_norm": 1.8431137882261313, + "learning_rate": 5.519785445401437e-10, + "loss": 0.3082, + "step": 7160 + }, + { + "epoch": 1.0, + "grad_norm": 2.5713938516167008, + "learning_rate": 5.190327367871062e-10, + "loss": 0.3013, + "step": 7161 + }, + { + "epoch": 1.0, + "grad_norm": 2.314730266317683, + "learning_rate": 4.871005771361903e-10, + "loss": 0.3098, + "step": 7162 + }, + { + "epoch": 1.0, + "grad_norm": 2.166613630761336, + "learning_rate": 4.5618207206166164e-10, + "loss": 0.3152, + "step": 7163 + }, + { + "epoch": 1.0, + "grad_norm": 1.9153095037112626, + "learning_rate": 4.2627722783183943e-10, + "loss": 0.3149, + "step": 7164 + }, + { + "epoch": 1.0, + "grad_norm": 2.434568543937153, + "learning_rate": 3.9738605051020675e-10, + "loss": 0.3031, + "step": 7165 + }, + { + "epoch": 1.0, + "grad_norm": 2.717287687705056, + "learning_rate": 3.695085459543002e-10, + "loss": 0.317, + "step": 7166 + }, + { + "epoch": 1.0, + "grad_norm": 2.5985939064503283, + "learning_rate": 3.4264471981626525e-10, + "loss": 0.3164, + "step": 7167 + }, + { + "epoch": 1.0, + "grad_norm": 2.9097337682678495, + "learning_rate": 3.1679457754341113e-10, + "loss": 0.2887, + "step": 7168 + }, + { + "epoch": 1.0, + "grad_norm": 3.9805314183352727, + "learning_rate": 2.919581243759906e-10, + "loss": 0.3002, + "step": 7169 + }, + { + "epoch": 1.0, + "grad_norm": 3.653913777945473, + "learning_rate": 2.681353653499752e-10, + "loss": 0.3242, + "step": 7170 + }, + { + "epoch": 1.0, + "grad_norm": 2.2277390211857617, + "learning_rate": 2.4532630529594537e-10, + "loss": 0.2896, + "step": 7171 + }, + { + "epoch": 1.0, + "grad_norm": 1.944132859148916, + "learning_rate": 2.235309488374249e-10, + "loss": 0.2822, + "step": 7172 + }, + { + "epoch": 1.0, + "grad_norm": 2.3855313041026434, + "learning_rate": 2.027493003942116e-10, + "loss": 0.2869, + "step": 7173 + }, + { + "epoch": 1.0, + "grad_norm": 2.2809956100249944, + "learning_rate": 1.8298136417960188e-10, + "loss": 0.3174, + "step": 7174 + }, + { + "epoch": 1.0, + "grad_norm": 3.245275427998087, + "learning_rate": 1.6422714420150088e-10, + "loss": 0.3089, + "step": 7175 + }, + { + "epoch": 1.0, + "grad_norm": 2.6754014449109227, + "learning_rate": 1.4648664426242243e-10, + "loss": 0.2691, + "step": 7176 + }, + { + "epoch": 1.0, + "grad_norm": 3.7512805646191545, + "learning_rate": 1.2975986795893402e-10, + "loss": 0.2911, + "step": 7177 + }, + { + "epoch": 1.0, + "grad_norm": 2.2528647996364786, + "learning_rate": 1.1404681868332212e-10, + "loss": 0.2846, + "step": 7178 + }, + { + "epoch": 1.0, + "grad_norm": 3.664348585619141, + "learning_rate": 9.934749962026147e-11, + "loss": 0.275, + "step": 7179 + }, + { + "epoch": 1.0, + "grad_norm": 2.16046131143751, + "learning_rate": 8.566191375070088e-11, + "loss": 0.264, + "step": 7180 + }, + { + "epoch": 1.0, + "grad_norm": 2.3393665665029477, + "learning_rate": 7.299006384908769e-11, + "loss": 0.2901, + "step": 7181 + }, + { + "epoch": 1.0, + "grad_norm": 2.4372231454957713, + "learning_rate": 6.133195248503309e-11, + "loss": 0.2857, + "step": 7182 + }, + { + "epoch": 1.0, + "grad_norm": 2.722801545164624, + "learning_rate": 5.068758202164681e-11, + "loss": 0.3053, + "step": 7183 + }, + { + "epoch": 1.0, + "grad_norm": 2.387713978354236, + "learning_rate": 4.105695461775749e-11, + "loss": 0.2977, + "step": 7184 + }, + { + "epoch": 1.0, + "grad_norm": 4.603568225137398, + "learning_rate": 3.244007222569234e-11, + "loss": 0.2985, + "step": 7185 + }, + { + "epoch": 1.0, + "grad_norm": 2.089999651609884, + "learning_rate": 2.4836936592387283e-11, + "loss": 0.3191, + "step": 7186 + }, + { + "epoch": 1.0, + "grad_norm": 2.679991116176632, + "learning_rate": 1.8247549259386988e-11, + "loss": 0.3011, + "step": 7187 + }, + { + "epoch": 1.0, + "grad_norm": 2.3873474467542537, + "learning_rate": 1.2671911563399975e-11, + "loss": 0.2954, + "step": 7188 + }, + { + "epoch": 1.0, + "grad_norm": 2.3553878001103272, + "learning_rate": 8.110024633523061e-12, + "loss": 0.2945, + "step": 7189 + }, + { + "epoch": 1.0, + "grad_norm": 3.6213258821979495, + "learning_rate": 4.561889396237362e-12, + "loss": 0.2824, + "step": 7190 + }, + { + "epoch": 1.0, + "grad_norm": 3.0854413203242235, + "learning_rate": 2.0275065693020622e-12, + "loss": 0.2797, + "step": 7191 + }, + { + "epoch": 1.0, + "grad_norm": 2.02595733874251, + "learning_rate": 5.068766678606452e-13, + "loss": 0.3073, + "step": 7192 + }, + { + "epoch": 1.0, + "grad_norm": 3.01222269822412, + "learning_rate": 0.0, + "loss": 0.3081, + "step": 7193 + }, + { + "epoch": 1.0, + "eval_EK100_accuracy": 0.6082442434210527, + "step": 7193 + }, + { + "epoch": 1.0, + "step": 7193, + "total_flos": 3552817014833152.0, + "train_loss": 0.04932548151499503, + "train_runtime": 16383.444, + "train_samples_per_second": 112.404, + "train_steps_per_second": 0.439 + } + ], + "logging_steps": 1.0, + "max_steps": 7193, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 2000, + "total_flos": 3552817014833152.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}