| { | |
| "best_metric": 4252.13720703125, | |
| "best_model_checkpoint": "./ckpts/tinyllama_v1.1/ternary-g128/checkpoint-200", | |
| "epoch": 2.0, | |
| "eval_steps": 40, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0, | |
| "loss": 11333.3047, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 2e-05, | |
| "loss": 12740.3945, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2e-05, | |
| "loss": 12751.8154, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2e-05, | |
| "loss": 13332.6777, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2e-05, | |
| "loss": 12041.1602, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2e-05, | |
| "loss": 11759.3701, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 2e-05, | |
| "loss": 7579.3691, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 2e-05, | |
| "loss": 7834.7402, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2e-05, | |
| "loss": 10647.7695, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2e-05, | |
| "loss": 7894.1113, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2e-05, | |
| "loss": 9336.8496, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2e-05, | |
| "loss": 9224.625, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2e-05, | |
| "loss": 8379.665, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 2e-05, | |
| "loss": 9123.9082, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2e-05, | |
| "loss": 10331.1914, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 2e-05, | |
| "loss": 8695.0479, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 2e-05, | |
| "loss": 7838.9502, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 2e-05, | |
| "loss": 7516.7783, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 2e-05, | |
| "loss": 8220.8701, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 2e-05, | |
| "loss": 6244.4844, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 2e-05, | |
| "loss": 7569.1367, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 2e-05, | |
| "loss": 6951.0103, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 2e-05, | |
| "loss": 7107.9727, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 2e-05, | |
| "loss": 7161.4819, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 2e-05, | |
| "loss": 7868.584, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 2e-05, | |
| "loss": 6625.4795, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 2e-05, | |
| "loss": 6222.7422, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 2e-05, | |
| "loss": 6319.2148, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 2e-05, | |
| "loss": 5998.3438, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 2e-05, | |
| "loss": 6406.7197, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 2e-05, | |
| "loss": 6865.812, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 2e-05, | |
| "loss": 5550.9609, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 2e-05, | |
| "loss": 6530.6548, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 2e-05, | |
| "loss": 5644.4668, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 2e-05, | |
| "loss": 6787.8086, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 2e-05, | |
| "loss": 6230.1162, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 2e-05, | |
| "loss": 5976.9404, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 2e-05, | |
| "loss": 5261.2661, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 2e-05, | |
| "loss": 5871.9727, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 2e-05, | |
| "loss": 5244.6387, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 5789.15625, | |
| "eval_runtime": 153.2221, | |
| "eval_samples_per_second": 10.416, | |
| "eval_steps_per_second": 0.653, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 2e-05, | |
| "loss": 5647.2305, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 2e-05, | |
| "loss": 5769.7246, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 2e-05, | |
| "loss": 5806.084, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 2e-05, | |
| "loss": 6338.6992, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 2e-05, | |
| "loss": 6663.2988, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2e-05, | |
| "loss": 5164.0898, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2e-05, | |
| "loss": 5250.1279, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2e-05, | |
| "loss": 4920.438, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2e-05, | |
| "loss": 5083.2578, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2e-05, | |
| "loss": 5441.7505, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2e-05, | |
| "loss": 5271.9443, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2e-05, | |
| "loss": 6324.0801, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2e-05, | |
| "loss": 5859.5479, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2e-05, | |
| "loss": 5835.2354, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2e-05, | |
| "loss": 5948.1357, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2e-05, | |
| "loss": 4974.9961, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2e-05, | |
| "loss": 5055.021, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2e-05, | |
| "loss": 5852.9814, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2e-05, | |
| "loss": 4996.5635, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2e-05, | |
| "loss": 4621.3154, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2e-05, | |
| "loss": 4878.3037, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2e-05, | |
| "loss": 4970.6401, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2e-05, | |
| "loss": 4931.6152, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 2e-05, | |
| "loss": 5476.6318, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 2e-05, | |
| "loss": 5228.8818, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 2e-05, | |
| "loss": 4760.2158, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 2e-05, | |
| "loss": 4873.3926, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 2e-05, | |
| "loss": 5457.2378, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 2e-05, | |
| "loss": 5250.7954, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 2e-05, | |
| "loss": 4680.854, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 2e-05, | |
| "loss": 4848.3828, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 2e-05, | |
| "loss": 5283.5181, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 2e-05, | |
| "loss": 4919.6475, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 2e-05, | |
| "loss": 5075.8682, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 2e-05, | |
| "loss": 4698.2148, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 2e-05, | |
| "loss": 4929.0264, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 2e-05, | |
| "loss": 5158.9775, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 2e-05, | |
| "loss": 5029.708, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 2e-05, | |
| "loss": 5271.4033, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 2e-05, | |
| "loss": 4820.6875, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 4961.72412109375, | |
| "eval_runtime": 153.3149, | |
| "eval_samples_per_second": 10.41, | |
| "eval_steps_per_second": 0.652, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 2e-05, | |
| "loss": 4819.8237, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2e-05, | |
| "loss": 4838.397, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 2e-05, | |
| "loss": 5228.1221, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 2e-05, | |
| "loss": 5043.3447, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 2e-05, | |
| "loss": 5215.2246, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2e-05, | |
| "loss": 4427.6909, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2e-05, | |
| "loss": 5120.7344, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2e-05, | |
| "loss": 5044.043, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 2e-05, | |
| "loss": 4083.1196, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 2e-05, | |
| "loss": 4404.4829, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2e-05, | |
| "loss": 5224.4292, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 2e-05, | |
| "loss": 5386.4004, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 2e-05, | |
| "loss": 5193.6807, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 2e-05, | |
| "loss": 5194.8838, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2e-05, | |
| "loss": 5112.0229, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2e-05, | |
| "loss": 4970.9795, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 2e-05, | |
| "loss": 4510.7227, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 2e-05, | |
| "loss": 5206.0742, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2e-05, | |
| "loss": 5560.2061, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 2e-05, | |
| "loss": 4309.6255, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 2e-05, | |
| "loss": 4270.1172, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 2e-05, | |
| "loss": 4307.147, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 2e-05, | |
| "loss": 5392.7363, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 2e-05, | |
| "loss": 4613.896, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 2e-05, | |
| "loss": 4642.5605, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 2e-05, | |
| "loss": 4990.2676, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 2e-05, | |
| "loss": 4562.7651, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 2e-05, | |
| "loss": 4420.7632, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 2e-05, | |
| "loss": 4600.2861, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 2e-05, | |
| "loss": 4772.0371, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 2e-05, | |
| "loss": 5246.8555, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 2e-05, | |
| "loss": 4700.6631, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 2e-05, | |
| "loss": 4570.7441, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 2e-05, | |
| "loss": 4589.0493, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 2e-05, | |
| "loss": 4846.958, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 2e-05, | |
| "loss": 4793.291, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 2e-05, | |
| "loss": 5359.541, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 2e-05, | |
| "loss": 4714.4346, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 2e-05, | |
| "loss": 4767.8203, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 2e-05, | |
| "loss": 5251.5449, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 4679.81640625, | |
| "eval_runtime": 153.1502, | |
| "eval_samples_per_second": 10.421, | |
| "eval_steps_per_second": 0.653, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 2e-05, | |
| "loss": 4597.8896, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 2e-05, | |
| "loss": 4525.1045, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 2e-05, | |
| "loss": 4740.7261, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 2e-05, | |
| "loss": 4635.7217, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 2e-05, | |
| "loss": 4253.7402, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 2e-05, | |
| "loss": 4565.4546, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 2e-05, | |
| "loss": 4249.1875, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 2e-05, | |
| "loss": 5051.9473, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 2e-05, | |
| "loss": 4128.4834, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 2e-05, | |
| "loss": 4775.8789, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 2e-05, | |
| "loss": 4661.0552, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 2e-05, | |
| "loss": 4616.229, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 2e-05, | |
| "loss": 5068.2056, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 2e-05, | |
| "loss": 5291.3818, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 2e-05, | |
| "loss": 4666.6348, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 2e-05, | |
| "loss": 5022.856, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 2e-05, | |
| "loss": 4555.9136, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 2e-05, | |
| "loss": 4543.4004, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 2e-05, | |
| "loss": 5208.0181, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2e-05, | |
| "loss": 4202.6709, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 2e-05, | |
| "loss": 4570.6206, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 2e-05, | |
| "loss": 5114.9287, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 2e-05, | |
| "loss": 4141.6299, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 2e-05, | |
| "loss": 4849.125, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 2e-05, | |
| "loss": 4281.4214, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 2e-05, | |
| "loss": 4833.1548, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 2e-05, | |
| "loss": 4661.8291, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 2e-05, | |
| "loss": 4485.2344, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 2e-05, | |
| "loss": 4128.439, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 2e-05, | |
| "loss": 4300.2852, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2e-05, | |
| "loss": 4689.8813, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 2e-05, | |
| "loss": 4261.6787, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 2e-05, | |
| "loss": 4410.7871, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 2e-05, | |
| "loss": 4859.6416, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 2e-05, | |
| "loss": 4945.5195, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 2e-05, | |
| "loss": 4227.752, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 2e-05, | |
| "loss": 4758.1689, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2e-05, | |
| "loss": 4772.7354, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 2e-05, | |
| "loss": 4114.6689, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 2e-05, | |
| "loss": 4792.8496, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 4522.2705078125, | |
| "eval_runtime": 153.0119, | |
| "eval_samples_per_second": 10.431, | |
| "eval_steps_per_second": 0.654, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 2e-05, | |
| "loss": 4813.7637, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 2e-05, | |
| "loss": 3838.3706, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 2e-05, | |
| "loss": 4562.4961, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2e-05, | |
| "loss": 4465.833, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2e-05, | |
| "loss": 4186.3252, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 2e-05, | |
| "loss": 4580.1899, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 2e-05, | |
| "loss": 4878.4688, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 2e-05, | |
| "loss": 4298.249, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 2e-05, | |
| "loss": 4254.103, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2e-05, | |
| "loss": 3883.5581, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 2e-05, | |
| "loss": 4455.5347, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 2e-05, | |
| "loss": 4322.6348, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 2e-05, | |
| "loss": 4249.686, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2e-05, | |
| "loss": 4670.4375, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 2e-05, | |
| "loss": 3984.6772, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 2e-05, | |
| "loss": 4424.6416, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 2e-05, | |
| "loss": 4487.6416, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 2e-05, | |
| "loss": 4786.0723, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 2e-05, | |
| "loss": 4320.7349, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 2e-05, | |
| "loss": 4392.415, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 2e-05, | |
| "loss": 4128.5469, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 2e-05, | |
| "loss": 3805.3416, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 2e-05, | |
| "loss": 5295.7549, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 2e-05, | |
| "loss": 4073.2388, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 2e-05, | |
| "loss": 4689.1528, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 2e-05, | |
| "loss": 4331.627, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 2e-05, | |
| "loss": 4956.3774, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 2e-05, | |
| "loss": 3998.3621, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 2e-05, | |
| "loss": 4095.9175, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 2e-05, | |
| "loss": 3817.9526, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 2e-05, | |
| "loss": 4632.4111, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 2e-05, | |
| "loss": 3853.741, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 2e-05, | |
| "loss": 3862.7646, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 2e-05, | |
| "loss": 4408.2993, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 2e-05, | |
| "loss": 3478.6504, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 2e-05, | |
| "loss": 4667.4932, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 2e-05, | |
| "loss": 3785.4988, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 2e-05, | |
| "loss": 4366.5952, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 2e-05, | |
| "loss": 4249.0308, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 2e-05, | |
| "loss": 4039.4888, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 4252.13720703125, | |
| "eval_runtime": 152.5389, | |
| "eval_samples_per_second": 10.463, | |
| "eval_steps_per_second": 0.656, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 40, | |
| "total_flos": 8.12049800185774e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |