{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9999270764967549,
  "eval_steps": 500,
  "global_step": 1714,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002916940129803836,
      "grad_norm": 9.421533600382833,
      "learning_rate": 1.1627906976744187e-07,
      "loss": 1.4153,
      "num_tokens": 671048.0,
      "step": 5
    },
    {
      "epoch": 0.005833880259607672,
      "grad_norm": 9.417087087487797,
      "learning_rate": 2.616279069767442e-07,
      "loss": 1.4115,
      "num_tokens": 1340514.0,
      "step": 10
    },
    {
      "epoch": 0.008750820389411508,
      "grad_norm": 8.64652450056725,
      "learning_rate": 4.0697674418604655e-07,
      "loss": 1.3864,
      "num_tokens": 1975142.0,
      "step": 15
    },
    {
      "epoch": 0.011667760519215344,
      "grad_norm": 6.364802582777174,
      "learning_rate": 5.523255813953489e-07,
      "loss": 1.2845,
      "num_tokens": 2629706.0,
      "step": 20
    },
    {
      "epoch": 0.01458470064901918,
      "grad_norm": 4.036625471123866,
      "learning_rate": 6.976744186046513e-07,
      "loss": 1.1622,
      "num_tokens": 3282682.0,
      "step": 25
    },
    {
      "epoch": 0.017501640778823015,
      "grad_norm": 4.1984081053855205,
      "learning_rate": 8.430232558139535e-07,
      "loss": 0.9647,
      "num_tokens": 3956656.0,
      "step": 30
    },
    {
      "epoch": 0.02041858090862685,
      "grad_norm": 2.9803387136629893,
      "learning_rate": 9.88372093023256e-07,
      "loss": 0.7913,
      "num_tokens": 4609315.0,
      "step": 35
    },
    {
      "epoch": 0.023335521038430687,
      "grad_norm": 1.5758517615202334,
      "learning_rate": 1.1337209302325581e-06,
      "loss": 0.6332,
      "num_tokens": 5258689.0,
      "step": 40
    },
    {
      "epoch": 0.02625246116823452,
      "grad_norm": 1.5452778456677885,
      "learning_rate": 1.2790697674418605e-06,
      "loss": 0.5782,
      "num_tokens": 5918842.0,
      "step": 45
    },
    {
      "epoch": 0.02916940129803836,
      "grad_norm": 1.3636179516378841,
      "learning_rate": 1.424418604651163e-06,
      "loss": 0.5198,
      "num_tokens": 6581062.0,
      "step": 50
    },
    {
      "epoch": 0.0320863414278422,
      "grad_norm": 1.3117664668985787,
      "learning_rate": 1.5697674418604653e-06,
      "loss": 0.4967,
      "num_tokens": 7229502.0,
      "step": 55
    },
    {
      "epoch": 0.03500328155764603,
      "grad_norm": 1.120351932885373,
      "learning_rate": 1.7151162790697675e-06,
      "loss": 0.4632,
      "num_tokens": 7873010.0,
      "step": 60
    },
    {
      "epoch": 0.037920221687449865,
      "grad_norm": 1.1652086123427123,
      "learning_rate": 1.86046511627907e-06,
      "loss": 0.4468,
      "num_tokens": 8521924.0,
      "step": 65
    },
    {
      "epoch": 0.0408371618172537,
      "grad_norm": 1.1356964733122763,
      "learning_rate": 2.005813953488372e-06,
      "loss": 0.4216,
      "num_tokens": 9195311.0,
      "step": 70
    },
    {
      "epoch": 0.04375410194705753,
      "grad_norm": 1.125126715027727,
      "learning_rate": 2.1511627906976745e-06,
      "loss": 0.4149,
      "num_tokens": 9857902.0,
      "step": 75
    },
    {
      "epoch": 0.046671042076861374,
      "grad_norm": 1.3774618307546735,
      "learning_rate": 2.296511627906977e-06,
      "loss": 0.4073,
      "num_tokens": 10508649.0,
      "step": 80
    },
    {
      "epoch": 0.04958798220666521,
      "grad_norm": 0.9977532994086189,
      "learning_rate": 2.4418604651162793e-06,
      "loss": 0.4007,
      "num_tokens": 11165778.0,
      "step": 85
    },
    {
      "epoch": 0.05250492233646904,
      "grad_norm": 1.1279002386313963,
      "learning_rate": 2.5872093023255817e-06,
      "loss": 0.3834,
      "num_tokens": 11832663.0,
      "step": 90
    },
    {
      "epoch": 0.05542186246627288,
      "grad_norm": 1.1949547587228786,
      "learning_rate": 2.7325581395348837e-06,
      "loss": 0.378,
      "num_tokens": 12491366.0,
      "step": 95
    },
    {
      "epoch": 0.05833880259607672,
      "grad_norm": 1.2272744151775523,
      "learning_rate": 2.8779069767441865e-06,
      "loss": 0.3711,
      "num_tokens": 13158513.0,
      "step": 100
    },
    {
      "epoch": 0.06125574272588055,
      "grad_norm": 1.1016643876078596,
      "learning_rate": 3.0232558139534885e-06,
      "loss": 0.3673,
      "num_tokens": 13815015.0,
      "step": 105
    },
    {
      "epoch": 0.0641726828556844,
      "grad_norm": 1.2643000713720414,
      "learning_rate": 3.168604651162791e-06,
      "loss": 0.3597,
      "num_tokens": 14469541.0,
      "step": 110
    },
    {
      "epoch": 0.06708962298548822,
      "grad_norm": 1.201838639094829,
      "learning_rate": 3.313953488372093e-06,
      "loss": 0.3555,
      "num_tokens": 15118625.0,
      "step": 115
    },
    {
      "epoch": 0.07000656311529206,
      "grad_norm": 1.109067555947549,
      "learning_rate": 3.4593023255813957e-06,
      "loss": 0.3533,
      "num_tokens": 15773909.0,
      "step": 120
    },
    {
      "epoch": 0.07292350324509589,
      "grad_norm": 1.1048072692029691,
      "learning_rate": 3.6046511627906977e-06,
      "loss": 0.347,
      "num_tokens": 16431558.0,
      "step": 125
    },
    {
      "epoch": 0.07584044337489973,
      "grad_norm": 1.1222392707741446,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.3561,
      "num_tokens": 17071991.0,
      "step": 130
    },
    {
      "epoch": 0.07875738350470357,
      "grad_norm": 1.268152454159826,
      "learning_rate": 3.8953488372093025e-06,
      "loss": 0.3534,
      "num_tokens": 17712413.0,
      "step": 135
    },
    {
      "epoch": 0.0816743236345074,
      "grad_norm": 1.2649313269721758,
      "learning_rate": 4.040697674418605e-06,
      "loss": 0.3299,
      "num_tokens": 18365792.0,
      "step": 140
    },
    {
      "epoch": 0.08459126376431124,
      "grad_norm": 1.1434149458336165,
      "learning_rate": 4.186046511627907e-06,
      "loss": 0.3306,
      "num_tokens": 19044956.0,
      "step": 145
    },
    {
      "epoch": 0.08750820389411507,
      "grad_norm": 1.2086728419334098,
      "learning_rate": 4.331395348837209e-06,
      "loss": 0.3374,
      "num_tokens": 19695160.0,
      "step": 150
    },
    {
      "epoch": 0.09042514402391891,
      "grad_norm": 1.1141028558068582,
      "learning_rate": 4.476744186046512e-06,
      "loss": 0.3329,
      "num_tokens": 20354718.0,
      "step": 155
    },
    {
      "epoch": 0.09334208415372275,
      "grad_norm": 1.1566468839887507,
      "learning_rate": 4.622093023255814e-06,
      "loss": 0.3341,
      "num_tokens": 21006198.0,
      "step": 160
    },
    {
      "epoch": 0.09625902428352658,
      "grad_norm": 1.1614415172217762,
      "learning_rate": 4.767441860465117e-06,
      "loss": 0.3443,
      "num_tokens": 21634895.0,
      "step": 165
    },
    {
      "epoch": 0.09917596441333042,
      "grad_norm": 1.2888710984530731,
      "learning_rate": 4.912790697674419e-06,
      "loss": 0.3332,
      "num_tokens": 22279296.0,
      "step": 170
    },
    {
      "epoch": 0.10209290454313426,
      "grad_norm": 1.066687159230309,
      "learning_rate": 4.999981321460686e-06,
      "loss": 0.33,
      "num_tokens": 22932800.0,
      "step": 175
    },
    {
      "epoch": 0.10500984467293809,
      "grad_norm": 1.0587240072052146,
      "learning_rate": 4.999771191454943e-06,
      "loss": 0.3227,
      "num_tokens": 23596098.0,
      "step": 180
    },
    {
      "epoch": 0.10792678480274193,
      "grad_norm": 1.0753866637705207,
      "learning_rate": 4.999327605146962e-06,
      "loss": 0.3265,
      "num_tokens": 24240753.0,
      "step": 185
    },
    {
      "epoch": 0.11084372493254575,
      "grad_norm": 1.1696062309729527,
      "learning_rate": 4.998650608567207e-06,
      "loss": 0.3209,
      "num_tokens": 24903847.0,
      "step": 190
    },
    {
      "epoch": 0.1137606650623496,
      "grad_norm": 1.0355502586963417,
      "learning_rate": 4.997740271966877e-06,
      "loss": 0.3146,
      "num_tokens": 25564496.0,
      "step": 195
    },
    {
      "epoch": 0.11667760519215344,
      "grad_norm": 1.119036751899036,
      "learning_rate": 4.996596689810608e-06,
      "loss": 0.315,
      "num_tokens": 26231532.0,
      "step": 200
    },
    {
      "epoch": 0.11959454532195726,
      "grad_norm": 1.088954101903663,
      "learning_rate": 4.995219980766672e-06,
      "loss": 0.3178,
      "num_tokens": 26876933.0,
      "step": 205
    },
    {
      "epoch": 0.1225114854517611,
      "grad_norm": 1.1468096627392024,
      "learning_rate": 4.993610287694663e-06,
      "loss": 0.3251,
      "num_tokens": 27510135.0,
      "step": 210
    },
    {
      "epoch": 0.12542842558156495,
      "grad_norm": 0.9548342307118838,
      "learning_rate": 4.991767777630679e-06,
      "loss": 0.3145,
      "num_tokens": 28154194.0,
      "step": 215
    },
    {
      "epoch": 0.1283453657113688,
      "grad_norm": 1.0591304201257135,
      "learning_rate": 4.989692641769976e-06,
      "loss": 0.3229,
      "num_tokens": 28801919.0,
      "step": 220
    },
    {
      "epoch": 0.1312623058411726,
      "grad_norm": 1.054145800912703,
      "learning_rate": 4.987385095447141e-06,
      "loss": 0.3208,
      "num_tokens": 29456315.0,
      "step": 225
    },
    {
      "epoch": 0.13417924597097644,
      "grad_norm": 1.1458373353014295,
      "learning_rate": 4.9848453781137355e-06,
      "loss": 0.3136,
      "num_tokens": 30111377.0,
      "step": 230
    },
    {
      "epoch": 0.13709618610078028,
      "grad_norm": 1.1484854864144058,
      "learning_rate": 4.982073753313459e-06,
      "loss": 0.3106,
      "num_tokens": 30769448.0,
      "step": 235
    },
    {
      "epoch": 0.14001312623058412,
      "grad_norm": 1.1136794674885304,
      "learning_rate": 4.979070508654793e-06,
      "loss": 0.3123,
      "num_tokens": 31423870.0,
      "step": 240
    },
    {
      "epoch": 0.14293006636038796,
      "grad_norm": 1.1270064595880545,
      "learning_rate": 4.975835955781159e-06,
      "loss": 0.3093,
      "num_tokens": 32085017.0,
      "step": 245
    },
    {
      "epoch": 0.14584700649019178,
      "grad_norm": 0.9197262026545928,
      "learning_rate": 4.97237043033858e-06,
      "loss": 0.3099,
      "num_tokens": 32748599.0,
      "step": 250
    },
    {
      "epoch": 0.14876394661999562,
      "grad_norm": 1.0031484672449607,
      "learning_rate": 4.968674291940851e-06,
      "loss": 0.3113,
      "num_tokens": 33396816.0,
      "step": 255
    },
    {
      "epoch": 0.15168088674979946,
      "grad_norm": 1.0106860095524073,
      "learning_rate": 4.964747924132222e-06,
      "loss": 0.3019,
      "num_tokens": 34048082.0,
      "step": 260
    },
    {
      "epoch": 0.1545978268796033,
      "grad_norm": 1.1390370333863529,
      "learning_rate": 4.9605917343475955e-06,
      "loss": 0.3094,
      "num_tokens": 34695245.0,
      "step": 265
    },
    {
      "epoch": 0.15751476700940714,
      "grad_norm": 0.9304690639529852,
      "learning_rate": 4.9562061538702525e-06,
      "loss": 0.3022,
      "num_tokens": 35357239.0,
      "step": 270
    },
    {
      "epoch": 0.16043170713921096,
      "grad_norm": 1.0043265999798545,
      "learning_rate": 4.9515916377870946e-06,
      "loss": 0.3089,
      "num_tokens": 36009775.0,
      "step": 275
    },
    {
      "epoch": 0.1633486472690148,
      "grad_norm": 1.2003647631325731,
      "learning_rate": 4.946748664941422e-06,
      "loss": 0.3011,
      "num_tokens": 36668698.0,
      "step": 280
    },
    {
      "epoch": 0.16626558739881864,
      "grad_norm": 0.8918137550325422,
      "learning_rate": 4.941677737883243e-06,
      "loss": 0.2965,
      "num_tokens": 37330940.0,
      "step": 285
    },
    {
      "epoch": 0.16918252752862248,
      "grad_norm": 0.9369985322231161,
      "learning_rate": 4.936379382817128e-06,
      "loss": 0.3018,
      "num_tokens": 37988190.0,
      "step": 290
    },
    {
      "epoch": 0.17209946765842632,
      "grad_norm": 0.9779525599227024,
      "learning_rate": 4.930854149547601e-06,
      "loss": 0.2918,
      "num_tokens": 38661283.0,
      "step": 295
    },
    {
      "epoch": 0.17501640778823013,
      "grad_norm": 1.1129159421490522,
      "learning_rate": 4.925102611422091e-06,
      "loss": 0.2927,
      "num_tokens": 39320436.0,
      "step": 300
    },
    {
      "epoch": 0.17793334791803397,
      "grad_norm": 0.9718156616589078,
      "learning_rate": 4.919125365271435e-06,
      "loss": 0.2965,
      "num_tokens": 39974872.0,
      "step": 305
    },
    {
      "epoch": 0.18085028804783781,
      "grad_norm": 0.997867663517212,
      "learning_rate": 4.912923031347944e-06,
      "loss": 0.2887,
      "num_tokens": 40624237.0,
      "step": 310
    },
    {
      "epoch": 0.18376722817764166,
      "grad_norm": 1.0604840903516273,
      "learning_rate": 4.9064962532610465e-06,
      "loss": 0.3093,
      "num_tokens": 41272880.0,
      "step": 315
    },
    {
      "epoch": 0.1866841683074455,
      "grad_norm": 0.9752847517901945,
      "learning_rate": 4.899845697910489e-06,
      "loss": 0.3023,
      "num_tokens": 41916518.0,
      "step": 320
    },
    {
      "epoch": 0.18960110843724934,
      "grad_norm": 0.9361317315869321,
      "learning_rate": 4.89297205541715e-06,
      "loss": 0.2906,
      "num_tokens": 42570198.0,
      "step": 325
    },
    {
      "epoch": 0.19251804856705315,
      "grad_norm": 0.924312875998321,
      "learning_rate": 4.885876039051408e-06,
      "loss": 0.2973,
      "num_tokens": 43235563.0,
      "step": 330
    },
    {
      "epoch": 0.195434988696857,
      "grad_norm": 0.9664895892779287,
      "learning_rate": 4.878558385159143e-06,
      "loss": 0.2932,
      "num_tokens": 43881944.0,
      "step": 335
    },
    {
      "epoch": 0.19835192882666083,
      "grad_norm": 1.1025545029030783,
      "learning_rate": 4.871019853085316e-06,
      "loss": 0.2969,
      "num_tokens": 44536573.0,
      "step": 340
    },
    {
      "epoch": 0.20126886895646467,
      "grad_norm": 1.1546365826334446,
      "learning_rate": 4.863261225095175e-06,
      "loss": 0.2975,
      "num_tokens": 45185298.0,
      "step": 345
    },
    {
      "epoch": 0.20418580908626852,
      "grad_norm": 0.8539653068120385,
      "learning_rate": 4.8552833062930845e-06,
      "loss": 0.2941,
      "num_tokens": 45833516.0,
      "step": 350
    },
    {
      "epoch": 0.20710274921607233,
      "grad_norm": 0.8887354370276694,
      "learning_rate": 4.847086924538975e-06,
      "loss": 0.2911,
      "num_tokens": 46492785.0,
      "step": 355
    },
    {
      "epoch": 0.21001968934587617,
      "grad_norm": 0.8965576316891322,
      "learning_rate": 4.838672930362438e-06,
      "loss": 0.2845,
      "num_tokens": 47147972.0,
      "step": 360
    },
    {
      "epoch": 0.21293662947568,
      "grad_norm": 1.0547471171000238,
      "learning_rate": 4.830042196874472e-06,
      "loss": 0.2889,
      "num_tokens": 47792460.0,
      "step": 365
    },
    {
      "epoch": 0.21585356960548385,
      "grad_norm": 0.9029069684538267,
      "learning_rate": 4.821195619676876e-06,
      "loss": 0.2935,
      "num_tokens": 48449608.0,
      "step": 370
    },
    {
      "epoch": 0.2187705097352877,
      "grad_norm": 1.1056029287041187,
      "learning_rate": 4.812134116769312e-06,
      "loss": 0.2927,
      "num_tokens": 49105080.0,
      "step": 375
    },
    {
      "epoch": 0.2216874498650915,
      "grad_norm": 0.8797309000795296,
      "learning_rate": 4.802858628454056e-06,
      "loss": 0.2895,
      "num_tokens": 49754281.0,
      "step": 380
    },
    {
      "epoch": 0.22460438999489535,
      "grad_norm": 1.044569804031736,
      "learning_rate": 4.793370117238408e-06,
      "loss": 0.2897,
      "num_tokens": 50392756.0,
      "step": 385
    },
    {
      "epoch": 0.2275213301246992,
      "grad_norm": 1.1130572813337458,
      "learning_rate": 4.783669567734826e-06,
      "loss": 0.2899,
      "num_tokens": 51048274.0,
      "step": 390
    },
    {
      "epoch": 0.23043827025450303,
      "grad_norm": 0.960716559832758,
      "learning_rate": 4.773757986558753e-06,
      "loss": 0.2917,
      "num_tokens": 51699260.0,
      "step": 395
    },
    {
      "epoch": 0.23335521038430687,
      "grad_norm": 0.9245960376672167,
      "learning_rate": 4.763636402224151e-06,
      "loss": 0.2897,
      "num_tokens": 52348343.0,
      "step": 400
    },
    {
      "epoch": 0.23627215051411068,
      "grad_norm": 0.8890810056058448,
      "learning_rate": 4.753305865036787e-06,
      "loss": 0.2885,
      "num_tokens": 53001056.0,
      "step": 405
    },
    {
      "epoch": 0.23918909064391453,
      "grad_norm": 0.8579871280345769,
      "learning_rate": 4.742767446985234e-06,
      "loss": 0.2861,
      "num_tokens": 53662504.0,
      "step": 410
    },
    {
      "epoch": 0.24210603077371837,
      "grad_norm": 0.8821250216443968,
      "learning_rate": 4.732022241629637e-06,
      "loss": 0.287,
      "num_tokens": 54318994.0,
      "step": 415
    },
    {
      "epoch": 0.2450229709035222,
      "grad_norm": 0.7974431662516955,
      "learning_rate": 4.721071363988235e-06,
      "loss": 0.2905,
      "num_tokens": 54969623.0,
      "step": 420
    },
    {
      "epoch": 0.24793991103332605,
      "grad_norm": 0.8353737566224922,
      "learning_rate": 4.709915950421653e-06,
      "loss": 0.2878,
      "num_tokens": 55620185.0,
      "step": 425
    },
    {
      "epoch": 0.2508568511631299,
      "grad_norm": 0.9111882828366351,
      "learning_rate": 4.698557158514988e-06,
      "loss": 0.2828,
      "num_tokens": 56274073.0,
      "step": 430
    },
    {
      "epoch": 0.2537737912929337,
      "grad_norm": 0.919475515341444,
      "learning_rate": 4.686996166957683e-06,
      "loss": 0.2822,
      "num_tokens": 56944526.0,
      "step": 435
    },
    {
      "epoch": 0.2566907314227376,
      "grad_norm": 0.8931932491434317,
      "learning_rate": 4.675234175421222e-06,
      "loss": 0.2892,
      "num_tokens": 57609076.0,
      "step": 440
    },
    {
      "epoch": 0.2596076715525414,
      "grad_norm": 0.9216610473053193,
      "learning_rate": 4.663272404434635e-06,
      "loss": 0.2774,
      "num_tokens": 58264579.0,
      "step": 445
    },
    {
      "epoch": 0.2625246116823452,
      "grad_norm": 0.9202742139945425,
      "learning_rate": 4.6511120952578484e-06,
      "loss": 0.2863,
      "num_tokens": 58902959.0,
      "step": 450
    },
    {
      "epoch": 0.26544155181214907,
      "grad_norm": 0.8943558537794027,
      "learning_rate": 4.638754509752878e-06,
      "loss": 0.2796,
      "num_tokens": 59564439.0,
      "step": 455
    },
    {
      "epoch": 0.2683584919419529,
      "grad_norm": 0.863810937001429,
      "learning_rate": 4.6262009302528915e-06,
      "loss": 0.2843,
      "num_tokens": 60214896.0,
      "step": 460
    },
    {
      "epoch": 0.27127543207175675,
      "grad_norm": 0.9983134689445968,
      "learning_rate": 4.61345265942914e-06,
      "loss": 0.2866,
      "num_tokens": 60876642.0,
      "step": 465
    },
    {
      "epoch": 0.27419237220156056,
      "grad_norm": 0.8983030440670102,
      "learning_rate": 4.600511020155778e-06,
      "loss": 0.2848,
      "num_tokens": 61518584.0,
      "step": 470
    },
    {
      "epoch": 0.2771093123313644,
      "grad_norm": 0.8031669478146738,
      "learning_rate": 4.587377355372596e-06,
      "loss": 0.2757,
      "num_tokens": 62183326.0,
      "step": 475
    },
    {
      "epoch": 0.28002625246116825,
      "grad_norm": 0.9265028610272964,
      "learning_rate": 4.57405302794566e-06,
      "loss": 0.2841,
      "num_tokens": 62825255.0,
      "step": 480
    },
    {
      "epoch": 0.28294319259097206,
      "grad_norm": 0.7881945775359938,
      "learning_rate": 4.5605394205258915e-06,
      "loss": 0.2784,
      "num_tokens": 63488470.0,
      "step": 485
    },
    {
      "epoch": 0.28586013272077593,
      "grad_norm": 0.890455779305603,
      "learning_rate": 4.546837935405591e-06,
      "loss": 0.2825,
      "num_tokens": 64144973.0,
      "step": 490
    },
    {
      "epoch": 0.28877707285057974,
      "grad_norm": 0.8245358042376737,
      "learning_rate": 4.532949994372922e-06,
      "loss": 0.2735,
      "num_tokens": 64800071.0,
      "step": 495
    },
    {
      "epoch": 0.29169401298038355,
      "grad_norm": 0.8022563146785544,
      "learning_rate": 4.518877038564374e-06,
      "loss": 0.2805,
      "step": 500
    },
    {
      "epoch": 0.29169401298038355,
      "eval_loss": 0.2839949429035187,
      "eval_num_tokens": 65444398.0,
      "eval_runtime": 611.5469,
      "eval_samples_per_second": 9.442,
      "eval_steps_per_second": 1.181,
      "step": 500
    },
    {
      "epoch": 0.2946109531101874,
      "grad_norm": 0.8282011678135937,
      "learning_rate": 4.50462052831522e-06,
      "loss": 0.2742,
      "num_tokens": 66102293.0,
      "step": 505
    },
    {
      "epoch": 0.29752789323999124,
      "grad_norm": 0.8459868862384502,
      "learning_rate": 4.4901819430079766e-06,
      "loss": 0.2797,
      "num_tokens": 66750025.0,
      "step": 510
    },
    {
      "epoch": 0.3004448333697951,
      "grad_norm": 1.0044989964568452,
      "learning_rate": 4.4755627809188885e-06,
      "loss": 0.2808,
      "num_tokens": 67398138.0,
      "step": 515
    },
    {
      "epoch": 0.3033617734995989,
      "grad_norm": 0.8645633533810008,
      "learning_rate": 4.460764559062458e-06,
      "loss": 0.2743,
      "num_tokens": 68064298.0,
      "step": 520
    },
    {
      "epoch": 0.30627871362940273,
      "grad_norm": 0.7902367580597776,
      "learning_rate": 4.445788813034024e-06,
      "loss": 0.2745,
      "num_tokens": 68730271.0,
      "step": 525
    },
    {
      "epoch": 0.3091956537592066,
      "grad_norm": 0.8074282979202131,
      "learning_rate": 4.430637096850415e-06,
      "loss": 0.278,
      "num_tokens": 69373992.0,
      "step": 530
    },
    {
      "epoch": 0.3121125938890104,
      "grad_norm": 0.7997689062881244,
      "learning_rate": 4.4153109827886894e-06,
      "loss": 0.2773,
      "num_tokens": 70013275.0,
      "step": 535
    },
    {
      "epoch": 0.3150295340188143,
      "grad_norm": 0.7763977877420887,
      "learning_rate": 4.399812061222983e-06,
      "loss": 0.2792,
      "num_tokens": 70667022.0,
      "step": 540
    },
    {
      "epoch": 0.3179464741486181,
      "grad_norm": 0.8027330754403003,
      "learning_rate": 4.384141940459482e-06,
      "loss": 0.2692,
      "num_tokens": 71329905.0,
      "step": 545
    },
    {
      "epoch": 0.3208634142784219,
      "grad_norm": 0.842566387021002,
      "learning_rate": 4.36830224656952e-06,
      "loss": 0.2777,
      "num_tokens": 71981698.0,
      "step": 550
    },
    {
      "epoch": 0.3237803544082258,
      "grad_norm": 0.760795844133808,
      "learning_rate": 4.352294623220853e-06,
      "loss": 0.2719,
      "num_tokens": 72641200.0,
      "step": 555
    },
    {
      "epoch": 0.3266972945380296,
      "grad_norm": 0.767412158526903,
      "learning_rate": 4.336120731507095e-06,
      "loss": 0.2786,
      "num_tokens": 73289056.0,
      "step": 560
    },
    {
      "epoch": 0.32961423466783346,
      "grad_norm": 0.839424142267615,
      "learning_rate": 4.319782249775343e-06,
      "loss": 0.2774,
      "num_tokens": 73940084.0,
      "step": 565
    },
    {
      "epoch": 0.3325311747976373,
      "grad_norm": 0.7953209168681115,
      "learning_rate": 4.303280873452022e-06,
      "loss": 0.2761,
      "num_tokens": 74602418.0,
      "step": 570
    },
    {
      "epoch": 0.3354481149274411,
      "grad_norm": 0.9703712347237416,
      "learning_rate": 4.286618314866953e-06,
      "loss": 0.2739,
      "num_tokens": 75260268.0,
      "step": 575
    },
    {
      "epoch": 0.33836505505724496,
      "grad_norm": 0.785005018552605,
      "learning_rate": 4.269796303075664e-06,
      "loss": 0.2719,
      "num_tokens": 75911746.0,
      "step": 580
    },
    {
      "epoch": 0.34128199518704877,
      "grad_norm": 0.9895796846796822,
      "learning_rate": 4.252816583679963e-06,
      "loss": 0.2703,
      "num_tokens": 76570208.0,
      "step": 585
    },
    {
      "epoch": 0.34419893531685264,
      "grad_norm": 0.8038675047644025,
      "learning_rate": 4.235680918646814e-06,
      "loss": 0.2733,
      "num_tokens": 77231968.0,
      "step": 590
    },
    {
      "epoch": 0.34711587544665645,
      "grad_norm": 0.8232561538467604,
      "learning_rate": 4.21839108612548e-06,
      "loss": 0.2753,
      "num_tokens": 77888752.0,
      "step": 595
    },
    {
      "epoch": 0.35003281557646027,
      "grad_norm": 0.8114352965857334,
      "learning_rate": 4.200948880263021e-06,
      "loss": 0.2764,
      "num_tokens": 78549259.0,
      "step": 600
    },
    {
      "epoch": 0.35294975570626413,
      "grad_norm": 0.918026001844272,
      "learning_rate": 4.1833561110181114e-06,
      "loss": 0.2696,
      "num_tokens": 79207039.0,
      "step": 605
    },
    {
      "epoch": 0.35586669583606795,
      "grad_norm": 0.8799204301828087,
      "learning_rate": 4.165614603973225e-06,
      "loss": 0.2663,
      "num_tokens": 79861860.0,
      "step": 610
    },
    {
      "epoch": 0.3587836359658718,
      "grad_norm": 0.87043880731276,
      "learning_rate": 4.147726200145192e-06,
      "loss": 0.2755,
      "num_tokens": 80518900.0,
      "step": 615
    },
    {
      "epoch": 0.36170057609567563,
      "grad_norm": 0.9211963447162147,
      "learning_rate": 4.1296927557941625e-06,
      "loss": 0.2767,
      "num_tokens": 81161582.0,
      "step": 620
    },
    {
      "epoch": 0.3646175162254795,
      "grad_norm": 0.8299533305119133,
      "learning_rate": 4.111516142230986e-06,
      "loss": 0.2696,
      "num_tokens": 81813290.0,
      "step": 625
    },
    {
      "epoch": 0.3675344563552833,
      "grad_norm": 0.8309388002459559,
      "learning_rate": 4.093198245623022e-06,
      "loss": 0.2718,
      "num_tokens": 82481134.0,
      "step": 630
    },
    {
      "epoch": 0.3704513964850871,
      "grad_norm": 0.7974940185655467,
      "learning_rate": 4.074740966798417e-06,
      "loss": 0.275,
      "num_tokens": 83121589.0,
      "step": 635
    },
    {
      "epoch": 0.373368336614891,
      "grad_norm": 0.7877568112781695,
      "learning_rate": 4.056146221048861e-06,
      "loss": 0.2689,
      "num_tokens": 83783858.0,
      "step": 640
    },
    {
      "epoch": 0.3762852767446948,
      "grad_norm": 0.7457416133818777,
      "learning_rate": 4.037415937930834e-06,
      "loss": 0.2698,
      "num_tokens": 84439289.0,
      "step": 645
    },
    {
      "epoch": 0.3792022168744987,
      "grad_norm": 0.804387030534503,
      "learning_rate": 4.018552061065385e-06,
      "loss": 0.2656,
      "num_tokens": 85096010.0,
      "step": 650
    },
    {
      "epoch": 0.3821191570043025,
      "grad_norm": 0.7866414763779639,
      "learning_rate": 3.999556547936433e-06,
      "loss": 0.2679,
      "num_tokens": 85763093.0,
      "step": 655
    },
    {
      "epoch": 0.3850360971341063,
      "grad_norm": 0.7975246358334895,
      "learning_rate": 3.980431369687657e-06,
      "loss": 0.2685,
      "num_tokens": 86414815.0,
      "step": 660
    },
    {
      "epoch": 0.38795303726391017,
      "grad_norm": 0.7541579622412656,
      "learning_rate": 3.961178510917938e-06,
      "loss": 0.2664,
      "num_tokens": 87072103.0,
      "step": 665
    },
    {
      "epoch": 0.390869977393714,
      "grad_norm": 0.8462615544123323,
      "learning_rate": 3.941799969475426e-06,
      "loss": 0.2697,
      "num_tokens": 87724238.0,
      "step": 670
    },
    {
      "epoch": 0.39378691752351785,
      "grad_norm": 0.719877168553545,
      "learning_rate": 3.922297756250231e-06,
      "loss": 0.2682,
      "num_tokens": 88385611.0,
      "step": 675
    },
    {
      "epoch": 0.39670385765332167,
      "grad_norm": 0.8061212707580667,
      "learning_rate": 3.902673894965739e-06,
      "loss": 0.2652,
      "num_tokens": 89047105.0,
      "step": 680
    },
    {
      "epoch": 0.3996207977831255,
      "grad_norm": 0.8326055553096022,
      "learning_rate": 3.88293042196863e-06,
      "loss": 0.2651,
      "num_tokens": 89703947.0,
      "step": 685
    },
    {
      "epoch": 0.40253773791292935,
      "grad_norm": 0.8409291362121083,
      "learning_rate": 3.863069386017559e-06,
      "loss": 0.2676,
      "num_tokens": 90375478.0,
      "step": 690
    },
    {
      "epoch": 0.40545467804273316,
      "grad_norm": 1.1749044956975412,
      "learning_rate": 3.8430928480705595e-06,
      "loss": 0.268,
      "num_tokens": 91041509.0,
      "step": 695
    },
    {
      "epoch": 0.40837161817253703,
      "grad_norm": 0.9615687348415685,
      "learning_rate": 3.823002881071182e-06,
      "loss": 0.2713,
      "num_tokens": 91687478.0,
      "step": 700
    },
    {
      "epoch": 0.41128855830234085,
      "grad_norm": 0.8598987130093434,
      "learning_rate": 3.802801569733385e-06,
      "loss": 0.2633,
      "num_tokens": 92350865.0,
      "step": 705
    },
    {
      "epoch": 0.41420549843214466,
      "grad_norm": 0.9022414410706127,
      "learning_rate": 3.7824910103252094e-06,
      "loss": 0.2705,
      "num_tokens": 93001864.0,
      "step": 710
    },
    {
      "epoch": 0.4171224385619485,
      "grad_norm": 0.8563602736920568,
      "learning_rate": 3.7620733104512457e-06,
      "loss": 0.2686,
      "num_tokens": 93646547.0,
      "step": 715
    },
    {
      "epoch": 0.42003937869175234,
      "grad_norm": 0.8403741919990089,
      "learning_rate": 3.741550588833938e-06,
      "loss": 0.261,
      "num_tokens": 94309272.0,
      "step": 720
    },
    {
      "epoch": 0.4229563188215562,
      "grad_norm": 0.8382205734712479,
      "learning_rate": 3.7209249750937194e-06,
      "loss": 0.2583,
      "num_tokens": 94974415.0,
      "step": 725
    },
    {
      "epoch": 0.42587325895136,
      "grad_norm": 0.7670934985699519,
      "learning_rate": 3.700198609528027e-06,
      "loss": 0.2621,
      "num_tokens": 95637473.0,
      "step": 730
    },
    {
      "epoch": 0.42879019908116384,
      "grad_norm": 0.889958478839124,
      "learning_rate": 3.679373642889205e-06,
      "loss": 0.263,
      "num_tokens": 96285419.0,
      "step": 735
    },
    {
      "epoch": 0.4317071392109677,
      "grad_norm": 0.8154007725747381,
      "learning_rate": 3.6584522361613227e-06,
      "loss": 0.2657,
      "num_tokens": 96926138.0,
      "step": 740
    },
    {
      "epoch": 0.4346240793407715,
      "grad_norm": 0.8307773360486326,
      "learning_rate": 3.6374365603359347e-06,
      "loss": 0.2662,
      "num_tokens": 97588450.0,
      "step": 745
    },
    {
      "epoch": 0.4375410194705754,
      "grad_norm": 0.8078529106621964,
      "learning_rate": 3.616328796186794e-06,
      "loss": 0.2681,
      "num_tokens": 98242023.0,
      "step": 750
    },
    {
      "epoch": 0.4404579596003792,
      "grad_norm": 0.9983558574344977,
      "learning_rate": 3.5951311340435597e-06,
      "loss": 0.2637,
      "num_tokens": 98902913.0,
      "step": 755
    },
    {
      "epoch": 0.443374899730183,
      "grad_norm": 0.9165975707045902,
      "learning_rate": 3.573845773564506e-06,
      "loss": 0.2727,
      "num_tokens": 99533885.0,
      "step": 760
    },
    {
      "epoch": 0.4462918398599869,
      "grad_norm": 0.7822942999479296,
      "learning_rate": 3.5524749235082728e-06,
      "loss": 0.2712,
      "num_tokens": 100188467.0,
      "step": 765
    },
    {
      "epoch": 0.4492087799897907,
      "grad_norm": 0.7478178653016506,
      "learning_rate": 3.5310208015046547e-06,
      "loss": 0.2661,
      "num_tokens": 100842626.0,
      "step": 770
    },
    {
      "epoch": 0.45212572011959457,
      "grad_norm": 0.8348739730561777,
      "learning_rate": 3.5094856338244882e-06,
      "loss": 0.2605,
      "num_tokens": 101509817.0,
      "step": 775
    },
    {
      "epoch": 0.4550426602493984,
      "grad_norm": 0.8650672549166954,
      "learning_rate": 3.4878716551486296e-06,
      "loss": 0.2626,
      "num_tokens": 102161268.0,
      "step": 780
    },
    {
      "epoch": 0.4579596003792022,
      "grad_norm": 0.8437538579535597,
      "learning_rate": 3.466181108336068e-06,
      "loss": 0.2631,
      "num_tokens": 102815935.0,
      "step": 785
    },
    {
      "epoch": 0.46087654050900606,
      "grad_norm": 0.8055131749035074,
      "learning_rate": 3.444416244191184e-06,
      "loss": 0.2548,
      "num_tokens": 103480501.0,
      "step": 790
    },
    {
      "epoch": 0.4637934806388099,
      "grad_norm": 0.7184736639862396,
      "learning_rate": 3.422579321230185e-06,
      "loss": 0.2615,
      "num_tokens": 104127953.0,
      "step": 795
    },
    {
      "epoch": 0.46671042076861374,
      "grad_norm": 0.7769160728536546,
      "learning_rate": 3.400672605446746e-06,
      "loss": 0.2649,
      "num_tokens": 104784029.0,
      "step": 800
    },
    {
      "epoch": 0.46962736089841756,
      "grad_norm": 0.7956791364511919,
      "learning_rate": 3.378698370076865e-06,
      "loss": 0.2585,
      "num_tokens": 105448623.0,
      "step": 805
    },
    {
      "epoch": 0.47254430102822137,
      "grad_norm": 0.7971292848423556,
      "learning_rate": 3.356658895362974e-06,
      "loss": 0.258,
      "num_tokens": 106102919.0,
      "step": 810
    },
    {
      "epoch": 0.47546124115802524,
      "grad_norm": 0.8024065230344523,
      "learning_rate": 3.334556468317322e-06,
      "loss": 0.2639,
      "num_tokens": 106766059.0,
      "step": 815
    },
    {
      "epoch": 0.47837818128782905,
      "grad_norm": 0.8627952778096136,
      "learning_rate": 3.3123933824846517e-06,
      "loss": 0.2637,
      "num_tokens": 107426299.0,
      "step": 820
    },
    {
      "epoch": 0.4812951214176329,
      "grad_norm": 0.8393622591727319,
      "learning_rate": 3.2901719377042003e-06,
      "loss": 0.2629,
      "num_tokens": 108065870.0,
      "step": 825
    },
    {
      "epoch": 0.48421206154743673,
      "grad_norm": 0.7495915565490366,
      "learning_rate": 3.2678944398710535e-06,
      "loss": 0.2624,
      "num_tokens": 108729197.0,
      "step": 830
    },
    {
      "epoch": 0.48712900167724055,
      "grad_norm": 0.7951127772717166,
      "learning_rate": 3.2455632006968564e-06,
      "loss": 0.2644,
      "num_tokens": 109377885.0,
      "step": 835
    },
    {
      "epoch": 0.4900459418070444,
      "grad_norm": 0.8730539703385664,
      "learning_rate": 3.223180537469938e-06,
      "loss": 0.2631,
      "num_tokens": 110046067.0,
      "step": 840
    },
    {
      "epoch": 0.49296288193684823,
      "grad_norm": 0.8086162516026355,
      "learning_rate": 3.200748772814844e-06,
      "loss": 0.2529,
      "num_tokens": 110715109.0,
      "step": 845
    },
    {
      "epoch": 0.4958798220666521,
      "grad_norm": 0.8485834526237742,
      "learning_rate": 3.1782702344513215e-06,
      "loss": 0.2607,
      "num_tokens": 111368579.0,
      "step": 850
    },
    {
      "epoch": 0.4987967621964559,
      "grad_norm": 0.79225012559078,
      "learning_rate": 3.1557472549527747e-06,
      "loss": 0.2618,
      "num_tokens": 112019671.0,
      "step": 855
    },
    {
      "epoch": 0.5017137023262598,
      "grad_norm": 0.8425810939639716,
      "learning_rate": 3.133182171504214e-06,
      "loss": 0.2627,
      "num_tokens": 112674101.0,
      "step": 860
    },
    {
      "epoch": 0.5046306424560636,
      "grad_norm": 0.744964617174656,
      "learning_rate": 3.110577325659734e-06,
      "loss": 0.2583,
      "num_tokens": 113327455.0,
      "step": 865
    },
    {
      "epoch": 0.5075475825858674,
      "grad_norm": 0.7709087869015558,
      "learning_rate": 3.0879350630995284e-06,
      "loss": 0.2631,
      "num_tokens": 113978391.0,
      "step": 870
    },
    {
      "epoch": 0.5104645227156712,
      "grad_norm": 0.8044798860684248,
      "learning_rate": 3.0652577333864812e-06,
      "loss": 0.2569,
      "num_tokens": 114620393.0,
      "step": 875
    },
    {
      "epoch": 0.5133814628454751,
      "grad_norm": 0.7842031411597096,
      "learning_rate": 3.0425476897223584e-06,
      "loss": 0.2611,
      "num_tokens": 115280804.0,
      "step": 880
    },
    {
      "epoch": 0.516298402975279,
      "grad_norm": 0.8035683788722667,
      "learning_rate": 3.019807288703615e-06,
      "loss": 0.2553,
      "num_tokens": 115935583.0,
      "step": 885
    },
    {
      "epoch": 0.5192153431050828,
      "grad_norm": 0.7544056798614338,
      "learning_rate": 2.99703889007686e-06,
      "loss": 0.2563,
      "num_tokens": 116605162.0,
      "step": 890
    },
    {
      "epoch": 0.5221322832348866,
      "grad_norm": 0.7470139344901265,
      "learning_rate": 2.9742448564939785e-06,
      "loss": 0.2566,
      "num_tokens": 117256010.0,
      "step": 895
    },
    {
      "epoch": 0.5250492233646904,
      "grad_norm": 0.7773494324524246,
      "learning_rate": 2.95142755326697e-06,
      "loss": 0.2557,
      "num_tokens": 117920457.0,
      "step": 900
    },
    {
      "epoch": 0.5279661634944943,
      "grad_norm": 0.908172722597612,
      "learning_rate": 2.9285893481224976e-06,
      "loss": 0.2548,
      "num_tokens": 118587118.0,
      "step": 905
    },
    {
      "epoch": 0.5308831036242981,
      "grad_norm": 0.7475823569135418,
      "learning_rate": 2.9057326109561955e-06,
      "loss": 0.2613,
      "num_tokens": 119238394.0,
      "step": 910
    },
    {
      "epoch": 0.533800043754102,
      "grad_norm": 0.7910527226931955,
      "learning_rate": 2.8828597135867446e-06,
      "loss": 0.2568,
      "num_tokens": 119902185.0,
      "step": 915
    },
    {
      "epoch": 0.5367169838839058,
      "grad_norm": 0.7067850895490586,
      "learning_rate": 2.859973029509753e-06,
      "loss": 0.2514,
      "num_tokens": 120578709.0,
      "step": 920
    },
    {
      "epoch": 0.5396339240137096,
      "grad_norm": 0.7318968008202511,
      "learning_rate": 2.83707493365146e-06,
      "loss": 0.2564,
      "num_tokens": 121237827.0,
      "step": 925
    },
    {
      "epoch": 0.5425508641435135,
      "grad_norm": 0.7705210064312026,
      "learning_rate": 2.8141678021222933e-06,
      "loss": 0.2525,
      "num_tokens": 121898853.0,
      "step": 930
    },
    {
      "epoch": 0.5454678042733173,
      "grad_norm": 0.8157907159758865,
      "learning_rate": 2.791254011970301e-06,
      "loss": 0.2558,
      "num_tokens": 122562375.0,
      "step": 935
    },
    {
      "epoch": 0.5483847444031211,
      "grad_norm": 0.8681587026037888,
      "learning_rate": 2.7683359409344905e-06,
      "loss": 0.2602,
      "num_tokens": 123211114.0,
      "step": 940
    },
    {
      "epoch": 0.5513016845329249,
      "grad_norm": 0.7556332005656767,
      "learning_rate": 2.745415967198093e-06,
      "loss": 0.254,
      "num_tokens": 123865098.0,
      "step": 945
    },
    {
      "epoch": 0.5542186246627288,
      "grad_norm": 0.7946261128648097,
      "learning_rate": 2.722496469141779e-06,
      "loss": 0.2586,
      "num_tokens": 124518258.0,
      "step": 950
    },
    {
      "epoch": 0.5571355647925327,
      "grad_norm": 0.7105424221615,
      "learning_rate": 2.699579825096857e-06,
      "loss": 0.2558,
      "num_tokens": 125177330.0,
      "step": 955
    },
    {
      "epoch": 0.5600525049223365,
      "grad_norm": 0.8047130073772089,
      "learning_rate": 2.6766684130984818e-06,
      "loss": 0.2518,
      "num_tokens": 125844180.0,
      "step": 960
    },
    {
      "epoch": 0.5629694450521403,
      "grad_norm": 0.9021937489211278,
      "learning_rate": 2.653764610638881e-06,
      "loss": 0.2622,
      "num_tokens": 126493064.0,
      "step": 965
    },
    {
      "epoch": 0.5658863851819441,
      "grad_norm": 0.8477417229161418,
      "learning_rate": 2.6308707944206487e-06,
      "loss": 0.256,
      "num_tokens": 127140885.0,
      "step": 970
    },
    {
      "epoch": 0.5688033253117479,
      "grad_norm": 0.707517618098058,
      "learning_rate": 2.607989340110121e-06,
      "loss": 0.247,
      "num_tokens": 127799249.0,
      "step": 975
    },
    {
      "epoch": 0.5717202654415519,
      "grad_norm": 0.7587943404137327,
      "learning_rate": 2.5851226220908504e-06,
      "loss": 0.2572,
      "num_tokens": 128456895.0,
      "step": 980
    },
    {
      "epoch": 0.5746372055713557,
      "grad_norm": 0.7584809790016415,
      "learning_rate": 2.562273013217218e-06,
      "loss": 0.2497,
      "num_tokens": 129122383.0,
      "step": 985
    },
    {
      "epoch": 0.5775541457011595,
      "grad_norm": 0.8214288920378848,
      "learning_rate": 2.539442884568211e-06,
      "loss": 0.2541,
      "num_tokens": 129779846.0,
      "step": 990
    },
    {
      "epoch": 0.5804710858309633,
      "grad_norm": 0.7290259201155611,
      "learning_rate": 2.5166346052013734e-06,
      "loss": 0.257,
      "num_tokens": 130425019.0,
      "step": 995
    },
    {
      "epoch": 0.5833880259607671,
      "grad_norm": 0.8673297271017076,
      "learning_rate": 2.4938505419069737e-06,
      "loss": 0.2524,
      "step": 1000
    },
    {
      "epoch": 0.5833880259607671,
      "eval_loss": 0.26100972294807434,
      "eval_num_tokens": 131094159.0,
      "eval_runtime": 607.9741,
      "eval_samples_per_second": 9.497,
      "eval_steps_per_second": 1.188,
      "step": 1000
    },
    {
      "epoch": 0.586304966090571,
      "grad_norm": 0.9024948847675129,
      "learning_rate": 2.4710930589624043e-06,
      "loss": 0.2543,
      "num_tokens": 131753695.0,
      "step": 1005
    },
    {
      "epoch": 0.5892219062203748,
      "grad_norm": 0.8681506257906868,
      "learning_rate": 2.4483645178868436e-06,
      "loss": 0.2554,
      "num_tokens": 132413825.0,
      "step": 1010
    },
    {
      "epoch": 0.5921388463501787,
      "grad_norm": 0.7962179648149258,
      "learning_rate": 2.425667277196205e-06,
      "loss": 0.2548,
      "num_tokens": 133068552.0,
      "step": 1015
    },
    {
      "epoch": 0.5950557864799825,
      "grad_norm": 0.804654580141575,
      "learning_rate": 2.4030036921583934e-06,
      "loss": 0.2564,
      "num_tokens": 133717600.0,
      "step": 1020
    },
    {
      "epoch": 0.5979727266097863,
      "grad_norm": 0.7346044185339496,
      "learning_rate": 2.380376114548905e-06,
      "loss": 0.2555,
      "num_tokens": 134372291.0,
      "step": 1025
    },
    {
      "epoch": 0.6008896667395902,
      "grad_norm": 0.8118188644094798,
      "learning_rate": 2.3577868924067838e-06,
      "loss": 0.2577,
      "num_tokens": 135017558.0,
      "step": 1030
    },
    {
      "epoch": 0.603806606869394,
      "grad_norm": 0.7221021652672162,
      "learning_rate": 2.3352383697909685e-06,
      "loss": 0.256,
      "num_tokens": 135666125.0,
      "step": 1035
    },
    {
      "epoch": 0.6067235469991978,
      "grad_norm": 0.7352033663094344,
      "learning_rate": 2.312732886537052e-06,
      "loss": 0.2575,
      "num_tokens": 136315008.0,
      "step": 1040
    },
    {
      "epoch": 0.6096404871290017,
      "grad_norm": 0.8607128532349083,
      "learning_rate": 2.29027277801448e-06,
      "loss": 0.2595,
      "num_tokens": 136959663.0,
      "step": 1045
    },
    {
      "epoch": 0.6125574272588055,
      "grad_norm": 0.7969656435537161,
      "learning_rate": 2.267860374884213e-06,
      "loss": 0.2541,
      "num_tokens": 137610280.0,
      "step": 1050
    },
    {
      "epoch": 0.6154743673886094,
      "grad_norm": 0.7672936399726409,
      "learning_rate": 2.245498002856874e-06,
      "loss": 0.2615,
      "num_tokens": 138253534.0,
      "step": 1055
    },
    {
      "epoch": 0.6183913075184132,
      "grad_norm": 0.8116002548995157,
      "learning_rate": 2.2231879824514114e-06,
      "loss": 0.2514,
      "num_tokens": 138918946.0,
      "step": 1060
    },
    {
      "epoch": 0.621308247648217,
      "grad_norm": 0.8343062523034186,
      "learning_rate": 2.2009326287543046e-06,
      "loss": 0.2549,
      "num_tokens": 139575919.0,
      "step": 1065
    },
    {
      "epoch": 0.6242251877780208,
      "grad_norm": 0.7030831591540176,
      "learning_rate": 2.1787342511793303e-06,
      "loss": 0.2543,
      "num_tokens": 140246394.0,
      "step": 1070
    },
    {
      "epoch": 0.6271421279078246,
      "grad_norm": 0.7676168932953417,
      "learning_rate": 2.156595153227911e-06,
      "loss": 0.2522,
      "num_tokens": 140908500.0,
      "step": 1075
    },
    {
      "epoch": 0.6300590680376286,
      "grad_norm": 0.7565551405042247,
      "learning_rate": 2.13451763225009e-06,
      "loss": 0.2538,
      "num_tokens": 141560588.0,
      "step": 1080
    },
    {
      "epoch": 0.6329760081674324,
      "grad_norm": 0.7523126517976036,
      "learning_rate": 2.1125039792061346e-06,
      "loss": 0.2507,
      "num_tokens": 142230913.0,
      "step": 1085
    },
    {
      "epoch": 0.6358929482972362,
      "grad_norm": 0.7547351807576873,
      "learning_rate": 2.0905564784288064e-06,
      "loss": 0.252,
      "num_tokens": 142880248.0,
      "step": 1090
    },
    {
      "epoch": 0.63880988842704,
      "grad_norm": 0.763981058542318,
      "learning_rate": 2.0686774073863183e-06,
      "loss": 0.2508,
      "num_tokens": 143538441.0,
      "step": 1095
    },
    {
      "epoch": 0.6417268285568438,
      "grad_norm": 0.7513777028630977,
      "learning_rate": 2.0468690364460032e-06,
      "loss": 0.2499,
      "num_tokens": 144195649.0,
      "step": 1100
    },
    {
      "epoch": 0.6446437686866477,
      "grad_norm": 0.6978388439562043,
      "learning_rate": 2.0251336286387246e-06,
      "loss": 0.2539,
      "num_tokens": 144840414.0,
      "step": 1105
    },
    {
      "epoch": 0.6475607088164516,
      "grad_norm": 0.7122399335583397,
      "learning_rate": 2.003473439424037e-06,
      "loss": 0.2516,
      "num_tokens": 145497670.0,
      "step": 1110
    },
    {
      "epoch": 0.6504776489462554,
      "grad_norm": 0.7662075027673452,
      "learning_rate": 1.9818907164561474e-06,
      "loss": 0.2508,
      "num_tokens": 146167309.0,
      "step": 1115
    },
    {
      "epoch": 0.6533945890760592,
      "grad_norm": 0.7079059380454418,
      "learning_rate": 1.960387699350673e-06,
      "loss": 0.2471,
      "num_tokens": 146836924.0,
      "step": 1120
    },
    {
      "epoch": 0.656311529205863,
      "grad_norm": 0.6995228988412967,
      "learning_rate": 1.9389666194522416e-06,
      "loss": 0.2477,
      "num_tokens": 147496049.0,
      "step": 1125
    },
    {
      "epoch": 0.6592284693356669,
      "grad_norm": 0.7018077763682937,
      "learning_rate": 1.9176296996029455e-06,
      "loss": 0.2505,
      "num_tokens": 148149073.0,
      "step": 1130
    },
    {
      "epoch": 0.6621454094654707,
      "grad_norm": 0.6955137369821246,
      "learning_rate": 1.8963791539116794e-06,
      "loss": 0.2589,
      "num_tokens": 148811679.0,
      "step": 1135
    },
    {
      "epoch": 0.6650623495952745,
      "grad_norm": 0.7175039810377445,
      "learning_rate": 1.8752171875243897e-06,
      "loss": 0.254,
      "num_tokens": 149453909.0,
      "step": 1140
    },
    {
      "epoch": 0.6679792897250784,
      "grad_norm": 0.7540266711347994,
      "learning_rate": 1.8541459963952401e-06,
      "loss": 0.2534,
      "num_tokens": 150113931.0,
      "step": 1145
    },
    {
      "epoch": 0.6708962298548822,
      "grad_norm": 0.6958604937899281,
      "learning_rate": 1.8331677670587489e-06,
      "loss": 0.2472,
      "num_tokens": 150775568.0,
      "step": 1150
    },
    {
      "epoch": 0.6738131699846861,
      "grad_norm": 0.7533171707259878,
      "learning_rate": 1.812284676402889e-06,
      "loss": 0.2532,
      "num_tokens": 151432991.0,
      "step": 1155
    },
    {
      "epoch": 0.6767301101144899,
      "grad_norm": 0.7169328151307629,
      "learning_rate": 1.7914988914431958e-06,
      "loss": 0.2515,
      "num_tokens": 152071671.0,
      "step": 1160
    },
    {
      "epoch": 0.6796470502442937,
      "grad_norm": 0.6814651782340072,
      "learning_rate": 1.7708125690978973e-06,
      "loss": 0.2577,
      "num_tokens": 152727004.0,
      "step": 1165
    },
    {
      "epoch": 0.6825639903740975,
      "grad_norm": 0.7553012186400901,
      "learning_rate": 1.7502278559641e-06,
      "loss": 0.246,
      "num_tokens": 153403038.0,
      "step": 1170
    },
    {
      "epoch": 0.6854809305039014,
      "grad_norm": 0.7916967627609816,
      "learning_rate": 1.7297468880950275e-06,
      "loss": 0.2456,
      "num_tokens": 154050570.0,
      "step": 1175
    },
    {
      "epoch": 0.6883978706337053,
      "grad_norm": 0.7570768314074023,
      "learning_rate": 1.7093717907783725e-06,
      "loss": 0.248,
      "num_tokens": 154707732.0,
      "step": 1180
    },
    {
      "epoch": 0.6913148107635091,
      "grad_norm": 0.7403254580619876,
      "learning_rate": 1.6891046783157577e-06,
      "loss": 0.2506,
      "num_tokens": 155357884.0,
      "step": 1185
    },
    {
      "epoch": 0.6942317508933129,
      "grad_norm": 0.7862768452334575,
      "learning_rate": 1.668947653803332e-06,
      "loss": 0.2478,
      "num_tokens": 156012053.0,
      "step": 1190
    },
    {
      "epoch": 0.6971486910231167,
      "grad_norm": 0.6944065718383872,
      "learning_rate": 1.6489028089135412e-06,
      "loss": 0.2403,
      "num_tokens": 156675319.0,
      "step": 1195
    },
    {
      "epoch": 0.7000656311529205,
      "grad_norm": 0.689978749467197,
      "learning_rate": 1.6289722236780708e-06,
      "loss": 0.2443,
      "num_tokens": 157341110.0,
      "step": 1200
    },
    {
      "epoch": 0.7029825712827245,
      "grad_norm": 0.7401973229886903,
      "learning_rate": 1.6091579662720085e-06,
      "loss": 0.2468,
      "num_tokens": 157999131.0,
      "step": 1205
    },
    {
      "epoch": 0.7058995114125283,
      "grad_norm": 0.7320601106731225,
      "learning_rate": 1.5894620927992305e-06,
      "loss": 0.2523,
      "num_tokens": 158639604.0,
      "step": 1210
    },
    {
      "epoch": 0.7088164515423321,
      "grad_norm": 0.7143585333776589,
      "learning_rate": 1.5698866470790408e-06,
      "loss": 0.2511,
      "num_tokens": 159287175.0,
      "step": 1215
    },
    {
      "epoch": 0.7117333916721359,
      "grad_norm": 0.805040758465756,
      "learning_rate": 1.5504336604340859e-06,
      "loss": 0.2527,
      "num_tokens": 159928090.0,
      "step": 1220
    },
    {
      "epoch": 0.7146503318019398,
      "grad_norm": 0.7525019096597073,
      "learning_rate": 1.5311051514795689e-06,
      "loss": 0.2524,
      "num_tokens": 160587939.0,
      "step": 1225
    },
    {
      "epoch": 0.7175672719317436,
      "grad_norm": 0.7709357498772824,
      "learning_rate": 1.5119031259137786e-06,
      "loss": 0.2495,
      "num_tokens": 161239965.0,
      "step": 1230
    },
    {
      "epoch": 0.7204842120615474,
      "grad_norm": 0.6930775959107991,
      "learning_rate": 1.4928295763099595e-06,
      "loss": 0.2471,
      "num_tokens": 161899763.0,
      "step": 1235
    },
    {
      "epoch": 0.7234011521913513,
      "grad_norm": 0.7146451766584444,
      "learning_rate": 1.4738864819095478e-06,
      "loss": 0.2468,
      "num_tokens": 162563978.0,
      "step": 1240
    },
    {
      "epoch": 0.7263180923211551,
      "grad_norm": 0.7233964208181292,
      "learning_rate": 1.4550758084167823e-06,
      "loss": 0.2468,
      "num_tokens": 163215392.0,
      "step": 1245
    },
    {
      "epoch": 0.729235032450959,
      "grad_norm": 0.7535387485017295,
      "learning_rate": 1.4363995077947318e-06,
      "loss": 0.2478,
      "num_tokens": 163876128.0,
      "step": 1250
    },
    {
      "epoch": 0.7321519725807628,
      "grad_norm": 0.710154633510974,
      "learning_rate": 1.417859518062738e-06,
      "loss": 0.2471,
      "num_tokens": 164536666.0,
      "step": 1255
    },
    {
      "epoch": 0.7350689127105666,
      "grad_norm": 0.7793325296651725,
      "learning_rate": 1.3994577630953085e-06,
      "loss": 0.2465,
      "num_tokens": 165186675.0,
      "step": 1260
    },
    {
      "epoch": 0.7379858528403704,
      "grad_norm": 0.7204453905441239,
      "learning_rate": 1.3811961524224838e-06,
      "loss": 0.2453,
      "num_tokens": 165828836.0,
      "step": 1265
    },
    {
      "epoch": 0.7409027929701743,
      "grad_norm": 0.6875863992968856,
      "learning_rate": 1.3630765810316799e-06,
      "loss": 0.2492,
      "num_tokens": 166480185.0,
      "step": 1270
    },
    {
      "epoch": 0.7438197330999782,
      "grad_norm": 0.7399219862362597,
      "learning_rate": 1.3451009291710542e-06,
      "loss": 0.2509,
      "num_tokens": 167127201.0,
      "step": 1275
    },
    {
      "epoch": 0.746736673229782,
      "grad_norm": 0.7264134555452614,
      "learning_rate": 1.3272710621543892e-06,
      "loss": 0.2488,
      "num_tokens": 167776261.0,
      "step": 1280
    },
    {
      "epoch": 0.7496536133595858,
      "grad_norm": 0.7732152888510079,
      "learning_rate": 1.309588830167536e-06,
      "loss": 0.2449,
      "num_tokens": 168426328.0,
      "step": 1285
    },
    {
      "epoch": 0.7525705534893896,
      "grad_norm": 0.7073729535659216,
      "learning_rate": 1.2920560680764165e-06,
      "loss": 0.2421,
      "num_tokens": 169092019.0,
      "step": 1290
    },
    {
      "epoch": 0.7554874936191934,
      "grad_norm": 0.7331560167444491,
      "learning_rate": 1.2746745952366275e-06,
      "loss": 0.2487,
      "num_tokens": 169741583.0,
      "step": 1295
    },
    {
      "epoch": 0.7584044337489974,
      "grad_norm": 0.6884922406400141,
      "learning_rate": 1.2574462153046441e-06,
      "loss": 0.2412,
      "num_tokens": 170389299.0,
      "step": 1300
    },
    {
      "epoch": 0.7613213738788012,
      "grad_norm": 0.781793314708951,
      "learning_rate": 1.2403727160506559e-06,
      "loss": 0.2444,
      "num_tokens": 171042211.0,
      "step": 1305
    },
    {
      "epoch": 0.764238314008605,
      "grad_norm": 0.7250494897320526,
      "learning_rate": 1.223455869173056e-06,
      "loss": 0.2438,
      "num_tokens": 171701780.0,
      "step": 1310
    },
    {
      "epoch": 0.7671552541384088,
      "grad_norm": 0.8740645184327206,
      "learning_rate": 1.2066974301145894e-06,
      "loss": 0.2456,
      "num_tokens": 172376337.0,
      "step": 1315
    },
    {
      "epoch": 0.7700721942682126,
      "grad_norm": 0.7443315304727108,
      "learning_rate": 1.1900991378801964e-06,
      "loss": 0.2439,
      "num_tokens": 173041371.0,
      "step": 1320
    },
    {
      "epoch": 0.7729891343980165,
      "grad_norm": 0.7116423082671238,
      "learning_rate": 1.1736627148565535e-06,
      "loss": 0.2531,
      "num_tokens": 173682308.0,
      "step": 1325
    },
    {
      "epoch": 0.7759060745278203,
      "grad_norm": 0.7077335089669503,
      "learning_rate": 1.157389866633348e-06,
      "loss": 0.246,
      "num_tokens": 174330336.0,
      "step": 1330
    },
    {
      "epoch": 0.7788230146576242,
      "grad_norm": 0.7308022672326984,
      "learning_rate": 1.1412822818262878e-06,
      "loss": 0.2524,
      "num_tokens": 174969619.0,
      "step": 1335
    },
    {
      "epoch": 0.781739954787428,
      "grad_norm": 0.6995063282185964,
      "learning_rate": 1.125341631901876e-06,
      "loss": 0.2406,
      "num_tokens": 175634291.0,
      "step": 1340
    },
    {
      "epoch": 0.7846568949172318,
      "grad_norm": 0.7514971805090548,
      "learning_rate": 1.1095695710039664e-06,
      "loss": 0.2403,
      "num_tokens": 176302019.0,
      "step": 1345
    },
    {
      "epoch": 0.7875738350470357,
      "grad_norm": 0.6772018040657682,
      "learning_rate": 1.0939677357821108e-06,
      "loss": 0.2445,
      "num_tokens": 176976275.0,
      "step": 1350
    },
    {
      "epoch": 0.7904907751768395,
      "grad_norm": 0.6817675548735839,
      "learning_rate": 1.0785377452217311e-06,
      "loss": 0.2503,
      "num_tokens": 177628328.0,
      "step": 1355
    },
    {
      "epoch": 0.7934077153066433,
      "grad_norm": 0.7346247121033264,
      "learning_rate": 1.0632812004761151e-06,
      "loss": 0.243,
      "num_tokens": 178288873.0,
      "step": 1360
    },
    {
      "epoch": 0.7963246554364471,
      "grad_norm": 0.7311842645346187,
      "learning_rate": 1.0481996847002676e-06,
      "loss": 0.2504,
      "num_tokens": 178942090.0,
      "step": 1365
    },
    {
      "epoch": 0.799241595566251,
      "grad_norm": 0.731394943481863,
      "learning_rate": 1.0332947628866273e-06,
      "loss": 0.2428,
      "num_tokens": 179600176.0,
      "step": 1370
    },
    {
      "epoch": 0.8021585356960549,
      "grad_norm": 0.7100948172511871,
      "learning_rate": 1.0185679817026715e-06,
      "loss": 0.2511,
      "num_tokens": 180244674.0,
      "step": 1375
    },
    {
      "epoch": 0.8050754758258587,
      "grad_norm": 0.708242217717899,
      "learning_rate": 1.0040208693304183e-06,
      "loss": 0.2456,
      "num_tokens": 180901761.0,
      "step": 1380
    },
    {
      "epoch": 0.8079924159556625,
      "grad_norm": 0.6803617534010951,
      "learning_rate": 9.89654935307848e-07,
      "loss": 0.2393,
      "num_tokens": 181564639.0,
      "step": 1385
    },
    {
      "epoch": 0.8109093560854663,
      "grad_norm": 0.6958034065313294,
      "learning_rate": 9.754716703722635e-07,
      "loss": 0.2347,
      "num_tokens": 182224730.0,
      "step": 1390
    },
    {
      "epoch": 0.8138262962152701,
      "grad_norm": 0.7165262095609373,
      "learning_rate": 9.614725463055931e-07,
      "loss": 0.2461,
      "num_tokens": 182883552.0,
      "step": 1395
    },
    {
      "epoch": 0.8167432363450741,
      "grad_norm": 0.7138480327038658,
      "learning_rate": 9.476590157816701e-07,
      "loss": 0.2411,
      "num_tokens": 183537453.0,
      "step": 1400
    },
    {
      "epoch": 0.8196601764748779,
      "grad_norm": 0.7133912213063753,
      "learning_rate": 9.340325122154878e-07,
      "loss": 0.2462,
      "num_tokens": 184199270.0,
      "step": 1405
    },
    {
      "epoch": 0.8225771166046817,
      "grad_norm": 0.6947806259213369,
      "learning_rate": 9.205944496144556e-07,
      "loss": 0.2486,
      "num_tokens": 184853282.0,
      "step": 1410
    },
    {
      "epoch": 0.8254940567344855,
      "grad_norm": 0.7272223644551763,
      "learning_rate": 9.073462224316707e-07,
      "loss": 0.2447,
      "num_tokens": 185514281.0,
      "step": 1415
    },
    {
      "epoch": 0.8284109968642893,
      "grad_norm": 0.7072312165887826,
      "learning_rate": 8.942892054212143e-07,
      "loss": 0.2397,
      "num_tokens": 186169820.0,
      "step": 1420
    },
    {
      "epoch": 0.8313279369940932,
      "grad_norm": 0.6903355667912614,
      "learning_rate": 8.814247534954983e-07,
      "loss": 0.2497,
      "num_tokens": 186817556.0,
      "step": 1425
    },
    {
      "epoch": 0.834244877123897,
      "grad_norm": 0.706857521915122,
      "learning_rate": 8.687542015846639e-07,
      "loss": 0.2438,
      "num_tokens": 187469631.0,
      "step": 1430
    },
    {
      "epoch": 0.8371618172537009,
      "grad_norm": 0.6988018386149026,
      "learning_rate": 8.562788644980624e-07,
      "loss": 0.2455,
      "num_tokens": 188113842.0,
      "step": 1435
    },
    {
      "epoch": 0.8400787573835047,
      "grad_norm": 0.673516286108323,
      "learning_rate": 8.440000367878115e-07,
      "loss": 0.2396,
      "num_tokens": 188782137.0,
      "step": 1440
    },
    {
      "epoch": 0.8429956975133085,
      "grad_norm": 0.699489145828766,
      "learning_rate": 8.319189926144688e-07,
      "loss": 0.2511,
      "num_tokens": 189419614.0,
      "step": 1445
    },
    {
      "epoch": 0.8459126376431124,
      "grad_norm": 0.7084426008074721,
      "learning_rate": 8.200369856148089e-07,
      "loss": 0.2457,
      "num_tokens": 190076342.0,
      "step": 1450
    },
    {
      "epoch": 0.8488295777729162,
      "grad_norm": 0.6739486565962028,
      "learning_rate": 8.083552487717358e-07,
      "loss": 0.2457,
      "num_tokens": 190726569.0,
      "step": 1455
    },
    {
      "epoch": 0.85174651790272,
      "grad_norm": 0.7095407182813062,
      "learning_rate": 7.968749942863385e-07,
      "loss": 0.2435,
      "num_tokens": 191375804.0,
      "step": 1460
    },
    {
      "epoch": 0.8546634580325239,
      "grad_norm": 0.6710001579802304,
      "learning_rate": 7.855974134520999e-07,
      "loss": 0.2405,
      "num_tokens": 192039149.0,
      "step": 1465
    },
    {
      "epoch": 0.8575803981623277,
      "grad_norm": 0.7079688435936735,
      "learning_rate": 7.745236765312819e-07,
      "loss": 0.2411,
      "num_tokens": 192695832.0,
      "step": 1470
    },
    {
      "epoch": 0.8604973382921316,
      "grad_norm": 0.6854485439890876,
      "learning_rate": 7.636549326334825e-07,
      "loss": 0.2417,
      "num_tokens": 193367048.0,
      "step": 1475
    },
    {
      "epoch": 0.8634142784219354,
      "grad_norm": 0.7184112786493275,
      "learning_rate": 7.529923095963999e-07,
      "loss": 0.2441,
      "num_tokens": 194017542.0,
      "step": 1480
    },
    {
      "epoch": 0.8663312185517392,
      "grad_norm": 0.6651542917903988,
      "learning_rate": 7.425369138687957e-07,
      "loss": 0.2422,
      "num_tokens": 194667338.0,
      "step": 1485
    },
    {
      "epoch": 0.869248158681543,
      "grad_norm": 0.6648797188770686,
      "learning_rate": 7.322898303956773e-07,
      "loss": 0.2416,
      "num_tokens": 195324822.0,
      "step": 1490
    },
    {
      "epoch": 0.8721650988113469,
      "grad_norm": 0.6743904697311516,
      "learning_rate": 7.222521225057187e-07,
      "loss": 0.2481,
      "num_tokens": 195972844.0,
      "step": 1495
    },
    {
      "epoch": 0.8750820389411508,
      "grad_norm": 0.6862148679988234,
      "learning_rate": 7.124248318009164e-07,
      "loss": 0.2426,
      "step": 1500
    },
    {
      "epoch": 0.8750820389411508,
      "eval_loss": 0.2491583675146103,
      "eval_num_tokens": 196624476.0,
      "eval_runtime": 607.7377,
      "eval_samples_per_second": 9.501,
      "eval_steps_per_second": 1.188,
      "step": 1500
    },
    {
      "epoch": 0.8779989790709546,
      "grad_norm": 0.7150785623554777,
      "learning_rate": 7.028089780485081e-07,
      "loss": 0.2437,
      "num_tokens": 197287303.0,
      "step": 1505
    },
    {
      "epoch": 0.8809159192007584,
      "grad_norm": 0.6922051228380665,
      "learning_rate": 6.934055590751461e-07,
      "loss": 0.237,
      "num_tokens": 197950146.0,
      "step": 1510
    },
    {
      "epoch": 0.8838328593305622,
      "grad_norm": 0.6814419896511972,
      "learning_rate": 6.842155506633598e-07,
      "loss": 0.2429,
      "num_tokens": 198606592.0,
      "step": 1515
    },
    {
      "epoch": 0.886749799460366,
      "grad_norm": 0.7528574216832195,
      "learning_rate": 6.752399064502959e-07,
      "loss": 0.2374,
      "num_tokens": 199260466.0,
      "step": 1520
    },
    {
      "epoch": 0.88966673959017,
      "grad_norm": 0.743850816504643,
      "learning_rate": 6.664795578287632e-07,
      "loss": 0.2498,
      "num_tokens": 199915169.0,
      "step": 1525
    },
    {
      "epoch": 0.8925836797199738,
      "grad_norm": 0.6659899742478199,
      "learning_rate": 6.579354138505817e-07,
      "loss": 0.239,
      "num_tokens": 200577384.0,
      "step": 1530
    },
    {
      "epoch": 0.8955006198497776,
      "grad_norm": 0.6989537096567529,
      "learning_rate": 6.496083611322503e-07,
      "loss": 0.2437,
      "num_tokens": 201229904.0,
      "step": 1535
    },
    {
      "epoch": 0.8984175599795814,
      "grad_norm": 0.6870906534872678,
      "learning_rate": 6.414992637629462e-07,
      "loss": 0.2459,
      "num_tokens": 201874398.0,
      "step": 1540
    },
    {
      "epoch": 0.9013345001093852,
      "grad_norm": 0.6851022314706521,
      "learning_rate": 6.336089632148566e-07,
      "loss": 0.2437,
      "num_tokens": 202514986.0,
      "step": 1545
    },
    {
      "epoch": 0.9042514402391891,
      "grad_norm": 0.708606924412846,
      "learning_rate": 6.259382782558623e-07,
      "loss": 0.2513,
      "num_tokens": 203157709.0,
      "step": 1550
    },
    {
      "epoch": 0.9071683803689929,
      "grad_norm": 0.6901795784562913,
      "learning_rate": 6.184880048645731e-07,
      "loss": 0.241,
      "num_tokens": 203820585.0,
      "step": 1555
    },
    {
      "epoch": 0.9100853204987968,
      "grad_norm": 0.6854819640738248,
      "learning_rate": 6.112589161477317e-07,
      "loss": 0.2474,
      "num_tokens": 204463922.0,
      "step": 1560
    },
    {
      "epoch": 0.9130022606286006,
      "grad_norm": 0.6490034167792925,
      "learning_rate": 6.042517622599872e-07,
      "loss": 0.2423,
      "num_tokens": 205116527.0,
      "step": 1565
    },
    {
      "epoch": 0.9159192007584044,
      "grad_norm": 0.6700066915913209,
      "learning_rate": 5.97467270326055e-07,
      "loss": 0.2399,
      "num_tokens": 205783661.0,
      "step": 1570
    },
    {
      "epoch": 0.9188361408882083,
      "grad_norm": 0.6666592595870292,
      "learning_rate": 5.909061443652619e-07,
      "loss": 0.2483,
      "num_tokens": 206431518.0,
      "step": 1575
    },
    {
      "epoch": 0.9217530810180121,
      "grad_norm": 0.6880885792235153,
      "learning_rate": 5.845690652184906e-07,
      "loss": 0.244,
      "num_tokens": 207082078.0,
      "step": 1580
    },
    {
      "epoch": 0.9246700211478159,
      "grad_norm": 0.7218901751907087,
      "learning_rate": 5.784566904775314e-07,
      "loss": 0.2439,
      "num_tokens": 207733633.0,
      "step": 1585
    },
    {
      "epoch": 0.9275869612776197,
      "grad_norm": 0.7224619275428422,
      "learning_rate": 5.725696544168431e-07,
      "loss": 0.2442,
      "num_tokens": 208384194.0,
      "step": 1590
    },
    {
      "epoch": 0.9305039014074236,
      "grad_norm": 0.685956055593419,
      "learning_rate": 5.669085679277353e-07,
      "loss": 0.2467,
      "num_tokens": 209017882.0,
      "step": 1595
    },
    {
      "epoch": 0.9334208415372275,
      "grad_norm": 0.6595594654962236,
      "learning_rate": 5.614740184549774e-07,
      "loss": 0.2385,
      "num_tokens": 209671100.0,
      "step": 1600
    },
    {
      "epoch": 0.9363377816670313,
      "grad_norm": 0.6635370096109607,
      "learning_rate": 5.562665699358395e-07,
      "loss": 0.2403,
      "num_tokens": 210332481.0,
      "step": 1605
    },
    {
      "epoch": 0.9392547217968351,
      "grad_norm": 0.6792107629780629,
      "learning_rate": 5.512867627415738e-07,
      "loss": 0.2485,
      "num_tokens": 210989185.0,
      "step": 1610
    },
    {
      "epoch": 0.9421716619266389,
      "grad_norm": 0.688506342901487,
      "learning_rate": 5.465351136213403e-07,
      "loss": 0.2404,
      "num_tokens": 211654349.0,
      "step": 1615
    },
    {
      "epoch": 0.9450886020564427,
      "grad_norm": 0.6957995520415324,
      "learning_rate": 5.420121156485843e-07,
      "loss": 0.2449,
      "num_tokens": 212306065.0,
      "step": 1620
    },
    {
      "epoch": 0.9480055421862467,
      "grad_norm": 0.6874524399723582,
      "learning_rate": 5.377182381698713e-07,
      "loss": 0.2461,
      "num_tokens": 212961559.0,
      "step": 1625
    },
    {
      "epoch": 0.9509224823160505,
      "grad_norm": 0.6531866036674674,
      "learning_rate": 5.336539267561834e-07,
      "loss": 0.2419,
      "num_tokens": 213623892.0,
      "step": 1630
    },
    {
      "epoch": 0.9538394224458543,
      "grad_norm": 0.6861036738287705,
      "learning_rate": 5.298196031566817e-07,
      "loss": 0.2397,
      "num_tokens": 214276109.0,
      "step": 1635
    },
    {
      "epoch": 0.9567563625756581,
      "grad_norm": 0.7385399272118951,
      "learning_rate": 5.262156652549434e-07,
      "loss": 0.2404,
      "num_tokens": 214935598.0,
      "step": 1640
    },
    {
      "epoch": 0.9596733027054619,
      "grad_norm": 0.7641216689081703,
      "learning_rate": 5.228424870276732e-07,
      "loss": 0.2413,
      "num_tokens": 215592154.0,
      "step": 1645
    },
    {
      "epoch": 0.9625902428352658,
      "grad_norm": 0.6626697793424543,
      "learning_rate": 5.197004185058957e-07,
      "loss": 0.2402,
      "num_tokens": 216248152.0,
      "step": 1650
    },
    {
      "epoch": 0.9655071829650697,
      "grad_norm": 0.6668250234146728,
      "learning_rate": 5.167897857386338e-07,
      "loss": 0.24,
      "num_tokens": 216908673.0,
      "step": 1655
    },
    {
      "epoch": 0.9684241230948735,
      "grad_norm": 0.6830282199194727,
      "learning_rate": 5.141108907590743e-07,
      "loss": 0.2447,
      "num_tokens": 217569951.0,
      "step": 1660
    },
    {
      "epoch": 0.9713410632246773,
      "grad_norm": 0.694769552967056,
      "learning_rate": 5.116640115532271e-07,
      "loss": 0.2463,
      "num_tokens": 218214661.0,
      "step": 1665
    },
    {
      "epoch": 0.9742580033544811,
      "grad_norm": 0.6576753862947748,
      "learning_rate": 5.09449402031078e-07,
      "loss": 0.2377,
      "num_tokens": 218875984.0,
      "step": 1670
    },
    {
      "epoch": 0.977174943484285,
      "grad_norm": 0.6885196158109818,
      "learning_rate": 5.074672920002409e-07,
      "loss": 0.2388,
      "num_tokens": 219529057.0,
      "step": 1675
    },
    {
      "epoch": 0.9800918836140888,
      "grad_norm": 0.6862150588167344,
      "learning_rate": 5.057178871421117e-07,
      "loss": 0.238,
      "num_tokens": 220197787.0,
      "step": 1680
    },
    {
      "epoch": 0.9830088237438926,
      "grad_norm": 0.6778601221453343,
      "learning_rate": 5.04201368990524e-07,
      "loss": 0.2513,
      "num_tokens": 220841743.0,
      "step": 1685
    },
    {
      "epoch": 0.9859257638736965,
      "grad_norm": 0.6899658979865586,
      "learning_rate": 5.029178949129118e-07,
      "loss": 0.243,
      "num_tokens": 221473805.0,
      "step": 1690
    },
    {
      "epoch": 0.9888427040035004,
      "grad_norm": 0.6528281983164207,
      "learning_rate": 5.018675980939805e-07,
      "loss": 0.2444,
      "num_tokens": 222120807.0,
      "step": 1695
    },
    {
      "epoch": 0.9917596441333042,
      "grad_norm": 0.6674621015647066,
      "learning_rate": 5.010505875218846e-07,
      "loss": 0.2386,
      "num_tokens": 222794425.0,
      "step": 1700
    },
    {
      "epoch": 0.994676584263108,
      "grad_norm": 0.7086141702806269,
      "learning_rate": 5.004669479769203e-07,
      "loss": 0.2405,
      "num_tokens": 223443814.0,
      "step": 1705
    },
    {
      "epoch": 0.9975935243929118,
      "grad_norm": 0.6795651043519514,
      "learning_rate": 5.001167400227263e-07,
      "loss": 0.2383,
      "num_tokens": 224098786.0,
      "step": 1710
    },
    {
      "epoch": 0.9999270764967549,
      "num_tokens": 224623551.0,
      "step": 1714,
      "total_flos": 9.914064352850215e+18,
      "train_loss": 0.2932530035076787,
      "train_runtime": 32864.6589,
      "train_samples_per_second": 3.338,
      "train_steps_per_second": 0.052
    }
  ],
  "logging_steps": 5,
  "max_steps": 1714,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.914064352850215e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}