{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1172,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008532423208191126,
      "grad_norm": 38.25,
      "learning_rate": 0.0,
      "loss": 1.6671178340911865,
      "mean_token_accuracy": 0.7226356267929077,
      "num_tokens": 1237132.0,
      "step": 1
    },
    {
      "epoch": 0.0017064846416382253,
      "grad_norm": 38.5,
      "learning_rate": 1.111111111111111e-06,
      "loss": 1.6624282598495483,
      "mean_token_accuracy": 0.7234913110733032,
      "num_tokens": 2454264.0,
      "step": 2
    },
    {
      "epoch": 0.002559726962457338,
      "grad_norm": 37.5,
      "learning_rate": 2.222222222222222e-06,
      "loss": 1.6899430751800537,
      "mean_token_accuracy": 0.7059391140937805,
      "num_tokens": 3745950.0,
      "step": 3
    },
    {
      "epoch": 0.0034129692832764505,
      "grad_norm": 36.25,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.6722791194915771,
      "mean_token_accuracy": 0.701712965965271,
      "num_tokens": 4926966.0,
      "step": 4
    },
    {
      "epoch": 0.004266211604095563,
      "grad_norm": 30.5,
      "learning_rate": 4.444444444444444e-06,
      "loss": 1.501098871231079,
      "mean_token_accuracy": 0.7184635400772095,
      "num_tokens": 6198918.0,
      "step": 5
    },
    {
      "epoch": 0.005119453924914676,
      "grad_norm": 26.625,
      "learning_rate": 5.555555555555557e-06,
      "loss": 1.4207741022109985,
      "mean_token_accuracy": 0.7093047499656677,
      "num_tokens": 7446013.0,
      "step": 6
    },
    {
      "epoch": 0.005972696245733789,
      "grad_norm": 19.25,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.1401493549346924,
      "mean_token_accuracy": 0.7296085357666016,
      "num_tokens": 8688952.0,
      "step": 7
    },
    {
      "epoch": 0.006825938566552901,
      "grad_norm": 13.625,
      "learning_rate": 7.77777777777778e-06,
      "loss": 1.027529001235962,
      "mean_token_accuracy": 0.7380459308624268,
      "num_tokens": 9940991.0,
      "step": 8
    },
    {
      "epoch": 0.007679180887372013,
      "grad_norm": 8.375,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.0067119598388672,
      "mean_token_accuracy": 0.7229213714599609,
      "num_tokens": 11182075.0,
      "step": 9
    },
    {
      "epoch": 0.008532423208191127,
      "grad_norm": 4.28125,
      "learning_rate": 1e-05,
      "loss": 0.8530652523040771,
      "mean_token_accuracy": 0.7559525966644287,
      "num_tokens": 12530861.0,
      "step": 10
    },
    {
      "epoch": 0.00938566552901024,
      "grad_norm": 4.125,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 0.8722737431526184,
      "mean_token_accuracy": 0.7519086599349976,
      "num_tokens": 13795400.0,
      "step": 11
    },
    {
      "epoch": 0.010238907849829351,
      "grad_norm": 3.5625,
      "learning_rate": 1.2222222222222224e-05,
      "loss": 0.7763490676879883,
      "mean_token_accuracy": 0.7738497257232666,
      "num_tokens": 14978469.0,
      "step": 12
    },
    {
      "epoch": 0.011092150170648464,
      "grad_norm": 3.15625,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.8119764924049377,
      "mean_token_accuracy": 0.7609139680862427,
      "num_tokens": 16180856.0,
      "step": 13
    },
    {
      "epoch": 0.011945392491467578,
      "grad_norm": 2.671875,
      "learning_rate": 1.4444444444444446e-05,
      "loss": 0.6962097883224487,
      "mean_token_accuracy": 0.7947577238082886,
      "num_tokens": 17483534.0,
      "step": 14
    },
    {
      "epoch": 0.012798634812286689,
      "grad_norm": 2.625,
      "learning_rate": 1.555555555555556e-05,
      "loss": 0.7247188091278076,
      "mean_token_accuracy": 0.7882294058799744,
      "num_tokens": 18803815.0,
      "step": 15
    },
    {
      "epoch": 0.013651877133105802,
      "grad_norm": 2.734375,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7438160181045532,
      "mean_token_accuracy": 0.7734143137931824,
      "num_tokens": 20123611.0,
      "step": 16
    },
    {
      "epoch": 0.014505119453924915,
      "grad_norm": 2.296875,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 0.7146241664886475,
      "mean_token_accuracy": 0.7854167222976685,
      "num_tokens": 21369531.0,
      "step": 17
    },
    {
      "epoch": 0.015358361774744027,
      "grad_norm": 2.40625,
      "learning_rate": 1.888888888888889e-05,
      "loss": 0.6799183487892151,
      "mean_token_accuracy": 0.7894506454467773,
      "num_tokens": 22527959.0,
      "step": 18
    },
    {
      "epoch": 0.016211604095563138,
      "grad_norm": 1.9765625,
      "learning_rate": 2e-05,
      "loss": 0.6314034461975098,
      "mean_token_accuracy": 0.8029162883758545,
      "num_tokens": 23774519.0,
      "step": 19
    },
    {
      "epoch": 0.017064846416382253,
      "grad_norm": 2.1875,
      "learning_rate": 2.1111111111111114e-05,
      "loss": 0.6711724996566772,
      "mean_token_accuracy": 0.7889853715896606,
      "num_tokens": 25005766.0,
      "step": 20
    },
    {
      "epoch": 0.017918088737201365,
      "grad_norm": 1.84375,
      "learning_rate": 2.2222222222222227e-05,
      "loss": 0.6222089529037476,
      "mean_token_accuracy": 0.8097364902496338,
      "num_tokens": 26327643.0,
      "step": 21
    },
    {
      "epoch": 0.01877133105802048,
      "grad_norm": 1.703125,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.5920464992523193,
      "mean_token_accuracy": 0.8147287964820862,
      "num_tokens": 27504227.0,
      "step": 22
    },
    {
      "epoch": 0.01962457337883959,
      "grad_norm": 1.8515625,
      "learning_rate": 2.444444444444445e-05,
      "loss": 0.5996453762054443,
      "mean_token_accuracy": 0.8111099004745483,
      "num_tokens": 28760113.0,
      "step": 23
    },
    {
      "epoch": 0.020477815699658702,
      "grad_norm": 1.734375,
      "learning_rate": 2.5555555555555554e-05,
      "loss": 0.6003448963165283,
      "mean_token_accuracy": 0.8131307363510132,
      "num_tokens": 29991718.0,
      "step": 24
    },
    {
      "epoch": 0.021331058020477817,
      "grad_norm": 1.859375,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.6071052551269531,
      "mean_token_accuracy": 0.8063894510269165,
      "num_tokens": 31297668.0,
      "step": 25
    },
    {
      "epoch": 0.02218430034129693,
      "grad_norm": 1.7890625,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6121305823326111,
      "mean_token_accuracy": 0.8066533207893372,
      "num_tokens": 32512717.0,
      "step": 26
    },
    {
      "epoch": 0.02303754266211604,
      "grad_norm": 1.578125,
      "learning_rate": 2.888888888888889e-05,
      "loss": 0.5630435347557068,
      "mean_token_accuracy": 0.8154351115226746,
      "num_tokens": 33779305.0,
      "step": 27
    },
    {
      "epoch": 0.023890784982935155,
      "grad_norm": 1.6015625,
      "learning_rate": 3.0000000000000004e-05,
      "loss": 0.5405244827270508,
      "mean_token_accuracy": 0.8229197263717651,
      "num_tokens": 34981151.0,
      "step": 28
    },
    {
      "epoch": 0.024744027303754267,
      "grad_norm": 2.328125,
      "learning_rate": 3.111111111111112e-05,
      "loss": 0.5952686667442322,
      "mean_token_accuracy": 0.8045962452888489,
      "num_tokens": 36150225.0,
      "step": 29
    },
    {
      "epoch": 0.025597269624573378,
      "grad_norm": 1.546875,
      "learning_rate": 3.222222222222223e-05,
      "loss": 0.5807881355285645,
      "mean_token_accuracy": 0.8127832412719727,
      "num_tokens": 37367923.0,
      "step": 30
    },
    {
      "epoch": 0.026450511945392493,
      "grad_norm": 1.703125,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.5666648745536804,
      "mean_token_accuracy": 0.8191201686859131,
      "num_tokens": 38577312.0,
      "step": 31
    },
    {
      "epoch": 0.027303754266211604,
      "grad_norm": 1.65625,
      "learning_rate": 3.444444444444445e-05,
      "loss": 0.570904552936554,
      "mean_token_accuracy": 0.8156166076660156,
      "num_tokens": 39894871.0,
      "step": 32
    },
    {
      "epoch": 0.028156996587030716,
      "grad_norm": 2.5,
      "learning_rate": 3.555555555555555e-05,
      "loss": 0.5892091989517212,
      "mean_token_accuracy": 0.8130444884300232,
      "num_tokens": 41154714.0,
      "step": 33
    },
    {
      "epoch": 0.02901023890784983,
      "grad_norm": 1.6953125,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.5757794380187988,
      "mean_token_accuracy": 0.814048171043396,
      "num_tokens": 42436712.0,
      "step": 34
    },
    {
      "epoch": 0.029863481228668942,
      "grad_norm": 1.6015625,
      "learning_rate": 3.777777777777778e-05,
      "loss": 0.5425353050231934,
      "mean_token_accuracy": 0.8228821754455566,
      "num_tokens": 43744206.0,
      "step": 35
    },
    {
      "epoch": 0.030716723549488054,
      "grad_norm": 1.53125,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.5665711760520935,
      "mean_token_accuracy": 0.8175134658813477,
      "num_tokens": 45040800.0,
      "step": 36
    },
    {
      "epoch": 0.031569965870307165,
      "grad_norm": 1.6015625,
      "learning_rate": 4e-05,
      "loss": 0.5284080505371094,
      "mean_token_accuracy": 0.8279036283493042,
      "num_tokens": 46299396.0,
      "step": 37
    },
    {
      "epoch": 0.032423208191126277,
      "grad_norm": 1.6796875,
      "learning_rate": 3.9999923520893434e-05,
      "loss": 0.5204638838768005,
      "mean_token_accuracy": 0.8313729763031006,
      "num_tokens": 47578969.0,
      "step": 38
    },
    {
      "epoch": 0.033276450511945395,
      "grad_norm": 1.7734375,
      "learning_rate": 3.999969408415864e-05,
      "loss": 0.5844517350196838,
      "mean_token_accuracy": 0.8101286292076111,
      "num_tokens": 48806745.0,
      "step": 39
    },
    {
      "epoch": 0.034129692832764506,
      "grad_norm": 1.4609375,
      "learning_rate": 3.999931169155032e-05,
      "loss": 0.5340092778205872,
      "mean_token_accuracy": 0.8268023729324341,
      "num_tokens": 50116992.0,
      "step": 40
    },
    {
      "epoch": 0.03498293515358362,
      "grad_norm": 1.5234375,
      "learning_rate": 3.999877634599299e-05,
      "loss": 0.5236454010009766,
      "mean_token_accuracy": 0.8232540488243103,
      "num_tokens": 51355741.0,
      "step": 41
    },
    {
      "epoch": 0.03583617747440273,
      "grad_norm": 1.4921875,
      "learning_rate": 3.9998088051580925e-05,
      "loss": 0.5255980491638184,
      "mean_token_accuracy": 0.8275700807571411,
      "num_tokens": 52542405.0,
      "step": 42
    },
    {
      "epoch": 0.03668941979522184,
      "grad_norm": 1.671875,
      "learning_rate": 3.999724681357813e-05,
      "loss": 0.546703577041626,
      "mean_token_accuracy": 0.8214466571807861,
      "num_tokens": 53762322.0,
      "step": 43
    },
    {
      "epoch": 0.03754266211604096,
      "grad_norm": 1.9140625,
      "learning_rate": 3.999625263841832e-05,
      "loss": 0.599216639995575,
      "mean_token_accuracy": 0.8036430478096008,
      "num_tokens": 55051866.0,
      "step": 44
    },
    {
      "epoch": 0.03839590443686007,
      "grad_norm": 1.5078125,
      "learning_rate": 3.999510553370487e-05,
      "loss": 0.5238852500915527,
      "mean_token_accuracy": 0.8318121433258057,
      "num_tokens": 56224455.0,
      "step": 45
    },
    {
      "epoch": 0.03924914675767918,
      "grad_norm": 1.3046875,
      "learning_rate": 3.999380550821071e-05,
      "loss": 0.5243685245513916,
      "mean_token_accuracy": 0.8268356323242188,
      "num_tokens": 57478088.0,
      "step": 46
    },
    {
      "epoch": 0.04010238907849829,
      "grad_norm": 1.4609375,
      "learning_rate": 3.999235257187835e-05,
      "loss": 0.5227737426757812,
      "mean_token_accuracy": 0.8307685256004333,
      "num_tokens": 58673473.0,
      "step": 47
    },
    {
      "epoch": 0.040955631399317405,
      "grad_norm": 1.3515625,
      "learning_rate": 3.999074673581969e-05,
      "loss": 0.5402700901031494,
      "mean_token_accuracy": 0.8232250809669495,
      "num_tokens": 59991318.0,
      "step": 48
    },
    {
      "epoch": 0.041808873720136516,
      "grad_norm": 1.21875,
      "learning_rate": 3.9988988012316035e-05,
      "loss": 0.47228097915649414,
      "mean_token_accuracy": 0.8405997157096863,
      "num_tokens": 61222524.0,
      "step": 49
    },
    {
      "epoch": 0.042662116040955635,
      "grad_norm": 1.4140625,
      "learning_rate": 3.998707641481795e-05,
      "loss": 0.5430753231048584,
      "mean_token_accuracy": 0.8243215084075928,
      "num_tokens": 62460491.0,
      "step": 50
    },
    {
      "epoch": 0.043515358361774746,
      "grad_norm": 1.328125,
      "learning_rate": 3.998501195794515e-05,
      "loss": 0.4768257141113281,
      "mean_token_accuracy": 0.8379573225975037,
      "num_tokens": 63675510.0,
      "step": 51
    },
    {
      "epoch": 0.04436860068259386,
      "grad_norm": 1.296875,
      "learning_rate": 3.998279465748642e-05,
      "loss": 0.5217003226280212,
      "mean_token_accuracy": 0.829858660697937,
      "num_tokens": 64845056.0,
      "step": 52
    },
    {
      "epoch": 0.04522184300341297,
      "grad_norm": 1.359375,
      "learning_rate": 3.9980424530399475e-05,
      "loss": 0.5585646629333496,
      "mean_token_accuracy": 0.8192288875579834,
      "num_tokens": 66154685.0,
      "step": 53
    },
    {
      "epoch": 0.04607508532423208,
      "grad_norm": 1.296875,
      "learning_rate": 3.997790159481084e-05,
      "loss": 0.4936741590499878,
      "mean_token_accuracy": 0.8303905725479126,
      "num_tokens": 67428502.0,
      "step": 54
    },
    {
      "epoch": 0.04692832764505119,
      "grad_norm": 1.8828125,
      "learning_rate": 3.9975225870015694e-05,
      "loss": 0.5263705253601074,
      "mean_token_accuracy": 0.8290708065032959,
      "num_tokens": 68745430.0,
      "step": 55
    },
    {
      "epoch": 0.04778156996587031,
      "grad_norm": 1.328125,
      "learning_rate": 3.997239737647775e-05,
      "loss": 0.4647553563117981,
      "mean_token_accuracy": 0.8405643105506897,
      "num_tokens": 69970914.0,
      "step": 56
    },
    {
      "epoch": 0.04863481228668942,
      "grad_norm": 1.2890625,
      "learning_rate": 3.9969416135829066e-05,
      "loss": 0.4889841377735138,
      "mean_token_accuracy": 0.839198112487793,
      "num_tokens": 71299554.0,
      "step": 57
    },
    {
      "epoch": 0.04948805460750853,
      "grad_norm": 1.40625,
      "learning_rate": 3.9966282170869905e-05,
      "loss": 0.5032781362533569,
      "mean_token_accuracy": 0.8334668874740601,
      "num_tokens": 72515024.0,
      "step": 58
    },
    {
      "epoch": 0.050341296928327645,
      "grad_norm": 1.4453125,
      "learning_rate": 3.996299550556856e-05,
      "loss": 0.5277822613716125,
      "mean_token_accuracy": 0.8230280876159668,
      "num_tokens": 73723696.0,
      "step": 59
    },
    {
      "epoch": 0.051194539249146756,
      "grad_norm": 1.2578125,
      "learning_rate": 3.9959556165061143e-05,
      "loss": 0.462985634803772,
      "mean_token_accuracy": 0.845061182975769,
      "num_tokens": 74993778.0,
      "step": 60
    },
    {
      "epoch": 0.05204778156996587,
      "grad_norm": 1.8046875,
      "learning_rate": 3.995596417565143e-05,
      "loss": 0.5597748756408691,
      "mean_token_accuracy": 0.8171534538269043,
      "num_tokens": 76255055.0,
      "step": 61
    },
    {
      "epoch": 0.052901023890784986,
      "grad_norm": 2.5625,
      "learning_rate": 3.995221956481063e-05,
      "loss": 0.4937603175640106,
      "mean_token_accuracy": 0.8328028917312622,
      "num_tokens": 77470293.0,
      "step": 62
    },
    {
      "epoch": 0.0537542662116041,
      "grad_norm": 1.3828125,
      "learning_rate": 3.9948322361177185e-05,
      "loss": 0.48047587275505066,
      "mean_token_accuracy": 0.833111584186554,
      "num_tokens": 78640060.0,
      "step": 63
    },
    {
      "epoch": 0.05460750853242321,
      "grad_norm": 1.71875,
      "learning_rate": 3.994427259455657e-05,
      "loss": 0.5005626082420349,
      "mean_token_accuracy": 0.8314197659492493,
      "num_tokens": 79869251.0,
      "step": 64
    },
    {
      "epoch": 0.05546075085324232,
      "grad_norm": 1.4453125,
      "learning_rate": 3.994007029592105e-05,
      "loss": 0.5642141103744507,
      "mean_token_accuracy": 0.8147813081741333,
      "num_tokens": 81136748.0,
      "step": 65
    },
    {
      "epoch": 0.05631399317406143,
      "grad_norm": 1.375,
      "learning_rate": 3.993571549740941e-05,
      "loss": 0.5149242281913757,
      "mean_token_accuracy": 0.8307733535766602,
      "num_tokens": 82323091.0,
      "step": 66
    },
    {
      "epoch": 0.05716723549488054,
      "grad_norm": 1.1640625,
      "learning_rate": 3.993120823232676e-05,
      "loss": 0.4595782160758972,
      "mean_token_accuracy": 0.8426697850227356,
      "num_tokens": 83597194.0,
      "step": 67
    },
    {
      "epoch": 0.05802047781569966,
      "grad_norm": 1.3359375,
      "learning_rate": 3.9926548535144285e-05,
      "loss": 0.5037087202072144,
      "mean_token_accuracy": 0.8330240249633789,
      "num_tokens": 84877078.0,
      "step": 68
    },
    {
      "epoch": 0.05887372013651877,
      "grad_norm": 1.28125,
      "learning_rate": 3.99217364414989e-05,
      "loss": 0.5255679488182068,
      "mean_token_accuracy": 0.8251068592071533,
      "num_tokens": 86129562.0,
      "step": 69
    },
    {
      "epoch": 0.059726962457337884,
      "grad_norm": 1.203125,
      "learning_rate": 3.9916771988193086e-05,
      "loss": 0.4897761940956116,
      "mean_token_accuracy": 0.8361822962760925,
      "num_tokens": 87366910.0,
      "step": 70
    },
    {
      "epoch": 0.060580204778156996,
      "grad_norm": 1.1484375,
      "learning_rate": 3.991165521319454e-05,
      "loss": 0.48412856459617615,
      "mean_token_accuracy": 0.8385250568389893,
      "num_tokens": 88679706.0,
      "step": 71
    },
    {
      "epoch": 0.06143344709897611,
      "grad_norm": 1.328125,
      "learning_rate": 3.990638615563588e-05,
      "loss": 0.5031937956809998,
      "mean_token_accuracy": 0.8315696716308594,
      "num_tokens": 89999187.0,
      "step": 72
    },
    {
      "epoch": 0.06228668941979522,
      "grad_norm": 1.390625,
      "learning_rate": 3.990096485581442e-05,
      "loss": 0.48631253838539124,
      "mean_token_accuracy": 0.8381010293960571,
      "num_tokens": 91288974.0,
      "step": 73
    },
    {
      "epoch": 0.06313993174061433,
      "grad_norm": 1.3671875,
      "learning_rate": 3.989539135519175e-05,
      "loss": 0.5479691028594971,
      "mean_token_accuracy": 0.8184207677841187,
      "num_tokens": 92577855.0,
      "step": 74
    },
    {
      "epoch": 0.06399317406143344,
      "grad_norm": 1.3203125,
      "learning_rate": 3.9889665696393514e-05,
      "loss": 0.5137754082679749,
      "mean_token_accuracy": 0.8310798406600952,
      "num_tokens": 93799232.0,
      "step": 75
    },
    {
      "epoch": 0.06484641638225255,
      "grad_norm": 1.28125,
      "learning_rate": 3.988378792320904e-05,
      "loss": 0.4834998548030853,
      "mean_token_accuracy": 0.8364094495773315,
      "num_tokens": 95008576.0,
      "step": 76
    },
    {
      "epoch": 0.06569965870307168,
      "grad_norm": 1.453125,
      "learning_rate": 3.9877758080591016e-05,
      "loss": 0.5211181044578552,
      "mean_token_accuracy": 0.8232471942901611,
      "num_tokens": 96274855.0,
      "step": 77
    },
    {
      "epoch": 0.06655290102389079,
      "grad_norm": 1.2578125,
      "learning_rate": 3.987157621465513e-05,
      "loss": 0.4809388816356659,
      "mean_token_accuracy": 0.8405702114105225,
      "num_tokens": 97521255.0,
      "step": 78
    },
    {
      "epoch": 0.0674061433447099,
      "grad_norm": 1.4375,
      "learning_rate": 3.9865242372679746e-05,
      "loss": 0.4696381092071533,
      "mean_token_accuracy": 0.8414092659950256,
      "num_tokens": 98665158.0,
      "step": 79
    },
    {
      "epoch": 0.06825938566552901,
      "grad_norm": 1.1875,
      "learning_rate": 3.985875660310552e-05,
      "loss": 0.4798639118671417,
      "mean_token_accuracy": 0.8337196111679077,
      "num_tokens": 99836874.0,
      "step": 80
    },
    {
      "epoch": 0.06911262798634812,
      "grad_norm": 1.28125,
      "learning_rate": 3.985211895553504e-05,
      "loss": 0.4820418357849121,
      "mean_token_accuracy": 0.837172269821167,
      "num_tokens": 101221945.0,
      "step": 81
    },
    {
      "epoch": 0.06996587030716724,
      "grad_norm": 1.3359375,
      "learning_rate": 3.984532948073244e-05,
      "loss": 0.528157651424408,
      "mean_token_accuracy": 0.8226147294044495,
      "num_tokens": 102500387.0,
      "step": 82
    },
    {
      "epoch": 0.07081911262798635,
      "grad_norm": 1.421875,
      "learning_rate": 3.983838823062301e-05,
      "loss": 0.5201035141944885,
      "mean_token_accuracy": 0.8270155787467957,
      "num_tokens": 103714842.0,
      "step": 83
    },
    {
      "epoch": 0.07167235494880546,
      "grad_norm": 1.3359375,
      "learning_rate": 3.983129525829282e-05,
      "loss": 0.508536696434021,
      "mean_token_accuracy": 0.8307348489761353,
      "num_tokens": 104975575.0,
      "step": 84
    },
    {
      "epoch": 0.07252559726962457,
      "grad_norm": 1.3671875,
      "learning_rate": 3.982405061798829e-05,
      "loss": 0.5323456525802612,
      "mean_token_accuracy": 0.819508969783783,
      "num_tokens": 106279121.0,
      "step": 85
    },
    {
      "epoch": 0.07337883959044368,
      "grad_norm": 1.2265625,
      "learning_rate": 3.981665436511578e-05,
      "loss": 0.47382795810699463,
      "mean_token_accuracy": 0.8388296961784363,
      "num_tokens": 107483844.0,
      "step": 86
    },
    {
      "epoch": 0.07423208191126279,
      "grad_norm": 1.2265625,
      "learning_rate": 3.980910655624117e-05,
      "loss": 0.458898663520813,
      "mean_token_accuracy": 0.8428360223770142,
      "num_tokens": 108804390.0,
      "step": 87
    },
    {
      "epoch": 0.07508532423208192,
      "grad_norm": 1.3046875,
      "learning_rate": 3.9801407249089426e-05,
      "loss": 0.4779137372970581,
      "mean_token_accuracy": 0.8370555639266968,
      "num_tokens": 110035704.0,
      "step": 88
    },
    {
      "epoch": 0.07593856655290103,
      "grad_norm": 1.3515625,
      "learning_rate": 3.979355650254416e-05,
      "loss": 0.5100655555725098,
      "mean_token_accuracy": 0.8307489156723022,
      "num_tokens": 111291332.0,
      "step": 89
    },
    {
      "epoch": 0.07679180887372014,
      "grad_norm": 1.28125,
      "learning_rate": 3.978555437664718e-05,
      "loss": 0.48221877217292786,
      "mean_token_accuracy": 0.8357584476470947,
      "num_tokens": 112619337.0,
      "step": 90
    },
    {
      "epoch": 0.07764505119453925,
      "grad_norm": 1.234375,
      "learning_rate": 3.977740093259804e-05,
      "loss": 0.4912947118282318,
      "mean_token_accuracy": 0.8347182273864746,
      "num_tokens": 113839541.0,
      "step": 91
    },
    {
      "epoch": 0.07849829351535836,
      "grad_norm": 1.234375,
      "learning_rate": 3.9769096232753535e-05,
      "loss": 0.5269333720207214,
      "mean_token_accuracy": 0.8210034370422363,
      "num_tokens": 115037265.0,
      "step": 92
    },
    {
      "epoch": 0.07935153583617748,
      "grad_norm": 1.3359375,
      "learning_rate": 3.976064034062728e-05,
      "loss": 0.5187820196151733,
      "mean_token_accuracy": 0.8273521065711975,
      "num_tokens": 116450852.0,
      "step": 93
    },
    {
      "epoch": 0.08020477815699659,
      "grad_norm": 1.15625,
      "learning_rate": 3.975203332088919e-05,
      "loss": 0.44256603717803955,
      "mean_token_accuracy": 0.8465872406959534,
      "num_tokens": 117713048.0,
      "step": 94
    },
    {
      "epoch": 0.0810580204778157,
      "grad_norm": 1.390625,
      "learning_rate": 3.974327523936496e-05,
      "loss": 0.46991395950317383,
      "mean_token_accuracy": 0.8397294878959656,
      "num_tokens": 118964212.0,
      "step": 95
    },
    {
      "epoch": 0.08191126279863481,
      "grad_norm": 1.4140625,
      "learning_rate": 3.9734366163035625e-05,
      "loss": 0.5207662582397461,
      "mean_token_accuracy": 0.8261317014694214,
      "num_tokens": 120204493.0,
      "step": 96
    },
    {
      "epoch": 0.08276450511945392,
      "grad_norm": 1.4140625,
      "learning_rate": 3.9725306160037016e-05,
      "loss": 0.4739900231361389,
      "mean_token_accuracy": 0.8358442187309265,
      "num_tokens": 121484251.0,
      "step": 97
    },
    {
      "epoch": 0.08361774744027303,
      "grad_norm": 1.265625,
      "learning_rate": 3.9716095299659213e-05,
      "loss": 0.4629378020763397,
      "mean_token_accuracy": 0.843445360660553,
      "num_tokens": 122664671.0,
      "step": 98
    },
    {
      "epoch": 0.08447098976109214,
      "grad_norm": 1.3671875,
      "learning_rate": 3.970673365234606e-05,
      "loss": 0.49342119693756104,
      "mean_token_accuracy": 0.8350244760513306,
      "num_tokens": 123889531.0,
      "step": 99
    },
    {
      "epoch": 0.08532423208191127,
      "grad_norm": 1.3984375,
      "learning_rate": 3.969722128969459e-05,
      "loss": 0.47572624683380127,
      "mean_token_accuracy": 0.8408681154251099,
      "num_tokens": 125117876.0,
      "step": 100
    },
    {
      "epoch": 0.08617747440273038,
      "grad_norm": 1.3515625,
      "learning_rate": 3.9687558284454515e-05,
      "loss": 0.470912367105484,
      "mean_token_accuracy": 0.8429086208343506,
      "num_tokens": 126364721.0,
      "step": 101
    },
    {
      "epoch": 0.08703071672354949,
      "grad_norm": 1.234375,
      "learning_rate": 3.9677744710527626e-05,
      "loss": 0.4839920997619629,
      "mean_token_accuracy": 0.8293547630310059,
      "num_tokens": 127695844.0,
      "step": 102
    },
    {
      "epoch": 0.0878839590443686,
      "grad_norm": 1.640625,
      "learning_rate": 3.966778064296727e-05,
      "loss": 0.4987448453903198,
      "mean_token_accuracy": 0.8310178518295288,
      "num_tokens": 128942437.0,
      "step": 103
    },
    {
      "epoch": 0.08873720136518772,
      "grad_norm": 1.2421875,
      "learning_rate": 3.965766615797774e-05,
      "loss": 0.4962223470211029,
      "mean_token_accuracy": 0.8318626284599304,
      "num_tokens": 130186213.0,
      "step": 104
    },
    {
      "epoch": 0.08959044368600683,
      "grad_norm": 1.34375,
      "learning_rate": 3.964740133291371e-05,
      "loss": 0.4899783730506897,
      "mean_token_accuracy": 0.8358625769615173,
      "num_tokens": 131462824.0,
      "step": 105
    },
    {
      "epoch": 0.09044368600682594,
      "grad_norm": 1.546875,
      "learning_rate": 3.963698624627965e-05,
      "loss": 0.4894953966140747,
      "mean_token_accuracy": 0.8351138830184937,
      "num_tokens": 132661217.0,
      "step": 106
    },
    {
      "epoch": 0.09129692832764505,
      "grad_norm": 1.3515625,
      "learning_rate": 3.962642097772921e-05,
      "loss": 0.5054884552955627,
      "mean_token_accuracy": 0.8324143290519714,
      "num_tokens": 133912918.0,
      "step": 107
    },
    {
      "epoch": 0.09215017064846416,
      "grad_norm": 1.53125,
      "learning_rate": 3.961570560806461e-05,
      "loss": 0.5067757368087769,
      "mean_token_accuracy": 0.8327956199645996,
      "num_tokens": 135178690.0,
      "step": 108
    },
    {
      "epoch": 0.09300341296928327,
      "grad_norm": 1.28125,
      "learning_rate": 3.960484021923606e-05,
      "loss": 0.5057029724121094,
      "mean_token_accuracy": 0.8299144506454468,
      "num_tokens": 136383190.0,
      "step": 109
    },
    {
      "epoch": 0.09385665529010238,
      "grad_norm": 1.2109375,
      "learning_rate": 3.959382489434107e-05,
      "loss": 0.4571252465248108,
      "mean_token_accuracy": 0.8422044515609741,
      "num_tokens": 137689738.0,
      "step": 110
    },
    {
      "epoch": 0.0947098976109215,
      "grad_norm": 1.1953125,
      "learning_rate": 3.958265971762386e-05,
      "loss": 0.4720349907875061,
      "mean_token_accuracy": 0.8394567966461182,
      "num_tokens": 138953965.0,
      "step": 111
    },
    {
      "epoch": 0.09556313993174062,
      "grad_norm": 1.296875,
      "learning_rate": 3.9571344774474715e-05,
      "loss": 0.47004902362823486,
      "mean_token_accuracy": 0.8400288224220276,
      "num_tokens": 140228424.0,
      "step": 112
    },
    {
      "epoch": 0.09641638225255973,
      "grad_norm": 1.3203125,
      "learning_rate": 3.95598801514293e-05,
      "loss": 0.4646698534488678,
      "mean_token_accuracy": 0.8433101773262024,
      "num_tokens": 141469984.0,
      "step": 113
    },
    {
      "epoch": 0.09726962457337884,
      "grad_norm": 1.734375,
      "learning_rate": 3.9548265936168026e-05,
      "loss": 0.45005208253860474,
      "mean_token_accuracy": 0.8463190793991089,
      "num_tokens": 142715767.0,
      "step": 114
    },
    {
      "epoch": 0.09812286689419795,
      "grad_norm": 1.375,
      "learning_rate": 3.953650221751538e-05,
      "loss": 0.4861631393432617,
      "mean_token_accuracy": 0.8390318155288696,
      "num_tokens": 143966239.0,
      "step": 115
    },
    {
      "epoch": 0.09897610921501707,
      "grad_norm": 1.40625,
      "learning_rate": 3.9524589085439225e-05,
      "loss": 0.47250357270240784,
      "mean_token_accuracy": 0.8411637544631958,
      "num_tokens": 145255277.0,
      "step": 116
    },
    {
      "epoch": 0.09982935153583618,
      "grad_norm": 1.28125,
      "learning_rate": 3.951252663105014e-05,
      "loss": 0.47703537344932556,
      "mean_token_accuracy": 0.8374214172363281,
      "num_tokens": 146459249.0,
      "step": 117
    },
    {
      "epoch": 0.10068259385665529,
      "grad_norm": 2.203125,
      "learning_rate": 3.950031494660069e-05,
      "loss": 0.45449715852737427,
      "mean_token_accuracy": 0.8462491631507874,
      "num_tokens": 147787272.0,
      "step": 118
    },
    {
      "epoch": 0.1015358361774744,
      "grad_norm": 1.390625,
      "learning_rate": 3.9487954125484746e-05,
      "loss": 0.48761698603630066,
      "mean_token_accuracy": 0.8339712023735046,
      "num_tokens": 149106271.0,
      "step": 119
    },
    {
      "epoch": 0.10238907849829351,
      "grad_norm": 1.2109375,
      "learning_rate": 3.947544426223677e-05,
      "loss": 0.4673748314380646,
      "mean_token_accuracy": 0.8375993967056274,
      "num_tokens": 150355961.0,
      "step": 120
    },
    {
      "epoch": 0.10324232081911262,
      "grad_norm": 1.2890625,
      "learning_rate": 3.9462785452531076e-05,
      "loss": 0.4913311302661896,
      "mean_token_accuracy": 0.8342646360397339,
      "num_tokens": 151565958.0,
      "step": 121
    },
    {
      "epoch": 0.10409556313993173,
      "grad_norm": 1.2890625,
      "learning_rate": 3.944997779318111e-05,
      "loss": 0.4819806218147278,
      "mean_token_accuracy": 0.8364108800888062,
      "num_tokens": 152849055.0,
      "step": 122
    },
    {
      "epoch": 0.10494880546075085,
      "grad_norm": 1.2109375,
      "learning_rate": 3.9437021382138704e-05,
      "loss": 0.4577251076698303,
      "mean_token_accuracy": 0.8420166969299316,
      "num_tokens": 154214669.0,
      "step": 123
    },
    {
      "epoch": 0.10580204778156997,
      "grad_norm": 1.5078125,
      "learning_rate": 3.942391631849333e-05,
      "loss": 0.47906461358070374,
      "mean_token_accuracy": 0.839522123336792,
      "num_tokens": 155444364.0,
      "step": 124
    },
    {
      "epoch": 0.10665529010238908,
      "grad_norm": 1.3203125,
      "learning_rate": 3.941066270247135e-05,
      "loss": 0.4790816307067871,
      "mean_token_accuracy": 0.834438681602478,
      "num_tokens": 156675699.0,
      "step": 125
    },
    {
      "epoch": 0.1075085324232082,
      "grad_norm": 1.234375,
      "learning_rate": 3.9397260635435234e-05,
      "loss": 0.4559358060359955,
      "mean_token_accuracy": 0.8446799516677856,
      "num_tokens": 157885391.0,
      "step": 126
    },
    {
      "epoch": 0.1083617747440273,
      "grad_norm": 1.390625,
      "learning_rate": 3.938371021988279e-05,
      "loss": 0.4998674988746643,
      "mean_token_accuracy": 0.8324342370033264,
      "num_tokens": 159177004.0,
      "step": 127
    },
    {
      "epoch": 0.10921501706484642,
      "grad_norm": 1.5,
      "learning_rate": 3.9370011559446384e-05,
      "loss": 0.4727720618247986,
      "mean_token_accuracy": 0.8394502401351929,
      "num_tokens": 160434576.0,
      "step": 128
    },
    {
      "epoch": 0.11006825938566553,
      "grad_norm": 1.25,
      "learning_rate": 3.935616475889216e-05,
      "loss": 0.4516422152519226,
      "mean_token_accuracy": 0.8440276980400085,
      "num_tokens": 161713456.0,
      "step": 129
    },
    {
      "epoch": 0.11092150170648464,
      "grad_norm": 1.2265625,
      "learning_rate": 3.934216992411919e-05,
      "loss": 0.4649214744567871,
      "mean_token_accuracy": 0.8417582511901855,
      "num_tokens": 162965987.0,
      "step": 130
    },
    {
      "epoch": 0.11177474402730375,
      "grad_norm": 1.265625,
      "learning_rate": 3.932802716215874e-05,
      "loss": 0.4920186996459961,
      "mean_token_accuracy": 0.8359864950180054,
      "num_tokens": 164280113.0,
      "step": 131
    },
    {
      "epoch": 0.11262798634812286,
      "grad_norm": 1.2578125,
      "learning_rate": 3.931373658117338e-05,
      "loss": 0.47814157605171204,
      "mean_token_accuracy": 0.8348724842071533,
      "num_tokens": 165546878.0,
      "step": 132
    },
    {
      "epoch": 0.11348122866894197,
      "grad_norm": 1.2578125,
      "learning_rate": 3.92992982904562e-05,
      "loss": 0.469256728887558,
      "mean_token_accuracy": 0.8385157585144043,
      "num_tokens": 166826850.0,
      "step": 133
    },
    {
      "epoch": 0.11433447098976109,
      "grad_norm": 1.1328125,
      "learning_rate": 3.928471240042995e-05,
      "loss": 0.4305429458618164,
      "mean_token_accuracy": 0.8522835969924927,
      "num_tokens": 168063694.0,
      "step": 134
    },
    {
      "epoch": 0.11518771331058021,
      "grad_norm": 1.375,
      "learning_rate": 3.9269979022646237e-05,
      "loss": 0.4919919967651367,
      "mean_token_accuracy": 0.8359761834144592,
      "num_tokens": 169403501.0,
      "step": 135
    },
    {
      "epoch": 0.11604095563139932,
      "grad_norm": 1.3046875,
      "learning_rate": 3.9255098269784585e-05,
      "loss": 0.4885086119174957,
      "mean_token_accuracy": 0.833430290222168,
      "num_tokens": 170656789.0,
      "step": 136
    },
    {
      "epoch": 0.11689419795221843,
      "grad_norm": 1.2109375,
      "learning_rate": 3.924007025565168e-05,
      "loss": 0.49744275212287903,
      "mean_token_accuracy": 0.8296787738800049,
      "num_tokens": 171923708.0,
      "step": 137
    },
    {
      "epoch": 0.11774744027303755,
      "grad_norm": 1.1953125,
      "learning_rate": 3.922489509518043e-05,
      "loss": 0.48081400990486145,
      "mean_token_accuracy": 0.8372715711593628,
      "num_tokens": 173184479.0,
      "step": 138
    },
    {
      "epoch": 0.11860068259385666,
      "grad_norm": 1.3515625,
      "learning_rate": 3.920957290442911e-05,
      "loss": 0.4128647446632385,
      "mean_token_accuracy": 0.8545809984207153,
      "num_tokens": 174414932.0,
      "step": 139
    },
    {
      "epoch": 0.11945392491467577,
      "grad_norm": 1.2578125,
      "learning_rate": 3.919410380058046e-05,
      "loss": 0.49558761715888977,
      "mean_token_accuracy": 0.8315925598144531,
      "num_tokens": 175613006.0,
      "step": 140
    },
    {
      "epoch": 0.12030716723549488,
      "grad_norm": 1.234375,
      "learning_rate": 3.91784879019408e-05,
      "loss": 0.45157134532928467,
      "mean_token_accuracy": 0.8429455757141113,
      "num_tokens": 176892998.0,
      "step": 141
    },
    {
      "epoch": 0.12116040955631399,
      "grad_norm": 1.375,
      "learning_rate": 3.9162725327939145e-05,
      "loss": 0.4698953330516815,
      "mean_token_accuracy": 0.840680718421936,
      "num_tokens": 178209002.0,
      "step": 142
    },
    {
      "epoch": 0.1220136518771331,
      "grad_norm": 1.2734375,
      "learning_rate": 3.914681619912624e-05,
      "loss": 0.4334169328212738,
      "mean_token_accuracy": 0.8479805588722229,
      "num_tokens": 179505044.0,
      "step": 143
    },
    {
      "epoch": 0.12286689419795221,
      "grad_norm": 1.09375,
      "learning_rate": 3.9130760637173676e-05,
      "loss": 0.4216551184654236,
      "mean_token_accuracy": 0.8515428304672241,
      "num_tokens": 180780894.0,
      "step": 144
    },
    {
      "epoch": 0.12372013651877133,
      "grad_norm": 1.2734375,
      "learning_rate": 3.9114558764872965e-05,
      "loss": 0.4744829833507538,
      "mean_token_accuracy": 0.8363406658172607,
      "num_tokens": 182019233.0,
      "step": 145
    },
    {
      "epoch": 0.12457337883959044,
      "grad_norm": 1.390625,
      "learning_rate": 3.909821070613458e-05,
      "loss": 0.46261051297187805,
      "mean_token_accuracy": 0.8410444259643555,
      "num_tokens": 183198103.0,
      "step": 146
    },
    {
      "epoch": 0.12542662116040956,
      "grad_norm": 1.2578125,
      "learning_rate": 3.9081716585987e-05,
      "loss": 0.4706021547317505,
      "mean_token_accuracy": 0.8416902422904968,
      "num_tokens": 184457397.0,
      "step": 147
    },
    {
      "epoch": 0.12627986348122866,
      "grad_norm": 1.4609375,
      "learning_rate": 3.90650765305758e-05,
      "loss": 0.49795109033584595,
      "mean_token_accuracy": 0.833423912525177,
      "num_tokens": 185716751.0,
      "step": 148
    },
    {
      "epoch": 0.12713310580204779,
      "grad_norm": 1.2734375,
      "learning_rate": 3.904829066716263e-05,
      "loss": 0.4738658368587494,
      "mean_token_accuracy": 0.839400589466095,
      "num_tokens": 186914442.0,
      "step": 149
    },
    {
      "epoch": 0.12798634812286688,
      "grad_norm": 1.296875,
      "learning_rate": 3.903135912412428e-05,
      "loss": 0.47307848930358887,
      "mean_token_accuracy": 0.8358522057533264,
      "num_tokens": 188189573.0,
      "step": 150
    },
    {
      "epoch": 0.128839590443686,
      "grad_norm": 1.5390625,
      "learning_rate": 3.9014282030951655e-05,
      "loss": 0.507688045501709,
      "mean_token_accuracy": 0.8262724876403809,
      "num_tokens": 189517163.0,
      "step": 151
    },
    {
      "epoch": 0.1296928327645051,
      "grad_norm": 1.34375,
      "learning_rate": 3.899705951824886e-05,
      "loss": 0.4883740246295929,
      "mean_token_accuracy": 0.8319364786148071,
      "num_tokens": 190782440.0,
      "step": 152
    },
    {
      "epoch": 0.13054607508532423,
      "grad_norm": 1.2734375,
      "learning_rate": 3.897969171773213e-05,
      "loss": 0.45081931352615356,
      "mean_token_accuracy": 0.8458638787269592,
      "num_tokens": 192071254.0,
      "step": 153
    },
    {
      "epoch": 0.13139931740614336,
      "grad_norm": 1.265625,
      "learning_rate": 3.896217876222885e-05,
      "loss": 0.46706950664520264,
      "mean_token_accuracy": 0.8416528701782227,
      "num_tokens": 193385171.0,
      "step": 154
    },
    {
      "epoch": 0.13225255972696245,
      "grad_norm": 1.3828125,
      "learning_rate": 3.8944520785676545e-05,
      "loss": 0.4930266737937927,
      "mean_token_accuracy": 0.8336679935455322,
      "num_tokens": 194699921.0,
      "step": 155
    },
    {
      "epoch": 0.13310580204778158,
      "grad_norm": 1.21875,
      "learning_rate": 3.892671792312182e-05,
      "loss": 0.4549766778945923,
      "mean_token_accuracy": 0.8450692892074585,
      "num_tokens": 195881190.0,
      "step": 156
    },
    {
      "epoch": 0.13395904436860068,
      "grad_norm": 1.296875,
      "learning_rate": 3.890877031071941e-05,
      "loss": 0.4789050221443176,
      "mean_token_accuracy": 0.8376064300537109,
      "num_tokens": 197189795.0,
      "step": 157
    },
    {
      "epoch": 0.1348122866894198,
      "grad_norm": 1.3671875,
      "learning_rate": 3.889067808573103e-05,
      "loss": 0.46956318616867065,
      "mean_token_accuracy": 0.8357698917388916,
      "num_tokens": 198398002.0,
      "step": 158
    },
    {
      "epoch": 0.1356655290102389,
      "grad_norm": 1.28125,
      "learning_rate": 3.887244138652439e-05,
      "loss": 0.4432661831378937,
      "mean_token_accuracy": 0.8456773161888123,
      "num_tokens": 199685935.0,
      "step": 159
    },
    {
      "epoch": 0.13651877133105803,
      "grad_norm": 1.09375,
      "learning_rate": 3.885406035257217e-05,
      "loss": 0.43149739503860474,
      "mean_token_accuracy": 0.8441345691680908,
      "num_tokens": 201017865.0,
      "step": 160
    },
    {
      "epoch": 0.13737201365187712,
      "grad_norm": 1.3671875,
      "learning_rate": 3.8835535124450845e-05,
      "loss": 0.4489964246749878,
      "mean_token_accuracy": 0.8476303219795227,
      "num_tokens": 202232847.0,
      "step": 161
    },
    {
      "epoch": 0.13822525597269625,
      "grad_norm": 1.2109375,
      "learning_rate": 3.8816865843839716e-05,
      "loss": 0.46801450848579407,
      "mean_token_accuracy": 0.8377986550331116,
      "num_tokens": 203543666.0,
      "step": 162
    },
    {
      "epoch": 0.13907849829351535,
      "grad_norm": 1.4375,
      "learning_rate": 3.879805265351977e-05,
      "loss": 0.47466951608657837,
      "mean_token_accuracy": 0.8353536128997803,
      "num_tokens": 204785770.0,
      "step": 163
    },
    {
      "epoch": 0.13993174061433447,
      "grad_norm": 1.3203125,
      "learning_rate": 3.877909569737262e-05,
      "loss": 0.49017807841300964,
      "mean_token_accuracy": 0.8298170566558838,
      "num_tokens": 206114448.0,
      "step": 164
    },
    {
      "epoch": 0.1407849829351536,
      "grad_norm": 1.25,
      "learning_rate": 3.8759995120379355e-05,
      "loss": 0.4916384220123291,
      "mean_token_accuracy": 0.8323613405227661,
      "num_tokens": 207382084.0,
      "step": 165
    },
    {
      "epoch": 0.1416382252559727,
      "grad_norm": 1.3359375,
      "learning_rate": 3.874075106861949e-05,
      "loss": 0.44442546367645264,
      "mean_token_accuracy": 0.8514591455459595,
      "num_tokens": 208641374.0,
      "step": 166
    },
    {
      "epoch": 0.14249146757679182,
      "grad_norm": 1.21875,
      "learning_rate": 3.8721363689269805e-05,
      "loss": 0.47127318382263184,
      "mean_token_accuracy": 0.8385240435600281,
      "num_tokens": 209954369.0,
      "step": 167
    },
    {
      "epoch": 0.14334470989761092,
      "grad_norm": 1.5625,
      "learning_rate": 3.870183313060326e-05,
      "loss": 0.4736810028553009,
      "mean_token_accuracy": 0.8350101709365845,
      "num_tokens": 211208589.0,
      "step": 168
    },
    {
      "epoch": 0.14419795221843004,
      "grad_norm": 1.078125,
      "learning_rate": 3.86821595419878e-05,
      "loss": 0.4203541874885559,
      "mean_token_accuracy": 0.8534204959869385,
      "num_tokens": 212579443.0,
      "step": 169
    },
    {
      "epoch": 0.14505119453924914,
      "grad_norm": 1.2578125,
      "learning_rate": 3.86623430738853e-05,
      "loss": 0.4757835566997528,
      "mean_token_accuracy": 0.8392912149429321,
      "num_tokens": 213820300.0,
      "step": 170
    },
    {
      "epoch": 0.14590443686006827,
      "grad_norm": 1.203125,
      "learning_rate": 3.864238387785031e-05,
      "loss": 0.4294852018356323,
      "mean_token_accuracy": 0.8515962362289429,
      "num_tokens": 215129215.0,
      "step": 171
    },
    {
      "epoch": 0.14675767918088736,
      "grad_norm": 1.21875,
      "learning_rate": 3.8622282106529005e-05,
      "loss": 0.4538024365901947,
      "mean_token_accuracy": 0.8449404835700989,
      "num_tokens": 216437979.0,
      "step": 172
    },
    {
      "epoch": 0.1476109215017065,
      "grad_norm": 1.28125,
      "learning_rate": 3.860203791365793e-05,
      "loss": 0.47619950771331787,
      "mean_token_accuracy": 0.8380199074745178,
      "num_tokens": 217733661.0,
      "step": 173
    },
    {
      "epoch": 0.14846416382252559,
      "grad_norm": 1.390625,
      "learning_rate": 3.858165145406285e-05,
      "loss": 0.49371758103370667,
      "mean_token_accuracy": 0.8303412199020386,
      "num_tokens": 219038902.0,
      "step": 174
    },
    {
      "epoch": 0.1493174061433447,
      "grad_norm": 1.3125,
      "learning_rate": 3.856112288365759e-05,
      "loss": 0.4793042838573456,
      "mean_token_accuracy": 0.8394200205802917,
      "num_tokens": 220277949.0,
      "step": 175
    },
    {
      "epoch": 0.15017064846416384,
      "grad_norm": 1.1328125,
      "learning_rate": 3.854045235944283e-05,
      "loss": 0.4266126751899719,
      "mean_token_accuracy": 0.8544156551361084,
      "num_tokens": 221529359.0,
      "step": 176
    },
    {
      "epoch": 0.15102389078498293,
      "grad_norm": 1.2265625,
      "learning_rate": 3.851964003950489e-05,
      "loss": 0.4735304117202759,
      "mean_token_accuracy": 0.8374233841896057,
      "num_tokens": 222817936.0,
      "step": 177
    },
    {
      "epoch": 0.15187713310580206,
      "grad_norm": 1.2421875,
      "learning_rate": 3.849868608301453e-05,
      "loss": 0.457353413105011,
      "mean_token_accuracy": 0.8445993661880493,
      "num_tokens": 224036549.0,
      "step": 178
    },
    {
      "epoch": 0.15273037542662116,
      "grad_norm": 1.1015625,
      "learning_rate": 3.8477590650225735e-05,
      "loss": 0.44284987449645996,
      "mean_token_accuracy": 0.8458787798881531,
      "num_tokens": 225256675.0,
      "step": 179
    },
    {
      "epoch": 0.15358361774744028,
      "grad_norm": 1.3203125,
      "learning_rate": 3.8456353902474504e-05,
      "loss": 0.46864694356918335,
      "mean_token_accuracy": 0.8409409523010254,
      "num_tokens": 226495829.0,
      "step": 180
    },
    {
      "epoch": 0.15443686006825938,
      "grad_norm": 1.21875,
      "learning_rate": 3.843497600217757e-05,
      "loss": 0.45589396357536316,
      "mean_token_accuracy": 0.8455904126167297,
      "num_tokens": 227790857.0,
      "step": 181
    },
    {
      "epoch": 0.1552901023890785,
      "grad_norm": 1.25,
      "learning_rate": 3.841345711283121e-05,
      "loss": 0.4705445170402527,
      "mean_token_accuracy": 0.8378282189369202,
      "num_tokens": 229128457.0,
      "step": 182
    },
    {
      "epoch": 0.1561433447098976,
      "grad_norm": 1.3125,
      "learning_rate": 3.839179739900998e-05,
      "loss": 0.5246204137802124,
      "mean_token_accuracy": 0.8269191980361938,
      "num_tokens": 230430413.0,
      "step": 183
    },
    {
      "epoch": 0.15699658703071673,
      "grad_norm": 1.3046875,
      "learning_rate": 3.836999702636541e-05,
      "loss": 0.4898459017276764,
      "mean_token_accuracy": 0.8311104774475098,
      "num_tokens": 231661339.0,
      "step": 184
    },
    {
      "epoch": 0.15784982935153583,
      "grad_norm": 1.2265625,
      "learning_rate": 3.834805616162483e-05,
      "loss": 0.4903266429901123,
      "mean_token_accuracy": 0.8318898677825928,
      "num_tokens": 232867843.0,
      "step": 185
    },
    {
      "epoch": 0.15870307167235495,
      "grad_norm": 1.2890625,
      "learning_rate": 3.832597497259e-05,
      "loss": 0.4606083333492279,
      "mean_token_accuracy": 0.8408007621765137,
      "num_tokens": 234143665.0,
      "step": 186
    },
    {
      "epoch": 0.15955631399317405,
      "grad_norm": 1.3984375,
      "learning_rate": 3.8303753628135865e-05,
      "loss": 0.46872371435165405,
      "mean_token_accuracy": 0.8402537703514099,
      "num_tokens": 235499713.0,
      "step": 187
    },
    {
      "epoch": 0.16040955631399317,
      "grad_norm": 1.1875,
      "learning_rate": 3.8281392298209313e-05,
      "loss": 0.4516822099685669,
      "mean_token_accuracy": 0.8455985188484192,
      "num_tokens": 236718260.0,
      "step": 188
    },
    {
      "epoch": 0.1612627986348123,
      "grad_norm": 1.2578125,
      "learning_rate": 3.825889115382777e-05,
      "loss": 0.45412927865982056,
      "mean_token_accuracy": 0.8479938507080078,
      "num_tokens": 237945410.0,
      "step": 189
    },
    {
      "epoch": 0.1621160409556314,
      "grad_norm": 1.1640625,
      "learning_rate": 3.8236250367078006e-05,
      "loss": 0.4712190628051758,
      "mean_token_accuracy": 0.8402495980262756,
      "num_tokens": 239144690.0,
      "step": 190
    },
    {
      "epoch": 0.16296928327645052,
      "grad_norm": 1.140625,
      "learning_rate": 3.821347011111472e-05,
      "loss": 0.44732701778411865,
      "mean_token_accuracy": 0.8475074768066406,
      "num_tokens": 240389866.0,
      "step": 191
    },
    {
      "epoch": 0.16382252559726962,
      "grad_norm": 1.1640625,
      "learning_rate": 3.819055056015926e-05,
      "loss": 0.4846896231174469,
      "mean_token_accuracy": 0.8350772261619568,
      "num_tokens": 241679572.0,
      "step": 192
    },
    {
      "epoch": 0.16467576791808874,
      "grad_norm": 1.2890625,
      "learning_rate": 3.8167491889498326e-05,
      "loss": 0.4617387056350708,
      "mean_token_accuracy": 0.8435602784156799,
      "num_tokens": 242917542.0,
      "step": 193
    },
    {
      "epoch": 0.16552901023890784,
      "grad_norm": 1.3203125,
      "learning_rate": 3.814429427548256e-05,
      "loss": 0.475017249584198,
      "mean_token_accuracy": 0.8369480967521667,
      "num_tokens": 244182351.0,
      "step": 194
    },
    {
      "epoch": 0.16638225255972697,
      "grad_norm": 1.265625,
      "learning_rate": 3.812095789552525e-05,
      "loss": 0.473707377910614,
      "mean_token_accuracy": 0.8352920413017273,
      "num_tokens": 245517877.0,
      "step": 195
    },
    {
      "epoch": 0.16723549488054607,
      "grad_norm": 1.21875,
      "learning_rate": 3.8097482928100936e-05,
      "loss": 0.43076425790786743,
      "mean_token_accuracy": 0.8486889600753784,
      "num_tokens": 246731178.0,
      "step": 196
    },
    {
      "epoch": 0.1680887372013652,
      "grad_norm": 1.2734375,
      "learning_rate": 3.807386955274408e-05,
      "loss": 0.4240972399711609,
      "mean_token_accuracy": 0.8534239530563354,
      "num_tokens": 248068087.0,
      "step": 197
    },
    {
      "epoch": 0.1689419795221843,
      "grad_norm": 1.1171875,
      "learning_rate": 3.805011795004766e-05,
      "loss": 0.3969980478286743,
      "mean_token_accuracy": 0.8631253242492676,
      "num_tokens": 249398377.0,
      "step": 198
    },
    {
      "epoch": 0.1697952218430034,
      "grad_norm": 1.25,
      "learning_rate": 3.802622830166181e-05,
      "loss": 0.4617486596107483,
      "mean_token_accuracy": 0.8423023223876953,
      "num_tokens": 250540747.0,
      "step": 199
    },
    {
      "epoch": 0.17064846416382254,
      "grad_norm": 1.4140625,
      "learning_rate": 3.800220079029243e-05,
      "loss": 0.4578988552093506,
      "mean_token_accuracy": 0.8419663906097412,
      "num_tokens": 251823915.0,
      "step": 200
    },
    {
      "epoch": 0.17150170648464164,
      "grad_norm": 1.1015625,
      "learning_rate": 3.797803559969979e-05,
      "loss": 0.4194392263889313,
      "mean_token_accuracy": 0.8565274477005005,
      "num_tokens": 253096924.0,
      "step": 201
    },
    {
      "epoch": 0.17235494880546076,
      "grad_norm": 1.1875,
      "learning_rate": 3.795373291469709e-05,
      "loss": 0.4757955074310303,
      "mean_token_accuracy": 0.8382071256637573,
      "num_tokens": 254413112.0,
      "step": 202
    },
    {
      "epoch": 0.17320819112627986,
      "grad_norm": 1.1484375,
      "learning_rate": 3.792929292114911e-05,
      "loss": 0.45573127269744873,
      "mean_token_accuracy": 0.8456025719642639,
      "num_tokens": 255705035.0,
      "step": 203
    },
    {
      "epoch": 0.17406143344709898,
      "grad_norm": 1.1953125,
      "learning_rate": 3.7904715805970725e-05,
      "loss": 0.47824519872665405,
      "mean_token_accuracy": 0.8377963304519653,
      "num_tokens": 256949552.0,
      "step": 204
    },
    {
      "epoch": 0.17491467576791808,
      "grad_norm": 1.359375,
      "learning_rate": 3.7880001757125515e-05,
      "loss": 0.430466890335083,
      "mean_token_accuracy": 0.8507641553878784,
      "num_tokens": 258141045.0,
      "step": 205
    },
    {
      "epoch": 0.1757679180887372,
      "grad_norm": 1.328125,
      "learning_rate": 3.7855150963624335e-05,
      "loss": 0.4850834012031555,
      "mean_token_accuracy": 0.8374762535095215,
      "num_tokens": 259503190.0,
      "step": 206
    },
    {
      "epoch": 0.1766211604095563,
      "grad_norm": 1.234375,
      "learning_rate": 3.7830163615523814e-05,
      "loss": 0.4386395215988159,
      "mean_token_accuracy": 0.8505452871322632,
      "num_tokens": 260807727.0,
      "step": 207
    },
    {
      "epoch": 0.17747440273037543,
      "grad_norm": 1.3515625,
      "learning_rate": 3.780503990392497e-05,
      "loss": 0.46706992387771606,
      "mean_token_accuracy": 0.8421170711517334,
      "num_tokens": 262028662.0,
      "step": 208
    },
    {
      "epoch": 0.17832764505119453,
      "grad_norm": 1.1796875,
      "learning_rate": 3.7779780020971694e-05,
      "loss": 0.47389906644821167,
      "mean_token_accuracy": 0.8383389115333557,
      "num_tokens": 263299871.0,
      "step": 209
    },
    {
      "epoch": 0.17918088737201365,
      "grad_norm": 1.25,
      "learning_rate": 3.775438415984932e-05,
      "loss": 0.49315619468688965,
      "mean_token_accuracy": 0.8354644775390625,
      "num_tokens": 264590479.0,
      "step": 210
    },
    {
      "epoch": 0.18003412969283278,
      "grad_norm": 1.140625,
      "learning_rate": 3.7728852514783135e-05,
      "loss": 0.4221291244029999,
      "mean_token_accuracy": 0.8525570034980774,
      "num_tokens": 265925500.0,
      "step": 211
    },
    {
      "epoch": 0.18088737201365188,
      "grad_norm": 1.3125,
      "learning_rate": 3.7703185281036866e-05,
      "loss": 0.4796106517314911,
      "mean_token_accuracy": 0.8365914821624756,
      "num_tokens": 267154555.0,
      "step": 212
    },
    {
      "epoch": 0.181740614334471,
      "grad_norm": 1.1328125,
      "learning_rate": 3.767738265491122e-05,
      "loss": 0.4542846083641052,
      "mean_token_accuracy": 0.844251275062561,
      "num_tokens": 268392393.0,
      "step": 213
    },
    {
      "epoch": 0.1825938566552901,
      "grad_norm": 1.265625,
      "learning_rate": 3.7651444833742394e-05,
      "loss": 0.4890654981136322,
      "mean_token_accuracy": 0.8337072730064392,
      "num_tokens": 269671693.0,
      "step": 214
    },
    {
      "epoch": 0.18344709897610922,
      "grad_norm": 1.265625,
      "learning_rate": 3.762537201590051e-05,
      "loss": 0.4308969974517822,
      "mean_token_accuracy": 0.8499782681465149,
      "num_tokens": 270922697.0,
      "step": 215
    },
    {
      "epoch": 0.18430034129692832,
      "grad_norm": 1.140625,
      "learning_rate": 3.759916440078816e-05,
      "loss": 0.4648487865924835,
      "mean_token_accuracy": 0.8449735641479492,
      "num_tokens": 272175857.0,
      "step": 216
    },
    {
      "epoch": 0.18515358361774745,
      "grad_norm": 1.15625,
      "learning_rate": 3.757282218883884e-05,
      "loss": 0.43143612146377563,
      "mean_token_accuracy": 0.8523366451263428,
      "num_tokens": 273469897.0,
      "step": 217
    },
    {
      "epoch": 0.18600682593856654,
      "grad_norm": 1.171875,
      "learning_rate": 3.7546345581515426e-05,
      "loss": 0.4495916962623596,
      "mean_token_accuracy": 0.8480303287506104,
      "num_tokens": 274642288.0,
      "step": 218
    },
    {
      "epoch": 0.18686006825938567,
      "grad_norm": 1.28125,
      "learning_rate": 3.751973478130865e-05,
      "loss": 0.4484361708164215,
      "mean_token_accuracy": 0.8469212651252747,
      "num_tokens": 275907047.0,
      "step": 219
    },
    {
      "epoch": 0.18771331058020477,
      "grad_norm": 1.46875,
      "learning_rate": 3.749298999173554e-05,
      "loss": 0.48330432176589966,
      "mean_token_accuracy": 0.8327616453170776,
      "num_tokens": 277186776.0,
      "step": 220
    },
    {
      "epoch": 0.1885665529010239,
      "grad_norm": 1.1796875,
      "learning_rate": 3.7466111417337857e-05,
      "loss": 0.4665176570415497,
      "mean_token_accuracy": 0.8397443294525146,
      "num_tokens": 278549677.0,
      "step": 221
    },
    {
      "epoch": 0.189419795221843,
      "grad_norm": 1.4140625,
      "learning_rate": 3.743909926368053e-05,
      "loss": 0.5152418613433838,
      "mean_token_accuracy": 0.8243826627731323,
      "num_tokens": 279864146.0,
      "step": 222
    },
    {
      "epoch": 0.19027303754266212,
      "grad_norm": 1.296875,
      "learning_rate": 3.74119537373501e-05,
      "loss": 0.47882330417633057,
      "mean_token_accuracy": 0.8406552076339722,
      "num_tokens": 281086763.0,
      "step": 223
    },
    {
      "epoch": 0.19112627986348124,
      "grad_norm": 1.15625,
      "learning_rate": 3.738467504595312e-05,
      "loss": 0.43267562985420227,
      "mean_token_accuracy": 0.8541044592857361,
      "num_tokens": 282370264.0,
      "step": 224
    },
    {
      "epoch": 0.19197952218430034,
      "grad_norm": 1.3125,
      "learning_rate": 3.73572633981146e-05,
      "loss": 0.4740004241466522,
      "mean_token_accuracy": 0.8391311168670654,
      "num_tokens": 283624469.0,
      "step": 225
    },
    {
      "epoch": 0.19283276450511946,
      "grad_norm": 1.2421875,
      "learning_rate": 3.732971900347637e-05,
      "loss": 0.4605056643486023,
      "mean_token_accuracy": 0.839042067527771,
      "num_tokens": 284897150.0,
      "step": 226
    },
    {
      "epoch": 0.19368600682593856,
      "grad_norm": 1.296875,
      "learning_rate": 3.730204207269548e-05,
      "loss": 0.47835296392440796,
      "mean_token_accuracy": 0.8362683057785034,
      "num_tokens": 286165790.0,
      "step": 227
    },
    {
      "epoch": 0.1945392491467577,
      "grad_norm": 1.203125,
      "learning_rate": 3.727423281744265e-05,
      "loss": 0.41937506198883057,
      "mean_token_accuracy": 0.8540375232696533,
      "num_tokens": 287384784.0,
      "step": 228
    },
    {
      "epoch": 0.19539249146757678,
      "grad_norm": 1.140625,
      "learning_rate": 3.724629145040056e-05,
      "loss": 0.434437096118927,
      "mean_token_accuracy": 0.848793625831604,
      "num_tokens": 288615699.0,
      "step": 229
    },
    {
      "epoch": 0.1962457337883959,
      "grad_norm": 1.484375,
      "learning_rate": 3.72182181852623e-05,
      "loss": 0.4756004214286804,
      "mean_token_accuracy": 0.838626503944397,
      "num_tokens": 289894361.0,
      "step": 230
    },
    {
      "epoch": 0.197098976109215,
      "grad_norm": 1.46875,
      "learning_rate": 3.7190013236729687e-05,
      "loss": 0.4835631251335144,
      "mean_token_accuracy": 0.8379888534545898,
      "num_tokens": 291150605.0,
      "step": 231
    },
    {
      "epoch": 0.19795221843003413,
      "grad_norm": 1.3828125,
      "learning_rate": 3.716167682051165e-05,
      "loss": 0.45977646112442017,
      "mean_token_accuracy": 0.8417062163352966,
      "num_tokens": 292499420.0,
      "step": 232
    },
    {
      "epoch": 0.19880546075085323,
      "grad_norm": 2.0625,
      "learning_rate": 3.7133209153322576e-05,
      "loss": 0.44524097442626953,
      "mean_token_accuracy": 0.8452048301696777,
      "num_tokens": 293751979.0,
      "step": 233
    },
    {
      "epoch": 0.19965870307167236,
      "grad_norm": 1.0859375,
      "learning_rate": 3.710461045288063e-05,
      "loss": 0.42973238229751587,
      "mean_token_accuracy": 0.8520611524581909,
      "num_tokens": 294937805.0,
      "step": 234
    },
    {
      "epoch": 0.20051194539249148,
      "grad_norm": 1.0546875,
      "learning_rate": 3.707588093790612e-05,
      "loss": 0.4376424551010132,
      "mean_token_accuracy": 0.8477351665496826,
      "num_tokens": 296178712.0,
      "step": 235
    },
    {
      "epoch": 0.20136518771331058,
      "grad_norm": 1.1640625,
      "learning_rate": 3.704702082811981e-05,
      "loss": 0.456723153591156,
      "mean_token_accuracy": 0.8462638258934021,
      "num_tokens": 297459792.0,
      "step": 236
    },
    {
      "epoch": 0.2022184300341297,
      "grad_norm": 1.234375,
      "learning_rate": 3.701803034424124e-05,
      "loss": 0.46412354707717896,
      "mean_token_accuracy": 0.8431435823440552,
      "num_tokens": 298713142.0,
      "step": 237
    },
    {
      "epoch": 0.2030716723549488,
      "grad_norm": 1.1796875,
      "learning_rate": 3.698890970798705e-05,
      "loss": 0.4519304633140564,
      "mean_token_accuracy": 0.846401572227478,
      "num_tokens": 299970347.0,
      "step": 238
    },
    {
      "epoch": 0.20392491467576793,
      "grad_norm": 1.34375,
      "learning_rate": 3.695965914206926e-05,
      "loss": 0.4444093704223633,
      "mean_token_accuracy": 0.8459943532943726,
      "num_tokens": 301241280.0,
      "step": 239
    },
    {
      "epoch": 0.20477815699658702,
      "grad_norm": 1.171875,
      "learning_rate": 3.693027887019357e-05,
      "loss": 0.4432458281517029,
      "mean_token_accuracy": 0.8485412001609802,
      "num_tokens": 302521277.0,
      "step": 240
    },
    {
      "epoch": 0.20563139931740615,
      "grad_norm": 1.0078125,
      "learning_rate": 3.6900769117057696e-05,
      "loss": 0.3962821662425995,
      "mean_token_accuracy": 0.8606544733047485,
      "num_tokens": 303769805.0,
      "step": 241
    },
    {
      "epoch": 0.20648464163822525,
      "grad_norm": 1.296875,
      "learning_rate": 3.6871130108349584e-05,
      "loss": 0.4304523169994354,
      "mean_token_accuracy": 0.8522424697875977,
      "num_tokens": 304974398.0,
      "step": 242
    },
    {
      "epoch": 0.20733788395904437,
      "grad_norm": 1.2109375,
      "learning_rate": 3.6841362070745715e-05,
      "loss": 0.43119847774505615,
      "mean_token_accuracy": 0.8495322465896606,
      "num_tokens": 306249154.0,
      "step": 243
    },
    {
      "epoch": 0.20819112627986347,
      "grad_norm": 1.046875,
      "learning_rate": 3.68114652319094e-05,
      "loss": 0.39746519923210144,
      "mean_token_accuracy": 0.8652471899986267,
      "num_tokens": 307478296.0,
      "step": 244
    },
    {
      "epoch": 0.2090443686006826,
      "grad_norm": 1.1171875,
      "learning_rate": 3.6781439820488974e-05,
      "loss": 0.4398203194141388,
      "mean_token_accuracy": 0.8470622301101685,
      "num_tokens": 308776394.0,
      "step": 245
    },
    {
      "epoch": 0.2098976109215017,
      "grad_norm": 1.2734375,
      "learning_rate": 3.6751286066116116e-05,
      "loss": 0.46068644523620605,
      "mean_token_accuracy": 0.8402528762817383,
      "num_tokens": 310079519.0,
      "step": 246
    },
    {
      "epoch": 0.21075085324232082,
      "grad_norm": 1.359375,
      "learning_rate": 3.672100419940403e-05,
      "loss": 0.4506421387195587,
      "mean_token_accuracy": 0.8491493463516235,
      "num_tokens": 311335639.0,
      "step": 247
    },
    {
      "epoch": 0.21160409556313994,
      "grad_norm": 1.15625,
      "learning_rate": 3.6690594451945746e-05,
      "loss": 0.46329864859580994,
      "mean_token_accuracy": 0.842716634273529,
      "num_tokens": 312661466.0,
      "step": 248
    },
    {
      "epoch": 0.21245733788395904,
      "grad_norm": 1.1171875,
      "learning_rate": 3.6660057056312276e-05,
      "loss": 0.421508252620697,
      "mean_token_accuracy": 0.8530087471008301,
      "num_tokens": 313921380.0,
      "step": 249
    },
    {
      "epoch": 0.21331058020477817,
      "grad_norm": 1.1484375,
      "learning_rate": 3.662939224605091e-05,
      "loss": 0.441190242767334,
      "mean_token_accuracy": 0.847051739692688,
      "num_tokens": 315165992.0,
      "step": 250
    },
    {
      "epoch": 0.21416382252559726,
      "grad_norm": 1.15625,
      "learning_rate": 3.6598600255683365e-05,
      "loss": 0.4475253224372864,
      "mean_token_accuracy": 0.8445391654968262,
      "num_tokens": 316513903.0,
      "step": 251
    },
    {
      "epoch": 0.2150170648464164,
      "grad_norm": 1.4375,
      "learning_rate": 3.6567681320704035e-05,
      "loss": 0.478191077709198,
      "mean_token_accuracy": 0.8374630808830261,
      "num_tokens": 317841722.0,
      "step": 252
    },
    {
      "epoch": 0.2158703071672355,
      "grad_norm": 1.2109375,
      "learning_rate": 3.653663567757818e-05,
      "loss": 0.47320500016212463,
      "mean_token_accuracy": 0.8391737937927246,
      "num_tokens": 319183726.0,
      "step": 253
    },
    {
      "epoch": 0.2167235494880546,
      "grad_norm": 1.28125,
      "learning_rate": 3.650546356374009e-05,
      "loss": 0.4273732900619507,
      "mean_token_accuracy": 0.850881814956665,
      "num_tokens": 320527357.0,
      "step": 254
    },
    {
      "epoch": 0.2175767918088737,
      "grad_norm": 1.1484375,
      "learning_rate": 3.647416521759133e-05,
      "loss": 0.433204710483551,
      "mean_token_accuracy": 0.8512927293777466,
      "num_tokens": 321768869.0,
      "step": 255
    },
    {
      "epoch": 0.21843003412969283,
      "grad_norm": 1.296875,
      "learning_rate": 3.6442740878498824e-05,
      "loss": 0.45229026675224304,
      "mean_token_accuracy": 0.8455748558044434,
      "num_tokens": 322994978.0,
      "step": 256
    },
    {
      "epoch": 0.21928327645051193,
      "grad_norm": 1.15625,
      "learning_rate": 3.641119078679314e-05,
      "loss": 0.4270596504211426,
      "mean_token_accuracy": 0.850013256072998,
      "num_tokens": 324318995.0,
      "step": 257
    },
    {
      "epoch": 0.22013651877133106,
      "grad_norm": 1.046875,
      "learning_rate": 3.637951518376654e-05,
      "loss": 0.40339896082878113,
      "mean_token_accuracy": 0.858301043510437,
      "num_tokens": 325572371.0,
      "step": 258
    },
    {
      "epoch": 0.22098976109215018,
      "grad_norm": 1.4375,
      "learning_rate": 3.634771431167122e-05,
      "loss": 0.45748043060302734,
      "mean_token_accuracy": 0.8435362577438354,
      "num_tokens": 326888886.0,
      "step": 259
    },
    {
      "epoch": 0.22184300341296928,
      "grad_norm": 1.1484375,
      "learning_rate": 3.6315788413717396e-05,
      "loss": 0.48206108808517456,
      "mean_token_accuracy": 0.8364836573600769,
      "num_tokens": 328128505.0,
      "step": 260
    },
    {
      "epoch": 0.2226962457337884,
      "grad_norm": 1.25,
      "learning_rate": 3.628373773407149e-05,
      "loss": 0.4826292395591736,
      "mean_token_accuracy": 0.8350750207901001,
      "num_tokens": 329341016.0,
      "step": 261
    },
    {
      "epoch": 0.2235494880546075,
      "grad_norm": 1.25,
      "learning_rate": 3.625156251785423e-05,
      "loss": 0.4518211781978607,
      "mean_token_accuracy": 0.841881513595581,
      "num_tokens": 330574363.0,
      "step": 262
    },
    {
      "epoch": 0.22440273037542663,
      "grad_norm": 1.1171875,
      "learning_rate": 3.621926301113881e-05,
      "loss": 0.46055471897125244,
      "mean_token_accuracy": 0.8434260487556458,
      "num_tokens": 331847706.0,
      "step": 263
    },
    {
      "epoch": 0.22525597269624573,
      "grad_norm": 1.2421875,
      "learning_rate": 3.618683946094896e-05,
      "loss": 0.4637811779975891,
      "mean_token_accuracy": 0.8413702249526978,
      "num_tokens": 333106707.0,
      "step": 264
    },
    {
      "epoch": 0.22610921501706485,
      "grad_norm": 1.3203125,
      "learning_rate": 3.615429211525709e-05,
      "loss": 0.4287470579147339,
      "mean_token_accuracy": 0.8475349545478821,
      "num_tokens": 334403439.0,
      "step": 265
    },
    {
      "epoch": 0.22696245733788395,
      "grad_norm": 1.2578125,
      "learning_rate": 3.6121621222982407e-05,
      "loss": 0.44202107191085815,
      "mean_token_accuracy": 0.8473933935165405,
      "num_tokens": 335643985.0,
      "step": 266
    },
    {
      "epoch": 0.22781569965870307,
      "grad_norm": 1.21875,
      "learning_rate": 3.6088827033988965e-05,
      "loss": 0.4498806297779083,
      "mean_token_accuracy": 0.8424839973449707,
      "num_tokens": 336879210.0,
      "step": 267
    },
    {
      "epoch": 0.22866894197952217,
      "grad_norm": 1.15625,
      "learning_rate": 3.605590979908379e-05,
      "loss": 0.42951101064682007,
      "mean_token_accuracy": 0.8490846157073975,
      "num_tokens": 338095428.0,
      "step": 268
    },
    {
      "epoch": 0.2295221843003413,
      "grad_norm": 1.1640625,
      "learning_rate": 3.6022869770014964e-05,
      "loss": 0.4599571228027344,
      "mean_token_accuracy": 0.8424212336540222,
      "num_tokens": 339387072.0,
      "step": 269
    },
    {
      "epoch": 0.23037542662116042,
      "grad_norm": 1.3515625,
      "learning_rate": 3.598970719946967e-05,
      "loss": 0.5019035935401917,
      "mean_token_accuracy": 0.828871488571167,
      "num_tokens": 340643777.0,
      "step": 270
    },
    {
      "epoch": 0.23122866894197952,
      "grad_norm": 1.109375,
      "learning_rate": 3.595642234107229e-05,
      "loss": 0.4214792251586914,
      "mean_token_accuracy": 0.8564902544021606,
      "num_tokens": 341928481.0,
      "step": 271
    },
    {
      "epoch": 0.23208191126279865,
      "grad_norm": 1.09375,
      "learning_rate": 3.592301544938244e-05,
      "loss": 0.4565838575363159,
      "mean_token_accuracy": 0.8502973318099976,
      "num_tokens": 343231688.0,
      "step": 272
    },
    {
      "epoch": 0.23293515358361774,
      "grad_norm": 1.1640625,
      "learning_rate": 3.588948677989305e-05,
      "loss": 0.3928641676902771,
      "mean_token_accuracy": 0.8626409769058228,
      "num_tokens": 344502941.0,
      "step": 273
    },
    {
      "epoch": 0.23378839590443687,
      "grad_norm": 1.1640625,
      "learning_rate": 3.585583658902838e-05,
      "loss": 0.44674593210220337,
      "mean_token_accuracy": 0.8468226194381714,
      "num_tokens": 345694093.0,
      "step": 274
    },
    {
      "epoch": 0.23464163822525597,
      "grad_norm": 1.2734375,
      "learning_rate": 3.582206513414209e-05,
      "loss": 0.4619629383087158,
      "mean_token_accuracy": 0.8395665884017944,
      "num_tokens": 346972889.0,
      "step": 275
    },
    {
      "epoch": 0.2354948805460751,
      "grad_norm": 1.171875,
      "learning_rate": 3.5788172673515254e-05,
      "loss": 0.46798717975616455,
      "mean_token_accuracy": 0.8402957320213318,
      "num_tokens": 348272336.0,
      "step": 276
    },
    {
      "epoch": 0.2363481228668942,
      "grad_norm": 1.125,
      "learning_rate": 3.575415946635437e-05,
      "loss": 0.4196368455886841,
      "mean_token_accuracy": 0.8541288375854492,
      "num_tokens": 349488795.0,
      "step": 277
    },
    {
      "epoch": 0.23720136518771331,
      "grad_norm": 1.09375,
      "learning_rate": 3.572002577278942e-05,
      "loss": 0.46687060594558716,
      "mean_token_accuracy": 0.8367984294891357,
      "num_tokens": 350854671.0,
      "step": 278
    },
    {
      "epoch": 0.2380546075085324,
      "grad_norm": 1.6484375,
      "learning_rate": 3.5685771853871834e-05,
      "loss": 0.464155375957489,
      "mean_token_accuracy": 0.8402614593505859,
      "num_tokens": 352262712.0,
      "step": 279
    },
    {
      "epoch": 0.23890784982935154,
      "grad_norm": 1.203125,
      "learning_rate": 3.565139797157253e-05,
      "loss": 0.47204524278640747,
      "mean_token_accuracy": 0.8390317559242249,
      "num_tokens": 353568013.0,
      "step": 280
    },
    {
      "epoch": 0.23976109215017063,
      "grad_norm": 1.21875,
      "learning_rate": 3.561690438877988e-05,
      "loss": 0.4276565909385681,
      "mean_token_accuracy": 0.8475570678710938,
      "num_tokens": 354900607.0,
      "step": 281
    },
    {
      "epoch": 0.24061433447098976,
      "grad_norm": 1.3984375,
      "learning_rate": 3.558229136929773e-05,
      "loss": 0.42133304476737976,
      "mean_token_accuracy": 0.8527024984359741,
      "num_tokens": 356255192.0,
      "step": 282
    },
    {
      "epoch": 0.24146757679180889,
      "grad_norm": 1.4453125,
      "learning_rate": 3.554755917784336e-05,
      "loss": 0.43990516662597656,
      "mean_token_accuracy": 0.8482701778411865,
      "num_tokens": 357520082.0,
      "step": 283
    },
    {
      "epoch": 0.24232081911262798,
      "grad_norm": 1.2265625,
      "learning_rate": 3.551270808004547e-05,
      "loss": 0.4381936490535736,
      "mean_token_accuracy": 0.8482813835144043,
      "num_tokens": 358793845.0,
      "step": 284
    },
    {
      "epoch": 0.2431740614334471,
      "grad_norm": 1.1875,
      "learning_rate": 3.547773834244214e-05,
      "loss": 0.42456915974617004,
      "mean_token_accuracy": 0.8502007722854614,
      "num_tokens": 360065205.0,
      "step": 285
    },
    {
      "epoch": 0.2440273037542662,
      "grad_norm": 1.21875,
      "learning_rate": 3.54426502324788e-05,
      "loss": 0.47686582803726196,
      "mean_token_accuracy": 0.8377107977867126,
      "num_tokens": 361349238.0,
      "step": 286
    },
    {
      "epoch": 0.24488054607508533,
      "grad_norm": 1.1953125,
      "learning_rate": 3.540744401850617e-05,
      "loss": 0.440234512090683,
      "mean_token_accuracy": 0.8463053703308105,
      "num_tokens": 362662043.0,
      "step": 287
    },
    {
      "epoch": 0.24573378839590443,
      "grad_norm": 1.21875,
      "learning_rate": 3.5372119969778245e-05,
      "loss": 0.4448232054710388,
      "mean_token_accuracy": 0.8479359149932861,
      "num_tokens": 363917425.0,
      "step": 288
    },
    {
      "epoch": 0.24658703071672355,
      "grad_norm": 1.328125,
      "learning_rate": 3.533667835645018e-05,
      "loss": 0.48730313777923584,
      "mean_token_accuracy": 0.8319054841995239,
      "num_tokens": 365119644.0,
      "step": 289
    },
    {
      "epoch": 0.24744027303754265,
      "grad_norm": 1.4140625,
      "learning_rate": 3.5301119449576274e-05,
      "loss": 0.4646996557712555,
      "mean_token_accuracy": 0.8400120139122009,
      "num_tokens": 366373886.0,
      "step": 290
    },
    {
      "epoch": 0.24829351535836178,
      "grad_norm": 1.0625,
      "learning_rate": 3.526544352110787e-05,
      "loss": 0.39567241072654724,
      "mean_token_accuracy": 0.8562027812004089,
      "num_tokens": 367574845.0,
      "step": 291
    },
    {
      "epoch": 0.24914675767918087,
      "grad_norm": 1.1328125,
      "learning_rate": 3.5229650843891284e-05,
      "loss": 0.3733476400375366,
      "mean_token_accuracy": 0.866489052772522,
      "num_tokens": 368764276.0,
      "step": 292
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2578125,
      "learning_rate": 3.51937416916657e-05,
      "loss": 0.4641663432121277,
      "mean_token_accuracy": 0.8396214246749878,
      "num_tokens": 370067586.0,
      "step": 293
    },
    {
      "epoch": 0.2508532423208191,
      "grad_norm": 1.1875,
      "learning_rate": 3.515771633906113e-05,
      "loss": 0.4344157874584198,
      "mean_token_accuracy": 0.8552476167678833,
      "num_tokens": 371275154.0,
      "step": 294
    },
    {
      "epoch": 0.25170648464163825,
      "grad_norm": 1.4140625,
      "learning_rate": 3.512157506159622e-05,
      "loss": 0.44393110275268555,
      "mean_token_accuracy": 0.8471590280532837,
      "num_tokens": 372583518.0,
      "step": 295
    },
    {
      "epoch": 0.2525597269624573,
      "grad_norm": 1.203125,
      "learning_rate": 3.5085318135676264e-05,
      "loss": 0.4249088764190674,
      "mean_token_accuracy": 0.8490287065505981,
      "num_tokens": 373761307.0,
      "step": 296
    },
    {
      "epoch": 0.25341296928327645,
      "grad_norm": 2.953125,
      "learning_rate": 3.5048945838590975e-05,
      "loss": 0.45666760206222534,
      "mean_token_accuracy": 0.8451513051986694,
      "num_tokens": 375010492.0,
      "step": 297
    },
    {
      "epoch": 0.25426621160409557,
      "grad_norm": 1.2734375,
      "learning_rate": 3.501245844851243e-05,
      "loss": 0.43689775466918945,
      "mean_token_accuracy": 0.8481562733650208,
      "num_tokens": 376213512.0,
      "step": 298
    },
    {
      "epoch": 0.2551194539249147,
      "grad_norm": 1.140625,
      "learning_rate": 3.4975856244492934e-05,
      "loss": 0.41200748085975647,
      "mean_token_accuracy": 0.8541728258132935,
      "num_tokens": 377498492.0,
      "step": 299
    },
    {
      "epoch": 0.25597269624573377,
      "grad_norm": 3.3125,
      "learning_rate": 3.493913950646287e-05,
      "loss": 0.41792792081832886,
      "mean_token_accuracy": 0.8538678884506226,
      "num_tokens": 378781394.0,
      "step": 300
    },
    {
      "epoch": 0.2568259385665529,
      "grad_norm": 1.25,
      "learning_rate": 3.4902308515228574e-05,
      "loss": 0.4667350649833679,
      "mean_token_accuracy": 0.8435813188552856,
      "num_tokens": 379974957.0,
      "step": 301
    },
    {
      "epoch": 0.257679180887372,
      "grad_norm": 1.09375,
      "learning_rate": 3.486536355247017e-05,
      "loss": 0.44550564885139465,
      "mean_token_accuracy": 0.8433248996734619,
      "num_tokens": 381258482.0,
      "step": 302
    },
    {
      "epoch": 0.25853242320819114,
      "grad_norm": 1.2578125,
      "learning_rate": 3.482830490073944e-05,
      "loss": 0.46867746114730835,
      "mean_token_accuracy": 0.8424823880195618,
      "num_tokens": 382476162.0,
      "step": 303
    },
    {
      "epoch": 0.2593856655290102,
      "grad_norm": 1.203125,
      "learning_rate": 3.4791132843457635e-05,
      "loss": 0.46217218041419983,
      "mean_token_accuracy": 0.8405507802963257,
      "num_tokens": 383702468.0,
      "step": 304
    },
    {
      "epoch": 0.26023890784982934,
      "grad_norm": 1.359375,
      "learning_rate": 3.475384766491333e-05,
      "loss": 0.45192646980285645,
      "mean_token_accuracy": 0.8431363105773926,
      "num_tokens": 384946448.0,
      "step": 305
    },
    {
      "epoch": 0.26109215017064846,
      "grad_norm": 1.0859375,
      "learning_rate": 3.471644965026024e-05,
      "loss": 0.4512714147567749,
      "mean_token_accuracy": 0.8416941165924072,
      "num_tokens": 386335707.0,
      "step": 306
    },
    {
      "epoch": 0.2619453924914676,
      "grad_norm": 1.1796875,
      "learning_rate": 3.467893908551505e-05,
      "loss": 0.45232245326042175,
      "mean_token_accuracy": 0.8442192077636719,
      "num_tokens": 387492990.0,
      "step": 307
    },
    {
      "epoch": 0.2627986348122867,
      "grad_norm": 1.2265625,
      "learning_rate": 3.464131625755519e-05,
      "loss": 0.476895272731781,
      "mean_token_accuracy": 0.8325604200363159,
      "num_tokens": 388757100.0,
      "step": 308
    },
    {
      "epoch": 0.2636518771331058,
      "grad_norm": 1.34375,
      "learning_rate": 3.460358145411669e-05,
      "loss": 0.48612940311431885,
      "mean_token_accuracy": 0.8361055850982666,
      "num_tokens": 390055803.0,
      "step": 309
    },
    {
      "epoch": 0.2645051194539249,
      "grad_norm": 1.1953125,
      "learning_rate": 3.4565734963791966e-05,
      "loss": 0.42867401242256165,
      "mean_token_accuracy": 0.8513859510421753,
      "num_tokens": 391317010.0,
      "step": 310
    },
    {
      "epoch": 0.26535836177474403,
      "grad_norm": 1.2265625,
      "learning_rate": 3.452777707602758e-05,
      "loss": 0.46520161628723145,
      "mean_token_accuracy": 0.8390433192253113,
      "num_tokens": 392519521.0,
      "step": 311
    },
    {
      "epoch": 0.26621160409556316,
      "grad_norm": 1.359375,
      "learning_rate": 3.4489708081122085e-05,
      "loss": 0.43540531396865845,
      "mean_token_accuracy": 0.8531479835510254,
      "num_tokens": 393782521.0,
      "step": 312
    },
    {
      "epoch": 0.26706484641638223,
      "grad_norm": 1.0546875,
      "learning_rate": 3.4451528270223734e-05,
      "loss": 0.4294455051422119,
      "mean_token_accuracy": 0.8524512052536011,
      "num_tokens": 395023126.0,
      "step": 313
    },
    {
      "epoch": 0.26791808873720135,
      "grad_norm": 1.1484375,
      "learning_rate": 3.4413237935328325e-05,
      "loss": 0.40167486667633057,
      "mean_token_accuracy": 0.8594574332237244,
      "num_tokens": 396406730.0,
      "step": 314
    },
    {
      "epoch": 0.2687713310580205,
      "grad_norm": 1.265625,
      "learning_rate": 3.437483736927691e-05,
      "loss": 0.4789988398551941,
      "mean_token_accuracy": 0.8385370373725891,
      "num_tokens": 397775227.0,
      "step": 315
    },
    {
      "epoch": 0.2696245733788396,
      "grad_norm": 1.1953125,
      "learning_rate": 3.433632686575359e-05,
      "loss": 0.447489857673645,
      "mean_token_accuracy": 0.8466324806213379,
      "num_tokens": 398971286.0,
      "step": 316
    },
    {
      "epoch": 0.27047781569965873,
      "grad_norm": 1.21875,
      "learning_rate": 3.4297706719283254e-05,
      "loss": 0.4246326684951782,
      "mean_token_accuracy": 0.8504009246826172,
      "num_tokens": 400138450.0,
      "step": 317
    },
    {
      "epoch": 0.2713310580204778,
      "grad_norm": 1.09375,
      "learning_rate": 3.4258977225229336e-05,
      "loss": 0.41263896226882935,
      "mean_token_accuracy": 0.8499135375022888,
      "num_tokens": 401481833.0,
      "step": 318
    },
    {
      "epoch": 0.2721843003412969,
      "grad_norm": 1.4765625,
      "learning_rate": 3.422013867979153e-05,
      "loss": 0.45532429218292236,
      "mean_token_accuracy": 0.8439770936965942,
      "num_tokens": 402780579.0,
      "step": 319
    },
    {
      "epoch": 0.27303754266211605,
      "grad_norm": 1.1484375,
      "learning_rate": 3.418119138000358e-05,
      "loss": 0.46896207332611084,
      "mean_token_accuracy": 0.8380571603775024,
      "num_tokens": 404031411.0,
      "step": 320
    },
    {
      "epoch": 0.2738907849829352,
      "grad_norm": 1.5625,
      "learning_rate": 3.4142135623730954e-05,
      "loss": 0.4720747470855713,
      "mean_token_accuracy": 0.8377820253372192,
      "num_tokens": 405318271.0,
      "step": 321
    },
    {
      "epoch": 0.27474402730375425,
      "grad_norm": 1.1640625,
      "learning_rate": 3.410297170966858e-05,
      "loss": 0.47692856192588806,
      "mean_token_accuracy": 0.8381778001785278,
      "num_tokens": 406649268.0,
      "step": 322
    },
    {
      "epoch": 0.27559726962457337,
      "grad_norm": 1.3203125,
      "learning_rate": 3.406369993733857e-05,
      "loss": 0.4394349753856659,
      "mean_token_accuracy": 0.842740535736084,
      "num_tokens": 407938261.0,
      "step": 323
    },
    {
      "epoch": 0.2764505119453925,
      "grad_norm": 1.4375,
      "learning_rate": 3.402432060708794e-05,
      "loss": 0.47366341948509216,
      "mean_token_accuracy": 0.841123640537262,
      "num_tokens": 409161990.0,
      "step": 324
    },
    {
      "epoch": 0.2773037542662116,
      "grad_norm": 1.0078125,
      "learning_rate": 3.398483402008629e-05,
      "loss": 0.4106162190437317,
      "mean_token_accuracy": 0.8591721653938293,
      "num_tokens": 410472444.0,
      "step": 325
    },
    {
      "epoch": 0.2781569965870307,
      "grad_norm": 1.1484375,
      "learning_rate": 3.3945240478323494e-05,
      "loss": 0.4619808793067932,
      "mean_token_accuracy": 0.8393364548683167,
      "num_tokens": 411725605.0,
      "step": 326
    },
    {
      "epoch": 0.2790102389078498,
      "grad_norm": 1.3359375,
      "learning_rate": 3.390554028460744e-05,
      "loss": 0.4833545386791229,
      "mean_token_accuracy": 0.8394773602485657,
      "num_tokens": 412927437.0,
      "step": 327
    },
    {
      "epoch": 0.27986348122866894,
      "grad_norm": 1.140625,
      "learning_rate": 3.3865733742561645e-05,
      "loss": 0.46363359689712524,
      "mean_token_accuracy": 0.8411970138549805,
      "num_tokens": 414206955.0,
      "step": 328
    },
    {
      "epoch": 0.28071672354948807,
      "grad_norm": 1.0546875,
      "learning_rate": 3.382582115662301e-05,
      "loss": 0.412569522857666,
      "mean_token_accuracy": 0.8506094217300415,
      "num_tokens": 415481324.0,
      "step": 329
    },
    {
      "epoch": 0.2815699658703072,
      "grad_norm": 1.140625,
      "learning_rate": 3.3785802832039403e-05,
      "loss": 0.43265873193740845,
      "mean_token_accuracy": 0.8524988889694214,
      "num_tokens": 416655234.0,
      "step": 330
    },
    {
      "epoch": 0.28242320819112626,
      "grad_norm": 1.2421875,
      "learning_rate": 3.374567907486741e-05,
      "loss": 0.47492289543151855,
      "mean_token_accuracy": 0.8412267565727234,
      "num_tokens": 417929578.0,
      "step": 331
    },
    {
      "epoch": 0.2832764505119454,
      "grad_norm": 2.0625,
      "learning_rate": 3.370545019196993e-05,
      "loss": 0.4286075830459595,
      "mean_token_accuracy": 0.8511423468589783,
      "num_tokens": 419176431.0,
      "step": 332
    },
    {
      "epoch": 0.2841296928327645,
      "grad_norm": 1.203125,
      "learning_rate": 3.3665116491013874e-05,
      "loss": 0.46797239780426025,
      "mean_token_accuracy": 0.8431693911552429,
      "num_tokens": 420464310.0,
      "step": 333
    },
    {
      "epoch": 0.28498293515358364,
      "grad_norm": 1.1484375,
      "learning_rate": 3.362467828046778e-05,
      "loss": 0.42530399560928345,
      "mean_token_accuracy": 0.8505164384841919,
      "num_tokens": 421722422.0,
      "step": 334
    },
    {
      "epoch": 0.2858361774744027,
      "grad_norm": 1.078125,
      "learning_rate": 3.358413586959947e-05,
      "loss": 0.41209936141967773,
      "mean_token_accuracy": 0.8566812872886658,
      "num_tokens": 422964216.0,
      "step": 335
    },
    {
      "epoch": 0.28668941979522183,
      "grad_norm": 1.328125,
      "learning_rate": 3.354348956847368e-05,
      "loss": 0.4449922442436218,
      "mean_token_accuracy": 0.8470195531845093,
      "num_tokens": 424185077.0,
      "step": 336
    },
    {
      "epoch": 0.28754266211604096,
      "grad_norm": 1.109375,
      "learning_rate": 3.3502739687949684e-05,
      "loss": 0.44218355417251587,
      "mean_token_accuracy": 0.8492416143417358,
      "num_tokens": 425424075.0,
      "step": 337
    },
    {
      "epoch": 0.2883959044368601,
      "grad_norm": 1.3515625,
      "learning_rate": 3.346188653967894e-05,
      "loss": 0.4787616431713104,
      "mean_token_accuracy": 0.8371739983558655,
      "num_tokens": 426664754.0,
      "step": 338
    },
    {
      "epoch": 0.28924914675767915,
      "grad_norm": 1.2421875,
      "learning_rate": 3.3420930436102664e-05,
      "loss": 0.4461944103240967,
      "mean_token_accuracy": 0.8490476608276367,
      "num_tokens": 427880580.0,
      "step": 339
    },
    {
      "epoch": 0.2901023890784983,
      "grad_norm": 1.1484375,
      "learning_rate": 3.3379871690449484e-05,
      "loss": 0.4721885025501251,
      "mean_token_accuracy": 0.838394284248352,
      "num_tokens": 429233613.0,
      "step": 340
    },
    {
      "epoch": 0.2909556313993174,
      "grad_norm": 1.171875,
      "learning_rate": 3.333871061673302e-05,
      "loss": 0.45633018016815186,
      "mean_token_accuracy": 0.8459765911102295,
      "num_tokens": 430477234.0,
      "step": 341
    },
    {
      "epoch": 0.29180887372013653,
      "grad_norm": 1.2421875,
      "learning_rate": 3.3297447529749475e-05,
      "loss": 0.4454781115055084,
      "mean_token_accuracy": 0.8477740287780762,
      "num_tokens": 431718113.0,
      "step": 342
    },
    {
      "epoch": 0.29266211604095566,
      "grad_norm": 1.2578125,
      "learning_rate": 3.3256082745075265e-05,
      "loss": 0.4570143222808838,
      "mean_token_accuracy": 0.8434370756149292,
      "num_tokens": 433000271.0,
      "step": 343
    },
    {
      "epoch": 0.2935153583617747,
      "grad_norm": 1.1015625,
      "learning_rate": 3.321461657906456e-05,
      "loss": 0.46686089038848877,
      "mean_token_accuracy": 0.841106116771698,
      "num_tokens": 434256828.0,
      "step": 344
    },
    {
      "epoch": 0.29436860068259385,
      "grad_norm": 1.3046875,
      "learning_rate": 3.31730493488469e-05,
      "loss": 0.4115692973136902,
      "mean_token_accuracy": 0.8575655221939087,
      "num_tokens": 435472800.0,
      "step": 345
    },
    {
      "epoch": 0.295221843003413,
      "grad_norm": 1.203125,
      "learning_rate": 3.313138137232473e-05,
      "loss": 0.46459341049194336,
      "mean_token_accuracy": 0.8405579328536987,
      "num_tokens": 436702564.0,
      "step": 346
    },
    {
      "epoch": 0.2960750853242321,
      "grad_norm": 1.109375,
      "learning_rate": 3.3089612968171034e-05,
      "loss": 0.3943824768066406,
      "mean_token_accuracy": 0.8597428202629089,
      "num_tokens": 438000314.0,
      "step": 347
    },
    {
      "epoch": 0.29692832764505117,
      "grad_norm": 1.1171875,
      "learning_rate": 3.304774445582682e-05,
      "loss": 0.4218485653400421,
      "mean_token_accuracy": 0.8537439107894897,
      "num_tokens": 439224390.0,
      "step": 348
    },
    {
      "epoch": 0.2977815699658703,
      "grad_norm": 1.3359375,
      "learning_rate": 3.300577615549874e-05,
      "loss": 0.47125938534736633,
      "mean_token_accuracy": 0.8394074440002441,
      "num_tokens": 440521343.0,
      "step": 349
    },
    {
      "epoch": 0.2986348122866894,
      "grad_norm": 1.078125,
      "learning_rate": 3.29637083881566e-05,
      "loss": 0.41932082176208496,
      "mean_token_accuracy": 0.8504601120948792,
      "num_tokens": 441823594.0,
      "step": 350
    },
    {
      "epoch": 0.29948805460750855,
      "grad_norm": 1.1171875,
      "learning_rate": 3.292154147553093e-05,
      "loss": 0.4191831946372986,
      "mean_token_accuracy": 0.8503984808921814,
      "num_tokens": 443140664.0,
      "step": 351
    },
    {
      "epoch": 0.3003412969283277,
      "grad_norm": 1.1484375,
      "learning_rate": 3.28792757401105e-05,
      "loss": 0.4543292224407196,
      "mean_token_accuracy": 0.8443950414657593,
      "num_tokens": 444388762.0,
      "step": 352
    },
    {
      "epoch": 0.30119453924914674,
      "grad_norm": 1.15625,
      "learning_rate": 3.283691150513989e-05,
      "loss": 0.4762479066848755,
      "mean_token_accuracy": 0.8386726379394531,
      "num_tokens": 445726287.0,
      "step": 353
    },
    {
      "epoch": 0.30204778156996587,
      "grad_norm": 1.125,
      "learning_rate": 3.279444909461698e-05,
      "loss": 0.4377902150154114,
      "mean_token_accuracy": 0.8500233888626099,
      "num_tokens": 446976261.0,
      "step": 354
    },
    {
      "epoch": 0.302901023890785,
      "grad_norm": 1.15625,
      "learning_rate": 3.275188883329048e-05,
      "loss": 0.4394272565841675,
      "mean_token_accuracy": 0.8427177667617798,
      "num_tokens": 448308457.0,
      "step": 355
    },
    {
      "epoch": 0.3037542662116041,
      "grad_norm": 1.234375,
      "learning_rate": 3.27092310466575e-05,
      "loss": 0.429455041885376,
      "mean_token_accuracy": 0.8468566536903381,
      "num_tokens": 449491682.0,
      "step": 356
    },
    {
      "epoch": 0.3046075085324232,
      "grad_norm": 1.1875,
      "learning_rate": 3.2666476060960945e-05,
      "loss": 0.44441354274749756,
      "mean_token_accuracy": 0.8462467789649963,
      "num_tokens": 450792647.0,
      "step": 357
    },
    {
      "epoch": 0.3054607508532423,
      "grad_norm": 1.3515625,
      "learning_rate": 3.262362420318714e-05,
      "loss": 0.4636932611465454,
      "mean_token_accuracy": 0.8381350040435791,
      "num_tokens": 452047038.0,
      "step": 358
    },
    {
      "epoch": 0.30631399317406144,
      "grad_norm": 1.1015625,
      "learning_rate": 3.2580675801063266e-05,
      "loss": 0.4268888235092163,
      "mean_token_accuracy": 0.8523919582366943,
      "num_tokens": 453300098.0,
      "step": 359
    },
    {
      "epoch": 0.30716723549488056,
      "grad_norm": 1.0546875,
      "learning_rate": 3.253763118305486e-05,
      "loss": 0.4382615089416504,
      "mean_token_accuracy": 0.8465285301208496,
      "num_tokens": 454597892.0,
      "step": 360
    },
    {
      "epoch": 0.30802047781569963,
      "grad_norm": 1.125,
      "learning_rate": 3.249449067836332e-05,
      "loss": 0.4473419487476349,
      "mean_token_accuracy": 0.8432580232620239,
      "num_tokens": 455860968.0,
      "step": 361
    },
    {
      "epoch": 0.30887372013651876,
      "grad_norm": 1.2421875,
      "learning_rate": 3.2451254616923375e-05,
      "loss": 0.45598119497299194,
      "mean_token_accuracy": 0.8376116752624512,
      "num_tokens": 457133319.0,
      "step": 362
    },
    {
      "epoch": 0.3097269624573379,
      "grad_norm": 1.140625,
      "learning_rate": 3.240792332940055e-05,
      "loss": 0.4473876953125,
      "mean_token_accuracy": 0.8438167572021484,
      "num_tokens": 458330981.0,
      "step": 363
    },
    {
      "epoch": 0.310580204778157,
      "grad_norm": 1.1015625,
      "learning_rate": 3.236449714718867e-05,
      "loss": 0.40998417139053345,
      "mean_token_accuracy": 0.8593680262565613,
      "num_tokens": 459627454.0,
      "step": 364
    },
    {
      "epoch": 0.31143344709897613,
      "grad_norm": 1.1953125,
      "learning_rate": 3.2320976402407286e-05,
      "loss": 0.4328822195529938,
      "mean_token_accuracy": 0.8478354215621948,
      "num_tokens": 460818874.0,
      "step": 365
    },
    {
      "epoch": 0.3122866894197952,
      "grad_norm": 1.21875,
      "learning_rate": 3.227736142789918e-05,
      "loss": 0.4507167637348175,
      "mean_token_accuracy": 0.845084011554718,
      "num_tokens": 462092284.0,
      "step": 366
    },
    {
      "epoch": 0.31313993174061433,
      "grad_norm": 1.21875,
      "learning_rate": 3.2233652557227765e-05,
      "loss": 0.4690914750099182,
      "mean_token_accuracy": 0.8463398218154907,
      "num_tokens": 463288434.0,
      "step": 367
    },
    {
      "epoch": 0.31399317406143346,
      "grad_norm": 1.2265625,
      "learning_rate": 3.218985012467459e-05,
      "loss": 0.46059438586235046,
      "mean_token_accuracy": 0.843924880027771,
      "num_tokens": 464489529.0,
      "step": 368
    },
    {
      "epoch": 0.3148464163822526,
      "grad_norm": 1.296875,
      "learning_rate": 3.214595446523674e-05,
      "loss": 0.46837687492370605,
      "mean_token_accuracy": 0.8426238298416138,
      "num_tokens": 465751836.0,
      "step": 369
    },
    {
      "epoch": 0.31569965870307165,
      "grad_norm": 1.125,
      "learning_rate": 3.2101965914624296e-05,
      "loss": 0.43735551834106445,
      "mean_token_accuracy": 0.8507262468338013,
      "num_tokens": 466984418.0,
      "step": 370
    },
    {
      "epoch": 0.3165529010238908,
      "grad_norm": 1.109375,
      "learning_rate": 3.205788480925777e-05,
      "loss": 0.41705870628356934,
      "mean_token_accuracy": 0.8561105728149414,
      "num_tokens": 468159410.0,
      "step": 371
    },
    {
      "epoch": 0.3174061433447099,
      "grad_norm": 1.140625,
      "learning_rate": 3.201371148626551e-05,
      "loss": 0.40593114495277405,
      "mean_token_accuracy": 0.8527323603630066,
      "num_tokens": 469434855.0,
      "step": 372
    },
    {
      "epoch": 0.318259385665529,
      "grad_norm": 1.078125,
      "learning_rate": 3.1969446283481157e-05,
      "loss": 0.4448232054710388,
      "mean_token_accuracy": 0.8464887142181396,
      "num_tokens": 470625842.0,
      "step": 373
    },
    {
      "epoch": 0.3191126279863481,
      "grad_norm": 2.265625,
      "learning_rate": 3.1925089539441e-05,
      "loss": 0.43496066331863403,
      "mean_token_accuracy": 0.8485532999038696,
      "num_tokens": 471938374.0,
      "step": 374
    },
    {
      "epoch": 0.3199658703071672,
      "grad_norm": 1.2265625,
      "learning_rate": 3.1880641593381484e-05,
      "loss": 0.45213890075683594,
      "mean_token_accuracy": 0.8437613248825073,
      "num_tokens": 473240618.0,
      "step": 375
    },
    {
      "epoch": 0.32081911262798635,
      "grad_norm": 1.140625,
      "learning_rate": 3.183610278523651e-05,
      "loss": 0.4411635994911194,
      "mean_token_accuracy": 0.8481951355934143,
      "num_tokens": 474573179.0,
      "step": 376
    },
    {
      "epoch": 0.3216723549488055,
      "grad_norm": 1.125,
      "learning_rate": 3.1791473455634914e-05,
      "loss": 0.4447738528251648,
      "mean_token_accuracy": 0.8416231274604797,
      "num_tokens": 475805197.0,
      "step": 377
    },
    {
      "epoch": 0.3225255972696246,
      "grad_norm": 1.53125,
      "learning_rate": 3.174675394589781e-05,
      "loss": 0.47181689739227295,
      "mean_token_accuracy": 0.8374716639518738,
      "num_tokens": 477107450.0,
      "step": 378
    },
    {
      "epoch": 0.32337883959044367,
      "grad_norm": 1.3203125,
      "learning_rate": 3.170194459803603e-05,
      "loss": 0.470605731010437,
      "mean_token_accuracy": 0.8383257985115051,
      "num_tokens": 478406928.0,
      "step": 379
    },
    {
      "epoch": 0.3242320819112628,
      "grad_norm": 1.109375,
      "learning_rate": 3.1657045754747456e-05,
      "loss": 0.426152765750885,
      "mean_token_accuracy": 0.8510598540306091,
      "num_tokens": 479674349.0,
      "step": 380
    },
    {
      "epoch": 0.3250853242320819,
      "grad_norm": 1.1796875,
      "learning_rate": 3.161205775941442e-05,
      "loss": 0.41141754388809204,
      "mean_token_accuracy": 0.8533170819282532,
      "num_tokens": 480810031.0,
      "step": 381
    },
    {
      "epoch": 0.32593856655290104,
      "grad_norm": 1.171875,
      "learning_rate": 3.15669809561011e-05,
      "loss": 0.4639444947242737,
      "mean_token_accuracy": 0.8422175049781799,
      "num_tokens": 482072893.0,
      "step": 382
    },
    {
      "epoch": 0.3267918088737201,
      "grad_norm": 1.1875,
      "learning_rate": 3.1521815689550855e-05,
      "loss": 0.4670180678367615,
      "mean_token_accuracy": 0.8400779962539673,
      "num_tokens": 483342234.0,
      "step": 383
    },
    {
      "epoch": 0.32764505119453924,
      "grad_norm": 1.0078125,
      "learning_rate": 3.147656230518362e-05,
      "loss": 0.39968037605285645,
      "mean_token_accuracy": 0.8614175915718079,
      "num_tokens": 484713783.0,
      "step": 384
    },
    {
      "epoch": 0.32849829351535836,
      "grad_norm": 1.171875,
      "learning_rate": 3.1431221149093224e-05,
      "loss": 0.44792523980140686,
      "mean_token_accuracy": 0.841788649559021,
      "num_tokens": 485955361.0,
      "step": 385
    },
    {
      "epoch": 0.3293515358361775,
      "grad_norm": 1.109375,
      "learning_rate": 3.138579256804478e-05,
      "loss": 0.413746178150177,
      "mean_token_accuracy": 0.8508884906768799,
      "num_tokens": 487276352.0,
      "step": 386
    },
    {
      "epoch": 0.3302047781569966,
      "grad_norm": 1.609375,
      "learning_rate": 3.1340276909472025e-05,
      "loss": 0.45872998237609863,
      "mean_token_accuracy": 0.8411749601364136,
      "num_tokens": 488542595.0,
      "step": 387
    },
    {
      "epoch": 0.3310580204778157,
      "grad_norm": 1.0546875,
      "learning_rate": 3.129467452147464e-05,
      "loss": 0.41522181034088135,
      "mean_token_accuracy": 0.8544907569885254,
      "num_tokens": 489702327.0,
      "step": 388
    },
    {
      "epoch": 0.3319112627986348,
      "grad_norm": 1.0703125,
      "learning_rate": 3.124898575281562e-05,
      "loss": 0.4458087682723999,
      "mean_token_accuracy": 0.8471222519874573,
      "num_tokens": 490969834.0,
      "step": 389
    },
    {
      "epoch": 0.33276450511945393,
      "grad_norm": 1.0859375,
      "learning_rate": 3.120321095291858e-05,
      "loss": 0.45720988512039185,
      "mean_token_accuracy": 0.8431023955345154,
      "num_tokens": 492245961.0,
      "step": 390
    },
    {
      "epoch": 0.33361774744027306,
      "grad_norm": 1.09375,
      "learning_rate": 3.11573504718651e-05,
      "loss": 0.46156805753707886,
      "mean_token_accuracy": 0.8429160118103027,
      "num_tokens": 493491648.0,
      "step": 391
    },
    {
      "epoch": 0.33447098976109213,
      "grad_norm": 1.1640625,
      "learning_rate": 3.111140466039205e-05,
      "loss": 0.4351236820220947,
      "mean_token_accuracy": 0.8476270437240601,
      "num_tokens": 494757660.0,
      "step": 392
    },
    {
      "epoch": 0.33532423208191126,
      "grad_norm": 1.171875,
      "learning_rate": 3.106537386988888e-05,
      "loss": 0.47244900465011597,
      "mean_token_accuracy": 0.8355914354324341,
      "num_tokens": 496103193.0,
      "step": 393
    },
    {
      "epoch": 0.3361774744027304,
      "grad_norm": 1.265625,
      "learning_rate": 3.101925845239498e-05,
      "loss": 0.47562137246131897,
      "mean_token_accuracy": 0.8334184288978577,
      "num_tokens": 497358269.0,
      "step": 394
    },
    {
      "epoch": 0.3370307167235495,
      "grad_norm": 1.21875,
      "learning_rate": 3.097305876059692e-05,
      "loss": 0.48175063729286194,
      "mean_token_accuracy": 0.8395435810089111,
      "num_tokens": 498652135.0,
      "step": 395
    },
    {
      "epoch": 0.3378839590443686,
      "grad_norm": 1.046875,
      "learning_rate": 3.092677514782583e-05,
      "loss": 0.4007660746574402,
      "mean_token_accuracy": 0.8568627238273621,
      "num_tokens": 499916726.0,
      "step": 396
    },
    {
      "epoch": 0.3387372013651877,
      "grad_norm": 1.09375,
      "learning_rate": 3.0880407968054646e-05,
      "loss": 0.42011481523513794,
      "mean_token_accuracy": 0.8503057956695557,
      "num_tokens": 501280424.0,
      "step": 397
    },
    {
      "epoch": 0.3395904436860068,
      "grad_norm": 1.421875,
      "learning_rate": 3.0833957575895416e-05,
      "loss": 0.4481493830680847,
      "mean_token_accuracy": 0.8443958759307861,
      "num_tokens": 502591835.0,
      "step": 398
    },
    {
      "epoch": 0.34044368600682595,
      "grad_norm": 1.15625,
      "learning_rate": 3.078742432659658e-05,
      "loss": 0.43925219774246216,
      "mean_token_accuracy": 0.846462607383728,
      "num_tokens": 503818613.0,
      "step": 399
    },
    {
      "epoch": 0.3412969283276451,
      "grad_norm": 1.4296875,
      "learning_rate": 3.0740808576040284e-05,
      "loss": 0.436448872089386,
      "mean_token_accuracy": 0.8460103273391724,
      "num_tokens": 505146490.0,
      "step": 400
    },
    {
      "epoch": 0.34215017064846415,
      "grad_norm": 1.234375,
      "learning_rate": 3.069411068073961e-05,
      "loss": 0.4347485303878784,
      "mean_token_accuracy": 0.8483883142471313,
      "num_tokens": 506487977.0,
      "step": 401
    },
    {
      "epoch": 0.3430034129692833,
      "grad_norm": 1.140625,
      "learning_rate": 3.06473309978359e-05,
      "loss": 0.436759889125824,
      "mean_token_accuracy": 0.8509647846221924,
      "num_tokens": 507838784.0,
      "step": 402
    },
    {
      "epoch": 0.3438566552901024,
      "grad_norm": 1.25,
      "learning_rate": 3.060046988509598e-05,
      "loss": 0.4016042947769165,
      "mean_token_accuracy": 0.8542234301567078,
      "num_tokens": 509059889.0,
      "step": 403
    },
    {
      "epoch": 0.3447098976109215,
      "grad_norm": 1.2109375,
      "learning_rate": 3.055352770090947e-05,
      "loss": 0.40167349576950073,
      "mean_token_accuracy": 0.8564907312393188,
      "num_tokens": 510382682.0,
      "step": 404
    },
    {
      "epoch": 0.3455631399317406,
      "grad_norm": 1.171875,
      "learning_rate": 3.0506504804285977e-05,
      "loss": 0.41943079233169556,
      "mean_token_accuracy": 0.855238676071167,
      "num_tokens": 511718120.0,
      "step": 405
    },
    {
      "epoch": 0.3464163822525597,
      "grad_norm": 1.2109375,
      "learning_rate": 3.0459401554852433e-05,
      "loss": 0.4303492307662964,
      "mean_token_accuracy": 0.8491234183311462,
      "num_tokens": 513003455.0,
      "step": 406
    },
    {
      "epoch": 0.34726962457337884,
      "grad_norm": 1.15625,
      "learning_rate": 3.0412218312850273e-05,
      "loss": 0.4734012484550476,
      "mean_token_accuracy": 0.8372538089752197,
      "num_tokens": 514223611.0,
      "step": 407
    },
    {
      "epoch": 0.34812286689419797,
      "grad_norm": 1.140625,
      "learning_rate": 3.036495543913272e-05,
      "loss": 0.4220488667488098,
      "mean_token_accuracy": 0.8502969145774841,
      "num_tokens": 515455106.0,
      "step": 408
    },
    {
      "epoch": 0.34897610921501704,
      "grad_norm": 1.0234375,
      "learning_rate": 3.0317613295162e-05,
      "loss": 0.4037281274795532,
      "mean_token_accuracy": 0.8589484691619873,
      "num_tokens": 516729516.0,
      "step": 409
    },
    {
      "epoch": 0.34982935153583616,
      "grad_norm": 1.140625,
      "learning_rate": 3.0270192243006616e-05,
      "loss": 0.446707546710968,
      "mean_token_accuracy": 0.845456600189209,
      "num_tokens": 517994574.0,
      "step": 410
    },
    {
      "epoch": 0.3506825938566553,
      "grad_norm": 0.96875,
      "learning_rate": 3.0222692645338525e-05,
      "loss": 0.39641469717025757,
      "mean_token_accuracy": 0.8604815006256104,
      "num_tokens": 519241220.0,
      "step": 411
    },
    {
      "epoch": 0.3515358361774744,
      "grad_norm": 1.0703125,
      "learning_rate": 3.017511486543041e-05,
      "loss": 0.4017910063266754,
      "mean_token_accuracy": 0.8559160232543945,
      "num_tokens": 520538922.0,
      "step": 412
    },
    {
      "epoch": 0.35238907849829354,
      "grad_norm": 1.1796875,
      "learning_rate": 3.012745926715288e-05,
      "loss": 0.4549269676208496,
      "mean_token_accuracy": 0.8460143804550171,
      "num_tokens": 521855766.0,
      "step": 413
    },
    {
      "epoch": 0.3532423208191126,
      "grad_norm": 1.5625,
      "learning_rate": 3.00797262149717e-05,
      "loss": 0.4478074014186859,
      "mean_token_accuracy": 0.8448442220687866,
      "num_tokens": 523132335.0,
      "step": 414
    },
    {
      "epoch": 0.35409556313993173,
      "grad_norm": 1.171875,
      "learning_rate": 3.0031916073944977e-05,
      "loss": 0.4623050391674042,
      "mean_token_accuracy": 0.8414711952209473,
      "num_tokens": 524430615.0,
      "step": 415
    },
    {
      "epoch": 0.35494880546075086,
      "grad_norm": 1.1171875,
      "learning_rate": 2.9984029209720407e-05,
      "loss": 0.39808952808380127,
      "mean_token_accuracy": 0.8584749698638916,
      "num_tokens": 525689382.0,
      "step": 416
    },
    {
      "epoch": 0.35580204778157,
      "grad_norm": 1.1875,
      "learning_rate": 2.9936065988532448e-05,
      "loss": 0.40475302934646606,
      "mean_token_accuracy": 0.8547696471214294,
      "num_tokens": 526869059.0,
      "step": 417
    },
    {
      "epoch": 0.35665529010238906,
      "grad_norm": 1.1796875,
      "learning_rate": 2.9888026777199526e-05,
      "loss": 0.4643518924713135,
      "mean_token_accuracy": 0.843845009803772,
      "num_tokens": 528119584.0,
      "step": 418
    },
    {
      "epoch": 0.3575085324232082,
      "grad_norm": 1.1171875,
      "learning_rate": 2.9839911943121244e-05,
      "loss": 0.43393051624298096,
      "mean_token_accuracy": 0.8489704132080078,
      "num_tokens": 529340561.0,
      "step": 419
    },
    {
      "epoch": 0.3583617747440273,
      "grad_norm": 1.25,
      "learning_rate": 2.9791721854275545e-05,
      "loss": 0.4566136598587036,
      "mean_token_accuracy": 0.84145587682724,
      "num_tokens": 530633524.0,
      "step": 420
    },
    {
      "epoch": 0.35921501706484643,
      "grad_norm": 1.0625,
      "learning_rate": 2.9743456879215934e-05,
      "loss": 0.46025416254997253,
      "mean_token_accuracy": 0.8431336879730225,
      "num_tokens": 531866874.0,
      "step": 421
    },
    {
      "epoch": 0.36006825938566556,
      "grad_norm": 1.0078125,
      "learning_rate": 2.9695117387068618e-05,
      "loss": 0.4101315140724182,
      "mean_token_accuracy": 0.8555788993835449,
      "num_tokens": 533141323.0,
      "step": 422
    },
    {
      "epoch": 0.3609215017064846,
      "grad_norm": 1.265625,
      "learning_rate": 2.9646703747529724e-05,
      "loss": 0.4559863805770874,
      "mean_token_accuracy": 0.8406989574432373,
      "num_tokens": 534485976.0,
      "step": 423
    },
    {
      "epoch": 0.36177474402730375,
      "grad_norm": 1.0859375,
      "learning_rate": 2.959821633086243e-05,
      "loss": 0.43788015842437744,
      "mean_token_accuracy": 0.8512685298919678,
      "num_tokens": 535794838.0,
      "step": 424
    },
    {
      "epoch": 0.3626279863481229,
      "grad_norm": 1.203125,
      "learning_rate": 2.9549655507894175e-05,
      "loss": 0.4572112560272217,
      "mean_token_accuracy": 0.8401914834976196,
      "num_tokens": 536976745.0,
      "step": 425
    },
    {
      "epoch": 0.363481228668942,
      "grad_norm": 1.109375,
      "learning_rate": 2.9501021650013797e-05,
      "loss": 0.4440622329711914,
      "mean_token_accuracy": 0.8460497260093689,
      "num_tokens": 538254115.0,
      "step": 426
    },
    {
      "epoch": 0.3643344709897611,
      "grad_norm": 1.3046875,
      "learning_rate": 2.945231512916869e-05,
      "loss": 0.4307566285133362,
      "mean_token_accuracy": 0.8482721447944641,
      "num_tokens": 539597655.0,
      "step": 427
    },
    {
      "epoch": 0.3651877133105802,
      "grad_norm": 1.1796875,
      "learning_rate": 2.9403536317861974e-05,
      "loss": 0.4430895447731018,
      "mean_token_accuracy": 0.8473089337348938,
      "num_tokens": 540832924.0,
      "step": 428
    },
    {
      "epoch": 0.3660409556313993,
      "grad_norm": 1.1640625,
      "learning_rate": 2.9354685589149637e-05,
      "loss": 0.45633232593536377,
      "mean_token_accuracy": 0.842922568321228,
      "num_tokens": 542091219.0,
      "step": 429
    },
    {
      "epoch": 0.36689419795221845,
      "grad_norm": 1.140625,
      "learning_rate": 2.93057633166377e-05,
      "loss": 0.4362283945083618,
      "mean_token_accuracy": 0.8469471335411072,
      "num_tokens": 543383802.0,
      "step": 430
    },
    {
      "epoch": 0.3677474402730375,
      "grad_norm": 1.125,
      "learning_rate": 2.9256769874479318e-05,
      "loss": 0.44247686862945557,
      "mean_token_accuracy": 0.84629887342453,
      "num_tokens": 544692782.0,
      "step": 431
    },
    {
      "epoch": 0.36860068259385664,
      "grad_norm": 1.1328125,
      "learning_rate": 2.920770563737197e-05,
      "loss": 0.43867504596710205,
      "mean_token_accuracy": 0.8485032320022583,
      "num_tokens": 545996916.0,
      "step": 432
    },
    {
      "epoch": 0.36945392491467577,
      "grad_norm": 1.1328125,
      "learning_rate": 2.915857098055455e-05,
      "loss": 0.4663504958152771,
      "mean_token_accuracy": 0.8379421830177307,
      "num_tokens": 547274829.0,
      "step": 433
    },
    {
      "epoch": 0.3703071672354949,
      "grad_norm": 1.6484375,
      "learning_rate": 2.9109366279804523e-05,
      "loss": 0.4782983958721161,
      "mean_token_accuracy": 0.8370530605316162,
      "num_tokens": 548575672.0,
      "step": 434
    },
    {
      "epoch": 0.371160409556314,
      "grad_norm": 1.2421875,
      "learning_rate": 2.9060091911435056e-05,
      "loss": 0.42879682779312134,
      "mean_token_accuracy": 0.8486077189445496,
      "num_tokens": 549850906.0,
      "step": 435
    },
    {
      "epoch": 0.3720136518771331,
      "grad_norm": 1.234375,
      "learning_rate": 2.9010748252292105e-05,
      "loss": 0.40358465909957886,
      "mean_token_accuracy": 0.8579407334327698,
      "num_tokens": 551127388.0,
      "step": 436
    },
    {
      "epoch": 0.3728668941979522,
      "grad_norm": 1.1015625,
      "learning_rate": 2.8961335679751573e-05,
      "loss": 0.447015643119812,
      "mean_token_accuracy": 0.8473055362701416,
      "num_tokens": 552443024.0,
      "step": 437
    },
    {
      "epoch": 0.37372013651877134,
      "grad_norm": 1.0078125,
      "learning_rate": 2.891185457171639e-05,
      "loss": 0.3943895399570465,
      "mean_token_accuracy": 0.86151522397995,
      "num_tokens": 553706648.0,
      "step": 438
    },
    {
      "epoch": 0.37457337883959047,
      "grad_norm": 1.046875,
      "learning_rate": 2.8862305306613665e-05,
      "loss": 0.4165075123310089,
      "mean_token_accuracy": 0.8554148077964783,
      "num_tokens": 554935909.0,
      "step": 439
    },
    {
      "epoch": 0.37542662116040953,
      "grad_norm": 1.0390625,
      "learning_rate": 2.8812688263391736e-05,
      "loss": 0.41519755125045776,
      "mean_token_accuracy": 0.8514286279678345,
      "num_tokens": 556239018.0,
      "step": 440
    },
    {
      "epoch": 0.37627986348122866,
      "grad_norm": 1.0625,
      "learning_rate": 2.8763003821517324e-05,
      "loss": 0.437913715839386,
      "mean_token_accuracy": 0.8508543968200684,
      "num_tokens": 557543945.0,
      "step": 441
    },
    {
      "epoch": 0.3771331058020478,
      "grad_norm": 1.0625,
      "learning_rate": 2.8713252360972598e-05,
      "loss": 0.40308839082717896,
      "mean_token_accuracy": 0.8582577109336853,
      "num_tokens": 558802855.0,
      "step": 442
    },
    {
      "epoch": 0.3779863481228669,
      "grad_norm": 1.140625,
      "learning_rate": 2.8663434262252284e-05,
      "loss": 0.43042272329330444,
      "mean_token_accuracy": 0.8500890135765076,
      "num_tokens": 560043258.0,
      "step": 443
    },
    {
      "epoch": 0.378839590443686,
      "grad_norm": 1.0,
      "learning_rate": 2.861354990636076e-05,
      "loss": 0.4142778217792511,
      "mean_token_accuracy": 0.8572806119918823,
      "num_tokens": 561315042.0,
      "step": 444
    },
    {
      "epoch": 0.3796928327645051,
      "grad_norm": 1.1796875,
      "learning_rate": 2.8563599674809105e-05,
      "loss": 0.45605820417404175,
      "mean_token_accuracy": 0.843693733215332,
      "num_tokens": 562549362.0,
      "step": 445
    },
    {
      "epoch": 0.38054607508532423,
      "grad_norm": 1.171875,
      "learning_rate": 2.8513583949612243e-05,
      "loss": 0.4716776907444,
      "mean_token_accuracy": 0.842058539390564,
      "num_tokens": 563827705.0,
      "step": 446
    },
    {
      "epoch": 0.38139931740614336,
      "grad_norm": 1.1875,
      "learning_rate": 2.8463503113285962e-05,
      "loss": 0.4476182460784912,
      "mean_token_accuracy": 0.8462197780609131,
      "num_tokens": 565172140.0,
      "step": 447
    },
    {
      "epoch": 0.3822525597269625,
      "grad_norm": 1.21875,
      "learning_rate": 2.841335754884403e-05,
      "loss": 0.41882842779159546,
      "mean_token_accuracy": 0.8537755012512207,
      "num_tokens": 566348206.0,
      "step": 448
    },
    {
      "epoch": 0.38310580204778155,
      "grad_norm": 1.109375,
      "learning_rate": 2.8363147639795235e-05,
      "loss": 0.41909196972846985,
      "mean_token_accuracy": 0.8506964445114136,
      "num_tokens": 567627113.0,
      "step": 449
    },
    {
      "epoch": 0.3839590443686007,
      "grad_norm": 1.1953125,
      "learning_rate": 2.8312873770140486e-05,
      "loss": 0.4459937810897827,
      "mean_token_accuracy": 0.8440756797790527,
      "num_tokens": 568914101.0,
      "step": 450
    },
    {
      "epoch": 0.3848122866894198,
      "grad_norm": 1.0703125,
      "learning_rate": 2.8262536324369842e-05,
      "loss": 0.44791150093078613,
      "mean_token_accuracy": 0.8464034199714661,
      "num_tokens": 570120130.0,
      "step": 451
    },
    {
      "epoch": 0.3856655290102389,
      "grad_norm": 1.1953125,
      "learning_rate": 2.8212135687459593e-05,
      "loss": 0.44444260001182556,
      "mean_token_accuracy": 0.8447258472442627,
      "num_tokens": 571432040.0,
      "step": 452
    },
    {
      "epoch": 0.386518771331058,
      "grad_norm": 1.0859375,
      "learning_rate": 2.81616722448693e-05,
      "loss": 0.4268878996372223,
      "mean_token_accuracy": 0.8525646924972534,
      "num_tokens": 572665584.0,
      "step": 453
    },
    {
      "epoch": 0.3873720136518771,
      "grad_norm": 1.1875,
      "learning_rate": 2.811114638253887e-05,
      "loss": 0.4648447334766388,
      "mean_token_accuracy": 0.8437522649765015,
      "num_tokens": 573808884.0,
      "step": 454
    },
    {
      "epoch": 0.38822525597269625,
      "grad_norm": 1.265625,
      "learning_rate": 2.8060558486885586e-05,
      "loss": 0.4978809058666229,
      "mean_token_accuracy": 0.828589916229248,
      "num_tokens": 575147918.0,
      "step": 455
    },
    {
      "epoch": 0.3890784982935154,
      "grad_norm": 1.1640625,
      "learning_rate": 2.800990894480115e-05,
      "loss": 0.42109155654907227,
      "mean_token_accuracy": 0.8551697134971619,
      "num_tokens": 576414849.0,
      "step": 456
    },
    {
      "epoch": 0.38993174061433444,
      "grad_norm": 1.1015625,
      "learning_rate": 2.7959198143648727e-05,
      "loss": 0.42847752571105957,
      "mean_token_accuracy": 0.8479151725769043,
      "num_tokens": 577625258.0,
      "step": 457
    },
    {
      "epoch": 0.39078498293515357,
      "grad_norm": 0.98046875,
      "learning_rate": 2.790842647126001e-05,
      "loss": 0.402197003364563,
      "mean_token_accuracy": 0.8549535870552063,
      "num_tokens": 578927142.0,
      "step": 458
    },
    {
      "epoch": 0.3916382252559727,
      "grad_norm": 1.09375,
      "learning_rate": 2.7857594315932196e-05,
      "loss": 0.43270331621170044,
      "mean_token_accuracy": 0.8466709852218628,
      "num_tokens": 580175423.0,
      "step": 459
    },
    {
      "epoch": 0.3924914675767918,
      "grad_norm": 1.078125,
      "learning_rate": 2.7806702066425087e-05,
      "loss": 0.44362926483154297,
      "mean_token_accuracy": 0.8444927930831909,
      "num_tokens": 581508751.0,
      "step": 460
    },
    {
      "epoch": 0.39334470989761094,
      "grad_norm": 1.109375,
      "learning_rate": 2.7755750111958043e-05,
      "loss": 0.4227468967437744,
      "mean_token_accuracy": 0.8500931262969971,
      "num_tokens": 582825922.0,
      "step": 461
    },
    {
      "epoch": 0.39419795221843,
      "grad_norm": 1.3515625,
      "learning_rate": 2.7704738842207075e-05,
      "loss": 0.4500751197338104,
      "mean_token_accuracy": 0.8451435565948486,
      "num_tokens": 584078659.0,
      "step": 462
    },
    {
      "epoch": 0.39505119453924914,
      "grad_norm": 1.0546875,
      "learning_rate": 2.7653668647301797e-05,
      "loss": 0.39167481660842896,
      "mean_token_accuracy": 0.8579052686691284,
      "num_tokens": 585292796.0,
      "step": 463
    },
    {
      "epoch": 0.39590443686006827,
      "grad_norm": 1.15625,
      "learning_rate": 2.760253991782251e-05,
      "loss": 0.41245365142822266,
      "mean_token_accuracy": 0.8564767241477966,
      "num_tokens": 586470921.0,
      "step": 464
    },
    {
      "epoch": 0.3967576791808874,
      "grad_norm": 1.1796875,
      "learning_rate": 2.7551353044797174e-05,
      "loss": 0.44880110025405884,
      "mean_token_accuracy": 0.8438332080841064,
      "num_tokens": 587744106.0,
      "step": 465
    },
    {
      "epoch": 0.39761092150170646,
      "grad_norm": 1.0703125,
      "learning_rate": 2.7500108419698408e-05,
      "loss": 0.4056968092918396,
      "mean_token_accuracy": 0.8615964651107788,
      "num_tokens": 588955437.0,
      "step": 466
    },
    {
      "epoch": 0.3984641638225256,
      "grad_norm": 1.0703125,
      "learning_rate": 2.7448806434440536e-05,
      "loss": 0.39632952213287354,
      "mean_token_accuracy": 0.8584248423576355,
      "num_tokens": 590190357.0,
      "step": 467
    },
    {
      "epoch": 0.3993174061433447,
      "grad_norm": 1.0,
      "learning_rate": 2.739744748137655e-05,
      "loss": 0.37714987993240356,
      "mean_token_accuracy": 0.8645815849304199,
      "num_tokens": 591440262.0,
      "step": 468
    },
    {
      "epoch": 0.40017064846416384,
      "grad_norm": 1.09375,
      "learning_rate": 2.734603195329514e-05,
      "loss": 0.4424920082092285,
      "mean_token_accuracy": 0.8466888666152954,
      "num_tokens": 592712739.0,
      "step": 469
    },
    {
      "epoch": 0.40102389078498296,
      "grad_norm": 1.1953125,
      "learning_rate": 2.729456024341767e-05,
      "loss": 0.4342699646949768,
      "mean_token_accuracy": 0.8513492941856384,
      "num_tokens": 593984043.0,
      "step": 470
    },
    {
      "epoch": 0.40187713310580203,
      "grad_norm": 1.140625,
      "learning_rate": 2.7243032745395174e-05,
      "loss": 0.4428744614124298,
      "mean_token_accuracy": 0.8454432487487793,
      "num_tokens": 595273312.0,
      "step": 471
    },
    {
      "epoch": 0.40273037542662116,
      "grad_norm": 1.015625,
      "learning_rate": 2.7191449853305358e-05,
      "loss": 0.3863757848739624,
      "mean_token_accuracy": 0.8646922707557678,
      "num_tokens": 596578058.0,
      "step": 472
    },
    {
      "epoch": 0.4035836177474403,
      "grad_norm": 1.109375,
      "learning_rate": 2.7139811961649577e-05,
      "loss": 0.4301474988460541,
      "mean_token_accuracy": 0.8499286770820618,
      "num_tokens": 597806833.0,
      "step": 473
    },
    {
      "epoch": 0.4044368600682594,
      "grad_norm": 1.265625,
      "learning_rate": 2.70881194653498e-05,
      "loss": 0.4132445454597473,
      "mean_token_accuracy": 0.8563423156738281,
      "num_tokens": 598967960.0,
      "step": 474
    },
    {
      "epoch": 0.4052901023890785,
      "grad_norm": 1.59375,
      "learning_rate": 2.703637275974563e-05,
      "loss": 0.43037357926368713,
      "mean_token_accuracy": 0.8484891653060913,
      "num_tokens": 600269858.0,
      "step": 475
    },
    {
      "epoch": 0.4061433447098976,
      "grad_norm": 1.15625,
      "learning_rate": 2.6984572240591248e-05,
      "loss": 0.3975830674171448,
      "mean_token_accuracy": 0.8553655743598938,
      "num_tokens": 601563531.0,
      "step": 476
    },
    {
      "epoch": 0.4069965870307167,
      "grad_norm": 1.3359375,
      "learning_rate": 2.6932718304052393e-05,
      "loss": 0.4108714163303375,
      "mean_token_accuracy": 0.859004020690918,
      "num_tokens": 602819102.0,
      "step": 477
    },
    {
      "epoch": 0.40784982935153585,
      "grad_norm": 3.0,
      "learning_rate": 2.6880811346703337e-05,
      "loss": 0.41669097542762756,
      "mean_token_accuracy": 0.8538253307342529,
      "num_tokens": 603999556.0,
      "step": 478
    },
    {
      "epoch": 0.4087030716723549,
      "grad_norm": 1.203125,
      "learning_rate": 2.682885176552386e-05,
      "loss": 0.4466935992240906,
      "mean_token_accuracy": 0.8470003008842468,
      "num_tokens": 605167423.0,
      "step": 479
    },
    {
      "epoch": 0.40955631399317405,
      "grad_norm": 1.140625,
      "learning_rate": 2.6776839957896185e-05,
      "loss": 0.4358293414115906,
      "mean_token_accuracy": 0.8486195802688599,
      "num_tokens": 606395556.0,
      "step": 480
    },
    {
      "epoch": 0.4104095563139932,
      "grad_norm": 1.15625,
      "learning_rate": 2.6724776321601984e-05,
      "loss": 0.4211139678955078,
      "mean_token_accuracy": 0.851761519908905,
      "num_tokens": 607693924.0,
      "step": 481
    },
    {
      "epoch": 0.4112627986348123,
      "grad_norm": 1.296875,
      "learning_rate": 2.6672661254819284e-05,
      "loss": 0.4738181233406067,
      "mean_token_accuracy": 0.8353275060653687,
      "num_tokens": 608970902.0,
      "step": 482
    },
    {
      "epoch": 0.4121160409556314,
      "grad_norm": 1.1484375,
      "learning_rate": 2.6620495156119468e-05,
      "loss": 0.4174370765686035,
      "mean_token_accuracy": 0.8550584316253662,
      "num_tokens": 610232422.0,
      "step": 483
    },
    {
      "epoch": 0.4129692832764505,
      "grad_norm": 1.2421875,
      "learning_rate": 2.656827842446419e-05,
      "loss": 0.46832579374313354,
      "mean_token_accuracy": 0.8372833728790283,
      "num_tokens": 611451846.0,
      "step": 484
    },
    {
      "epoch": 0.4138225255972696,
      "grad_norm": 1.2421875,
      "learning_rate": 2.6516011459202356e-05,
      "loss": 0.4629147946834564,
      "mean_token_accuracy": 0.8414827585220337,
      "num_tokens": 612708902.0,
      "step": 485
    },
    {
      "epoch": 0.41467576791808874,
      "grad_norm": 1.0859375,
      "learning_rate": 2.6463694660067042e-05,
      "loss": 0.404660701751709,
      "mean_token_accuracy": 0.8569943904876709,
      "num_tokens": 614019361.0,
      "step": 486
    },
    {
      "epoch": 0.41552901023890787,
      "grad_norm": 0.98046875,
      "learning_rate": 2.6411328427172453e-05,
      "loss": 0.37041449546813965,
      "mean_token_accuracy": 0.8675882816314697,
      "num_tokens": 615312886.0,
      "step": 487
    },
    {
      "epoch": 0.41638225255972694,
      "grad_norm": 1.34375,
      "learning_rate": 2.6358913161010865e-05,
      "loss": 0.4414120614528656,
      "mean_token_accuracy": 0.8444743752479553,
      "num_tokens": 616569137.0,
      "step": 488
    },
    {
      "epoch": 0.41723549488054607,
      "grad_norm": 1.15625,
      "learning_rate": 2.6306449262449544e-05,
      "loss": 0.4539835453033447,
      "mean_token_accuracy": 0.8416935205459595,
      "num_tokens": 617890583.0,
      "step": 489
    },
    {
      "epoch": 0.4180887372013652,
      "grad_norm": 1.109375,
      "learning_rate": 2.6253937132727705e-05,
      "loss": 0.42433610558509827,
      "mean_token_accuracy": 0.8531235456466675,
      "num_tokens": 619147957.0,
      "step": 490
    },
    {
      "epoch": 0.4189419795221843,
      "grad_norm": 1.1796875,
      "learning_rate": 2.620137717345342e-05,
      "loss": 0.43854761123657227,
      "mean_token_accuracy": 0.8476085662841797,
      "num_tokens": 620445191.0,
      "step": 491
    },
    {
      "epoch": 0.4197952218430034,
      "grad_norm": 1.0234375,
      "learning_rate": 2.614876978660057e-05,
      "loss": 0.4274616837501526,
      "mean_token_accuracy": 0.854659914970398,
      "num_tokens": 621711964.0,
      "step": 492
    },
    {
      "epoch": 0.4206484641638225,
      "grad_norm": 1.40625,
      "learning_rate": 2.609611537450573e-05,
      "loss": 0.43805333971977234,
      "mean_token_accuracy": 0.8476394414901733,
      "num_tokens": 622948952.0,
      "step": 493
    },
    {
      "epoch": 0.42150170648464164,
      "grad_norm": 1.0859375,
      "learning_rate": 2.604341433986517e-05,
      "loss": 0.458620548248291,
      "mean_token_accuracy": 0.8395786285400391,
      "num_tokens": 624178281.0,
      "step": 494
    },
    {
      "epoch": 0.42235494880546076,
      "grad_norm": 1.15625,
      "learning_rate": 2.5990667085731662e-05,
      "loss": 0.442231148481369,
      "mean_token_accuracy": 0.8464713096618652,
      "num_tokens": 625439812.0,
      "step": 495
    },
    {
      "epoch": 0.4232081911262799,
      "grad_norm": 1.1875,
      "learning_rate": 2.593787401551151e-05,
      "loss": 0.46863991022109985,
      "mean_token_accuracy": 0.839333176612854,
      "num_tokens": 626702313.0,
      "step": 496
    },
    {
      "epoch": 0.42406143344709896,
      "grad_norm": 1.015625,
      "learning_rate": 2.5885035532961404e-05,
      "loss": 0.4399535059928894,
      "mean_token_accuracy": 0.8453229069709778,
      "num_tokens": 627900553.0,
      "step": 497
    },
    {
      "epoch": 0.4249146757679181,
      "grad_norm": 1.1171875,
      "learning_rate": 2.5832152042185322e-05,
      "loss": 0.4482826292514801,
      "mean_token_accuracy": 0.8447152376174927,
      "num_tokens": 629240421.0,
      "step": 498
    },
    {
      "epoch": 0.4257679180887372,
      "grad_norm": 1.0546875,
      "learning_rate": 2.577922394763149e-05,
      "loss": 0.41260039806365967,
      "mean_token_accuracy": 0.8538705110549927,
      "num_tokens": 630596485.0,
      "step": 499
    },
    {
      "epoch": 0.42662116040955633,
      "grad_norm": 1.1640625,
      "learning_rate": 2.5726251654089237e-05,
      "loss": 0.44060420989990234,
      "mean_token_accuracy": 0.8438305854797363,
      "num_tokens": 631845206.0,
      "step": 500
    },
    {
      "epoch": 0.4274744027303754,
      "grad_norm": 1.125,
      "learning_rate": 2.5673235566685935e-05,
      "loss": 0.4429669976234436,
      "mean_token_accuracy": 0.8433709144592285,
      "num_tokens": 633067915.0,
      "step": 501
    },
    {
      "epoch": 0.4283276450511945,
      "grad_norm": 1.15625,
      "learning_rate": 2.5620176090883888e-05,
      "loss": 0.43634021282196045,
      "mean_token_accuracy": 0.848242998123169,
      "num_tokens": 634274445.0,
      "step": 502
    },
    {
      "epoch": 0.42918088737201365,
      "grad_norm": 1.1484375,
      "learning_rate": 2.556707363247722e-05,
      "loss": 0.4042015075683594,
      "mean_token_accuracy": 0.8564639091491699,
      "num_tokens": 635536078.0,
      "step": 503
    },
    {
      "epoch": 0.4300341296928328,
      "grad_norm": 1.4765625,
      "learning_rate": 2.551392859758879e-05,
      "loss": 0.4341173768043518,
      "mean_token_accuracy": 0.8485914468765259,
      "num_tokens": 636810931.0,
      "step": 504
    },
    {
      "epoch": 0.4308873720136519,
      "grad_norm": 1.234375,
      "learning_rate": 2.5460741392667077e-05,
      "loss": 0.4423428177833557,
      "mean_token_accuracy": 0.845649242401123,
      "num_tokens": 638010249.0,
      "step": 505
    },
    {
      "epoch": 0.431740614334471,
      "grad_norm": 1.1640625,
      "learning_rate": 2.540751242448308e-05,
      "loss": 0.44387924671173096,
      "mean_token_accuracy": 0.8456257581710815,
      "num_tokens": 639259300.0,
      "step": 506
    },
    {
      "epoch": 0.4325938566552901,
      "grad_norm": 1.09375,
      "learning_rate": 2.5354242100127175e-05,
      "loss": 0.40269935131073,
      "mean_token_accuracy": 0.8553915023803711,
      "num_tokens": 640603521.0,
      "step": 507
    },
    {
      "epoch": 0.4334470989761092,
      "grad_norm": 1.234375,
      "learning_rate": 2.5300930827006065e-05,
      "loss": 0.4233725070953369,
      "mean_token_accuracy": 0.8519324064254761,
      "num_tokens": 641801091.0,
      "step": 508
    },
    {
      "epoch": 0.43430034129692835,
      "grad_norm": 1.2109375,
      "learning_rate": 2.5247579012839584e-05,
      "loss": 0.42693066596984863,
      "mean_token_accuracy": 0.8510384559631348,
      "num_tokens": 642992995.0,
      "step": 509
    },
    {
      "epoch": 0.4351535836177474,
      "grad_norm": 0.984375,
      "learning_rate": 2.5194187065657652e-05,
      "loss": 0.40046393871307373,
      "mean_token_accuracy": 0.8601737022399902,
      "num_tokens": 644297348.0,
      "step": 510
    },
    {
      "epoch": 0.43600682593856654,
      "grad_norm": 1.125,
      "learning_rate": 2.5140755393797105e-05,
      "loss": 0.4545167088508606,
      "mean_token_accuracy": 0.8444639444351196,
      "num_tokens": 645553287.0,
      "step": 511
    },
    {
      "epoch": 0.43686006825938567,
      "grad_norm": 1.1640625,
      "learning_rate": 2.5087284405898594e-05,
      "loss": 0.427469402551651,
      "mean_token_accuracy": 0.8488271236419678,
      "num_tokens": 646789451.0,
      "step": 512
    },
    {
      "epoch": 0.4377133105802048,
      "grad_norm": 1.1640625,
      "learning_rate": 2.503377451090347e-05,
      "loss": 0.4326542317867279,
      "mean_token_accuracy": 0.848865270614624,
      "num_tokens": 648019256.0,
      "step": 513
    },
    {
      "epoch": 0.43856655290102387,
      "grad_norm": 1.1640625,
      "learning_rate": 2.4980226118050616e-05,
      "loss": 0.45150989294052124,
      "mean_token_accuracy": 0.8423042297363281,
      "num_tokens": 649267207.0,
      "step": 514
    },
    {
      "epoch": 0.439419795221843,
      "grad_norm": 1.0859375,
      "learning_rate": 2.4926639636873363e-05,
      "loss": 0.4263092279434204,
      "mean_token_accuracy": 0.8500568270683289,
      "num_tokens": 650565995.0,
      "step": 515
    },
    {
      "epoch": 0.4402730375426621,
      "grad_norm": 1.140625,
      "learning_rate": 2.4873015477196324e-05,
      "loss": 0.436440110206604,
      "mean_token_accuracy": 0.8467062711715698,
      "num_tokens": 651853143.0,
      "step": 516
    },
    {
      "epoch": 0.44112627986348124,
      "grad_norm": 1.3046875,
      "learning_rate": 2.481935404913229e-05,
      "loss": 0.42966410517692566,
      "mean_token_accuracy": 0.8503332734107971,
      "num_tokens": 653148366.0,
      "step": 517
    },
    {
      "epoch": 0.44197952218430037,
      "grad_norm": 1.0703125,
      "learning_rate": 2.4765655763079063e-05,
      "loss": 0.4635200500488281,
      "mean_token_accuracy": 0.841242253780365,
      "num_tokens": 654435167.0,
      "step": 518
    },
    {
      "epoch": 0.44283276450511944,
      "grad_norm": 1.078125,
      "learning_rate": 2.4711921029716342e-05,
      "loss": 0.4216068387031555,
      "mean_token_accuracy": 0.8490701913833618,
      "num_tokens": 655723669.0,
      "step": 519
    },
    {
      "epoch": 0.44368600682593856,
      "grad_norm": 1.1875,
      "learning_rate": 2.465815026000256e-05,
      "loss": 0.43572860956192017,
      "mean_token_accuracy": 0.8476413488388062,
      "num_tokens": 657010065.0,
      "step": 520
    },
    {
      "epoch": 0.4445392491467577,
      "grad_norm": 1.1171875,
      "learning_rate": 2.460434386517177e-05,
      "loss": 0.4369256794452667,
      "mean_token_accuracy": 0.8476170301437378,
      "num_tokens": 658234081.0,
      "step": 521
    },
    {
      "epoch": 0.4453924914675768,
      "grad_norm": 1.21875,
      "learning_rate": 2.455050225673046e-05,
      "loss": 0.4450273811817169,
      "mean_token_accuracy": 0.8473008275032043,
      "num_tokens": 659501745.0,
      "step": 522
    },
    {
      "epoch": 0.4462457337883959,
      "grad_norm": 1.1328125,
      "learning_rate": 2.4496625846454448e-05,
      "loss": 0.4302975535392761,
      "mean_token_accuracy": 0.8479271531105042,
      "num_tokens": 660798789.0,
      "step": 523
    },
    {
      "epoch": 0.447098976109215,
      "grad_norm": 1.1484375,
      "learning_rate": 2.4442715046385703e-05,
      "loss": 0.41797730326652527,
      "mean_token_accuracy": 0.8505008816719055,
      "num_tokens": 662090041.0,
      "step": 524
    },
    {
      "epoch": 0.44795221843003413,
      "grad_norm": 1.5390625,
      "learning_rate": 2.4388770268829217e-05,
      "loss": 0.42019855976104736,
      "mean_token_accuracy": 0.8510664701461792,
      "num_tokens": 663421206.0,
      "step": 525
    },
    {
      "epoch": 0.44880546075085326,
      "grad_norm": 1.1015625,
      "learning_rate": 2.4334791926349812e-05,
      "loss": 0.4029104709625244,
      "mean_token_accuracy": 0.858415961265564,
      "num_tokens": 664677877.0,
      "step": 526
    },
    {
      "epoch": 0.4496587030716723,
      "grad_norm": 1.0234375,
      "learning_rate": 2.4280780431769046e-05,
      "loss": 0.4028194844722748,
      "mean_token_accuracy": 0.856376051902771,
      "num_tokens": 665994653.0,
      "step": 527
    },
    {
      "epoch": 0.45051194539249145,
      "grad_norm": 1.1640625,
      "learning_rate": 2.4226736198161994e-05,
      "loss": 0.446869432926178,
      "mean_token_accuracy": 0.842737078666687,
      "num_tokens": 667238166.0,
      "step": 528
    },
    {
      "epoch": 0.4513651877133106,
      "grad_norm": 1.03125,
      "learning_rate": 2.417265963885413e-05,
      "loss": 0.42114901542663574,
      "mean_token_accuracy": 0.8574761748313904,
      "num_tokens": 668409282.0,
      "step": 529
    },
    {
      "epoch": 0.4522184300341297,
      "grad_norm": 1.1015625,
      "learning_rate": 2.4118551167418148e-05,
      "loss": 0.42626696825027466,
      "mean_token_accuracy": 0.8492233753204346,
      "num_tokens": 669688238.0,
      "step": 530
    },
    {
      "epoch": 0.45307167235494883,
      "grad_norm": 1.0078125,
      "learning_rate": 2.40644111976708e-05,
      "loss": 0.39950352907180786,
      "mean_token_accuracy": 0.8607754707336426,
      "num_tokens": 670877415.0,
      "step": 531
    },
    {
      "epoch": 0.4539249146757679,
      "grad_norm": 1.125,
      "learning_rate": 2.401024014366974e-05,
      "loss": 0.4736454486846924,
      "mean_token_accuracy": 0.8440489172935486,
      "num_tokens": 672136390.0,
      "step": 532
    },
    {
      "epoch": 0.454778156996587,
      "grad_norm": 1.2421875,
      "learning_rate": 2.3956038419710348e-05,
      "loss": 0.4395982623100281,
      "mean_token_accuracy": 0.8492919206619263,
      "num_tokens": 673376289.0,
      "step": 533
    },
    {
      "epoch": 0.45563139931740615,
      "grad_norm": 1.1171875,
      "learning_rate": 2.390180644032257e-05,
      "loss": 0.3990391492843628,
      "mean_token_accuracy": 0.8561760187149048,
      "num_tokens": 674579659.0,
      "step": 534
    },
    {
      "epoch": 0.4564846416382253,
      "grad_norm": 1.1484375,
      "learning_rate": 2.384754462026773e-05,
      "loss": 0.41264504194259644,
      "mean_token_accuracy": 0.8490070700645447,
      "num_tokens": 675889682.0,
      "step": 535
    },
    {
      "epoch": 0.45733788395904434,
      "grad_norm": 1.1796875,
      "learning_rate": 2.3793253374535397e-05,
      "loss": 0.45611372590065,
      "mean_token_accuracy": 0.8432552814483643,
      "num_tokens": 677131932.0,
      "step": 536
    },
    {
      "epoch": 0.45819112627986347,
      "grad_norm": 1.1875,
      "learning_rate": 2.3738933118340146e-05,
      "loss": 0.4510513246059418,
      "mean_token_accuracy": 0.8447238802909851,
      "num_tokens": 678471558.0,
      "step": 537
    },
    {
      "epoch": 0.4590443686006826,
      "grad_norm": 1.1484375,
      "learning_rate": 2.368458426711846e-05,
      "loss": 0.4205324947834015,
      "mean_token_accuracy": 0.8514668941497803,
      "num_tokens": 679753032.0,
      "step": 538
    },
    {
      "epoch": 0.4598976109215017,
      "grad_norm": 1.09375,
      "learning_rate": 2.3630207236525492e-05,
      "loss": 0.40956348180770874,
      "mean_token_accuracy": 0.8553588390350342,
      "num_tokens": 681087417.0,
      "step": 539
    },
    {
      "epoch": 0.46075085324232085,
      "grad_norm": 1.1015625,
      "learning_rate": 2.3575802442431913e-05,
      "loss": 0.4314265847206116,
      "mean_token_accuracy": 0.8470524549484253,
      "num_tokens": 682357695.0,
      "step": 540
    },
    {
      "epoch": 0.4616040955631399,
      "grad_norm": 1.1015625,
      "learning_rate": 2.3521370300920726e-05,
      "loss": 0.3984050452709198,
      "mean_token_accuracy": 0.8621013164520264,
      "num_tokens": 683717193.0,
      "step": 541
    },
    {
      "epoch": 0.46245733788395904,
      "grad_norm": 1.1484375,
      "learning_rate": 2.3466911228284094e-05,
      "loss": 0.3952445685863495,
      "mean_token_accuracy": 0.8577967882156372,
      "num_tokens": 685070114.0,
      "step": 542
    },
    {
      "epoch": 0.46331058020477817,
      "grad_norm": 1.1953125,
      "learning_rate": 2.3412425641020123e-05,
      "loss": 0.4872712790966034,
      "mean_token_accuracy": 0.8367127180099487,
      "num_tokens": 686365732.0,
      "step": 543
    },
    {
      "epoch": 0.4641638225255973,
      "grad_norm": 1.09375,
      "learning_rate": 2.3357913955829736e-05,
      "loss": 0.41109561920166016,
      "mean_token_accuracy": 0.8511577844619751,
      "num_tokens": 687696017.0,
      "step": 544
    },
    {
      "epoch": 0.46501706484641636,
      "grad_norm": 1.2421875,
      "learning_rate": 2.3303376589613413e-05,
      "loss": 0.41507503390312195,
      "mean_token_accuracy": 0.8545355796813965,
      "num_tokens": 689006415.0,
      "step": 545
    },
    {
      "epoch": 0.4658703071672355,
      "grad_norm": 1.0625,
      "learning_rate": 2.3248813959468076e-05,
      "loss": 0.42896437644958496,
      "mean_token_accuracy": 0.84931480884552,
      "num_tokens": 690350447.0,
      "step": 546
    },
    {
      "epoch": 0.4667235494880546,
      "grad_norm": 1.1796875,
      "learning_rate": 2.319422648268383e-05,
      "loss": 0.45196855068206787,
      "mean_token_accuracy": 0.8422336578369141,
      "num_tokens": 691600215.0,
      "step": 547
    },
    {
      "epoch": 0.46757679180887374,
      "grad_norm": 1.078125,
      "learning_rate": 2.3139614576740834e-05,
      "loss": 0.38236483931541443,
      "mean_token_accuracy": 0.8636423349380493,
      "num_tokens": 692800347.0,
      "step": 548
    },
    {
      "epoch": 0.4684300341296928,
      "grad_norm": 1.2265625,
      "learning_rate": 2.3084978659306048e-05,
      "loss": 0.4078746438026428,
      "mean_token_accuracy": 0.857507050037384,
      "num_tokens": 694005212.0,
      "step": 549
    },
    {
      "epoch": 0.46928327645051193,
      "grad_norm": 1.1171875,
      "learning_rate": 2.3030319148230106e-05,
      "loss": 0.47553297877311707,
      "mean_token_accuracy": 0.8339481353759766,
      "num_tokens": 695282388.0,
      "step": 550
    },
    {
      "epoch": 0.47013651877133106,
      "grad_norm": 1.0390625,
      "learning_rate": 2.297563646154405e-05,
      "loss": 0.38933587074279785,
      "mean_token_accuracy": 0.8610739707946777,
      "num_tokens": 696573899.0,
      "step": 551
    },
    {
      "epoch": 0.4709897610921502,
      "grad_norm": 1.171875,
      "learning_rate": 2.2920931017456195e-05,
      "loss": 0.45291998982429504,
      "mean_token_accuracy": 0.8422815203666687,
      "num_tokens": 697771569.0,
      "step": 552
    },
    {
      "epoch": 0.4718430034129693,
      "grad_norm": 1.046875,
      "learning_rate": 2.286620323434888e-05,
      "loss": 0.38468024134635925,
      "mean_token_accuracy": 0.8639845848083496,
      "num_tokens": 699068133.0,
      "step": 553
    },
    {
      "epoch": 0.4726962457337884,
      "grad_norm": 1.2265625,
      "learning_rate": 2.281145353077531e-05,
      "loss": 0.4618445634841919,
      "mean_token_accuracy": 0.8397313356399536,
      "num_tokens": 700339240.0,
      "step": 554
    },
    {
      "epoch": 0.4735494880546075,
      "grad_norm": 1.0625,
      "learning_rate": 2.2756682325456316e-05,
      "loss": 0.4045029878616333,
      "mean_token_accuracy": 0.854323148727417,
      "num_tokens": 701533321.0,
      "step": 555
    },
    {
      "epoch": 0.47440273037542663,
      "grad_norm": 1.265625,
      "learning_rate": 2.270189003727719e-05,
      "loss": 0.4212821424007416,
      "mean_token_accuracy": 0.850151777267456,
      "num_tokens": 702777584.0,
      "step": 556
    },
    {
      "epoch": 0.47525597269624575,
      "grad_norm": 1.0703125,
      "learning_rate": 2.2647077085284453e-05,
      "loss": 0.4083683490753174,
      "mean_token_accuracy": 0.8562377691268921,
      "num_tokens": 704019908.0,
      "step": 557
    },
    {
      "epoch": 0.4761092150170648,
      "grad_norm": 1.265625,
      "learning_rate": 2.2592243888682663e-05,
      "loss": 0.43752285838127136,
      "mean_token_accuracy": 0.8504473567008972,
      "num_tokens": 705256083.0,
      "step": 558
    },
    {
      "epoch": 0.47696245733788395,
      "grad_norm": 2.375,
      "learning_rate": 2.2537390866831216e-05,
      "loss": 0.4114428162574768,
      "mean_token_accuracy": 0.8523138761520386,
      "num_tokens": 706421291.0,
      "step": 559
    },
    {
      "epoch": 0.4778156996587031,
      "grad_norm": 1.1015625,
      "learning_rate": 2.2482518439241118e-05,
      "loss": 0.42962080240249634,
      "mean_token_accuracy": 0.8544087409973145,
      "num_tokens": 707782449.0,
      "step": 560
    },
    {
      "epoch": 0.4786689419795222,
      "grad_norm": 1.125,
      "learning_rate": 2.2427627025571795e-05,
      "loss": 0.4353313446044922,
      "mean_token_accuracy": 0.8487526178359985,
      "num_tokens": 709080080.0,
      "step": 561
    },
    {
      "epoch": 0.47952218430034127,
      "grad_norm": 1.125,
      "learning_rate": 2.237271704562787e-05,
      "loss": 0.4453912675380707,
      "mean_token_accuracy": 0.8455314040184021,
      "num_tokens": 710318186.0,
      "step": 562
    },
    {
      "epoch": 0.4803754266211604,
      "grad_norm": 1.1796875,
      "learning_rate": 2.2317788919355967e-05,
      "loss": 0.4446667432785034,
      "mean_token_accuracy": 0.8491466641426086,
      "num_tokens": 711595902.0,
      "step": 563
    },
    {
      "epoch": 0.4812286689419795,
      "grad_norm": 1.171875,
      "learning_rate": 2.226284306684148e-05,
      "loss": 0.4418070316314697,
      "mean_token_accuracy": 0.8459998369216919,
      "num_tokens": 712787556.0,
      "step": 564
    },
    {
      "epoch": 0.48208191126279865,
      "grad_norm": 1.0546875,
      "learning_rate": 2.22078799083054e-05,
      "loss": 0.4194298982620239,
      "mean_token_accuracy": 0.8513416051864624,
      "num_tokens": 714034982.0,
      "step": 565
    },
    {
      "epoch": 0.48293515358361777,
      "grad_norm": 1.0859375,
      "learning_rate": 2.2152899864101034e-05,
      "loss": 0.41106969118118286,
      "mean_token_accuracy": 0.8552191853523254,
      "num_tokens": 715291056.0,
      "step": 566
    },
    {
      "epoch": 0.48378839590443684,
      "grad_norm": 1.1328125,
      "learning_rate": 2.209790335471085e-05,
      "loss": 0.38512974977493286,
      "mean_token_accuracy": 0.868931770324707,
      "num_tokens": 716514234.0,
      "step": 567
    },
    {
      "epoch": 0.48464163822525597,
      "grad_norm": 1.2109375,
      "learning_rate": 2.2042890800743252e-05,
      "loss": 0.440070778131485,
      "mean_token_accuracy": 0.8477826118469238,
      "num_tokens": 717768994.0,
      "step": 568
    },
    {
      "epoch": 0.4854948805460751,
      "grad_norm": 1.15625,
      "learning_rate": 2.198786262292932e-05,
      "loss": 0.3991817235946655,
      "mean_token_accuracy": 0.8579760789871216,
      "num_tokens": 719061801.0,
      "step": 569
    },
    {
      "epoch": 0.4863481228668942,
      "grad_norm": 1.4921875,
      "learning_rate": 2.1932819242119655e-05,
      "loss": 0.45343518257141113,
      "mean_token_accuracy": 0.8409927487373352,
      "num_tokens": 720359704.0,
      "step": 570
    },
    {
      "epoch": 0.4872013651877133,
      "grad_norm": 1.5859375,
      "learning_rate": 2.187776107928111e-05,
      "loss": 0.4231145977973938,
      "mean_token_accuracy": 0.852569580078125,
      "num_tokens": 721624804.0,
      "step": 571
    },
    {
      "epoch": 0.4880546075085324,
      "grad_norm": 1.1875,
      "learning_rate": 2.1822688555493596e-05,
      "loss": 0.4735916256904602,
      "mean_token_accuracy": 0.8356332778930664,
      "num_tokens": 722878455.0,
      "step": 572
    },
    {
      "epoch": 0.48890784982935154,
      "grad_norm": 1.0625,
      "learning_rate": 2.176760209194685e-05,
      "loss": 0.44428515434265137,
      "mean_token_accuracy": 0.8493218421936035,
      "num_tokens": 724187652.0,
      "step": 573
    },
    {
      "epoch": 0.48976109215017066,
      "grad_norm": 1.09375,
      "learning_rate": 2.171250210993723e-05,
      "loss": 0.4318031966686249,
      "mean_token_accuracy": 0.8505599498748779,
      "num_tokens": 725451202.0,
      "step": 574
    },
    {
      "epoch": 0.4906143344709898,
      "grad_norm": 1.109375,
      "learning_rate": 2.1657389030864478e-05,
      "loss": 0.456437349319458,
      "mean_token_accuracy": 0.8432179093360901,
      "num_tokens": 726697181.0,
      "step": 575
    },
    {
      "epoch": 0.49146757679180886,
      "grad_norm": 1.21875,
      "learning_rate": 2.1602263276228493e-05,
      "loss": 0.4419264793395996,
      "mean_token_accuracy": 0.8494125604629517,
      "num_tokens": 728014567.0,
      "step": 576
    },
    {
      "epoch": 0.492320819112628,
      "grad_norm": 1.28125,
      "learning_rate": 2.1547125267626122e-05,
      "loss": 0.45526617765426636,
      "mean_token_accuracy": 0.8454446792602539,
      "num_tokens": 729237000.0,
      "step": 577
    },
    {
      "epoch": 0.4931740614334471,
      "grad_norm": 1.328125,
      "learning_rate": 2.1491975426747923e-05,
      "loss": 0.40408092737197876,
      "mean_token_accuracy": 0.8550388813018799,
      "num_tokens": 730613254.0,
      "step": 578
    },
    {
      "epoch": 0.49402730375426623,
      "grad_norm": 1.0546875,
      "learning_rate": 2.1436814175374973e-05,
      "loss": 0.41774892807006836,
      "mean_token_accuracy": 0.8501543998718262,
      "num_tokens": 731949117.0,
      "step": 579
    },
    {
      "epoch": 0.4948805460750853,
      "grad_norm": 1.09375,
      "learning_rate": 2.1381641935375564e-05,
      "loss": 0.45763707160949707,
      "mean_token_accuracy": 0.844201922416687,
      "num_tokens": 733215332.0,
      "step": 580
    },
    {
      "epoch": 0.49573378839590443,
      "grad_norm": 1.0859375,
      "learning_rate": 2.132645912870208e-05,
      "loss": 0.3960700035095215,
      "mean_token_accuracy": 0.8583261966705322,
      "num_tokens": 734408025.0,
      "step": 581
    },
    {
      "epoch": 0.49658703071672355,
      "grad_norm": 1.0546875,
      "learning_rate": 2.1271266177387687e-05,
      "loss": 0.41157662868499756,
      "mean_token_accuracy": 0.8542569875717163,
      "num_tokens": 735750297.0,
      "step": 582
    },
    {
      "epoch": 0.4974402730375427,
      "grad_norm": 1.125,
      "learning_rate": 2.1216063503543154e-05,
      "loss": 0.4445127844810486,
      "mean_token_accuracy": 0.8431657552719116,
      "num_tokens": 737002756.0,
      "step": 583
    },
    {
      "epoch": 0.49829351535836175,
      "grad_norm": 1.0703125,
      "learning_rate": 2.116085152935359e-05,
      "loss": 0.44030505418777466,
      "mean_token_accuracy": 0.8462070226669312,
      "num_tokens": 738322539.0,
      "step": 584
    },
    {
      "epoch": 0.4991467576791809,
      "grad_norm": 1.2265625,
      "learning_rate": 2.1105630677075243e-05,
      "loss": 0.41604071855545044,
      "mean_token_accuracy": 0.8539321422576904,
      "num_tokens": 739574224.0,
      "step": 585
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.1484375,
      "learning_rate": 2.1050401369032262e-05,
      "loss": 0.44255733489990234,
      "mean_token_accuracy": 0.8446137309074402,
      "num_tokens": 740777836.0,
      "step": 586
    },
    {
      "epoch": 0.5008532423208191,
      "grad_norm": 1.0703125,
      "learning_rate": 2.0995164027613447e-05,
      "loss": 0.40306901931762695,
      "mean_token_accuracy": 0.8550693988800049,
      "num_tokens": 742042608.0,
      "step": 587
    },
    {
      "epoch": 0.5017064846416383,
      "grad_norm": 1.1328125,
      "learning_rate": 2.093991907526907e-05,
      "loss": 0.4629879593849182,
      "mean_token_accuracy": 0.8397067785263062,
      "num_tokens": 743332410.0,
      "step": 588
    },
    {
      "epoch": 0.5025597269624573,
      "grad_norm": 1.125,
      "learning_rate": 2.088466693450758e-05,
      "loss": 0.4537632167339325,
      "mean_token_accuracy": 0.8416218757629395,
      "num_tokens": 744673768.0,
      "step": 589
    },
    {
      "epoch": 0.5034129692832765,
      "grad_norm": 1.1015625,
      "learning_rate": 2.082940802789241e-05,
      "loss": 0.39999040961265564,
      "mean_token_accuracy": 0.8609871864318848,
      "num_tokens": 745875215.0,
      "step": 590
    },
    {
      "epoch": 0.5042662116040956,
      "grad_norm": 1.0546875,
      "learning_rate": 2.0774142778038742e-05,
      "loss": 0.4068806767463684,
      "mean_token_accuracy": 0.8563341498374939,
      "num_tokens": 747190883.0,
      "step": 591
    },
    {
      "epoch": 0.5051194539249146,
      "grad_norm": 1.1796875,
      "learning_rate": 2.071887160761028e-05,
      "loss": 0.4329739809036255,
      "mean_token_accuracy": 0.8509830236434937,
      "num_tokens": 748526829.0,
      "step": 592
    },
    {
      "epoch": 0.5059726962457338,
      "grad_norm": 1.1171875,
      "learning_rate": 2.0663594939315984e-05,
      "loss": 0.4518081545829773,
      "mean_token_accuracy": 0.8447908163070679,
      "num_tokens": 749878121.0,
      "step": 593
    },
    {
      "epoch": 0.5068259385665529,
      "grad_norm": 1.1015625,
      "learning_rate": 2.0608313195906885e-05,
      "loss": 0.3930965065956116,
      "mean_token_accuracy": 0.862498939037323,
      "num_tokens": 751065334.0,
      "step": 594
    },
    {
      "epoch": 0.507679180887372,
      "grad_norm": 1.1328125,
      "learning_rate": 2.0553026800172807e-05,
      "loss": 0.416619211435318,
      "mean_token_accuracy": 0.8533003330230713,
      "num_tokens": 752321862.0,
      "step": 595
    },
    {
      "epoch": 0.5085324232081911,
      "grad_norm": 1.1328125,
      "learning_rate": 2.0497736174939172e-05,
      "loss": 0.433879017829895,
      "mean_token_accuracy": 0.8493025302886963,
      "num_tokens": 753598616.0,
      "step": 596
    },
    {
      "epoch": 0.5093856655290102,
      "grad_norm": 1.1796875,
      "learning_rate": 2.044244174306374e-05,
      "loss": 0.42270320653915405,
      "mean_token_accuracy": 0.8514422178268433,
      "num_tokens": 754837274.0,
      "step": 597
    },
    {
      "epoch": 0.5102389078498294,
      "grad_norm": 1.09375,
      "learning_rate": 2.038714392743339e-05,
      "loss": 0.4209945797920227,
      "mean_token_accuracy": 0.8533947467803955,
      "num_tokens": 756106048.0,
      "step": 598
    },
    {
      "epoch": 0.5110921501706485,
      "grad_norm": 0.95703125,
      "learning_rate": 2.0331843150960866e-05,
      "loss": 0.377909779548645,
      "mean_token_accuracy": 0.866572916507721,
      "num_tokens": 757330738.0,
      "step": 599
    },
    {
      "epoch": 0.5119453924914675,
      "grad_norm": 1.171875,
      "learning_rate": 2.0276539836581575e-05,
      "loss": 0.4012511372566223,
      "mean_token_accuracy": 0.8576227426528931,
      "num_tokens": 758537759.0,
      "step": 600
    },
    {
      "epoch": 0.5127986348122867,
      "grad_norm": 1.1484375,
      "learning_rate": 2.022123440725032e-05,
      "loss": 0.44054651260375977,
      "mean_token_accuracy": 0.8492074012756348,
      "num_tokens": 759864454.0,
      "step": 601
    },
    {
      "epoch": 0.5136518771331058,
      "grad_norm": 1.078125,
      "learning_rate": 2.0165927285938082e-05,
      "loss": 0.43531936407089233,
      "mean_token_accuracy": 0.8477919101715088,
      "num_tokens": 761118163.0,
      "step": 602
    },
    {
      "epoch": 0.514505119453925,
      "grad_norm": 1.0859375,
      "learning_rate": 2.011061889562879e-05,
      "loss": 0.41037750244140625,
      "mean_token_accuracy": 0.8557760715484619,
      "num_tokens": 762386750.0,
      "step": 603
    },
    {
      "epoch": 0.515358361774744,
      "grad_norm": 1.0234375,
      "learning_rate": 2.0055309659316062e-05,
      "loss": 0.3891329765319824,
      "mean_token_accuracy": 0.8615739345550537,
      "num_tokens": 763642703.0,
      "step": 604
    },
    {
      "epoch": 0.5162116040955631,
      "grad_norm": 1.2109375,
      "learning_rate": 2e-05,
      "loss": 0.42349815368652344,
      "mean_token_accuracy": 0.8538278341293335,
      "num_tokens": 764834474.0,
      "step": 605
    },
    {
      "epoch": 0.5170648464163823,
      "grad_norm": 1.078125,
      "learning_rate": 1.9944690340683944e-05,
      "loss": 0.42084023356437683,
      "mean_token_accuracy": 0.8533438444137573,
      "num_tokens": 766062154.0,
      "step": 606
    },
    {
      "epoch": 0.5179180887372014,
      "grad_norm": 1.1640625,
      "learning_rate": 1.9889381104371214e-05,
      "loss": 0.4517775774002075,
      "mean_token_accuracy": 0.8447047472000122,
      "num_tokens": 767292231.0,
      "step": 607
    },
    {
      "epoch": 0.5187713310580204,
      "grad_norm": 1.15625,
      "learning_rate": 1.9834072714061928e-05,
      "loss": 0.4177018105983734,
      "mean_token_accuracy": 0.8489707708358765,
      "num_tokens": 768529800.0,
      "step": 608
    },
    {
      "epoch": 0.5196245733788396,
      "grad_norm": 1.03125,
      "learning_rate": 1.9778765592749688e-05,
      "loss": 0.41561830043792725,
      "mean_token_accuracy": 0.853600025177002,
      "num_tokens": 769810442.0,
      "step": 609
    },
    {
      "epoch": 0.5204778156996587,
      "grad_norm": 1.125,
      "learning_rate": 1.972346016341843e-05,
      "loss": 0.43313920497894287,
      "mean_token_accuracy": 0.8483097553253174,
      "num_tokens": 771062061.0,
      "step": 610
    },
    {
      "epoch": 0.5213310580204779,
      "grad_norm": 1.1015625,
      "learning_rate": 1.9668156849039137e-05,
      "loss": 0.4079813063144684,
      "mean_token_accuracy": 0.8549296855926514,
      "num_tokens": 772298527.0,
      "step": 611
    },
    {
      "epoch": 0.5221843003412969,
      "grad_norm": 0.98828125,
      "learning_rate": 1.9612856072566623e-05,
      "loss": 0.40580442547798157,
      "mean_token_accuracy": 0.8527456521987915,
      "num_tokens": 773577866.0,
      "step": 612
    },
    {
      "epoch": 0.523037542662116,
      "grad_norm": 1.15625,
      "learning_rate": 1.9557558256936266e-05,
      "loss": 0.45018213987350464,
      "mean_token_accuracy": 0.8437914848327637,
      "num_tokens": 774875543.0,
      "step": 613
    },
    {
      "epoch": 0.5238907849829352,
      "grad_norm": 1.1875,
      "learning_rate": 1.950226382506083e-05,
      "loss": 0.4362746775150299,
      "mean_token_accuracy": 0.8512569665908813,
      "num_tokens": 776189975.0,
      "step": 614
    },
    {
      "epoch": 0.5247440273037542,
      "grad_norm": 1.0703125,
      "learning_rate": 1.9446973199827196e-05,
      "loss": 0.4370247721672058,
      "mean_token_accuracy": 0.8484543561935425,
      "num_tokens": 777570823.0,
      "step": 615
    },
    {
      "epoch": 0.5255972696245734,
      "grad_norm": 1.140625,
      "learning_rate": 1.9391686804093125e-05,
      "loss": 0.4168882369995117,
      "mean_token_accuracy": 0.8524504899978638,
      "num_tokens": 778832763.0,
      "step": 616
    },
    {
      "epoch": 0.5264505119453925,
      "grad_norm": 1.1875,
      "learning_rate": 1.933640506068402e-05,
      "loss": 0.45345214009284973,
      "mean_token_accuracy": 0.8458211421966553,
      "num_tokens": 780090113.0,
      "step": 617
    },
    {
      "epoch": 0.5273037542662116,
      "grad_norm": 1.125,
      "learning_rate": 1.9281128392389725e-05,
      "loss": 0.39708369970321655,
      "mean_token_accuracy": 0.8600180149078369,
      "num_tokens": 781309537.0,
      "step": 618
    },
    {
      "epoch": 0.5281569965870307,
      "grad_norm": 1.0390625,
      "learning_rate": 1.9225857221961254e-05,
      "loss": 0.3814324140548706,
      "mean_token_accuracy": 0.8605290055274963,
      "num_tokens": 782622777.0,
      "step": 619
    },
    {
      "epoch": 0.5290102389078498,
      "grad_norm": 1.1015625,
      "learning_rate": 1.91705919721076e-05,
      "loss": 0.4082639217376709,
      "mean_token_accuracy": 0.8546578884124756,
      "num_tokens": 783891375.0,
      "step": 620
    },
    {
      "epoch": 0.5298634812286689,
      "grad_norm": 1.0078125,
      "learning_rate": 1.9115333065492428e-05,
      "loss": 0.3964448869228363,
      "mean_token_accuracy": 0.8572026491165161,
      "num_tokens": 785153269.0,
      "step": 621
    },
    {
      "epoch": 0.5307167235494881,
      "grad_norm": 1.0859375,
      "learning_rate": 1.9060080924730933e-05,
      "loss": 0.3856889605522156,
      "mean_token_accuracy": 0.8616636395454407,
      "num_tokens": 786456145.0,
      "step": 622
    },
    {
      "epoch": 0.5315699658703071,
      "grad_norm": 1.1640625,
      "learning_rate": 1.900483597238655e-05,
      "loss": 0.42238372564315796,
      "mean_token_accuracy": 0.8500760793685913,
      "num_tokens": 787809181.0,
      "step": 623
    },
    {
      "epoch": 0.5324232081911263,
      "grad_norm": 1.109375,
      "learning_rate": 1.8949598630967748e-05,
      "loss": 0.4335959255695343,
      "mean_token_accuracy": 0.8504573702812195,
      "num_tokens": 789024897.0,
      "step": 624
    },
    {
      "epoch": 0.5332764505119454,
      "grad_norm": 1.140625,
      "learning_rate": 1.8894369322924764e-05,
      "loss": 0.42161279916763306,
      "mean_token_accuracy": 0.8498580455780029,
      "num_tokens": 790348984.0,
      "step": 625
    },
    {
      "epoch": 0.5341296928327645,
      "grad_norm": 1.015625,
      "learning_rate": 1.8839148470646415e-05,
      "loss": 0.39598071575164795,
      "mean_token_accuracy": 0.8566062450408936,
      "num_tokens": 791667528.0,
      "step": 626
    },
    {
      "epoch": 0.5349829351535836,
      "grad_norm": 1.03125,
      "learning_rate": 1.878393649645685e-05,
      "loss": 0.3877071142196655,
      "mean_token_accuracy": 0.8589860200881958,
      "num_tokens": 792983216.0,
      "step": 627
    },
    {
      "epoch": 0.5358361774744027,
      "grad_norm": 1.1328125,
      "learning_rate": 1.8728733822612316e-05,
      "loss": 0.4434708058834076,
      "mean_token_accuracy": 0.8459131121635437,
      "num_tokens": 794285375.0,
      "step": 628
    },
    {
      "epoch": 0.5366894197952219,
      "grad_norm": 1.1015625,
      "learning_rate": 1.8673540871297927e-05,
      "loss": 0.4079039692878723,
      "mean_token_accuracy": 0.8572409749031067,
      "num_tokens": 795520869.0,
      "step": 629
    },
    {
      "epoch": 0.537542662116041,
      "grad_norm": 1.0625,
      "learning_rate": 1.8618358064624443e-05,
      "loss": 0.40205514430999756,
      "mean_token_accuracy": 0.8597898483276367,
      "num_tokens": 796928474.0,
      "step": 630
    },
    {
      "epoch": 0.53839590443686,
      "grad_norm": 1.0625,
      "learning_rate": 1.8563185824625037e-05,
      "loss": 0.4180516004562378,
      "mean_token_accuracy": 0.8505590558052063,
      "num_tokens": 798273455.0,
      "step": 631
    },
    {
      "epoch": 0.5392491467576792,
      "grad_norm": 1.03125,
      "learning_rate": 1.8508024573252073e-05,
      "loss": 0.3760383725166321,
      "mean_token_accuracy": 0.8679073452949524,
      "num_tokens": 799512612.0,
      "step": 632
    },
    {
      "epoch": 0.5401023890784983,
      "grad_norm": 1.015625,
      "learning_rate": 1.8452874732373885e-05,
      "loss": 0.40165501832962036,
      "mean_token_accuracy": 0.8572343587875366,
      "num_tokens": 800782799.0,
      "step": 633
    },
    {
      "epoch": 0.5409556313993175,
      "grad_norm": 1.1484375,
      "learning_rate": 1.8397736723771517e-05,
      "loss": 0.4315105676651001,
      "mean_token_accuracy": 0.8489906787872314,
      "num_tokens": 802151893.0,
      "step": 634
    },
    {
      "epoch": 0.5418088737201365,
      "grad_norm": 1.1640625,
      "learning_rate": 1.8342610969135525e-05,
      "loss": 0.42733538150787354,
      "mean_token_accuracy": 0.8513486981391907,
      "num_tokens": 803421833.0,
      "step": 635
    },
    {
      "epoch": 0.5426621160409556,
      "grad_norm": 1.0859375,
      "learning_rate": 1.8287497890062772e-05,
      "loss": 0.4186640977859497,
      "mean_token_accuracy": 0.8524532914161682,
      "num_tokens": 804700141.0,
      "step": 636
    },
    {
      "epoch": 0.5435153583617748,
      "grad_norm": 1.0859375,
      "learning_rate": 1.8232397908053154e-05,
      "loss": 0.4399576783180237,
      "mean_token_accuracy": 0.8451616168022156,
      "num_tokens": 805968394.0,
      "step": 637
    },
    {
      "epoch": 0.5443686006825939,
      "grad_norm": 1.2890625,
      "learning_rate": 1.817731144450641e-05,
      "loss": 0.4379298985004425,
      "mean_token_accuracy": 0.8490935564041138,
      "num_tokens": 807167629.0,
      "step": 638
    },
    {
      "epoch": 0.5452218430034129,
      "grad_norm": 1.1484375,
      "learning_rate": 1.8122238920718896e-05,
      "loss": 0.46577590703964233,
      "mean_token_accuracy": 0.838182806968689,
      "num_tokens": 808491019.0,
      "step": 639
    },
    {
      "epoch": 0.5460750853242321,
      "grad_norm": 1.21875,
      "learning_rate": 1.806718075788035e-05,
      "loss": 0.4561329483985901,
      "mean_token_accuracy": 0.8396494388580322,
      "num_tokens": 809764797.0,
      "step": 640
    },
    {
      "epoch": 0.5469283276450512,
      "grad_norm": 1.140625,
      "learning_rate": 1.8012137377070683e-05,
      "loss": 0.41830697655677795,
      "mean_token_accuracy": 0.8527405261993408,
      "num_tokens": 810934573.0,
      "step": 641
    },
    {
      "epoch": 0.5477815699658704,
      "grad_norm": 1.0703125,
      "learning_rate": 1.7957109199256755e-05,
      "loss": 0.44165655970573425,
      "mean_token_accuracy": 0.8456424474716187,
      "num_tokens": 812197336.0,
      "step": 642
    },
    {
      "epoch": 0.5486348122866894,
      "grad_norm": 1.328125,
      "learning_rate": 1.7902096645289152e-05,
      "loss": 0.45015034079551697,
      "mean_token_accuracy": 0.842876672744751,
      "num_tokens": 813493162.0,
      "step": 643
    },
    {
      "epoch": 0.5494880546075085,
      "grad_norm": 1.1953125,
      "learning_rate": 1.7847100135898973e-05,
      "loss": 0.4328192472457886,
      "mean_token_accuracy": 0.8485882878303528,
      "num_tokens": 814710823.0,
      "step": 644
    },
    {
      "epoch": 0.5503412969283277,
      "grad_norm": 1.21875,
      "learning_rate": 1.779212009169461e-05,
      "loss": 0.4178626537322998,
      "mean_token_accuracy": 0.853665828704834,
      "num_tokens": 815951671.0,
      "step": 645
    },
    {
      "epoch": 0.5511945392491467,
      "grad_norm": 1.234375,
      "learning_rate": 1.7737156933158522e-05,
      "loss": 0.44559425115585327,
      "mean_token_accuracy": 0.8465609550476074,
      "num_tokens": 817201191.0,
      "step": 646
    },
    {
      "epoch": 0.5520477815699659,
      "grad_norm": 1.0390625,
      "learning_rate": 1.7682211080644037e-05,
      "loss": 0.3857070803642273,
      "mean_token_accuracy": 0.863402247428894,
      "num_tokens": 818451546.0,
      "step": 647
    },
    {
      "epoch": 0.552901023890785,
      "grad_norm": 1.2578125,
      "learning_rate": 1.7627282954372138e-05,
      "loss": 0.4235992729663849,
      "mean_token_accuracy": 0.8513301610946655,
      "num_tokens": 819699350.0,
      "step": 648
    },
    {
      "epoch": 0.5537542662116041,
      "grad_norm": 1.0859375,
      "learning_rate": 1.7572372974428215e-05,
      "loss": 0.4067445397377014,
      "mean_token_accuracy": 0.8585854172706604,
      "num_tokens": 820982667.0,
      "step": 649
    },
    {
      "epoch": 0.5546075085324232,
      "grad_norm": 1.2421875,
      "learning_rate": 1.7517481560758885e-05,
      "loss": 0.454000324010849,
      "mean_token_accuracy": 0.8414619565010071,
      "num_tokens": 822214158.0,
      "step": 650
    },
    {
      "epoch": 0.5554607508532423,
      "grad_norm": 1.1640625,
      "learning_rate": 1.7462609133168788e-05,
      "loss": 0.4347715675830841,
      "mean_token_accuracy": 0.8499559164047241,
      "num_tokens": 823433566.0,
      "step": 651
    },
    {
      "epoch": 0.5563139931740614,
      "grad_norm": 1.109375,
      "learning_rate": 1.7407756111317337e-05,
      "loss": 0.4325719475746155,
      "mean_token_accuracy": 0.8497698307037354,
      "num_tokens": 824766069.0,
      "step": 652
    },
    {
      "epoch": 0.5571672354948806,
      "grad_norm": 1.09375,
      "learning_rate": 1.7352922914715553e-05,
      "loss": 0.43888285756111145,
      "mean_token_accuracy": 0.8494541049003601,
      "num_tokens": 825948554.0,
      "step": 653
    },
    {
      "epoch": 0.5580204778156996,
      "grad_norm": 1.0625,
      "learning_rate": 1.729810996272282e-05,
      "loss": 0.40496283769607544,
      "mean_token_accuracy": 0.8585351705551147,
      "num_tokens": 827229102.0,
      "step": 654
    },
    {
      "epoch": 0.5588737201365188,
      "grad_norm": 1.125,
      "learning_rate": 1.7243317674543688e-05,
      "loss": 0.4012455940246582,
      "mean_token_accuracy": 0.8577613830566406,
      "num_tokens": 828468097.0,
      "step": 655
    },
    {
      "epoch": 0.5597269624573379,
      "grad_norm": 1.890625,
      "learning_rate": 1.7188546469224695e-05,
      "loss": 0.42292094230651855,
      "mean_token_accuracy": 0.8536627292633057,
      "num_tokens": 829680294.0,
      "step": 656
    },
    {
      "epoch": 0.560580204778157,
      "grad_norm": 1.1171875,
      "learning_rate": 1.7133796765651117e-05,
      "loss": 0.4334273934364319,
      "mean_token_accuracy": 0.8470739126205444,
      "num_tokens": 830965074.0,
      "step": 657
    },
    {
      "epoch": 0.5614334470989761,
      "grad_norm": 1.15625,
      "learning_rate": 1.7079068982543812e-05,
      "loss": 0.4390108287334442,
      "mean_token_accuracy": 0.8470625877380371,
      "num_tokens": 832238820.0,
      "step": 658
    },
    {
      "epoch": 0.5622866894197952,
      "grad_norm": 1.1484375,
      "learning_rate": 1.7024363538455952e-05,
      "loss": 0.4335821866989136,
      "mean_token_accuracy": 0.8456578850746155,
      "num_tokens": 833530900.0,
      "step": 659
    },
    {
      "epoch": 0.5631399317406144,
      "grad_norm": 1.3125,
      "learning_rate": 1.6969680851769897e-05,
      "loss": 0.43871545791625977,
      "mean_token_accuracy": 0.8440375328063965,
      "num_tokens": 834801750.0,
      "step": 660
    },
    {
      "epoch": 0.5639931740614335,
      "grad_norm": 1.0859375,
      "learning_rate": 1.6915021340693952e-05,
      "loss": 0.39312106370925903,
      "mean_token_accuracy": 0.8628737926483154,
      "num_tokens": 836064617.0,
      "step": 661
    },
    {
      "epoch": 0.5648464163822525,
      "grad_norm": 1.1640625,
      "learning_rate": 1.6860385423259177e-05,
      "loss": 0.4192885160446167,
      "mean_token_accuracy": 0.8542699813842773,
      "num_tokens": 837302305.0,
      "step": 662
    },
    {
      "epoch": 0.5656996587030717,
      "grad_norm": 1.2109375,
      "learning_rate": 1.6805773517316174e-05,
      "loss": 0.427473247051239,
      "mean_token_accuracy": 0.8486703634262085,
      "num_tokens": 838582562.0,
      "step": 663
    },
    {
      "epoch": 0.5665529010238908,
      "grad_norm": 1.140625,
      "learning_rate": 1.6751186040531927e-05,
      "loss": 0.4230491816997528,
      "mean_token_accuracy": 0.8529016971588135,
      "num_tokens": 839878429.0,
      "step": 664
    },
    {
      "epoch": 0.5674061433447098,
      "grad_norm": 1.0546875,
      "learning_rate": 1.6696623410386587e-05,
      "loss": 0.4039490222930908,
      "mean_token_accuracy": 0.8552538156509399,
      "num_tokens": 841161369.0,
      "step": 665
    },
    {
      "epoch": 0.568259385665529,
      "grad_norm": 1.0390625,
      "learning_rate": 1.6642086044170277e-05,
      "loss": 0.4188658595085144,
      "mean_token_accuracy": 0.8512685298919678,
      "num_tokens": 842403114.0,
      "step": 666
    },
    {
      "epoch": 0.5691126279863481,
      "grad_norm": 1.0546875,
      "learning_rate": 1.658757435897988e-05,
      "loss": 0.41954144835472107,
      "mean_token_accuracy": 0.8521789312362671,
      "num_tokens": 843645878.0,
      "step": 667
    },
    {
      "epoch": 0.5699658703071673,
      "grad_norm": 1.0859375,
      "learning_rate": 1.6533088771715916e-05,
      "loss": 0.41389015316963196,
      "mean_token_accuracy": 0.854069709777832,
      "num_tokens": 844886195.0,
      "step": 668
    },
    {
      "epoch": 0.5708191126279863,
      "grad_norm": 1.1484375,
      "learning_rate": 1.6478629699079278e-05,
      "loss": 0.42659062147140503,
      "mean_token_accuracy": 0.8456200361251831,
      "num_tokens": 846131168.0,
      "step": 669
    },
    {
      "epoch": 0.5716723549488054,
      "grad_norm": 1.171875,
      "learning_rate": 1.6424197557568097e-05,
      "loss": 0.43074071407318115,
      "mean_token_accuracy": 0.8515996932983398,
      "num_tokens": 847403054.0,
      "step": 670
    },
    {
      "epoch": 0.5725255972696246,
      "grad_norm": 1.015625,
      "learning_rate": 1.6369792763474514e-05,
      "loss": 0.4161721467971802,
      "mean_token_accuracy": 0.8545560836791992,
      "num_tokens": 848718163.0,
      "step": 671
    },
    {
      "epoch": 0.5733788395904437,
      "grad_norm": 1.265625,
      "learning_rate": 1.6315415732881543e-05,
      "loss": 0.43793487548828125,
      "mean_token_accuracy": 0.8452411890029907,
      "num_tokens": 850059443.0,
      "step": 672
    },
    {
      "epoch": 0.5742320819112628,
      "grad_norm": 1.078125,
      "learning_rate": 1.6261066881659854e-05,
      "loss": 0.4101327955722809,
      "mean_token_accuracy": 0.8611778020858765,
      "num_tokens": 851359605.0,
      "step": 673
    },
    {
      "epoch": 0.5750853242320819,
      "grad_norm": 1.1328125,
      "learning_rate": 1.6206746625464616e-05,
      "loss": 0.4308299422264099,
      "mean_token_accuracy": 0.8513889312744141,
      "num_tokens": 852729239.0,
      "step": 674
    },
    {
      "epoch": 0.575938566552901,
      "grad_norm": 1.109375,
      "learning_rate": 1.6152455379732275e-05,
      "loss": 0.4416044056415558,
      "mean_token_accuracy": 0.848764181137085,
      "num_tokens": 854014774.0,
      "step": 675
    },
    {
      "epoch": 0.5767918088737202,
      "grad_norm": 1.1484375,
      "learning_rate": 1.609819355967744e-05,
      "loss": 0.42548659443855286,
      "mean_token_accuracy": 0.8489372730255127,
      "num_tokens": 855309256.0,
      "step": 676
    },
    {
      "epoch": 0.5776450511945392,
      "grad_norm": 1.1484375,
      "learning_rate": 1.6043961580289656e-05,
      "loss": 0.43047136068344116,
      "mean_token_accuracy": 0.8447110056877136,
      "num_tokens": 856573324.0,
      "step": 677
    },
    {
      "epoch": 0.5784982935153583,
      "grad_norm": 1.171875,
      "learning_rate": 1.5989759856330264e-05,
      "loss": 0.4528496265411377,
      "mean_token_accuracy": 0.8409486413002014,
      "num_tokens": 857895255.0,
      "step": 678
    },
    {
      "epoch": 0.5793515358361775,
      "grad_norm": 1.375,
      "learning_rate": 1.593558880232921e-05,
      "loss": 0.4279690086841583,
      "mean_token_accuracy": 0.8467363119125366,
      "num_tokens": 859114312.0,
      "step": 679
    },
    {
      "epoch": 0.5802047781569966,
      "grad_norm": 1.1640625,
      "learning_rate": 1.5881448832581862e-05,
      "loss": 0.4421594738960266,
      "mean_token_accuracy": 0.8433245420455933,
      "num_tokens": 860439685.0,
      "step": 680
    },
    {
      "epoch": 0.5810580204778157,
      "grad_norm": 1.140625,
      "learning_rate": 1.5827340361145873e-05,
      "loss": 0.4308147430419922,
      "mean_token_accuracy": 0.8499049544334412,
      "num_tokens": 861657337.0,
      "step": 681
    },
    {
      "epoch": 0.5819112627986348,
      "grad_norm": 1.1484375,
      "learning_rate": 1.577326380183801e-05,
      "loss": 0.41357746720314026,
      "mean_token_accuracy": 0.8544108867645264,
      "num_tokens": 862951727.0,
      "step": 682
    },
    {
      "epoch": 0.5827645051194539,
      "grad_norm": 1.203125,
      "learning_rate": 1.571921956823096e-05,
      "loss": 0.46013498306274414,
      "mean_token_accuracy": 0.8414944410324097,
      "num_tokens": 864294184.0,
      "step": 683
    },
    {
      "epoch": 0.5836177474402731,
      "grad_norm": 1.1484375,
      "learning_rate": 1.566520807365019e-05,
      "loss": 0.4412294030189514,
      "mean_token_accuracy": 0.8421302437782288,
      "num_tokens": 865635274.0,
      "step": 684
    },
    {
      "epoch": 0.5844709897610921,
      "grad_norm": 1.0703125,
      "learning_rate": 1.5611229731170793e-05,
      "loss": 0.43168479204177856,
      "mean_token_accuracy": 0.8477981090545654,
      "num_tokens": 866962855.0,
      "step": 685
    },
    {
      "epoch": 0.5853242320819113,
      "grad_norm": 1.125,
      "learning_rate": 1.5557284953614297e-05,
      "loss": 0.42754048109054565,
      "mean_token_accuracy": 0.8470189571380615,
      "num_tokens": 868228418.0,
      "step": 686
    },
    {
      "epoch": 0.5861774744027304,
      "grad_norm": 14.125,
      "learning_rate": 1.5503374153545562e-05,
      "loss": 0.442811518907547,
      "mean_token_accuracy": 0.84596848487854,
      "num_tokens": 869551783.0,
      "step": 687
    },
    {
      "epoch": 0.5870307167235495,
      "grad_norm": 1.0703125,
      "learning_rate": 1.544949774326955e-05,
      "loss": 0.41952311992645264,
      "mean_token_accuracy": 0.8526431322097778,
      "num_tokens": 870852391.0,
      "step": 688
    },
    {
      "epoch": 0.5878839590443686,
      "grad_norm": 1.109375,
      "learning_rate": 1.539565613482824e-05,
      "loss": 0.45333993434906006,
      "mean_token_accuracy": 0.8428258299827576,
      "num_tokens": 872129994.0,
      "step": 689
    },
    {
      "epoch": 0.5887372013651877,
      "grad_norm": 1.0703125,
      "learning_rate": 1.5341849739997443e-05,
      "loss": 0.38925057649612427,
      "mean_token_accuracy": 0.8625512719154358,
      "num_tokens": 873348211.0,
      "step": 690
    },
    {
      "epoch": 0.5895904436860068,
      "grad_norm": 1.03125,
      "learning_rate": 1.5288078970283668e-05,
      "loss": 0.40125030279159546,
      "mean_token_accuracy": 0.8591695427894592,
      "num_tokens": 874539075.0,
      "step": 691
    },
    {
      "epoch": 0.590443686006826,
      "grad_norm": 1.1484375,
      "learning_rate": 1.5234344236920944e-05,
      "loss": 0.41985952854156494,
      "mean_token_accuracy": 0.8554584980010986,
      "num_tokens": 875728246.0,
      "step": 692
    },
    {
      "epoch": 0.591296928327645,
      "grad_norm": 1.078125,
      "learning_rate": 1.5180645950867714e-05,
      "loss": 0.4123210310935974,
      "mean_token_accuracy": 0.8537267446517944,
      "num_tokens": 877051484.0,
      "step": 693
    },
    {
      "epoch": 0.5921501706484642,
      "grad_norm": 1.078125,
      "learning_rate": 1.5126984522803679e-05,
      "loss": 0.423831582069397,
      "mean_token_accuracy": 0.8533201217651367,
      "num_tokens": 878338192.0,
      "step": 694
    },
    {
      "epoch": 0.5930034129692833,
      "grad_norm": 1.1796875,
      "learning_rate": 1.5073360363126647e-05,
      "loss": 0.42436668276786804,
      "mean_token_accuracy": 0.8561064004898071,
      "num_tokens": 879668253.0,
      "step": 695
    },
    {
      "epoch": 0.5938566552901023,
      "grad_norm": 1.046875,
      "learning_rate": 1.5019773881949391e-05,
      "loss": 0.4175211191177368,
      "mean_token_accuracy": 0.8512572646141052,
      "num_tokens": 880906375.0,
      "step": 696
    },
    {
      "epoch": 0.5947098976109215,
      "grad_norm": 1.1015625,
      "learning_rate": 1.4966225489096535e-05,
      "loss": 0.42132091522216797,
      "mean_token_accuracy": 0.8512777090072632,
      "num_tokens": 882200830.0,
      "step": 697
    },
    {
      "epoch": 0.5955631399317406,
      "grad_norm": 1.1328125,
      "learning_rate": 1.4912715594101404e-05,
      "loss": 0.4065746068954468,
      "mean_token_accuracy": 0.8582139015197754,
      "num_tokens": 883518269.0,
      "step": 698
    },
    {
      "epoch": 0.5964163822525598,
      "grad_norm": 1.1171875,
      "learning_rate": 1.4859244606202904e-05,
      "loss": 0.39612188935279846,
      "mean_token_accuracy": 0.8601369857788086,
      "num_tokens": 884850518.0,
      "step": 699
    },
    {
      "epoch": 0.5972696245733788,
      "grad_norm": 1.09375,
      "learning_rate": 1.4805812934342357e-05,
      "loss": 0.41263896226882935,
      "mean_token_accuracy": 0.8519245982170105,
      "num_tokens": 886164775.0,
      "step": 700
    },
    {
      "epoch": 0.5981228668941979,
      "grad_norm": 1.0703125,
      "learning_rate": 1.4752420987160423e-05,
      "loss": 0.39621469378471375,
      "mean_token_accuracy": 0.857099711894989,
      "num_tokens": 887467156.0,
      "step": 701
    },
    {
      "epoch": 0.5989761092150171,
      "grad_norm": 1.2109375,
      "learning_rate": 1.4699069172993943e-05,
      "loss": 0.4363038241863251,
      "mean_token_accuracy": 0.8497233390808105,
      "num_tokens": 888712488.0,
      "step": 702
    },
    {
      "epoch": 0.5998293515358362,
      "grad_norm": 1.015625,
      "learning_rate": 1.4645757899872823e-05,
      "loss": 0.3981558680534363,
      "mean_token_accuracy": 0.8595453500747681,
      "num_tokens": 890003853.0,
      "step": 703
    },
    {
      "epoch": 0.6006825938566553,
      "grad_norm": 1.1484375,
      "learning_rate": 1.4592487575516926e-05,
      "loss": 0.44280076026916504,
      "mean_token_accuracy": 0.8427318334579468,
      "num_tokens": 891323172.0,
      "step": 704
    },
    {
      "epoch": 0.6015358361774744,
      "grad_norm": 1.1796875,
      "learning_rate": 1.4539258607332926e-05,
      "loss": 0.42328640818595886,
      "mean_token_accuracy": 0.8533531427383423,
      "num_tokens": 892616072.0,
      "step": 705
    },
    {
      "epoch": 0.6023890784982935,
      "grad_norm": 1.1796875,
      "learning_rate": 1.4486071402411215e-05,
      "loss": 0.4273563623428345,
      "mean_token_accuracy": 0.8480685949325562,
      "num_tokens": 893882175.0,
      "step": 706
    },
    {
      "epoch": 0.6032423208191127,
      "grad_norm": 1.09375,
      "learning_rate": 1.4432926367522784e-05,
      "loss": 0.4258284568786621,
      "mean_token_accuracy": 0.8485459089279175,
      "num_tokens": 895203308.0,
      "step": 707
    },
    {
      "epoch": 0.6040955631399317,
      "grad_norm": 1.1640625,
      "learning_rate": 1.4379823909116119e-05,
      "loss": 0.4295752942562103,
      "mean_token_accuracy": 0.8487343192100525,
      "num_tokens": 896498471.0,
      "step": 708
    },
    {
      "epoch": 0.6049488054607508,
      "grad_norm": 1.1875,
      "learning_rate": 1.4326764433314066e-05,
      "loss": 0.4181467294692993,
      "mean_token_accuracy": 0.8553112745285034,
      "num_tokens": 897763501.0,
      "step": 709
    },
    {
      "epoch": 0.60580204778157,
      "grad_norm": 1.0859375,
      "learning_rate": 1.427374834591077e-05,
      "loss": 0.4517369568347931,
      "mean_token_accuracy": 0.8421539068222046,
      "num_tokens": 899056602.0,
      "step": 710
    },
    {
      "epoch": 0.606655290102389,
      "grad_norm": 1.078125,
      "learning_rate": 1.4220776052368517e-05,
      "loss": 0.40613463521003723,
      "mean_token_accuracy": 0.8589709401130676,
      "num_tokens": 900277800.0,
      "step": 711
    },
    {
      "epoch": 0.6075085324232082,
      "grad_norm": 1.0859375,
      "learning_rate": 1.4167847957814683e-05,
      "loss": 0.4038735330104828,
      "mean_token_accuracy": 0.8599674105644226,
      "num_tokens": 901537744.0,
      "step": 712
    },
    {
      "epoch": 0.6083617747440273,
      "grad_norm": 1.078125,
      "learning_rate": 1.4114964467038603e-05,
      "loss": 0.4354320764541626,
      "mean_token_accuracy": 0.84878009557724,
      "num_tokens": 902839000.0,
      "step": 713
    },
    {
      "epoch": 0.6092150170648464,
      "grad_norm": 1.0703125,
      "learning_rate": 1.4062125984488492e-05,
      "loss": 0.4507269263267517,
      "mean_token_accuracy": 0.846520185470581,
      "num_tokens": 904169636.0,
      "step": 714
    },
    {
      "epoch": 0.6100682593856656,
      "grad_norm": 1.1328125,
      "learning_rate": 1.4009332914268343e-05,
      "loss": 0.432539701461792,
      "mean_token_accuracy": 0.8466730713844299,
      "num_tokens": 905415234.0,
      "step": 715
    },
    {
      "epoch": 0.6109215017064846,
      "grad_norm": 1.0859375,
      "learning_rate": 1.3956585660134843e-05,
      "loss": 0.3945189118385315,
      "mean_token_accuracy": 0.8565716743469238,
      "num_tokens": 906581981.0,
      "step": 716
    },
    {
      "epoch": 0.6117747440273038,
      "grad_norm": 0.99609375,
      "learning_rate": 1.390388462549427e-05,
      "loss": 0.3951007127761841,
      "mean_token_accuracy": 0.8592360615730286,
      "num_tokens": 907802029.0,
      "step": 717
    },
    {
      "epoch": 0.6126279863481229,
      "grad_norm": 1.0390625,
      "learning_rate": 1.3851230213399435e-05,
      "loss": 0.41636306047439575,
      "mean_token_accuracy": 0.8538328409194946,
      "num_tokens": 909121659.0,
      "step": 718
    },
    {
      "epoch": 0.613481228668942,
      "grad_norm": 1.265625,
      "learning_rate": 1.3798622826546583e-05,
      "loss": 0.41935044527053833,
      "mean_token_accuracy": 0.8520289659500122,
      "num_tokens": 910394166.0,
      "step": 719
    },
    {
      "epoch": 0.6143344709897611,
      "grad_norm": 1.2265625,
      "learning_rate": 1.3746062867272301e-05,
      "loss": 0.43312495946884155,
      "mean_token_accuracy": 0.8498440980911255,
      "num_tokens": 911593197.0,
      "step": 720
    },
    {
      "epoch": 0.6151877133105802,
      "grad_norm": 1.0546875,
      "learning_rate": 1.3693550737550461e-05,
      "loss": 0.4256262183189392,
      "mean_token_accuracy": 0.8515105843544006,
      "num_tokens": 912890171.0,
      "step": 721
    },
    {
      "epoch": 0.6160409556313993,
      "grad_norm": 1.5625,
      "learning_rate": 1.364108683898914e-05,
      "loss": 0.4089497923851013,
      "mean_token_accuracy": 0.854285717010498,
      "num_tokens": 914163365.0,
      "step": 722
    },
    {
      "epoch": 0.6168941979522184,
      "grad_norm": 1.0390625,
      "learning_rate": 1.358867157282755e-05,
      "loss": 0.39454782009124756,
      "mean_token_accuracy": 0.8586098551750183,
      "num_tokens": 915374622.0,
      "step": 723
    },
    {
      "epoch": 0.6177474402730375,
      "grad_norm": 1.1015625,
      "learning_rate": 1.3536305339932965e-05,
      "loss": 0.4278569221496582,
      "mean_token_accuracy": 0.8471983671188354,
      "num_tokens": 916675256.0,
      "step": 724
    },
    {
      "epoch": 0.6186006825938567,
      "grad_norm": 1.078125,
      "learning_rate": 1.3483988540797652e-05,
      "loss": 0.43968355655670166,
      "mean_token_accuracy": 0.8464534282684326,
      "num_tokens": 917973144.0,
      "step": 725
    },
    {
      "epoch": 0.6194539249146758,
      "grad_norm": 1.15625,
      "learning_rate": 1.3431721575535816e-05,
      "loss": 0.4086441397666931,
      "mean_token_accuracy": 0.8545673489570618,
      "num_tokens": 919254705.0,
      "step": 726
    },
    {
      "epoch": 0.6203071672354948,
      "grad_norm": 1.0859375,
      "learning_rate": 1.337950484388054e-05,
      "loss": 0.4210573434829712,
      "mean_token_accuracy": 0.8555586338043213,
      "num_tokens": 920529683.0,
      "step": 727
    },
    {
      "epoch": 0.621160409556314,
      "grad_norm": 1.078125,
      "learning_rate": 1.3327338745180718e-05,
      "loss": 0.4225504398345947,
      "mean_token_accuracy": 0.8536198139190674,
      "num_tokens": 921785631.0,
      "step": 728
    },
    {
      "epoch": 0.6220136518771331,
      "grad_norm": 1.203125,
      "learning_rate": 1.3275223678398026e-05,
      "loss": 0.42212188243865967,
      "mean_token_accuracy": 0.8522540330886841,
      "num_tokens": 922945397.0,
      "step": 729
    },
    {
      "epoch": 0.6228668941979523,
      "grad_norm": 1.03125,
      "learning_rate": 1.322316004210382e-05,
      "loss": 0.41758665442466736,
      "mean_token_accuracy": 0.8533700704574585,
      "num_tokens": 924195599.0,
      "step": 730
    },
    {
      "epoch": 0.6237201365187713,
      "grad_norm": 1.1328125,
      "learning_rate": 1.3171148234476148e-05,
      "loss": 0.42089471220970154,
      "mean_token_accuracy": 0.8504568934440613,
      "num_tokens": 925489876.0,
      "step": 731
    },
    {
      "epoch": 0.6245733788395904,
      "grad_norm": 1.109375,
      "learning_rate": 1.3119188653296666e-05,
      "loss": 0.4102407991886139,
      "mean_token_accuracy": 0.855168342590332,
      "num_tokens": 926764823.0,
      "step": 732
    },
    {
      "epoch": 0.6254266211604096,
      "grad_norm": 0.99609375,
      "learning_rate": 1.3067281695947619e-05,
      "loss": 0.3836125135421753,
      "mean_token_accuracy": 0.8646092414855957,
      "num_tokens": 927991959.0,
      "step": 733
    },
    {
      "epoch": 0.6262798634812287,
      "grad_norm": 2.671875,
      "learning_rate": 1.301542775940876e-05,
      "loss": 0.4087880253791809,
      "mean_token_accuracy": 0.8540505170822144,
      "num_tokens": 929203744.0,
      "step": 734
    },
    {
      "epoch": 0.6271331058020477,
      "grad_norm": 1.125,
      "learning_rate": 1.2963627240254373e-05,
      "loss": 0.4346690773963928,
      "mean_token_accuracy": 0.8516008853912354,
      "num_tokens": 930480245.0,
      "step": 735
    },
    {
      "epoch": 0.6279863481228669,
      "grad_norm": 1.0859375,
      "learning_rate": 1.2911880534650201e-05,
      "loss": 0.43417131900787354,
      "mean_token_accuracy": 0.8494067192077637,
      "num_tokens": 931720198.0,
      "step": 736
    },
    {
      "epoch": 0.628839590443686,
      "grad_norm": 1.25,
      "learning_rate": 1.2860188038350435e-05,
      "loss": 0.43970224261283875,
      "mean_token_accuracy": 0.8447173833847046,
      "num_tokens": 932944205.0,
      "step": 737
    },
    {
      "epoch": 0.6296928327645052,
      "grad_norm": 1.0078125,
      "learning_rate": 1.2808550146694647e-05,
      "loss": 0.40661025047302246,
      "mean_token_accuracy": 0.8551698923110962,
      "num_tokens": 934238368.0,
      "step": 738
    },
    {
      "epoch": 0.6305460750853242,
      "grad_norm": 1.03125,
      "learning_rate": 1.275696725460483e-05,
      "loss": 0.39733102917671204,
      "mean_token_accuracy": 0.8601347208023071,
      "num_tokens": 935468733.0,
      "step": 739
    },
    {
      "epoch": 0.6313993174061433,
      "grad_norm": 1.171875,
      "learning_rate": 1.2705439756582334e-05,
      "loss": 0.4371893107891083,
      "mean_token_accuracy": 0.8477665781974792,
      "num_tokens": 936738177.0,
      "step": 740
    },
    {
      "epoch": 0.6322525597269625,
      "grad_norm": 1.0546875,
      "learning_rate": 1.265396804670487e-05,
      "loss": 0.4017401933670044,
      "mean_token_accuracy": 0.8528299331665039,
      "num_tokens": 938028230.0,
      "step": 741
    },
    {
      "epoch": 0.6331058020477816,
      "grad_norm": 1.0859375,
      "learning_rate": 1.2602552518623456e-05,
      "loss": 0.39727726578712463,
      "mean_token_accuracy": 0.8560886383056641,
      "num_tokens": 939221575.0,
      "step": 742
    },
    {
      "epoch": 0.6339590443686007,
      "grad_norm": 1.0859375,
      "learning_rate": 1.2551193565559469e-05,
      "loss": 0.43056029081344604,
      "mean_token_accuracy": 0.8495368361473083,
      "num_tokens": 940453714.0,
      "step": 743
    },
    {
      "epoch": 0.6348122866894198,
      "grad_norm": 1.2421875,
      "learning_rate": 1.2499891580301595e-05,
      "loss": 0.4344298541545868,
      "mean_token_accuracy": 0.8479069471359253,
      "num_tokens": 941742926.0,
      "step": 744
    },
    {
      "epoch": 0.6356655290102389,
      "grad_norm": 1.0859375,
      "learning_rate": 1.2448646955202835e-05,
      "loss": 0.3753342926502228,
      "mean_token_accuracy": 0.8663811683654785,
      "num_tokens": 942924929.0,
      "step": 745
    },
    {
      "epoch": 0.636518771331058,
      "grad_norm": 1.109375,
      "learning_rate": 1.2397460082177493e-05,
      "loss": 0.41401368379592896,
      "mean_token_accuracy": 0.8546045422554016,
      "num_tokens": 944257081.0,
      "step": 746
    },
    {
      "epoch": 0.6373720136518771,
      "grad_norm": 1.0859375,
      "learning_rate": 1.2346331352698206e-05,
      "loss": 0.4342406392097473,
      "mean_token_accuracy": 0.850756824016571,
      "num_tokens": 945550859.0,
      "step": 747
    },
    {
      "epoch": 0.6382252559726962,
      "grad_norm": 1.1484375,
      "learning_rate": 1.2295261157792933e-05,
      "loss": 0.4053502678871155,
      "mean_token_accuracy": 0.8545789122581482,
      "num_tokens": 946834513.0,
      "step": 748
    },
    {
      "epoch": 0.6390784982935154,
      "grad_norm": 1.0703125,
      "learning_rate": 1.2244249888041955e-05,
      "loss": 0.43143171072006226,
      "mean_token_accuracy": 0.8496818542480469,
      "num_tokens": 948061790.0,
      "step": 749
    },
    {
      "epoch": 0.6399317406143344,
      "grad_norm": 1.296875,
      "learning_rate": 1.2193297933574921e-05,
      "loss": 0.4166182279586792,
      "mean_token_accuracy": 0.8499701023101807,
      "num_tokens": 949235516.0,
      "step": 750
    },
    {
      "epoch": 0.6407849829351536,
      "grad_norm": 1.125,
      "learning_rate": 1.2142405684067806e-05,
      "loss": 0.3996542990207672,
      "mean_token_accuracy": 0.8604468107223511,
      "num_tokens": 950445001.0,
      "step": 751
    },
    {
      "epoch": 0.6416382252559727,
      "grad_norm": 0.98828125,
      "learning_rate": 1.2091573528739998e-05,
      "loss": 0.38675546646118164,
      "mean_token_accuracy": 0.8662380576133728,
      "num_tokens": 951705156.0,
      "step": 752
    },
    {
      "epoch": 0.6424914675767918,
      "grad_norm": 1.0625,
      "learning_rate": 1.2040801856351275e-05,
      "loss": 0.4324323832988739,
      "mean_token_accuracy": 0.8497709035873413,
      "num_tokens": 953038356.0,
      "step": 753
    },
    {
      "epoch": 0.643344709897611,
      "grad_norm": 2.4375,
      "learning_rate": 1.1990091055198865e-05,
      "loss": 0.3989053964614868,
      "mean_token_accuracy": 0.8604363203048706,
      "num_tokens": 954294118.0,
      "step": 754
    },
    {
      "epoch": 0.64419795221843,
      "grad_norm": 1.0859375,
      "learning_rate": 1.1939441513114423e-05,
      "loss": 0.4067000150680542,
      "mean_token_accuracy": 0.8513942956924438,
      "num_tokens": 955533341.0,
      "step": 755
    },
    {
      "epoch": 0.6450511945392492,
      "grad_norm": 1.125,
      "learning_rate": 1.1888853617461134e-05,
      "loss": 0.41696566343307495,
      "mean_token_accuracy": 0.8550819158554077,
      "num_tokens": 956817201.0,
      "step": 756
    },
    {
      "epoch": 0.6459044368600683,
      "grad_norm": 1.2109375,
      "learning_rate": 1.1838327755130701e-05,
      "loss": 0.40972623229026794,
      "mean_token_accuracy": 0.8551120758056641,
      "num_tokens": 958085835.0,
      "step": 757
    },
    {
      "epoch": 0.6467576791808873,
      "grad_norm": 1.1484375,
      "learning_rate": 1.1787864312540415e-05,
      "loss": 0.4300018846988678,
      "mean_token_accuracy": 0.8487316370010376,
      "num_tokens": 959330228.0,
      "step": 758
    },
    {
      "epoch": 0.6476109215017065,
      "grad_norm": 1.1875,
      "learning_rate": 1.1737463675630163e-05,
      "loss": 0.4143313765525818,
      "mean_token_accuracy": 0.8553274869918823,
      "num_tokens": 960552642.0,
      "step": 759
    },
    {
      "epoch": 0.6484641638225256,
      "grad_norm": 1.2421875,
      "learning_rate": 1.1687126229859515e-05,
      "loss": 0.43225210905075073,
      "mean_token_accuracy": 0.8495205640792847,
      "num_tokens": 961843303.0,
      "step": 760
    },
    {
      "epoch": 0.6493174061433447,
      "grad_norm": 1.046875,
      "learning_rate": 1.1636852360204765e-05,
      "loss": 0.410086452960968,
      "mean_token_accuracy": 0.8551639914512634,
      "num_tokens": 963166377.0,
      "step": 761
    },
    {
      "epoch": 0.6501706484641638,
      "grad_norm": 1.4140625,
      "learning_rate": 1.158664245115598e-05,
      "loss": 0.4341108798980713,
      "mean_token_accuracy": 0.8505324721336365,
      "num_tokens": 964473738.0,
      "step": 762
    },
    {
      "epoch": 0.6510238907849829,
      "grad_norm": 1.1953125,
      "learning_rate": 1.1536496886714047e-05,
      "loss": 0.3812198340892792,
      "mean_token_accuracy": 0.8653488755226135,
      "num_tokens": 965712554.0,
      "step": 763
    },
    {
      "epoch": 0.6518771331058021,
      "grad_norm": 1.1953125,
      "learning_rate": 1.148641605038776e-05,
      "loss": 0.43712347745895386,
      "mean_token_accuracy": 0.8434666395187378,
      "num_tokens": 966981169.0,
      "step": 764
    },
    {
      "epoch": 0.6527303754266212,
      "grad_norm": 1.203125,
      "learning_rate": 1.1436400325190898e-05,
      "loss": 0.428911417722702,
      "mean_token_accuracy": 0.8550093173980713,
      "num_tokens": 968199304.0,
      "step": 765
    },
    {
      "epoch": 0.6535836177474402,
      "grad_norm": 1.1640625,
      "learning_rate": 1.1386450093639252e-05,
      "loss": 0.4197339713573456,
      "mean_token_accuracy": 0.8498107194900513,
      "num_tokens": 969482555.0,
      "step": 766
    },
    {
      "epoch": 0.6544368600682594,
      "grad_norm": 1.0546875,
      "learning_rate": 1.133656573774772e-05,
      "loss": 0.4075573682785034,
      "mean_token_accuracy": 0.8554080128669739,
      "num_tokens": 970726533.0,
      "step": 767
    },
    {
      "epoch": 0.6552901023890785,
      "grad_norm": 1.0859375,
      "learning_rate": 1.1286747639027407e-05,
      "loss": 0.39426279067993164,
      "mean_token_accuracy": 0.8597115278244019,
      "num_tokens": 972040730.0,
      "step": 768
    },
    {
      "epoch": 0.6561433447098977,
      "grad_norm": 1.078125,
      "learning_rate": 1.123699617848268e-05,
      "loss": 0.4162084758281708,
      "mean_token_accuracy": 0.8560755848884583,
      "num_tokens": 973292105.0,
      "step": 769
    },
    {
      "epoch": 0.6569965870307167,
      "grad_norm": 1.203125,
      "learning_rate": 1.1187311736608275e-05,
      "loss": 0.42430153489112854,
      "mean_token_accuracy": 0.8535183072090149,
      "num_tokens": 974573637.0,
      "step": 770
    },
    {
      "epoch": 0.6578498293515358,
      "grad_norm": 1.1171875,
      "learning_rate": 1.1137694693386345e-05,
      "loss": 0.3872300982475281,
      "mean_token_accuracy": 0.8624697923660278,
      "num_tokens": 975834709.0,
      "step": 771
    },
    {
      "epoch": 0.658703071672355,
      "grad_norm": 1.1328125,
      "learning_rate": 1.1088145428283612e-05,
      "loss": 0.39715999364852905,
      "mean_token_accuracy": 0.8602268695831299,
      "num_tokens": 977034553.0,
      "step": 772
    },
    {
      "epoch": 0.659556313993174,
      "grad_norm": 1.3046875,
      "learning_rate": 1.103866432024843e-05,
      "loss": 0.4081801176071167,
      "mean_token_accuracy": 0.8584600687026978,
      "num_tokens": 978311454.0,
      "step": 773
    },
    {
      "epoch": 0.6604095563139932,
      "grad_norm": 1.125,
      "learning_rate": 1.0989251747707896e-05,
      "loss": 0.4222116470336914,
      "mean_token_accuracy": 0.8506911396980286,
      "num_tokens": 979480268.0,
      "step": 774
    },
    {
      "epoch": 0.6612627986348123,
      "grad_norm": 1.0546875,
      "learning_rate": 1.093990808856495e-05,
      "loss": 0.40232789516448975,
      "mean_token_accuracy": 0.8575530052185059,
      "num_tokens": 980793231.0,
      "step": 775
    },
    {
      "epoch": 0.6621160409556314,
      "grad_norm": 1.046875,
      "learning_rate": 1.089063372019548e-05,
      "loss": 0.40536290407180786,
      "mean_token_accuracy": 0.8582439422607422,
      "num_tokens": 982100344.0,
      "step": 776
    },
    {
      "epoch": 0.6629692832764505,
      "grad_norm": 1.140625,
      "learning_rate": 1.0841429019445458e-05,
      "loss": 0.4440503418445587,
      "mean_token_accuracy": 0.84715735912323,
      "num_tokens": 983347982.0,
      "step": 777
    },
    {
      "epoch": 0.6638225255972696,
      "grad_norm": 1.0703125,
      "learning_rate": 1.0792294362628036e-05,
      "loss": 0.4436691701412201,
      "mean_token_accuracy": 0.8440119028091431,
      "num_tokens": 984642637.0,
      "step": 778
    },
    {
      "epoch": 0.6646757679180887,
      "grad_norm": 1.390625,
      "learning_rate": 1.0743230125520692e-05,
      "loss": 0.43782490491867065,
      "mean_token_accuracy": 0.8461368083953857,
      "num_tokens": 985822938.0,
      "step": 779
    },
    {
      "epoch": 0.6655290102389079,
      "grad_norm": 1.125,
      "learning_rate": 1.069423668336231e-05,
      "loss": 0.4207554757595062,
      "mean_token_accuracy": 0.8514894247055054,
      "num_tokens": 987110438.0,
      "step": 780
    },
    {
      "epoch": 0.6663822525597269,
      "grad_norm": 1.0078125,
      "learning_rate": 1.0645314410850368e-05,
      "loss": 0.36480164527893066,
      "mean_token_accuracy": 0.8708664178848267,
      "num_tokens": 988355209.0,
      "step": 781
    },
    {
      "epoch": 0.6672354948805461,
      "grad_norm": 1.1640625,
      "learning_rate": 1.0596463682138034e-05,
      "loss": 0.4153715968132019,
      "mean_token_accuracy": 0.8527963161468506,
      "num_tokens": 989440955.0,
      "step": 782
    },
    {
      "epoch": 0.6680887372013652,
      "grad_norm": 1.09375,
      "learning_rate": 1.0547684870831322e-05,
      "loss": 0.42345380783081055,
      "mean_token_accuracy": 0.8559162616729736,
      "num_tokens": 990707709.0,
      "step": 783
    },
    {
      "epoch": 0.6689419795221843,
      "grad_norm": 1.25,
      "learning_rate": 1.0498978349986211e-05,
      "loss": 0.4291629195213318,
      "mean_token_accuracy": 0.8488958477973938,
      "num_tokens": 992009876.0,
      "step": 784
    },
    {
      "epoch": 0.6697952218430034,
      "grad_norm": 1.1015625,
      "learning_rate": 1.0450344492105828e-05,
      "loss": 0.42940378189086914,
      "mean_token_accuracy": 0.849949061870575,
      "num_tokens": 993283841.0,
      "step": 785
    },
    {
      "epoch": 0.6706484641638225,
      "grad_norm": 1.03125,
      "learning_rate": 1.0401783669137573e-05,
      "loss": 0.42250925302505493,
      "mean_token_accuracy": 0.8505643606185913,
      "num_tokens": 994566640.0,
      "step": 786
    },
    {
      "epoch": 0.6715017064846417,
      "grad_norm": 0.9609375,
      "learning_rate": 1.0353296252470288e-05,
      "loss": 0.3954888582229614,
      "mean_token_accuracy": 0.8602569103240967,
      "num_tokens": 995866637.0,
      "step": 787
    },
    {
      "epoch": 0.6723549488054608,
      "grad_norm": 1.1015625,
      "learning_rate": 1.0304882612931388e-05,
      "loss": 0.4021111726760864,
      "mean_token_accuracy": 0.8549807071685791,
      "num_tokens": 997077987.0,
      "step": 788
    },
    {
      "epoch": 0.6732081911262798,
      "grad_norm": 1.0,
      "learning_rate": 1.0256543120784074e-05,
      "loss": 0.3830087184906006,
      "mean_token_accuracy": 0.8604985475540161,
      "num_tokens": 998443468.0,
      "step": 789
    },
    {
      "epoch": 0.674061433447099,
      "grad_norm": 1.0390625,
      "learning_rate": 1.0208278145724458e-05,
      "loss": 0.38845381140708923,
      "mean_token_accuracy": 0.8632364273071289,
      "num_tokens": 999764691.0,
      "step": 790
    },
    {
      "epoch": 0.6749146757679181,
      "grad_norm": 1.296875,
      "learning_rate": 1.0160088056878767e-05,
      "loss": 0.4149647057056427,
      "mean_token_accuracy": 0.8558659553527832,
      "num_tokens": 1001039811.0,
      "step": 791
    },
    {
      "epoch": 0.6757679180887372,
      "grad_norm": 1.2734375,
      "learning_rate": 1.0111973222800482e-05,
      "loss": 0.3725243806838989,
      "mean_token_accuracy": 0.8679097294807434,
      "num_tokens": 1002313221.0,
      "step": 792
    },
    {
      "epoch": 0.6766211604095563,
      "grad_norm": 1.1171875,
      "learning_rate": 1.006393401146756e-05,
      "loss": 0.40442317724227905,
      "mean_token_accuracy": 0.8580458164215088,
      "num_tokens": 1003608212.0,
      "step": 793
    },
    {
      "epoch": 0.6774744027303754,
      "grad_norm": 1.0390625,
      "learning_rate": 1.0015970790279597e-05,
      "loss": 0.39823365211486816,
      "mean_token_accuracy": 0.860603392124176,
      "num_tokens": 1004957806.0,
      "step": 794
    },
    {
      "epoch": 0.6783276450511946,
      "grad_norm": 1.203125,
      "learning_rate": 9.968083926055025e-06,
      "loss": 0.37618565559387207,
      "mean_token_accuracy": 0.8631827235221863,
      "num_tokens": 1006271928.0,
      "step": 795
    },
    {
      "epoch": 0.6791808873720137,
      "grad_norm": 1.1171875,
      "learning_rate": 9.920273785028311e-06,
      "loss": 0.44252946972846985,
      "mean_token_accuracy": 0.8491775989532471,
      "num_tokens": 1007590711.0,
      "step": 796
    },
    {
      "epoch": 0.6800341296928327,
      "grad_norm": 1.1328125,
      "learning_rate": 9.872540732847126e-06,
      "loss": 0.37631094455718994,
      "mean_token_accuracy": 0.8628123998641968,
      "num_tokens": 1008807597.0,
      "step": 797
    },
    {
      "epoch": 0.6808873720136519,
      "grad_norm": 1.0625,
      "learning_rate": 9.824885134569597e-06,
      "loss": 0.42826157808303833,
      "mean_token_accuracy": 0.850793182849884,
      "num_tokens": 1010095849.0,
      "step": 798
    },
    {
      "epoch": 0.681740614334471,
      "grad_norm": 1.0859375,
      "learning_rate": 9.777307354661483e-06,
      "loss": 0.4039434492588043,
      "mean_token_accuracy": 0.8521813154220581,
      "num_tokens": 1011331341.0,
      "step": 799
    },
    {
      "epoch": 0.6825938566552902,
      "grad_norm": 1.1171875,
      "learning_rate": 9.729807756993396e-06,
      "loss": 0.44468891620635986,
      "mean_token_accuracy": 0.8554663062095642,
      "num_tokens": 1012564122.0,
      "step": 800
    },
    {
      "epoch": 0.6834470989761092,
      "grad_norm": 1.2421875,
      "learning_rate": 9.682386704838007e-06,
      "loss": 0.36254632472991943,
      "mean_token_accuracy": 0.8676851987838745,
      "num_tokens": 1013786423.0,
      "step": 801
    },
    {
      "epoch": 0.6843003412969283,
      "grad_norm": 1.4296875,
      "learning_rate": 9.635044560867288e-06,
      "loss": 0.43980151414871216,
      "mean_token_accuracy": 0.8510434627532959,
      "num_tokens": 1015042747.0,
      "step": 802
    },
    {
      "epoch": 0.6851535836177475,
      "grad_norm": 1.1328125,
      "learning_rate": 9.58778168714973e-06,
      "loss": 0.4100112318992615,
      "mean_token_accuracy": 0.8597438335418701,
      "num_tokens": 1016347774.0,
      "step": 803
    },
    {
      "epoch": 0.6860068259385665,
      "grad_norm": 1.7265625,
      "learning_rate": 9.540598445147577e-06,
      "loss": 0.38256925344467163,
      "mean_token_accuracy": 0.8663723468780518,
      "num_tokens": 1017543007.0,
      "step": 804
    },
    {
      "epoch": 0.6868600682593856,
      "grad_norm": 1.0625,
      "learning_rate": 9.493495195714028e-06,
      "loss": 0.41075658798217773,
      "mean_token_accuracy": 0.8533617258071899,
      "num_tokens": 1018809447.0,
      "step": 805
    },
    {
      "epoch": 0.6877133105802048,
      "grad_norm": 2.65625,
      "learning_rate": 9.44647229909054e-06,
      "loss": 0.4418414831161499,
      "mean_token_accuracy": 0.8463571071624756,
      "num_tokens": 1020031643.0,
      "step": 806
    },
    {
      "epoch": 0.6885665529010239,
      "grad_norm": 0.99609375,
      "learning_rate": 9.399530114904023e-06,
      "loss": 0.3819839358329773,
      "mean_token_accuracy": 0.8650012016296387,
      "num_tokens": 1021286345.0,
      "step": 807
    },
    {
      "epoch": 0.689419795221843,
      "grad_norm": 1.046875,
      "learning_rate": 9.352669002164106e-06,
      "loss": 0.41330528259277344,
      "mean_token_accuracy": 0.8534460067749023,
      "num_tokens": 1022550216.0,
      "step": 808
    },
    {
      "epoch": 0.6902730375426621,
      "grad_norm": 1.015625,
      "learning_rate": 9.305889319260399e-06,
      "loss": 0.3930574953556061,
      "mean_token_accuracy": 0.8597910404205322,
      "num_tokens": 1023950169.0,
      "step": 809
    },
    {
      "epoch": 0.6911262798634812,
      "grad_norm": 1.234375,
      "learning_rate": 9.259191423959726e-06,
      "loss": 0.4142628312110901,
      "mean_token_accuracy": 0.8532405495643616,
      "num_tokens": 1025241444.0,
      "step": 810
    },
    {
      "epoch": 0.6919795221843004,
      "grad_norm": 1.15625,
      "learning_rate": 9.212575673403425e-06,
      "loss": 0.4250876009464264,
      "mean_token_accuracy": 0.8504900932312012,
      "num_tokens": 1026510304.0,
      "step": 811
    },
    {
      "epoch": 0.6928327645051194,
      "grad_norm": 1.109375,
      "learning_rate": 9.166042424104589e-06,
      "loss": 0.4497537612915039,
      "mean_token_accuracy": 0.8404860496520996,
      "num_tokens": 1027725525.0,
      "step": 812
    },
    {
      "epoch": 0.6936860068259386,
      "grad_norm": 1.09375,
      "learning_rate": 9.119592031945354e-06,
      "loss": 0.4331575036048889,
      "mean_token_accuracy": 0.8457444906234741,
      "num_tokens": 1028973372.0,
      "step": 813
    },
    {
      "epoch": 0.6945392491467577,
      "grad_norm": 1.1796875,
      "learning_rate": 9.073224852174177e-06,
      "loss": 0.4472113847732544,
      "mean_token_accuracy": 0.8471002578735352,
      "num_tokens": 1030355744.0,
      "step": 814
    },
    {
      "epoch": 0.6953924914675768,
      "grad_norm": 1.015625,
      "learning_rate": 9.026941239403087e-06,
      "loss": 0.38642019033432007,
      "mean_token_accuracy": 0.8599845767021179,
      "num_tokens": 1031557412.0,
      "step": 815
    },
    {
      "epoch": 0.6962457337883959,
      "grad_norm": 1.140625,
      "learning_rate": 8.98074154760503e-06,
      "loss": 0.4230565130710602,
      "mean_token_accuracy": 0.8483209013938904,
      "num_tokens": 1032909777.0,
      "step": 816
    },
    {
      "epoch": 0.697098976109215,
      "grad_norm": 1.125,
      "learning_rate": 8.934626130111122e-06,
      "loss": 0.42412126064300537,
      "mean_token_accuracy": 0.851210355758667,
      "num_tokens": 1034173470.0,
      "step": 817
    },
    {
      "epoch": 0.6979522184300341,
      "grad_norm": 1.0546875,
      "learning_rate": 8.888595339607961e-06,
      "loss": 0.4262489080429077,
      "mean_token_accuracy": 0.8562053442001343,
      "num_tokens": 1035475740.0,
      "step": 818
    },
    {
      "epoch": 0.6988054607508533,
      "grad_norm": 1.125,
      "learning_rate": 8.842649528134909e-06,
      "loss": 0.4197317361831665,
      "mean_token_accuracy": 0.8518404960632324,
      "num_tokens": 1036700768.0,
      "step": 819
    },
    {
      "epoch": 0.6996587030716723,
      "grad_norm": 1.03125,
      "learning_rate": 8.796789047081429e-06,
      "loss": 0.37740328907966614,
      "mean_token_accuracy": 0.8661898374557495,
      "num_tokens": 1038018219.0,
      "step": 820
    },
    {
      "epoch": 0.7005119453924915,
      "grad_norm": 1.09375,
      "learning_rate": 8.751014247184388e-06,
      "loss": 0.4073388874530792,
      "mean_token_accuracy": 0.8573184609413147,
      "num_tokens": 1039251925.0,
      "step": 821
    },
    {
      "epoch": 0.7013651877133106,
      "grad_norm": 1.0859375,
      "learning_rate": 8.705325478525364e-06,
      "loss": 0.398876816034317,
      "mean_token_accuracy": 0.8603102564811707,
      "num_tokens": 1040461500.0,
      "step": 822
    },
    {
      "epoch": 0.7022184300341296,
      "grad_norm": 1.140625,
      "learning_rate": 8.659723090527984e-06,
      "loss": 0.4210307002067566,
      "mean_token_accuracy": 0.8557091355323792,
      "num_tokens": 1041682836.0,
      "step": 823
    },
    {
      "epoch": 0.7030716723549488,
      "grad_norm": 1.1796875,
      "learning_rate": 8.614207431955224e-06,
      "loss": 0.4209953844547272,
      "mean_token_accuracy": 0.8449440002441406,
      "num_tokens": 1042950061.0,
      "step": 824
    },
    {
      "epoch": 0.7039249146757679,
      "grad_norm": 1.140625,
      "learning_rate": 8.568778850906781e-06,
      "loss": 0.4312911629676819,
      "mean_token_accuracy": 0.8492382168769836,
      "num_tokens": 1044209921.0,
      "step": 825
    },
    {
      "epoch": 0.7047781569965871,
      "grad_norm": 1.1484375,
      "learning_rate": 8.523437694816383e-06,
      "loss": 0.38010480999946594,
      "mean_token_accuracy": 0.8646488785743713,
      "num_tokens": 1045518210.0,
      "step": 826
    },
    {
      "epoch": 0.7056313993174061,
      "grad_norm": 2.15625,
      "learning_rate": 8.478184310449145e-06,
      "loss": 0.4344509243965149,
      "mean_token_accuracy": 0.8512169122695923,
      "num_tokens": 1046786448.0,
      "step": 827
    },
    {
      "epoch": 0.7064846416382252,
      "grad_norm": 1.40625,
      "learning_rate": 8.43301904389891e-06,
      "loss": 0.4187681972980499,
      "mean_token_accuracy": 0.8543927669525146,
      "num_tokens": 1047972761.0,
      "step": 828
    },
    {
      "epoch": 0.7073378839590444,
      "grad_norm": 1.1328125,
      "learning_rate": 8.387942240585587e-06,
      "loss": 0.4113813042640686,
      "mean_token_accuracy": 0.8526463508605957,
      "num_tokens": 1049257872.0,
      "step": 829
    },
    {
      "epoch": 0.7081911262798635,
      "grad_norm": 1.375,
      "learning_rate": 8.342954245252553e-06,
      "loss": 0.4016193747520447,
      "mean_token_accuracy": 0.8582109212875366,
      "num_tokens": 1050599591.0,
      "step": 830
    },
    {
      "epoch": 0.7090443686006825,
      "grad_norm": 1.0546875,
      "learning_rate": 8.298055401963969e-06,
      "loss": 0.40881067514419556,
      "mean_token_accuracy": 0.8524539470672607,
      "num_tokens": 1051798514.0,
      "step": 831
    },
    {
      "epoch": 0.7098976109215017,
      "grad_norm": 1.109375,
      "learning_rate": 8.253246054102185e-06,
      "loss": 0.43954432010650635,
      "mean_token_accuracy": 0.8480045795440674,
      "num_tokens": 1053094326.0,
      "step": 832
    },
    {
      "epoch": 0.7107508532423208,
      "grad_norm": 1.0546875,
      "learning_rate": 8.208526544365092e-06,
      "loss": 0.4087798595428467,
      "mean_token_accuracy": 0.8563933968544006,
      "num_tokens": 1054457411.0,
      "step": 833
    },
    {
      "epoch": 0.71160409556314,
      "grad_norm": 1.0703125,
      "learning_rate": 8.163897214763493e-06,
      "loss": 0.4090237021446228,
      "mean_token_accuracy": 0.857208251953125,
      "num_tokens": 1055777660.0,
      "step": 834
    },
    {
      "epoch": 0.712457337883959,
      "grad_norm": 1.1171875,
      "learning_rate": 8.119358406618521e-06,
      "loss": 0.42780041694641113,
      "mean_token_accuracy": 0.8459384441375732,
      "num_tokens": 1057091948.0,
      "step": 835
    },
    {
      "epoch": 0.7133105802047781,
      "grad_norm": 1.1328125,
      "learning_rate": 8.074910460558998e-06,
      "loss": 0.4324817657470703,
      "mean_token_accuracy": 0.8458794951438904,
      "num_tokens": 1058341222.0,
      "step": 836
    },
    {
      "epoch": 0.7141638225255973,
      "grad_norm": 1.203125,
      "learning_rate": 8.030553716518853e-06,
      "loss": 0.39742356538772583,
      "mean_token_accuracy": 0.8609949350357056,
      "num_tokens": 1059615221.0,
      "step": 837
    },
    {
      "epoch": 0.7150170648464164,
      "grad_norm": 1.5078125,
      "learning_rate": 7.986288513734492e-06,
      "loss": 0.42181551456451416,
      "mean_token_accuracy": 0.8513354063034058,
      "num_tokens": 1060926863.0,
      "step": 838
    },
    {
      "epoch": 0.7158703071672355,
      "grad_norm": 1.125,
      "learning_rate": 7.942115190742232e-06,
      "loss": 0.41256269812583923,
      "mean_token_accuracy": 0.8566028475761414,
      "num_tokens": 1062180442.0,
      "step": 839
    },
    {
      "epoch": 0.7167235494880546,
      "grad_norm": 1.1328125,
      "learning_rate": 7.898034085375703e-06,
      "loss": 0.45280733704566956,
      "mean_token_accuracy": 0.8393298983573914,
      "num_tokens": 1063388746.0,
      "step": 840
    },
    {
      "epoch": 0.7175767918088737,
      "grad_norm": 1.0703125,
      "learning_rate": 7.854045534763268e-06,
      "loss": 0.38201114535331726,
      "mean_token_accuracy": 0.8618240356445312,
      "num_tokens": 1064569188.0,
      "step": 841
    },
    {
      "epoch": 0.7184300341296929,
      "grad_norm": 1.1875,
      "learning_rate": 7.810149875325417e-06,
      "loss": 0.44304075837135315,
      "mean_token_accuracy": 0.8445332050323486,
      "num_tokens": 1065865827.0,
      "step": 842
    },
    {
      "epoch": 0.7192832764505119,
      "grad_norm": 1.203125,
      "learning_rate": 7.76634744277224e-06,
      "loss": 0.42808371782302856,
      "mean_token_accuracy": 0.8523356318473816,
      "num_tokens": 1067161837.0,
      "step": 843
    },
    {
      "epoch": 0.7201365187713311,
      "grad_norm": 1.296875,
      "learning_rate": 7.722638572100827e-06,
      "loss": 0.40428364276885986,
      "mean_token_accuracy": 0.8537883758544922,
      "num_tokens": 1068438570.0,
      "step": 844
    },
    {
      "epoch": 0.7209897610921502,
      "grad_norm": 1.078125,
      "learning_rate": 7.679023597592713e-06,
      "loss": 0.4180000126361847,
      "mean_token_accuracy": 0.8517836332321167,
      "num_tokens": 1069708577.0,
      "step": 845
    },
    {
      "epoch": 0.7218430034129693,
      "grad_norm": 1.09375,
      "learning_rate": 7.635502852811339e-06,
      "loss": 0.3939325511455536,
      "mean_token_accuracy": 0.8606342077255249,
      "num_tokens": 1070933510.0,
      "step": 846
    },
    {
      "epoch": 0.7226962457337884,
      "grad_norm": 1.0859375,
      "learning_rate": 7.592076670599453e-06,
      "loss": 0.4094383418560028,
      "mean_token_accuracy": 0.8583788871765137,
      "num_tokens": 1072210927.0,
      "step": 847
    },
    {
      "epoch": 0.7235494880546075,
      "grad_norm": 1.1171875,
      "learning_rate": 7.548745383076628e-06,
      "loss": 0.4269757866859436,
      "mean_token_accuracy": 0.8512881994247437,
      "num_tokens": 1073457513.0,
      "step": 848
    },
    {
      "epoch": 0.7244027303754266,
      "grad_norm": 1.1875,
      "learning_rate": 7.505509321636677e-06,
      "loss": 0.39594176411628723,
      "mean_token_accuracy": 0.8600922226905823,
      "num_tokens": 1074718943.0,
      "step": 849
    },
    {
      "epoch": 0.7252559726962458,
      "grad_norm": 1.09375,
      "learning_rate": 7.462368816945145e-06,
      "loss": 0.38474512100219727,
      "mean_token_accuracy": 0.862127959728241,
      "num_tokens": 1075958668.0,
      "step": 850
    },
    {
      "epoch": 0.7261092150170648,
      "grad_norm": 1.1171875,
      "learning_rate": 7.419324198936739e-06,
      "loss": 0.3880165219306946,
      "mean_token_accuracy": 0.8613038659095764,
      "num_tokens": 1077227735.0,
      "step": 851
    },
    {
      "epoch": 0.726962457337884,
      "grad_norm": 1.1953125,
      "learning_rate": 7.3763757968128646e-06,
      "loss": 0.4291451871395111,
      "mean_token_accuracy": 0.8512101769447327,
      "num_tokens": 1078547905.0,
      "step": 852
    },
    {
      "epoch": 0.7278156996587031,
      "grad_norm": 1.125,
      "learning_rate": 7.333523939039057e-06,
      "loss": 0.3926810622215271,
      "mean_token_accuracy": 0.8568257093429565,
      "num_tokens": 1079768085.0,
      "step": 853
    },
    {
      "epoch": 0.7286689419795221,
      "grad_norm": 1.1484375,
      "learning_rate": 7.2907689533425105e-06,
      "loss": 0.4551146626472473,
      "mean_token_accuracy": 0.8406661748886108,
      "num_tokens": 1081126004.0,
      "step": 854
    },
    {
      "epoch": 0.7295221843003413,
      "grad_norm": 1.125,
      "learning_rate": 7.248111166709517e-06,
      "loss": 0.42530155181884766,
      "mean_token_accuracy": 0.8527835607528687,
      "num_tokens": 1082386772.0,
      "step": 855
    },
    {
      "epoch": 0.7303754266211604,
      "grad_norm": 1.1796875,
      "learning_rate": 7.205550905383025e-06,
      "loss": 0.4087600111961365,
      "mean_token_accuracy": 0.855338990688324,
      "num_tokens": 1083684341.0,
      "step": 856
    },
    {
      "epoch": 0.7312286689419796,
      "grad_norm": 1.71875,
      "learning_rate": 7.163088494860113e-06,
      "loss": 0.4335992932319641,
      "mean_token_accuracy": 0.8463903069496155,
      "num_tokens": 1084930305.0,
      "step": 857
    },
    {
      "epoch": 0.7320819112627986,
      "grad_norm": 0.93359375,
      "learning_rate": 7.120724259889507e-06,
      "loss": 0.3836511969566345,
      "mean_token_accuracy": 0.8640062808990479,
      "num_tokens": 1086218234.0,
      "step": 858
    },
    {
      "epoch": 0.7329351535836177,
      "grad_norm": 1.1875,
      "learning_rate": 7.078458524469078e-06,
      "loss": 0.414017915725708,
      "mean_token_accuracy": 0.8593389391899109,
      "num_tokens": 1087507899.0,
      "step": 859
    },
    {
      "epoch": 0.7337883959044369,
      "grad_norm": 1.0546875,
      "learning_rate": 7.036291611843402e-06,
      "loss": 0.39484792947769165,
      "mean_token_accuracy": 0.8615354299545288,
      "num_tokens": 1088735449.0,
      "step": 860
    },
    {
      "epoch": 0.734641638225256,
      "grad_norm": 1.1171875,
      "learning_rate": 6.99422384450126e-06,
      "loss": 0.40726131200790405,
      "mean_token_accuracy": 0.8563534021377563,
      "num_tokens": 1089969907.0,
      "step": 861
    },
    {
      "epoch": 0.735494880546075,
      "grad_norm": 1.09375,
      "learning_rate": 6.9522555441731835e-06,
      "loss": 0.42039239406585693,
      "mean_token_accuracy": 0.8530364036560059,
      "num_tokens": 1091251975.0,
      "step": 862
    },
    {
      "epoch": 0.7363481228668942,
      "grad_norm": 1.0703125,
      "learning_rate": 6.910387031828971e-06,
      "loss": 0.37323445081710815,
      "mean_token_accuracy": 0.8653548955917358,
      "num_tokens": 1092490417.0,
      "step": 863
    },
    {
      "epoch": 0.7372013651877133,
      "grad_norm": 1.171875,
      "learning_rate": 6.868618627675272e-06,
      "loss": 0.4401625394821167,
      "mean_token_accuracy": 0.844910740852356,
      "num_tokens": 1093761282.0,
      "step": 864
    },
    {
      "epoch": 0.7380546075085325,
      "grad_norm": 1.1328125,
      "learning_rate": 6.826950651153108e-06,
      "loss": 0.41529425978660583,
      "mean_token_accuracy": 0.8534643650054932,
      "num_tokens": 1094982231.0,
      "step": 865
    },
    {
      "epoch": 0.7389078498293515,
      "grad_norm": 1.0,
      "learning_rate": 6.785383420935441e-06,
      "loss": 0.34643200039863586,
      "mean_token_accuracy": 0.8717191815376282,
      "num_tokens": 1096201079.0,
      "step": 866
    },
    {
      "epoch": 0.7397610921501706,
      "grad_norm": 1.1015625,
      "learning_rate": 6.7439172549247415e-06,
      "loss": 0.4155789017677307,
      "mean_token_accuracy": 0.8521484136581421,
      "num_tokens": 1097563572.0,
      "step": 867
    },
    {
      "epoch": 0.7406143344709898,
      "grad_norm": 1.1640625,
      "learning_rate": 6.7025524702505275e-06,
      "loss": 0.35580557584762573,
      "mean_token_accuracy": 0.8691191673278809,
      "num_tokens": 1098866038.0,
      "step": 868
    },
    {
      "epoch": 0.7414675767918089,
      "grad_norm": 1.1171875,
      "learning_rate": 6.661289383266984e-06,
      "loss": 0.42278242111206055,
      "mean_token_accuracy": 0.8494211435317993,
      "num_tokens": 1100086238.0,
      "step": 869
    },
    {
      "epoch": 0.742320819112628,
      "grad_norm": 1.15625,
      "learning_rate": 6.620128309550515e-06,
      "loss": 0.4138052463531494,
      "mean_token_accuracy": 0.8531437516212463,
      "num_tokens": 1101375706.0,
      "step": 870
    },
    {
      "epoch": 0.7431740614334471,
      "grad_norm": 1.046875,
      "learning_rate": 6.57906956389734e-06,
      "loss": 0.40648433566093445,
      "mean_token_accuracy": 0.8587487936019897,
      "num_tokens": 1102625188.0,
      "step": 871
    },
    {
      "epoch": 0.7440273037542662,
      "grad_norm": 1.0625,
      "learning_rate": 6.538113460321065e-06,
      "loss": 0.3804891109466553,
      "mean_token_accuracy": 0.8683966994285583,
      "num_tokens": 1103837854.0,
      "step": 872
    },
    {
      "epoch": 0.7448805460750854,
      "grad_norm": 1.078125,
      "learning_rate": 6.497260312050318e-06,
      "loss": 0.4331369400024414,
      "mean_token_accuracy": 0.8510516285896301,
      "num_tokens": 1105097861.0,
      "step": 873
    },
    {
      "epoch": 0.7457337883959044,
      "grad_norm": 1.0703125,
      "learning_rate": 6.456510431526324e-06,
      "loss": 0.42266231775283813,
      "mean_token_accuracy": 0.8549261093139648,
      "num_tokens": 1106425161.0,
      "step": 874
    },
    {
      "epoch": 0.7465870307167235,
      "grad_norm": 1.109375,
      "learning_rate": 6.415864130400537e-06,
      "loss": 0.44113433361053467,
      "mean_token_accuracy": 0.8469362854957581,
      "num_tokens": 1107688815.0,
      "step": 875
    },
    {
      "epoch": 0.7474402730375427,
      "grad_norm": 0.94921875,
      "learning_rate": 6.375321719532226e-06,
      "loss": 0.40756452083587646,
      "mean_token_accuracy": 0.8592301607131958,
      "num_tokens": 1108919922.0,
      "step": 876
    },
    {
      "epoch": 0.7482935153583617,
      "grad_norm": 1.2109375,
      "learning_rate": 6.334883508986132e-06,
      "loss": 0.42040860652923584,
      "mean_token_accuracy": 0.8562484979629517,
      "num_tokens": 1110179622.0,
      "step": 877
    },
    {
      "epoch": 0.7491467576791809,
      "grad_norm": 1.0625,
      "learning_rate": 6.294549808030073e-06,
      "loss": 0.40860050916671753,
      "mean_token_accuracy": 0.8558719158172607,
      "num_tokens": 1111415501.0,
      "step": 878
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.93359375,
      "learning_rate": 6.254320925132602e-06,
      "loss": 0.38524919748306274,
      "mean_token_accuracy": 0.8626477122306824,
      "num_tokens": 1112672121.0,
      "step": 879
    },
    {
      "epoch": 0.7508532423208191,
      "grad_norm": 1.0625,
      "learning_rate": 6.214197167960602e-06,
      "loss": 0.38657626509666443,
      "mean_token_accuracy": 0.8609930276870728,
      "num_tokens": 1114016443.0,
      "step": 880
    },
    {
      "epoch": 0.7517064846416383,
      "grad_norm": 1.2265625,
      "learning_rate": 6.174178843376999e-06,
      "loss": 0.41322121024131775,
      "mean_token_accuracy": 0.8608590364456177,
      "num_tokens": 1115257668.0,
      "step": 881
    },
    {
      "epoch": 0.7525597269624573,
      "grad_norm": 1.0390625,
      "learning_rate": 6.1342662574383546e-06,
      "loss": 0.36880770325660706,
      "mean_token_accuracy": 0.8674668073654175,
      "num_tokens": 1116443328.0,
      "step": 882
    },
    {
      "epoch": 0.7534129692832765,
      "grad_norm": 1.1640625,
      "learning_rate": 6.094459715392571e-06,
      "loss": 0.4255765974521637,
      "mean_token_accuracy": 0.8526012897491455,
      "num_tokens": 1117691647.0,
      "step": 883
    },
    {
      "epoch": 0.7542662116040956,
      "grad_norm": 1.203125,
      "learning_rate": 6.054759521676512e-06,
      "loss": 0.45156174898147583,
      "mean_token_accuracy": 0.8378329277038574,
      "num_tokens": 1118962834.0,
      "step": 884
    },
    {
      "epoch": 0.7551194539249146,
      "grad_norm": 1.0859375,
      "learning_rate": 6.01516597991372e-06,
      "loss": 0.43079066276550293,
      "mean_token_accuracy": 0.8493002653121948,
      "num_tokens": 1120199153.0,
      "step": 885
    },
    {
      "epoch": 0.7559726962457338,
      "grad_norm": 1.1328125,
      "learning_rate": 5.975679392912061e-06,
      "loss": 0.43787503242492676,
      "mean_token_accuracy": 0.845841646194458,
      "num_tokens": 1121472969.0,
      "step": 886
    },
    {
      "epoch": 0.7568259385665529,
      "grad_norm": 1.0859375,
      "learning_rate": 5.936300062661435e-06,
      "loss": 0.4467785656452179,
      "mean_token_accuracy": 0.8420335054397583,
      "num_tokens": 1122699591.0,
      "step": 887
    },
    {
      "epoch": 0.757679180887372,
      "grad_norm": 0.9921875,
      "learning_rate": 5.897028290331428e-06,
      "loss": 0.38398295640945435,
      "mean_token_accuracy": 0.8625005483627319,
      "num_tokens": 1123919320.0,
      "step": 888
    },
    {
      "epoch": 0.7585324232081911,
      "grad_norm": 1.0390625,
      "learning_rate": 5.857864376269051e-06,
      "loss": 0.42201560735702515,
      "mean_token_accuracy": 0.8513107895851135,
      "num_tokens": 1125191913.0,
      "step": 889
    },
    {
      "epoch": 0.7593856655290102,
      "grad_norm": 1.125,
      "learning_rate": 5.818808619996421e-06,
      "loss": 0.40835830569267273,
      "mean_token_accuracy": 0.8562037944793701,
      "num_tokens": 1126430838.0,
      "step": 890
    },
    {
      "epoch": 0.7602389078498294,
      "grad_norm": 1.0625,
      "learning_rate": 5.779861320208471e-06,
      "loss": 0.41223305463790894,
      "mean_token_accuracy": 0.857351541519165,
      "num_tokens": 1127702972.0,
      "step": 891
    },
    {
      "epoch": 0.7610921501706485,
      "grad_norm": 1.078125,
      "learning_rate": 5.741022774770677e-06,
      "loss": 0.3904264569282532,
      "mean_token_accuracy": 0.8604041337966919,
      "num_tokens": 1128900540.0,
      "step": 892
    },
    {
      "epoch": 0.7619453924914675,
      "grad_norm": 1.1328125,
      "learning_rate": 5.702293280716753e-06,
      "loss": 0.4162328541278839,
      "mean_token_accuracy": 0.8519338369369507,
      "num_tokens": 1130181587.0,
      "step": 893
    },
    {
      "epoch": 0.7627986348122867,
      "grad_norm": 1.0625,
      "learning_rate": 5.663673134246414e-06,
      "loss": 0.43056660890579224,
      "mean_token_accuracy": 0.8501580953598022,
      "num_tokens": 1131471749.0,
      "step": 894
    },
    {
      "epoch": 0.7636518771331058,
      "grad_norm": 1.109375,
      "learning_rate": 5.625162630723091e-06,
      "loss": 0.43100273609161377,
      "mean_token_accuracy": 0.8481286764144897,
      "num_tokens": 1132671507.0,
      "step": 895
    },
    {
      "epoch": 0.764505119453925,
      "grad_norm": 1.0625,
      "learning_rate": 5.586762064671683e-06,
      "loss": 0.4080134630203247,
      "mean_token_accuracy": 0.8588391542434692,
      "num_tokens": 1134020848.0,
      "step": 896
    },
    {
      "epoch": 0.765358361774744,
      "grad_norm": 1.09375,
      "learning_rate": 5.54847172977627e-06,
      "loss": 0.4136390686035156,
      "mean_token_accuracy": 0.8567047119140625,
      "num_tokens": 1135253010.0,
      "step": 897
    },
    {
      "epoch": 0.7662116040955631,
      "grad_norm": 1.1328125,
      "learning_rate": 5.510291918877921e-06,
      "loss": 0.41519707441329956,
      "mean_token_accuracy": 0.8530816435813904,
      "num_tokens": 1136513186.0,
      "step": 898
    },
    {
      "epoch": 0.7670648464163823,
      "grad_norm": 1.0234375,
      "learning_rate": 5.472222923972421e-06,
      "loss": 0.38287484645843506,
      "mean_token_accuracy": 0.8622580766677856,
      "num_tokens": 1137783807.0,
      "step": 899
    },
    {
      "epoch": 0.7679180887372014,
      "grad_norm": 1.140625,
      "learning_rate": 5.434265036208046e-06,
      "loss": 0.38058948516845703,
      "mean_token_accuracy": 0.8642020225524902,
      "num_tokens": 1139058815.0,
      "step": 900
    },
    {
      "epoch": 0.7687713310580204,
      "grad_norm": 1.078125,
      "learning_rate": 5.396418545883318e-06,
      "loss": 0.4565323293209076,
      "mean_token_accuracy": 0.8410910964012146,
      "num_tokens": 1140384645.0,
      "step": 901
    },
    {
      "epoch": 0.7696245733788396,
      "grad_norm": 1.1328125,
      "learning_rate": 5.358683742444822e-06,
      "loss": 0.36041566729545593,
      "mean_token_accuracy": 0.8725635409355164,
      "num_tokens": 1141602748.0,
      "step": 902
    },
    {
      "epoch": 0.7704778156996587,
      "grad_norm": 1.0546875,
      "learning_rate": 5.321060914484959e-06,
      "loss": 0.38665345311164856,
      "mean_token_accuracy": 0.8619093894958496,
      "num_tokens": 1142813571.0,
      "step": 903
    },
    {
      "epoch": 0.7713310580204779,
      "grad_norm": 3.015625,
      "learning_rate": 5.283550349739764e-06,
      "loss": 0.402282178401947,
      "mean_token_accuracy": 0.8564893007278442,
      "num_tokens": 1144100384.0,
      "step": 904
    },
    {
      "epoch": 0.7721843003412969,
      "grad_norm": 1.1953125,
      "learning_rate": 5.246152335086674e-06,
      "loss": 0.4080726206302643,
      "mean_token_accuracy": 0.8562952280044556,
      "num_tokens": 1145344982.0,
      "step": 905
    },
    {
      "epoch": 0.773037542662116,
      "grad_norm": 1.234375,
      "learning_rate": 5.20886715654237e-06,
      "loss": 0.437753289937973,
      "mean_token_accuracy": 0.8525388240814209,
      "num_tokens": 1146558695.0,
      "step": 906
    },
    {
      "epoch": 0.7738907849829352,
      "grad_norm": 1.1015625,
      "learning_rate": 5.171695099260563e-06,
      "loss": 0.35097581148147583,
      "mean_token_accuracy": 0.8769875764846802,
      "num_tokens": 1147878200.0,
      "step": 907
    },
    {
      "epoch": 0.7747440273037542,
      "grad_norm": 1.2109375,
      "learning_rate": 5.1346364475298374e-06,
      "loss": 0.4062862992286682,
      "mean_token_accuracy": 0.8551477789878845,
      "num_tokens": 1149157776.0,
      "step": 908
    },
    {
      "epoch": 0.7755972696245734,
      "grad_norm": 1.265625,
      "learning_rate": 5.097691484771434e-06,
      "loss": 0.398479700088501,
      "mean_token_accuracy": 0.8599405884742737,
      "num_tokens": 1150365692.0,
      "step": 909
    },
    {
      "epoch": 0.7764505119453925,
      "grad_norm": 1.0390625,
      "learning_rate": 5.060860493537134e-06,
      "loss": 0.412792444229126,
      "mean_token_accuracy": 0.8542439937591553,
      "num_tokens": 1151649224.0,
      "step": 910
    },
    {
      "epoch": 0.7773037542662116,
      "grad_norm": 1.0,
      "learning_rate": 5.024143755507069e-06,
      "loss": 0.3787228465080261,
      "mean_token_accuracy": 0.8606414794921875,
      "num_tokens": 1152961565.0,
      "step": 911
    },
    {
      "epoch": 0.7781569965870307,
      "grad_norm": 1.0703125,
      "learning_rate": 4.987541551487578e-06,
      "loss": 0.39150702953338623,
      "mean_token_accuracy": 0.8609663248062134,
      "num_tokens": 1154163866.0,
      "step": 912
    },
    {
      "epoch": 0.7790102389078498,
      "grad_norm": 1.0234375,
      "learning_rate": 4.951054161409035e-06,
      "loss": 0.4048803746700287,
      "mean_token_accuracy": 0.8535192012786865,
      "num_tokens": 1155478956.0,
      "step": 913
    },
    {
      "epoch": 0.7798634812286689,
      "grad_norm": 1.0390625,
      "learning_rate": 4.91468186432374e-06,
      "loss": 0.39697980880737305,
      "mean_token_accuracy": 0.8583462834358215,
      "num_tokens": 1156696518.0,
      "step": 914
    },
    {
      "epoch": 0.7807167235494881,
      "grad_norm": 1.0703125,
      "learning_rate": 4.878424938403778e-06,
      "loss": 0.40347033739089966,
      "mean_token_accuracy": 0.8608555793762207,
      "num_tokens": 1158038007.0,
      "step": 915
    },
    {
      "epoch": 0.7815699658703071,
      "grad_norm": 1.1875,
      "learning_rate": 4.842283660938878e-06,
      "loss": 0.4442577362060547,
      "mean_token_accuracy": 0.8456679582595825,
      "num_tokens": 1159367332.0,
      "step": 916
    },
    {
      "epoch": 0.7824232081911263,
      "grad_norm": 1.1796875,
      "learning_rate": 4.806258308334306e-06,
      "loss": 0.4301019012928009,
      "mean_token_accuracy": 0.8498237133026123,
      "num_tokens": 1160587064.0,
      "step": 917
    },
    {
      "epoch": 0.7832764505119454,
      "grad_norm": 1.671875,
      "learning_rate": 4.770349156108727e-06,
      "loss": 0.4383105933666229,
      "mean_token_accuracy": 0.8487505912780762,
      "num_tokens": 1161809798.0,
      "step": 918
    },
    {
      "epoch": 0.7841296928327645,
      "grad_norm": 1.15625,
      "learning_rate": 4.734556478892136e-06,
      "loss": 0.43311381340026855,
      "mean_token_accuracy": 0.849035382270813,
      "num_tokens": 1163115958.0,
      "step": 919
    },
    {
      "epoch": 0.7849829351535836,
      "grad_norm": 1.2109375,
      "learning_rate": 4.69888055042373e-06,
      "loss": 0.41883188486099243,
      "mean_token_accuracy": 0.8517540693283081,
      "num_tokens": 1164379078.0,
      "step": 920
    },
    {
      "epoch": 0.7858361774744027,
      "grad_norm": 1.046875,
      "learning_rate": 4.66332164354983e-06,
      "loss": 0.3931668698787689,
      "mean_token_accuracy": 0.8556213974952698,
      "num_tokens": 1165671160.0,
      "step": 921
    },
    {
      "epoch": 0.7866894197952219,
      "grad_norm": 1.1328125,
      "learning_rate": 4.627880030221765e-06,
      "loss": 0.40344688296318054,
      "mean_token_accuracy": 0.8569764494895935,
      "num_tokens": 1166979849.0,
      "step": 922
    },
    {
      "epoch": 0.787542662116041,
      "grad_norm": 1.109375,
      "learning_rate": 4.592555981493836e-06,
      "loss": 0.4423007369041443,
      "mean_token_accuracy": 0.8470763564109802,
      "num_tokens": 1168364132.0,
      "step": 923
    },
    {
      "epoch": 0.78839590443686,
      "grad_norm": 1.453125,
      "learning_rate": 4.557349767521208e-06,
      "loss": 0.4301292896270752,
      "mean_token_accuracy": 0.8504316210746765,
      "num_tokens": 1169597575.0,
      "step": 924
    },
    {
      "epoch": 0.7892491467576792,
      "grad_norm": 1.265625,
      "learning_rate": 4.5222616575578695e-06,
      "loss": 0.4010273218154907,
      "mean_token_accuracy": 0.8599902391433716,
      "num_tokens": 1170853523.0,
      "step": 925
    },
    {
      "epoch": 0.7901023890784983,
      "grad_norm": 1.1015625,
      "learning_rate": 4.487291919954535e-06,
      "loss": 0.4090416729450226,
      "mean_token_accuracy": 0.8562982082366943,
      "num_tokens": 1172169575.0,
      "step": 926
    },
    {
      "epoch": 0.7909556313993175,
      "grad_norm": 0.93359375,
      "learning_rate": 4.452440822156643e-06,
      "loss": 0.35915493965148926,
      "mean_token_accuracy": 0.8716059923171997,
      "num_tokens": 1173418890.0,
      "step": 927
    },
    {
      "epoch": 0.7918088737201365,
      "grad_norm": 1.125,
      "learning_rate": 4.4177086307022736e-06,
      "loss": 0.4389945864677429,
      "mean_token_accuracy": 0.8478891849517822,
      "num_tokens": 1174752673.0,
      "step": 928
    },
    {
      "epoch": 0.7926621160409556,
      "grad_norm": 1.171875,
      "learning_rate": 4.383095611220132e-06,
      "loss": 0.4104952812194824,
      "mean_token_accuracy": 0.8529853820800781,
      "num_tokens": 1175978887.0,
      "step": 929
    },
    {
      "epoch": 0.7935153583617748,
      "grad_norm": 1.1328125,
      "learning_rate": 4.348602028427482e-06,
      "loss": 0.40251606702804565,
      "mean_token_accuracy": 0.8556443452835083,
      "num_tokens": 1177282127.0,
      "step": 930
    },
    {
      "epoch": 0.7943686006825939,
      "grad_norm": 1.1484375,
      "learning_rate": 4.314228146128174e-06,
      "loss": 0.43880361318588257,
      "mean_token_accuracy": 0.8461077809333801,
      "num_tokens": 1178686721.0,
      "step": 931
    },
    {
      "epoch": 0.7952218430034129,
      "grad_norm": 1.1171875,
      "learning_rate": 4.279974227210588e-06,
      "loss": 0.3915668725967407,
      "mean_token_accuracy": 0.8625930547714233,
      "num_tokens": 1179926172.0,
      "step": 932
    },
    {
      "epoch": 0.7960750853242321,
      "grad_norm": 1.0859375,
      "learning_rate": 4.2458405336456395e-06,
      "loss": 0.37605857849121094,
      "mean_token_accuracy": 0.8657439947128296,
      "num_tokens": 1181121096.0,
      "step": 933
    },
    {
      "epoch": 0.7969283276450512,
      "grad_norm": 1.0390625,
      "learning_rate": 4.211827326484756e-06,
      "loss": 0.39829251170158386,
      "mean_token_accuracy": 0.8628296852111816,
      "num_tokens": 1182405585.0,
      "step": 934
    },
    {
      "epoch": 0.7977815699658704,
      "grad_norm": 1.078125,
      "learning_rate": 4.177934865857916e-06,
      "loss": 0.417942076921463,
      "mean_token_accuracy": 0.8537082672119141,
      "num_tokens": 1183647547.0,
      "step": 935
    },
    {
      "epoch": 0.7986348122866894,
      "grad_norm": 1.1171875,
      "learning_rate": 4.144163410971625e-06,
      "loss": 0.42460399866104126,
      "mean_token_accuracy": 0.8483660221099854,
      "num_tokens": 1184976834.0,
      "step": 936
    },
    {
      "epoch": 0.7994880546075085,
      "grad_norm": 1.109375,
      "learning_rate": 4.110513220106957e-06,
      "loss": 0.41980448365211487,
      "mean_token_accuracy": 0.8548370003700256,
      "num_tokens": 1186287284.0,
      "step": 937
    },
    {
      "epoch": 0.8003412969283277,
      "grad_norm": 1.09375,
      "learning_rate": 4.0769845506175685e-06,
      "loss": 0.40496665239334106,
      "mean_token_accuracy": 0.8563840389251709,
      "num_tokens": 1187545936.0,
      "step": 938
    },
    {
      "epoch": 0.8011945392491467,
      "grad_norm": 1.09375,
      "learning_rate": 4.043577658927718e-06,
      "loss": 0.3844448924064636,
      "mean_token_accuracy": 0.8645089864730835,
      "num_tokens": 1188797744.0,
      "step": 939
    },
    {
      "epoch": 0.8020477815699659,
      "grad_norm": 1.09375,
      "learning_rate": 4.010292800530333e-06,
      "loss": 0.34539860486984253,
      "mean_token_accuracy": 0.872482180595398,
      "num_tokens": 1190038787.0,
      "step": 940
    },
    {
      "epoch": 0.802901023890785,
      "grad_norm": 1.140625,
      "learning_rate": 3.977130229985038e-06,
      "loss": 0.44747087359428406,
      "mean_token_accuracy": 0.8472003936767578,
      "num_tokens": 1191253117.0,
      "step": 941
    },
    {
      "epoch": 0.8037542662116041,
      "grad_norm": 1.15625,
      "learning_rate": 3.944090200916215e-06,
      "loss": 0.4132612943649292,
      "mean_token_accuracy": 0.8540629148483276,
      "num_tokens": 1192537149.0,
      "step": 942
    },
    {
      "epoch": 0.8046075085324232,
      "grad_norm": 1.546875,
      "learning_rate": 3.911172966011043e-06,
      "loss": 0.3817031979560852,
      "mean_token_accuracy": 0.869147539138794,
      "num_tokens": 1193823573.0,
      "step": 943
    },
    {
      "epoch": 0.8054607508532423,
      "grad_norm": 1.0859375,
      "learning_rate": 3.878378777017602e-06,
      "loss": 0.3914313018321991,
      "mean_token_accuracy": 0.8579877018928528,
      "num_tokens": 1195060662.0,
      "step": 944
    },
    {
      "epoch": 0.8063139931740614,
      "grad_norm": 1.328125,
      "learning_rate": 3.845707884742913e-06,
      "loss": 0.4297591745853424,
      "mean_token_accuracy": 0.846786379814148,
      "num_tokens": 1196339826.0,
      "step": 945
    },
    {
      "epoch": 0.8071672354948806,
      "grad_norm": 1.1328125,
      "learning_rate": 3.813160539051053e-06,
      "loss": 0.4121960997581482,
      "mean_token_accuracy": 0.8552278876304626,
      "num_tokens": 1197630420.0,
      "step": 946
    },
    {
      "epoch": 0.8080204778156996,
      "grad_norm": 1.046875,
      "learning_rate": 3.7807369888612e-06,
      "loss": 0.3562585711479187,
      "mean_token_accuracy": 0.8758576512336731,
      "num_tokens": 1198874553.0,
      "step": 947
    },
    {
      "epoch": 0.8088737201365188,
      "grad_norm": 1.0546875,
      "learning_rate": 3.7484374821457725e-06,
      "loss": 0.4012739360332489,
      "mean_token_accuracy": 0.8609949946403503,
      "num_tokens": 1200176484.0,
      "step": 948
    },
    {
      "epoch": 0.8097269624573379,
      "grad_norm": 1.171875,
      "learning_rate": 3.7162622659285185e-06,
      "loss": 0.39989155530929565,
      "mean_token_accuracy": 0.8597093820571899,
      "num_tokens": 1201451019.0,
      "step": 949
    },
    {
      "epoch": 0.810580204778157,
      "grad_norm": 1.15625,
      "learning_rate": 3.684211586282611e-06,
      "loss": 0.42083740234375,
      "mean_token_accuracy": 0.8499430418014526,
      "num_tokens": 1202699734.0,
      "step": 950
    },
    {
      "epoch": 0.8114334470989761,
      "grad_norm": 1.0546875,
      "learning_rate": 3.6522856883287914e-06,
      "loss": 0.39814963936805725,
      "mean_token_accuracy": 0.857201337814331,
      "num_tokens": 1203943851.0,
      "step": 951
    },
    {
      "epoch": 0.8122866894197952,
      "grad_norm": 1.5859375,
      "learning_rate": 3.620484816233467e-06,
      "loss": 0.4180968999862671,
      "mean_token_accuracy": 0.8534501791000366,
      "num_tokens": 1205212333.0,
      "step": 952
    },
    {
      "epoch": 0.8131399317406144,
      "grad_norm": 1.078125,
      "learning_rate": 3.5888092132068676e-06,
      "loss": 0.4013528823852539,
      "mean_token_accuracy": 0.860994279384613,
      "num_tokens": 1206523742.0,
      "step": 953
    },
    {
      "epoch": 0.8139931740614335,
      "grad_norm": 1.0859375,
      "learning_rate": 3.5572591215011776e-06,
      "loss": 0.4065794348716736,
      "mean_token_accuracy": 0.8556302785873413,
      "num_tokens": 1207745467.0,
      "step": 954
    },
    {
      "epoch": 0.8148464163822525,
      "grad_norm": 0.9453125,
      "learning_rate": 3.5258347824086794e-06,
      "loss": 0.35373976826667786,
      "mean_token_accuracy": 0.8675440549850464,
      "num_tokens": 1209116128.0,
      "step": 955
    },
    {
      "epoch": 0.8156996587030717,
      "grad_norm": 1.109375,
      "learning_rate": 3.4945364362599144e-06,
      "loss": 0.4176662564277649,
      "mean_token_accuracy": 0.8541394472122192,
      "num_tokens": 1210407421.0,
      "step": 956
    },
    {
      "epoch": 0.8165529010238908,
      "grad_norm": 1.3359375,
      "learning_rate": 3.463364322421827e-06,
      "loss": 0.4394294023513794,
      "mean_token_accuracy": 0.8524001836776733,
      "num_tokens": 1211623959.0,
      "step": 957
    },
    {
      "epoch": 0.8174061433447098,
      "grad_norm": 1.0390625,
      "learning_rate": 3.4323186792959697e-06,
      "loss": 0.3878791928291321,
      "mean_token_accuracy": 0.862446665763855,
      "num_tokens": 1212904838.0,
      "step": 958
    },
    {
      "epoch": 0.818259385665529,
      "grad_norm": 1.2890625,
      "learning_rate": 3.40139974431664e-06,
      "loss": 0.42137008905410767,
      "mean_token_accuracy": 0.8541736602783203,
      "num_tokens": 1214178834.0,
      "step": 959
    },
    {
      "epoch": 0.8191126279863481,
      "grad_norm": 1.4375,
      "learning_rate": 3.3706077539490933e-06,
      "loss": 0.42008331418037415,
      "mean_token_accuracy": 0.8494697213172913,
      "num_tokens": 1215493562.0,
      "step": 960
    },
    {
      "epoch": 0.8199658703071673,
      "grad_norm": 1.03125,
      "learning_rate": 3.3399429436877285e-06,
      "loss": 0.4019176661968231,
      "mean_token_accuracy": 0.8599542379379272,
      "num_tokens": 1216704979.0,
      "step": 961
    },
    {
      "epoch": 0.8208191126279863,
      "grad_norm": 1.1953125,
      "learning_rate": 3.3094055480542608e-06,
      "loss": 0.43573957681655884,
      "mean_token_accuracy": 0.8482577800750732,
      "num_tokens": 1217916334.0,
      "step": 962
    },
    {
      "epoch": 0.8216723549488054,
      "grad_norm": 0.984375,
      "learning_rate": 3.2789958005959697e-06,
      "loss": 0.37323445081710815,
      "mean_token_accuracy": 0.866051435470581,
      "num_tokens": 1219164414.0,
      "step": 963
    },
    {
      "epoch": 0.8225255972696246,
      "grad_norm": 1.1015625,
      "learning_rate": 3.248713933883887e-06,
      "loss": 0.40365615487098694,
      "mean_token_accuracy": 0.8599064946174622,
      "num_tokens": 1220435836.0,
      "step": 964
    },
    {
      "epoch": 0.8233788395904437,
      "grad_norm": 2.515625,
      "learning_rate": 3.21856017951103e-06,
      "loss": 0.4225703179836273,
      "mean_token_accuracy": 0.8513222932815552,
      "num_tokens": 1221781481.0,
      "step": 965
    },
    {
      "epoch": 0.8242320819112628,
      "grad_norm": 1.046875,
      "learning_rate": 3.1885347680906075e-06,
      "loss": 0.38411808013916016,
      "mean_token_accuracy": 0.8658238649368286,
      "num_tokens": 1223006387.0,
      "step": 966
    },
    {
      "epoch": 0.8250853242320819,
      "grad_norm": 1.359375,
      "learning_rate": 3.158637929254287e-06,
      "loss": 0.44442933797836304,
      "mean_token_accuracy": 0.8445950746536255,
      "num_tokens": 1224324421.0,
      "step": 967
    },
    {
      "epoch": 0.825938566552901,
      "grad_norm": 1.2890625,
      "learning_rate": 3.1288698916504234e-06,
      "loss": 0.4349685609340668,
      "mean_token_accuracy": 0.8481582999229431,
      "num_tokens": 1225596864.0,
      "step": 968
    },
    {
      "epoch": 0.8267918088737202,
      "grad_norm": 1.0859375,
      "learning_rate": 3.099230882942306e-06,
      "loss": 0.40553486347198486,
      "mean_token_accuracy": 0.8576364517211914,
      "num_tokens": 1226843533.0,
      "step": 969
    },
    {
      "epoch": 0.8276450511945392,
      "grad_norm": 1.1328125,
      "learning_rate": 3.0697211298064333e-06,
      "loss": 0.4234634339809418,
      "mean_token_accuracy": 0.8476954102516174,
      "num_tokens": 1228053730.0,
      "step": 970
    },
    {
      "epoch": 0.8284982935153583,
      "grad_norm": 1.1015625,
      "learning_rate": 3.040340857930748e-06,
      "loss": 0.42053669691085815,
      "mean_token_accuracy": 0.853062629699707,
      "num_tokens": 1229306889.0,
      "step": 971
    },
    {
      "epoch": 0.8293515358361775,
      "grad_norm": 1.2265625,
      "learning_rate": 3.0110902920129505e-06,
      "loss": 0.39836177229881287,
      "mean_token_accuracy": 0.8588363528251648,
      "num_tokens": 1230482124.0,
      "step": 972
    },
    {
      "epoch": 0.8302047781569966,
      "grad_norm": 1.078125,
      "learning_rate": 2.981969655758756e-06,
      "loss": 0.4152633845806122,
      "mean_token_accuracy": 0.8553322553634644,
      "num_tokens": 1231746823.0,
      "step": 973
    },
    {
      "epoch": 0.8310580204778157,
      "grad_norm": 1.2421875,
      "learning_rate": 2.9529791718801903e-06,
      "loss": 0.4209632873535156,
      "mean_token_accuracy": 0.85206139087677,
      "num_tokens": 1233117986.0,
      "step": 974
    },
    {
      "epoch": 0.8319112627986348,
      "grad_norm": 1.1171875,
      "learning_rate": 2.9241190620938843e-06,
      "loss": 0.42453381419181824,
      "mean_token_accuracy": 0.8534893989562988,
      "num_tokens": 1234331897.0,
      "step": 975
    },
    {
      "epoch": 0.8327645051194539,
      "grad_norm": 1.0234375,
      "learning_rate": 2.895389547119376e-06,
      "loss": 0.4009641706943512,
      "mean_token_accuracy": 0.8575571179389954,
      "num_tokens": 1235589165.0,
      "step": 976
    },
    {
      "epoch": 0.8336177474402731,
      "grad_norm": 1.078125,
      "learning_rate": 2.8667908466774276e-06,
      "loss": 0.3543548583984375,
      "mean_token_accuracy": 0.866879403591156,
      "num_tokens": 1236841197.0,
      "step": 977
    },
    {
      "epoch": 0.8344709897610921,
      "grad_norm": 1.1484375,
      "learning_rate": 2.838323179488349e-06,
      "loss": 0.40018177032470703,
      "mean_token_accuracy": 0.857872724533081,
      "num_tokens": 1238066007.0,
      "step": 978
    },
    {
      "epoch": 0.8353242320819113,
      "grad_norm": 1.1640625,
      "learning_rate": 2.809986763270318e-06,
      "loss": 0.44221389293670654,
      "mean_token_accuracy": 0.8449406027793884,
      "num_tokens": 1239351884.0,
      "step": 979
    },
    {
      "epoch": 0.8361774744027304,
      "grad_norm": 1.1015625,
      "learning_rate": 2.7817818147377052e-06,
      "loss": 0.43132883310317993,
      "mean_token_accuracy": 0.8490229249000549,
      "num_tokens": 1240694543.0,
      "step": 980
    },
    {
      "epoch": 0.8370307167235495,
      "grad_norm": 1.1015625,
      "learning_rate": 2.753708549599443e-06,
      "loss": 0.3917606472969055,
      "mean_token_accuracy": 0.8625988960266113,
      "num_tokens": 1242003614.0,
      "step": 981
    },
    {
      "epoch": 0.8378839590443686,
      "grad_norm": 1.171875,
      "learning_rate": 2.725767182557355e-06,
      "loss": 0.41746461391448975,
      "mean_token_accuracy": 0.8557800650596619,
      "num_tokens": 1243242943.0,
      "step": 982
    },
    {
      "epoch": 0.8387372013651877,
      "grad_norm": 1.0859375,
      "learning_rate": 2.6979579273045242e-06,
      "loss": 0.3857690095901489,
      "mean_token_accuracy": 0.8643182516098022,
      "num_tokens": 1244477669.0,
      "step": 983
    },
    {
      "epoch": 0.8395904436860068,
      "grad_norm": 1.0625,
      "learning_rate": 2.6702809965236398e-06,
      "loss": 0.4208175539970398,
      "mean_token_accuracy": 0.8556832075119019,
      "num_tokens": 1245680016.0,
      "step": 984
    },
    {
      "epoch": 0.840443686006826,
      "grad_norm": 1.125,
      "learning_rate": 2.642736601885401e-06,
      "loss": 0.41642889380455017,
      "mean_token_accuracy": 0.8503470420837402,
      "num_tokens": 1246835939.0,
      "step": 985
    },
    {
      "epoch": 0.841296928327645,
      "grad_norm": 1.15625,
      "learning_rate": 2.615324954046876e-06,
      "loss": 0.44190600514411926,
      "mean_token_accuracy": 0.8462138175964355,
      "num_tokens": 1248079665.0,
      "step": 986
    },
    {
      "epoch": 0.8421501706484642,
      "grad_norm": 1.1171875,
      "learning_rate": 2.5880462626499016e-06,
      "loss": 0.4308258295059204,
      "mean_token_accuracy": 0.8490502834320068,
      "num_tokens": 1249312810.0,
      "step": 987
    },
    {
      "epoch": 0.8430034129692833,
      "grad_norm": 1.03125,
      "learning_rate": 2.5609007363194738e-06,
      "loss": 0.412384033203125,
      "mean_token_accuracy": 0.8515262007713318,
      "num_tokens": 1250648782.0,
      "step": 988
    },
    {
      "epoch": 0.8438566552901023,
      "grad_norm": 1.2734375,
      "learning_rate": 2.533888582662145e-06,
      "loss": 0.4279044270515442,
      "mean_token_accuracy": 0.8533958792686462,
      "num_tokens": 1251871954.0,
      "step": 989
    },
    {
      "epoch": 0.8447098976109215,
      "grad_norm": 1.0546875,
      "learning_rate": 2.5070100082644587e-06,
      "loss": 0.40935346484184265,
      "mean_token_accuracy": 0.8566968441009521,
      "num_tokens": 1253136080.0,
      "step": 990
    },
    {
      "epoch": 0.8455631399317406,
      "grad_norm": 1.0390625,
      "learning_rate": 2.4802652186913466e-06,
      "loss": 0.418900728225708,
      "mean_token_accuracy": 0.8517787456512451,
      "num_tokens": 1254369083.0,
      "step": 991
    },
    {
      "epoch": 0.8464163822525598,
      "grad_norm": 1.15625,
      "learning_rate": 2.453654418484579e-06,
      "loss": 0.39912766218185425,
      "mean_token_accuracy": 0.8599436283111572,
      "num_tokens": 1255592025.0,
      "step": 992
    },
    {
      "epoch": 0.8472696245733788,
      "grad_norm": 1.15625,
      "learning_rate": 2.427177811161168e-06,
      "loss": 0.440998375415802,
      "mean_token_accuracy": 0.8465694189071655,
      "num_tokens": 1256878183.0,
      "step": 993
    },
    {
      "epoch": 0.8481228668941979,
      "grad_norm": 1.171875,
      "learning_rate": 2.4008355992118417e-06,
      "loss": 0.4249975085258484,
      "mean_token_accuracy": 0.8504351377487183,
      "num_tokens": 1258200630.0,
      "step": 994
    },
    {
      "epoch": 0.8489761092150171,
      "grad_norm": 1.109375,
      "learning_rate": 2.374627984099489e-06,
      "loss": 0.44977042078971863,
      "mean_token_accuracy": 0.8427711725234985,
      "num_tokens": 1259470268.0,
      "step": 995
    },
    {
      "epoch": 0.8498293515358362,
      "grad_norm": 1.1796875,
      "learning_rate": 2.348555166257611e-06,
      "loss": 0.4250715970993042,
      "mean_token_accuracy": 0.8523235321044922,
      "num_tokens": 1260697331.0,
      "step": 996
    },
    {
      "epoch": 0.8506825938566553,
      "grad_norm": 1.046875,
      "learning_rate": 2.322617345088778e-06,
      "loss": 0.3951297402381897,
      "mean_token_accuracy": 0.8583654165267944,
      "num_tokens": 1261897853.0,
      "step": 997
    },
    {
      "epoch": 0.8515358361774744,
      "grad_norm": 1.234375,
      "learning_rate": 2.2968147189631383e-06,
      "loss": 0.4574107527732849,
      "mean_token_accuracy": 0.8438763618469238,
      "num_tokens": 1263125664.0,
      "step": 998
    },
    {
      "epoch": 0.8523890784982935,
      "grad_norm": 1.625,
      "learning_rate": 2.271147485216867e-06,
      "loss": 0.4166111648082733,
      "mean_token_accuracy": 0.8523375391960144,
      "num_tokens": 1264364524.0,
      "step": 999
    },
    {
      "epoch": 0.8532423208191127,
      "grad_norm": 1.1015625,
      "learning_rate": 2.2456158401506787e-06,
      "loss": 0.4304125905036926,
      "mean_token_accuracy": 0.8463189601898193,
      "num_tokens": 1265586220.0,
      "step": 1000
    },
    {
      "epoch": 0.8540955631399317,
      "grad_norm": 1.515625,
      "learning_rate": 2.2202199790283084e-06,
      "loss": 0.4086989760398865,
      "mean_token_accuracy": 0.8569351434707642,
      "num_tokens": 1266891211.0,
      "step": 1001
    },
    {
      "epoch": 0.8549488054607508,
      "grad_norm": 1.03125,
      "learning_rate": 2.194960096075036e-06,
      "loss": 0.4013961851596832,
      "mean_token_accuracy": 0.8564854860305786,
      "num_tokens": 1268123789.0,
      "step": 1002
    },
    {
      "epoch": 0.85580204778157,
      "grad_norm": 1.1171875,
      "learning_rate": 2.1698363844761894e-06,
      "loss": 0.4133029878139496,
      "mean_token_accuracy": 0.8547881245613098,
      "num_tokens": 1269312381.0,
      "step": 1003
    },
    {
      "epoch": 0.856655290102389,
      "grad_norm": 1.0703125,
      "learning_rate": 2.1448490363756737e-06,
      "loss": 0.3964317739009857,
      "mean_token_accuracy": 0.8576831817626953,
      "num_tokens": 1270587937.0,
      "step": 1004
    },
    {
      "epoch": 0.8575085324232082,
      "grad_norm": 1.0390625,
      "learning_rate": 2.119998242874488e-06,
      "loss": 0.4124843180179596,
      "mean_token_accuracy": 0.8553404211997986,
      "num_tokens": 1271817590.0,
      "step": 1005
    },
    {
      "epoch": 0.8583617747440273,
      "grad_norm": 1.140625,
      "learning_rate": 2.0952841940292812e-06,
      "loss": 0.42988139390945435,
      "mean_token_accuracy": 0.8529845476150513,
      "num_tokens": 1273104877.0,
      "step": 1006
    },
    {
      "epoch": 0.8592150170648464,
      "grad_norm": 1.140625,
      "learning_rate": 2.0707070788508954e-06,
      "loss": 0.4052852988243103,
      "mean_token_accuracy": 0.8555690050125122,
      "num_tokens": 1274460837.0,
      "step": 1007
    },
    {
      "epoch": 0.8600682593856656,
      "grad_norm": 1.078125,
      "learning_rate": 2.046267085302911e-06,
      "loss": 0.3990178108215332,
      "mean_token_accuracy": 0.8576197028160095,
      "num_tokens": 1275726752.0,
      "step": 1008
    },
    {
      "epoch": 0.8609215017064846,
      "grad_norm": 0.98046875,
      "learning_rate": 2.021964400300216e-06,
      "loss": 0.39786186814308167,
      "mean_token_accuracy": 0.857766330242157,
      "num_tokens": 1277000762.0,
      "step": 1009
    },
    {
      "epoch": 0.8617747440273038,
      "grad_norm": 1.109375,
      "learning_rate": 1.9977992097075696e-06,
      "loss": 0.43053025007247925,
      "mean_token_accuracy": 0.8473188281059265,
      "num_tokens": 1278227807.0,
      "step": 1010
    },
    {
      "epoch": 0.8626279863481229,
      "grad_norm": 1.125,
      "learning_rate": 1.9737716983381894e-06,
      "loss": 0.4364871382713318,
      "mean_token_accuracy": 0.8484715223312378,
      "num_tokens": 1279422117.0,
      "step": 1011
    },
    {
      "epoch": 0.863481228668942,
      "grad_norm": 1.015625,
      "learning_rate": 1.949882049952343e-06,
      "loss": 0.36093631386756897,
      "mean_token_accuracy": 0.8710545301437378,
      "num_tokens": 1280720511.0,
      "step": 1012
    },
    {
      "epoch": 0.8643344709897611,
      "grad_norm": 1.3125,
      "learning_rate": 1.926130447255925e-06,
      "loss": 0.4127407670021057,
      "mean_token_accuracy": 0.8526710867881775,
      "num_tokens": 1281961693.0,
      "step": 1013
    },
    {
      "epoch": 0.8651877133105802,
      "grad_norm": 1.09375,
      "learning_rate": 1.9025170718990638e-06,
      "loss": 0.42149120569229126,
      "mean_token_accuracy": 0.8562800884246826,
      "num_tokens": 1283261194.0,
      "step": 1014
    },
    {
      "epoch": 0.8660409556313993,
      "grad_norm": 1.0546875,
      "learning_rate": 1.879042104474751e-06,
      "loss": 0.43093425035476685,
      "mean_token_accuracy": 0.8494300842285156,
      "num_tokens": 1284522007.0,
      "step": 1015
    },
    {
      "epoch": 0.8668941979522184,
      "grad_norm": 1.0703125,
      "learning_rate": 1.8557057245174381e-06,
      "loss": 0.400747686624527,
      "mean_token_accuracy": 0.8594410419464111,
      "num_tokens": 1285818166.0,
      "step": 1016
    },
    {
      "epoch": 0.8677474402730375,
      "grad_norm": 1.1640625,
      "learning_rate": 1.8325081105016784e-06,
      "loss": 0.41760867834091187,
      "mean_token_accuracy": 0.8533364534378052,
      "num_tokens": 1287145127.0,
      "step": 1017
    },
    {
      "epoch": 0.8686006825938567,
      "grad_norm": 1.1328125,
      "learning_rate": 1.8094494398407425e-06,
      "loss": 0.41143038868904114,
      "mean_token_accuracy": 0.857535183429718,
      "num_tokens": 1288414097.0,
      "step": 1018
    },
    {
      "epoch": 0.8694539249146758,
      "grad_norm": 1.0546875,
      "learning_rate": 1.786529888885289e-06,
      "loss": 0.43416517972946167,
      "mean_token_accuracy": 0.8471888899803162,
      "num_tokens": 1289723148.0,
      "step": 1019
    },
    {
      "epoch": 0.8703071672354948,
      "grad_norm": 1.125,
      "learning_rate": 1.763749632921996e-06,
      "loss": 0.40093326568603516,
      "mean_token_accuracy": 0.8597019910812378,
      "num_tokens": 1291080884.0,
      "step": 1020
    },
    {
      "epoch": 0.871160409556314,
      "grad_norm": 0.98828125,
      "learning_rate": 1.7411088461722281e-06,
      "loss": 0.39587530493736267,
      "mean_token_accuracy": 0.8568847179412842,
      "num_tokens": 1292345992.0,
      "step": 1021
    },
    {
      "epoch": 0.8720136518771331,
      "grad_norm": 1.0546875,
      "learning_rate": 1.7186077017906932e-06,
      "loss": 0.41542696952819824,
      "mean_token_accuracy": 0.8570137023925781,
      "num_tokens": 1293528220.0,
      "step": 1022
    },
    {
      "epoch": 0.8728668941979523,
      "grad_norm": 1.109375,
      "learning_rate": 1.6962463718641364e-06,
      "loss": 0.43250441551208496,
      "mean_token_accuracy": 0.8443676233291626,
      "num_tokens": 1294739908.0,
      "step": 1023
    },
    {
      "epoch": 0.8737201365187713,
      "grad_norm": 1.1015625,
      "learning_rate": 1.6740250274100112e-06,
      "loss": 0.43916285037994385,
      "mean_token_accuracy": 0.8469216227531433,
      "num_tokens": 1296016971.0,
      "step": 1024
    },
    {
      "epoch": 0.8745733788395904,
      "grad_norm": 1.2734375,
      "learning_rate": 1.6519438383751784e-06,
      "loss": 0.40925899147987366,
      "mean_token_accuracy": 0.858592689037323,
      "num_tokens": 1297329228.0,
      "step": 1025
    },
    {
      "epoch": 0.8754266211604096,
      "grad_norm": 1.015625,
      "learning_rate": 1.6300029736345903e-06,
      "loss": 0.38524508476257324,
      "mean_token_accuracy": 0.8611131906509399,
      "num_tokens": 1298600921.0,
      "step": 1026
    },
    {
      "epoch": 0.8762798634812287,
      "grad_norm": 1.1953125,
      "learning_rate": 1.6082026009900276e-06,
      "loss": 0.42935067415237427,
      "mean_token_accuracy": 0.8497625589370728,
      "num_tokens": 1299880939.0,
      "step": 1027
    },
    {
      "epoch": 0.8771331058020477,
      "grad_norm": 1.1484375,
      "learning_rate": 1.5865428871687893e-06,
      "loss": 0.39867091178894043,
      "mean_token_accuracy": 0.860588014125824,
      "num_tokens": 1301158125.0,
      "step": 1028
    },
    {
      "epoch": 0.8779863481228669,
      "grad_norm": 1.171875,
      "learning_rate": 1.5650239978224346e-06,
      "loss": 0.4330579340457916,
      "mean_token_accuracy": 0.8511508107185364,
      "num_tokens": 1302423724.0,
      "step": 1029
    },
    {
      "epoch": 0.878839590443686,
      "grad_norm": 1.1015625,
      "learning_rate": 1.543646097525502e-06,
      "loss": 0.4383401870727539,
      "mean_token_accuracy": 0.8486262559890747,
      "num_tokens": 1303647150.0,
      "step": 1030
    },
    {
      "epoch": 0.8796928327645052,
      "grad_norm": 1.1484375,
      "learning_rate": 1.5224093497742654e-06,
      "loss": 0.4085264801979065,
      "mean_token_accuracy": 0.8545894622802734,
      "num_tokens": 1304894413.0,
      "step": 1031
    },
    {
      "epoch": 0.8805460750853242,
      "grad_norm": 1.046875,
      "learning_rate": 1.501313916985474e-06,
      "loss": 0.38453659415245056,
      "mean_token_accuracy": 0.8609013557434082,
      "num_tokens": 1306172123.0,
      "step": 1032
    },
    {
      "epoch": 0.8813993174061433,
      "grad_norm": 1.140625,
      "learning_rate": 1.4803599604951148e-06,
      "loss": 0.3898594379425049,
      "mean_token_accuracy": 0.8580502867698669,
      "num_tokens": 1307371807.0,
      "step": 1033
    },
    {
      "epoch": 0.8822525597269625,
      "grad_norm": 1.0546875,
      "learning_rate": 1.459547640557175e-06,
      "loss": 0.39247265458106995,
      "mean_token_accuracy": 0.8626889586448669,
      "num_tokens": 1308593560.0,
      "step": 1034
    },
    {
      "epoch": 0.8831058020477816,
      "grad_norm": 1.1015625,
      "learning_rate": 1.4388771163424142e-06,
      "loss": 0.4038049876689911,
      "mean_token_accuracy": 0.8562614917755127,
      "num_tokens": 1309787549.0,
      "step": 1035
    },
    {
      "epoch": 0.8839590443686007,
      "grad_norm": 1.0859375,
      "learning_rate": 1.418348545937156e-06,
      "loss": 0.4087333083152771,
      "mean_token_accuracy": 0.8542405366897583,
      "num_tokens": 1311041899.0,
      "step": 1036
    },
    {
      "epoch": 0.8848122866894198,
      "grad_norm": 1.1953125,
      "learning_rate": 1.397962086342075e-06,
      "loss": 0.4089187681674957,
      "mean_token_accuracy": 0.8553030490875244,
      "num_tokens": 1312315811.0,
      "step": 1037
    },
    {
      "epoch": 0.8856655290102389,
      "grad_norm": 1.03125,
      "learning_rate": 1.3777178934709956e-06,
      "loss": 0.40214112401008606,
      "mean_token_accuracy": 0.8563858270645142,
      "num_tokens": 1313538828.0,
      "step": 1038
    },
    {
      "epoch": 0.886518771331058,
      "grad_norm": 1.109375,
      "learning_rate": 1.3576161221496874e-06,
      "loss": 0.388289213180542,
      "mean_token_accuracy": 0.8620519042015076,
      "num_tokens": 1314799368.0,
      "step": 1039
    },
    {
      "epoch": 0.8873720136518771,
      "grad_norm": 1.0234375,
      "learning_rate": 1.3376569261147077e-06,
      "loss": 0.4019544720649719,
      "mean_token_accuracy": 0.8554121255874634,
      "num_tokens": 1316018177.0,
      "step": 1040
    },
    {
      "epoch": 0.8882252559726962,
      "grad_norm": 1.234375,
      "learning_rate": 1.317840458012203e-06,
      "loss": 0.4305703639984131,
      "mean_token_accuracy": 0.847967803478241,
      "num_tokens": 1317196543.0,
      "step": 1041
    },
    {
      "epoch": 0.8890784982935154,
      "grad_norm": 1.1171875,
      "learning_rate": 1.298166869396753e-06,
      "loss": 0.4524891674518585,
      "mean_token_accuracy": 0.8421063423156738,
      "num_tokens": 1318514980.0,
      "step": 1042
    },
    {
      "epoch": 0.8899317406143344,
      "grad_norm": 1.140625,
      "learning_rate": 1.2786363107302014e-06,
      "loss": 0.4293751120567322,
      "mean_token_accuracy": 0.8512202501296997,
      "num_tokens": 1319800709.0,
      "step": 1043
    },
    {
      "epoch": 0.8907849829351536,
      "grad_norm": 1.25,
      "learning_rate": 1.259248931380519e-06,
      "loss": 0.4416761100292206,
      "mean_token_accuracy": 0.8447776436805725,
      "num_tokens": 1321004506.0,
      "step": 1044
    },
    {
      "epoch": 0.8916382252559727,
      "grad_norm": 1.078125,
      "learning_rate": 1.240004879620651e-06,
      "loss": 0.37688252329826355,
      "mean_token_accuracy": 0.866811990737915,
      "num_tokens": 1322240311.0,
      "step": 1045
    },
    {
      "epoch": 0.8924914675767918,
      "grad_norm": 0.98828125,
      "learning_rate": 1.2209043026273882e-06,
      "loss": 0.39067643880844116,
      "mean_token_accuracy": 0.8634235858917236,
      "num_tokens": 1323580756.0,
      "step": 1046
    },
    {
      "epoch": 0.893344709897611,
      "grad_norm": 1.0859375,
      "learning_rate": 1.20194734648023e-06,
      "loss": 0.3935970067977905,
      "mean_token_accuracy": 0.8653815984725952,
      "num_tokens": 1324802879.0,
      "step": 1047
    },
    {
      "epoch": 0.89419795221843,
      "grad_norm": 1.265625,
      "learning_rate": 1.1831341561602882e-06,
      "loss": 0.4394611716270447,
      "mean_token_accuracy": 0.8460792303085327,
      "num_tokens": 1326052523.0,
      "step": 1048
    },
    {
      "epoch": 0.8950511945392492,
      "grad_norm": 1.15625,
      "learning_rate": 1.1644648755491583e-06,
      "loss": 0.43212664127349854,
      "mean_token_accuracy": 0.8505580425262451,
      "num_tokens": 1327366450.0,
      "step": 1049
    },
    {
      "epoch": 0.8959044368600683,
      "grad_norm": 1.0703125,
      "learning_rate": 1.1459396474278383e-06,
      "loss": 0.38802093267440796,
      "mean_token_accuracy": 0.8616904020309448,
      "num_tokens": 1328590316.0,
      "step": 1050
    },
    {
      "epoch": 0.8967576791808873,
      "grad_norm": 1.046875,
      "learning_rate": 1.12755861347561e-06,
      "loss": 0.41259703040122986,
      "mean_token_accuracy": 0.8539952039718628,
      "num_tokens": 1329818588.0,
      "step": 1051
    },
    {
      "epoch": 0.8976109215017065,
      "grad_norm": 1.2890625,
      "learning_rate": 1.1093219142689793e-06,
      "loss": 0.3986208438873291,
      "mean_token_accuracy": 0.8598235845565796,
      "num_tokens": 1331111874.0,
      "step": 1052
    },
    {
      "epoch": 0.8984641638225256,
      "grad_norm": 1.2421875,
      "learning_rate": 1.0912296892805952e-06,
      "loss": 0.4144875407218933,
      "mean_token_accuracy": 0.8540969491004944,
      "num_tokens": 1332330197.0,
      "step": 1053
    },
    {
      "epoch": 0.8993174061433447,
      "grad_norm": 1.0234375,
      "learning_rate": 1.0732820768781794e-06,
      "loss": 0.3969424068927765,
      "mean_token_accuracy": 0.8627074956893921,
      "num_tokens": 1333620432.0,
      "step": 1054
    },
    {
      "epoch": 0.9001706484641638,
      "grad_norm": 1.0234375,
      "learning_rate": 1.0554792143234627e-06,
      "loss": 0.39462679624557495,
      "mean_token_accuracy": 0.8640392422676086,
      "num_tokens": 1334835225.0,
      "step": 1055
    },
    {
      "epoch": 0.9010238907849829,
      "grad_norm": 1.0234375,
      "learning_rate": 1.0378212377711505e-06,
      "loss": 0.40885818004608154,
      "mean_token_accuracy": 0.8579808473587036,
      "num_tokens": 1336119543.0,
      "step": 1056
    },
    {
      "epoch": 0.9018771331058021,
      "grad_norm": 1.1015625,
      "learning_rate": 1.0203082822678701e-06,
      "loss": 0.4025481343269348,
      "mean_token_accuracy": 0.8612590432167053,
      "num_tokens": 1337366317.0,
      "step": 1057
    },
    {
      "epoch": 0.9027303754266212,
      "grad_norm": 1.1640625,
      "learning_rate": 1.00294048175114e-06,
      "loss": 0.39342623949050903,
      "mean_token_accuracy": 0.8591340780258179,
      "num_tokens": 1338591620.0,
      "step": 1058
    },
    {
      "epoch": 0.9035836177474402,
      "grad_norm": 1.03125,
      "learning_rate": 9.8571796904835e-07,
      "loss": 0.41190406680107117,
      "mean_token_accuracy": 0.8539581298828125,
      "num_tokens": 1339847563.0,
      "step": 1059
    },
    {
      "epoch": 0.9044368600682594,
      "grad_norm": 1.1484375,
      "learning_rate": 9.686408758757326e-07,
      "loss": 0.4164518713951111,
      "mean_token_accuracy": 0.8528033494949341,
      "num_tokens": 1341151104.0,
      "step": 1060
    },
    {
      "epoch": 0.9052901023890785,
      "grad_norm": 1.0546875,
      "learning_rate": 9.517093328373739e-07,
      "loss": 0.40417972207069397,
      "mean_token_accuracy": 0.8572325706481934,
      "num_tokens": 1342434376.0,
      "step": 1061
    },
    {
      "epoch": 0.9061433447098977,
      "grad_norm": 1.1015625,
      "learning_rate": 9.349234694242026e-07,
      "loss": 0.41262781620025635,
      "mean_token_accuracy": 0.8551879525184631,
      "num_tokens": 1343786103.0,
      "step": 1062
    },
    {
      "epoch": 0.9069965870307167,
      "grad_norm": 1.1796875,
      "learning_rate": 9.182834140130037e-07,
      "loss": 0.41547274589538574,
      "mean_token_accuracy": 0.8521822094917297,
      "num_tokens": 1344957840.0,
      "step": 1063
    },
    {
      "epoch": 0.9078498293515358,
      "grad_norm": 1.296875,
      "learning_rate": 9.017892938654293e-07,
      "loss": 0.44339555501937866,
      "mean_token_accuracy": 0.8452877998352051,
      "num_tokens": 1346229605.0,
      "step": 1064
    },
    {
      "epoch": 0.908703071672355,
      "grad_norm": 1.125,
      "learning_rate": 8.854412351270402e-07,
      "loss": 0.41897261142730713,
      "mean_token_accuracy": 0.8486167192459106,
      "num_tokens": 1347326472.0,
      "step": 1065
    },
    {
      "epoch": 0.909556313993174,
      "grad_norm": 1.0078125,
      "learning_rate": 8.692393628263285e-07,
      "loss": 0.4091816544532776,
      "mean_token_accuracy": 0.859234094619751,
      "num_tokens": 1348574021.0,
      "step": 1066
    },
    {
      "epoch": 0.9104095563139932,
      "grad_norm": 1.0859375,
      "learning_rate": 8.531838008737692e-07,
      "loss": 0.40822353959083557,
      "mean_token_accuracy": 0.8555095195770264,
      "num_tokens": 1349926589.0,
      "step": 1067
    },
    {
      "epoch": 0.9112627986348123,
      "grad_norm": 0.97265625,
      "learning_rate": 8.372746720608594e-07,
      "loss": 0.3842357397079468,
      "mean_token_accuracy": 0.8605822324752808,
      "num_tokens": 1351166510.0,
      "step": 1068
    },
    {
      "epoch": 0.9121160409556314,
      "grad_norm": 0.9765625,
      "learning_rate": 8.215120980591984e-07,
      "loss": 0.39240604639053345,
      "mean_token_accuracy": 0.8588828444480896,
      "num_tokens": 1352471118.0,
      "step": 1069
    },
    {
      "epoch": 0.9129692832764505,
      "grad_norm": 1.015625,
      "learning_rate": 8.058961994195447e-07,
      "loss": 0.3981377184391022,
      "mean_token_accuracy": 0.865482747554779,
      "num_tokens": 1353728444.0,
      "step": 1070
    },
    {
      "epoch": 0.9138225255972696,
      "grad_norm": 1.171875,
      "learning_rate": 7.904270955708959e-07,
      "loss": 0.44214263558387756,
      "mean_token_accuracy": 0.8446980118751526,
      "num_tokens": 1355049529.0,
      "step": 1071
    },
    {
      "epoch": 0.9146757679180887,
      "grad_norm": 1.0546875,
      "learning_rate": 7.751049048195725e-07,
      "loss": 0.40574491024017334,
      "mean_token_accuracy": 0.8560816049575806,
      "num_tokens": 1356321317.0,
      "step": 1072
    },
    {
      "epoch": 0.9155290102389079,
      "grad_norm": 1.015625,
      "learning_rate": 7.599297443483244e-07,
      "loss": 0.405978798866272,
      "mean_token_accuracy": 0.8522521257400513,
      "num_tokens": 1357572796.0,
      "step": 1073
    },
    {
      "epoch": 0.9163822525597269,
      "grad_norm": 1.078125,
      "learning_rate": 7.449017302154193e-07,
      "loss": 0.40216460824012756,
      "mean_token_accuracy": 0.859719455242157,
      "num_tokens": 1358766851.0,
      "step": 1074
    },
    {
      "epoch": 0.9172354948805461,
      "grad_norm": 1.09375,
      "learning_rate": 7.300209773537737e-07,
      "loss": 0.39764493703842163,
      "mean_token_accuracy": 0.8603819608688354,
      "num_tokens": 1360068992.0,
      "step": 1075
    },
    {
      "epoch": 0.9180887372013652,
      "grad_norm": 1.1484375,
      "learning_rate": 7.15287599570047e-07,
      "loss": 0.40513768792152405,
      "mean_token_accuracy": 0.8540180921554565,
      "num_tokens": 1361338436.0,
      "step": 1076
    },
    {
      "epoch": 0.9189419795221843,
      "grad_norm": 1.0234375,
      "learning_rate": 7.007017095438029e-07,
      "loss": 0.3950877785682678,
      "mean_token_accuracy": 0.863223671913147,
      "num_tokens": 1362647565.0,
      "step": 1077
    },
    {
      "epoch": 0.9197952218430034,
      "grad_norm": 1.078125,
      "learning_rate": 6.862634188266204e-07,
      "loss": 0.4004179537296295,
      "mean_token_accuracy": 0.8556629419326782,
      "num_tokens": 1363877117.0,
      "step": 1078
    },
    {
      "epoch": 0.9206484641638225,
      "grad_norm": 1.1171875,
      "learning_rate": 6.719728378412615e-07,
      "loss": 0.40483012795448303,
      "mean_token_accuracy": 0.8518319129943848,
      "num_tokens": 1365188315.0,
      "step": 1079
    },
    {
      "epoch": 0.9215017064846417,
      "grad_norm": 1.1328125,
      "learning_rate": 6.57830075880812e-07,
      "loss": 0.405880331993103,
      "mean_token_accuracy": 0.857725977897644,
      "num_tokens": 1366548260.0,
      "step": 1080
    },
    {
      "epoch": 0.9223549488054608,
      "grad_norm": 1.109375,
      "learning_rate": 6.438352411078486e-07,
      "loss": 0.38886964321136475,
      "mean_token_accuracy": 0.8613330125808716,
      "num_tokens": 1367789887.0,
      "step": 1081
    },
    {
      "epoch": 0.9232081911262798,
      "grad_norm": 1.1171875,
      "learning_rate": 6.299884405536172e-07,
      "loss": 0.41736093163490295,
      "mean_token_accuracy": 0.851046085357666,
      "num_tokens": 1369127170.0,
      "step": 1082
    },
    {
      "epoch": 0.924061433447099,
      "grad_norm": 1.0546875,
      "learning_rate": 6.16289780117214e-07,
      "loss": 0.361106812953949,
      "mean_token_accuracy": 0.8683589696884155,
      "num_tokens": 1370434303.0,
      "step": 1083
    },
    {
      "epoch": 0.9249146757679181,
      "grad_norm": 1.1640625,
      "learning_rate": 6.027393645647728e-07,
      "loss": 0.4086286425590515,
      "mean_token_accuracy": 0.858412504196167,
      "num_tokens": 1371695892.0,
      "step": 1084
    },
    {
      "epoch": 0.9257679180887372,
      "grad_norm": 1.0546875,
      "learning_rate": 5.89337297528656e-07,
      "loss": 0.40580248832702637,
      "mean_token_accuracy": 0.8547213077545166,
      "num_tokens": 1372980153.0,
      "step": 1085
    },
    {
      "epoch": 0.9266211604095563,
      "grad_norm": 1.03125,
      "learning_rate": 5.760836815066739e-07,
      "loss": 0.3815538287162781,
      "mean_token_accuracy": 0.8602560758590698,
      "num_tokens": 1374201337.0,
      "step": 1086
    },
    {
      "epoch": 0.9274744027303754,
      "grad_norm": 1.125,
      "learning_rate": 5.629786178613028e-07,
      "loss": 0.43255531787872314,
      "mean_token_accuracy": 0.8486387729644775,
      "num_tokens": 1375514375.0,
      "step": 1087
    },
    {
      "epoch": 0.9283276450511946,
      "grad_norm": 1.1171875,
      "learning_rate": 5.500222068188988e-07,
      "loss": 0.4407997131347656,
      "mean_token_accuracy": 0.8480604290962219,
      "num_tokens": 1376791785.0,
      "step": 1088
    },
    {
      "epoch": 0.9291808873720137,
      "grad_norm": 1.078125,
      "learning_rate": 5.372145474689295e-07,
      "loss": 0.42206019163131714,
      "mean_token_accuracy": 0.8518239855766296,
      "num_tokens": 1378111071.0,
      "step": 1089
    },
    {
      "epoch": 0.9300341296928327,
      "grad_norm": 1.15625,
      "learning_rate": 5.245557377632327e-07,
      "loss": 0.4325522780418396,
      "mean_token_accuracy": 0.8482134938240051,
      "num_tokens": 1379434865.0,
      "step": 1090
    },
    {
      "epoch": 0.9308873720136519,
      "grad_norm": 1.2734375,
      "learning_rate": 5.120458745152568e-07,
      "loss": 0.43331050872802734,
      "mean_token_accuracy": 0.8479718565940857,
      "num_tokens": 1380759638.0,
      "step": 1091
    },
    {
      "epoch": 0.931740614334471,
      "grad_norm": 1.2109375,
      "learning_rate": 4.996850533993147e-07,
      "loss": 0.4390457272529602,
      "mean_token_accuracy": 0.8461377024650574,
      "num_tokens": 1382064644.0,
      "step": 1092
    },
    {
      "epoch": 0.9325938566552902,
      "grad_norm": 1.0234375,
      "learning_rate": 4.874733689498645e-07,
      "loss": 0.419075608253479,
      "mean_token_accuracy": 0.8520498275756836,
      "num_tokens": 1383288837.0,
      "step": 1093
    },
    {
      "epoch": 0.9334470989761092,
      "grad_norm": 1.125,
      "learning_rate": 4.754109145607766e-07,
      "loss": 0.41389209032058716,
      "mean_token_accuracy": 0.8561879396438599,
      "num_tokens": 1384547270.0,
      "step": 1094
    },
    {
      "epoch": 0.9343003412969283,
      "grad_norm": 1.1015625,
      "learning_rate": 4.6349778248462317e-07,
      "loss": 0.4235087037086487,
      "mean_token_accuracy": 0.8499725461006165,
      "num_tokens": 1385849167.0,
      "step": 1095
    },
    {
      "epoch": 0.9351535836177475,
      "grad_norm": 1.296875,
      "learning_rate": 4.5173406383197673e-07,
      "loss": 0.42252159118652344,
      "mean_token_accuracy": 0.8510680198669434,
      "num_tokens": 1387196563.0,
      "step": 1096
    },
    {
      "epoch": 0.9360068259385665,
      "grad_norm": 1.2578125,
      "learning_rate": 4.4011984857070367e-07,
      "loss": 0.3967806100845337,
      "mean_token_accuracy": 0.8574002981185913,
      "num_tokens": 1388491657.0,
      "step": 1097
    },
    {
      "epoch": 0.9368600682593856,
      "grad_norm": 1.1875,
      "learning_rate": 4.286552255252874e-07,
      "loss": 0.4319368600845337,
      "mean_token_accuracy": 0.850172758102417,
      "num_tokens": 1389684515.0,
      "step": 1098
    },
    {
      "epoch": 0.9377133105802048,
      "grad_norm": 1.3046875,
      "learning_rate": 4.1734028237613967e-07,
      "loss": 0.45046818256378174,
      "mean_token_accuracy": 0.8437384366989136,
      "num_tokens": 1390990258.0,
      "step": 1099
    },
    {
      "epoch": 0.9385665529010239,
      "grad_norm": 1.2890625,
      "learning_rate": 4.061751056589325e-07,
      "loss": 0.37912851572036743,
      "mean_token_accuracy": 0.8644329309463501,
      "num_tokens": 1392280988.0,
      "step": 1100
    },
    {
      "epoch": 0.939419795221843,
      "grad_norm": 1.0546875,
      "learning_rate": 3.9515978076394293e-07,
      "loss": 0.403181791305542,
      "mean_token_accuracy": 0.8551227450370789,
      "num_tokens": 1393481467.0,
      "step": 1101
    },
    {
      "epoch": 0.9402730375426621,
      "grad_norm": 1.078125,
      "learning_rate": 3.842943919353914e-07,
      "loss": 0.38875749707221985,
      "mean_token_accuracy": 0.8651366233825684,
      "num_tokens": 1394785429.0,
      "step": 1102
    },
    {
      "epoch": 0.9411262798634812,
      "grad_norm": 1.0625,
      "learning_rate": 3.7357902227080023e-07,
      "loss": 0.39551645517349243,
      "mean_token_accuracy": 0.8595495223999023,
      "num_tokens": 1395982075.0,
      "step": 1103
    },
    {
      "epoch": 0.9419795221843004,
      "grad_norm": 1.1953125,
      "learning_rate": 3.630137537203582e-07,
      "loss": 0.42296576499938965,
      "mean_token_accuracy": 0.8496309518814087,
      "num_tokens": 1397188308.0,
      "step": 1104
    },
    {
      "epoch": 0.9428327645051194,
      "grad_norm": 1.25,
      "learning_rate": 3.5259866708629466e-07,
      "loss": 0.3805265426635742,
      "mean_token_accuracy": 0.8676981329917908,
      "num_tokens": 1398440035.0,
      "step": 1105
    },
    {
      "epoch": 0.9436860068259386,
      "grad_norm": 1.1015625,
      "learning_rate": 3.423338420222644e-07,
      "loss": 0.3920983076095581,
      "mean_token_accuracy": 0.8604702949523926,
      "num_tokens": 1399635056.0,
      "step": 1106
    },
    {
      "epoch": 0.9445392491467577,
      "grad_norm": 1.0546875,
      "learning_rate": 3.3221935703273257e-07,
      "loss": 0.36700010299682617,
      "mean_token_accuracy": 0.8664853572845459,
      "num_tokens": 1400815188.0,
      "step": 1107
    },
    {
      "epoch": 0.9453924914675768,
      "grad_norm": 1.09375,
      "learning_rate": 3.2225528947237295e-07,
      "loss": 0.40616387128829956,
      "mean_token_accuracy": 0.8542220592498779,
      "num_tokens": 1402004764.0,
      "step": 1108
    },
    {
      "epoch": 0.9462457337883959,
      "grad_norm": 0.98046875,
      "learning_rate": 3.124417155454884e-07,
      "loss": 0.3705618679523468,
      "mean_token_accuracy": 0.8690600395202637,
      "num_tokens": 1403314047.0,
      "step": 1109
    },
    {
      "epoch": 0.947098976109215,
      "grad_norm": 1.078125,
      "learning_rate": 3.027787103054114e-07,
      "loss": 0.44086068868637085,
      "mean_token_accuracy": 0.8475100994110107,
      "num_tokens": 1404569573.0,
      "step": 1110
    },
    {
      "epoch": 0.9479522184300341,
      "grad_norm": 1.171875,
      "learning_rate": 2.9326634765394656e-07,
      "loss": 0.4369352459907532,
      "mean_token_accuracy": 0.8510620594024658,
      "num_tokens": 1405880332.0,
      "step": 1111
    },
    {
      "epoch": 0.9488054607508533,
      "grad_norm": 1.1328125,
      "learning_rate": 2.839047003407913e-07,
      "loss": 0.43473589420318604,
      "mean_token_accuracy": 0.8479146957397461,
      "num_tokens": 1407138588.0,
      "step": 1112
    },
    {
      "epoch": 0.9496587030716723,
      "grad_norm": 1.0703125,
      "learning_rate": 2.7469383996298946e-07,
      "loss": 0.4125126004219055,
      "mean_token_accuracy": 0.856917679309845,
      "num_tokens": 1408369120.0,
      "step": 1113
    },
    {
      "epoch": 0.9505119453924915,
      "grad_norm": 1.2578125,
      "learning_rate": 2.6563383696437626e-07,
      "loss": 0.4540451765060425,
      "mean_token_accuracy": 0.8411432504653931,
      "num_tokens": 1409677457.0,
      "step": 1114
    },
    {
      "epoch": 0.9513651877133106,
      "grad_norm": 1.046875,
      "learning_rate": 2.5672476063504537e-07,
      "loss": 0.4244677722454071,
      "mean_token_accuracy": 0.8493010997772217,
      "num_tokens": 1411047214.0,
      "step": 1115
    },
    {
      "epoch": 0.9522184300341296,
      "grad_norm": 1.1171875,
      "learning_rate": 2.4796667911081815e-07,
      "loss": 0.40864086151123047,
      "mean_token_accuracy": 0.8567209243774414,
      "num_tokens": 1412335864.0,
      "step": 1116
    },
    {
      "epoch": 0.9530716723549488,
      "grad_norm": 1.0234375,
      "learning_rate": 2.393596593727199e-07,
      "loss": 0.3834322690963745,
      "mean_token_accuracy": 0.8647650480270386,
      "num_tokens": 1413649816.0,
      "step": 1117
    },
    {
      "epoch": 0.9539249146757679,
      "grad_norm": 1.09375,
      "learning_rate": 2.3090376724646646e-07,
      "loss": 0.407406747341156,
      "mean_token_accuracy": 0.8564303517341614,
      "num_tokens": 1414829236.0,
      "step": 1118
    },
    {
      "epoch": 0.9547781569965871,
      "grad_norm": 1.140625,
      "learning_rate": 2.2259906740196512e-07,
      "loss": 0.43111148476600647,
      "mean_token_accuracy": 0.8479621410369873,
      "num_tokens": 1416077695.0,
      "step": 1119
    },
    {
      "epoch": 0.9556313993174061,
      "grad_norm": 1.2109375,
      "learning_rate": 2.1444562335282138e-07,
      "loss": 0.4619360864162445,
      "mean_token_accuracy": 0.839044988155365,
      "num_tokens": 1417316738.0,
      "step": 1120
    },
    {
      "epoch": 0.9564846416382252,
      "grad_norm": 1.2734375,
      "learning_rate": 2.064434974558438e-07,
      "loss": 0.4368628263473511,
      "mean_token_accuracy": 0.8498209714889526,
      "num_tokens": 1418616982.0,
      "step": 1121
    },
    {
      "epoch": 0.9573378839590444,
      "grad_norm": 1.1171875,
      "learning_rate": 1.9859275091058006e-07,
      "loss": 0.3795967102050781,
      "mean_token_accuracy": 0.8643100261688232,
      "num_tokens": 1419868981.0,
      "step": 1122
    },
    {
      "epoch": 0.9581911262798635,
      "grad_norm": 1.0625,
      "learning_rate": 1.908934437588328e-07,
      "loss": 0.39072978496551514,
      "mean_token_accuracy": 0.8628364205360413,
      "num_tokens": 1421071568.0,
      "step": 1123
    },
    {
      "epoch": 0.9590443686006825,
      "grad_norm": 1.046875,
      "learning_rate": 1.8334563488421997e-07,
      "loss": 0.39901405572891235,
      "mean_token_accuracy": 0.8591890335083008,
      "num_tokens": 1422353593.0,
      "step": 1124
    },
    {
      "epoch": 0.9598976109215017,
      "grad_norm": 1.0703125,
      "learning_rate": 1.7594938201170863e-07,
      "loss": 0.4175514578819275,
      "mean_token_accuracy": 0.8537093997001648,
      "num_tokens": 1423702549.0,
      "step": 1125
    },
    {
      "epoch": 0.9607508532423208,
      "grad_norm": 1.0859375,
      "learning_rate": 1.6870474170717745e-07,
      "loss": 0.4174620807170868,
      "mean_token_accuracy": 0.8536412715911865,
      "num_tokens": 1425010694.0,
      "step": 1126
    },
    {
      "epoch": 0.96160409556314,
      "grad_norm": 1.1484375,
      "learning_rate": 1.616117693769903e-07,
      "loss": 0.4365914463996887,
      "mean_token_accuracy": 0.8449963927268982,
      "num_tokens": 1426330660.0,
      "step": 1127
    },
    {
      "epoch": 0.962457337883959,
      "grad_norm": 1.0390625,
      "learning_rate": 1.5467051926756572e-07,
      "loss": 0.4015546441078186,
      "mean_token_accuracy": 0.8528693914413452,
      "num_tokens": 1427643776.0,
      "step": 1128
    },
    {
      "epoch": 0.9633105802047781,
      "grad_norm": 1.1328125,
      "learning_rate": 1.4788104446496365e-07,
      "loss": 0.4203071594238281,
      "mean_token_accuracy": 0.8529801368713379,
      "num_tokens": 1428997313.0,
      "step": 1129
    },
    {
      "epoch": 0.9641638225255973,
      "grad_norm": 1.140625,
      "learning_rate": 1.4124339689448373e-07,
      "loss": 0.4250871539115906,
      "mean_token_accuracy": 0.852069616317749,
      "num_tokens": 1430209491.0,
      "step": 1130
    },
    {
      "epoch": 0.9650170648464164,
      "grad_norm": 1.046875,
      "learning_rate": 1.3475762732025666e-07,
      "loss": 0.38243743777275085,
      "mean_token_accuracy": 0.8616844415664673,
      "num_tokens": 1431484102.0,
      "step": 1131
    },
    {
      "epoch": 0.9658703071672355,
      "grad_norm": 1.046875,
      "learning_rate": 1.284237853448711e-07,
      "loss": 0.4068406820297241,
      "mean_token_accuracy": 0.8553925156593323,
      "num_tokens": 1432760435.0,
      "step": 1132
    },
    {
      "epoch": 0.9667235494880546,
      "grad_norm": 1.2265625,
      "learning_rate": 1.2224191940898521e-07,
      "loss": 0.4483400583267212,
      "mean_token_accuracy": 0.844413161277771,
      "num_tokens": 1434123883.0,
      "step": 1133
    },
    {
      "epoch": 0.9675767918088737,
      "grad_norm": 1.078125,
      "learning_rate": 1.1621207679095802e-07,
      "loss": 0.41551607847213745,
      "mean_token_accuracy": 0.8519910573959351,
      "num_tokens": 1435371796.0,
      "step": 1134
    },
    {
      "epoch": 0.9684300341296929,
      "grad_norm": 1.140625,
      "learning_rate": 1.1033430360648523e-07,
      "loss": 0.3878856897354126,
      "mean_token_accuracy": 0.8570556640625,
      "num_tokens": 1436706578.0,
      "step": 1135
    },
    {
      "epoch": 0.9692832764505119,
      "grad_norm": 1.125,
      "learning_rate": 1.0460864480825062e-07,
      "loss": 0.45177680253982544,
      "mean_token_accuracy": 0.8437227010726929,
      "num_tokens": 1437952214.0,
      "step": 1136
    },
    {
      "epoch": 0.9701365187713311,
      "grad_norm": 1.109375,
      "learning_rate": 9.903514418558191e-08,
      "loss": 0.41595199704170227,
      "mean_token_accuracy": 0.8535334467887878,
      "num_tokens": 1439260540.0,
      "step": 1137
    },
    {
      "epoch": 0.9709897610921502,
      "grad_norm": 1.0234375,
      "learning_rate": 9.361384436411547e-08,
      "loss": 0.4120619297027588,
      "mean_token_accuracy": 0.853062629699707,
      "num_tokens": 1440587559.0,
      "step": 1138
    },
    {
      "epoch": 0.9718430034129693,
      "grad_norm": 1.0234375,
      "learning_rate": 8.834478680546543e-08,
      "loss": 0.4178963899612427,
      "mean_token_accuracy": 0.8523726463317871,
      "num_tokens": 1441817817.0,
      "step": 1139
    },
    {
      "epoch": 0.9726962457337884,
      "grad_norm": 1.2421875,
      "learning_rate": 8.32280118069151e-08,
      "loss": 0.41996288299560547,
      "mean_token_accuracy": 0.8542212247848511,
      "num_tokens": 1443037949.0,
      "step": 1140
    },
    {
      "epoch": 0.9735494880546075,
      "grad_norm": 1.0390625,
      "learning_rate": 7.826355850110378e-08,
      "loss": 0.404221773147583,
      "mean_token_accuracy": 0.8596411943435669,
      "num_tokens": 1444257454.0,
      "step": 1141
    },
    {
      "epoch": 0.9744027303754266,
      "grad_norm": 1.0703125,
      "learning_rate": 7.345146485572274e-08,
      "loss": 0.4095650017261505,
      "mean_token_accuracy": 0.8574566841125488,
      "num_tokens": 1445525675.0,
      "step": 1142
    },
    {
      "epoch": 0.9752559726962458,
      "grad_norm": 1.0234375,
      "learning_rate": 6.879176767323747e-08,
      "loss": 0.4228940010070801,
      "mean_token_accuracy": 0.8558884859085083,
      "num_tokens": 1446767651.0,
      "step": 1143
    },
    {
      "epoch": 0.9761092150170648,
      "grad_norm": 1.375,
      "learning_rate": 6.42845025905947e-08,
      "loss": 0.3945685029029846,
      "mean_token_accuracy": 0.8624492883682251,
      "num_tokens": 1448028691.0,
      "step": 1144
    },
    {
      "epoch": 0.976962457337884,
      "grad_norm": 1.2109375,
      "learning_rate": 5.992970407895371e-08,
      "loss": 0.44147419929504395,
      "mean_token_accuracy": 0.8473066091537476,
      "num_tokens": 1449281991.0,
      "step": 1145
    },
    {
      "epoch": 0.9778156996587031,
      "grad_norm": 1.0859375,
      "learning_rate": 5.5727405443428696e-08,
      "loss": 0.397579550743103,
      "mean_token_accuracy": 0.856451153755188,
      "num_tokens": 1450543606.0,
      "step": 1146
    },
    {
      "epoch": 0.9786689419795221,
      "grad_norm": 1.1171875,
      "learning_rate": 5.1677638822817954e-08,
      "loss": 0.4168291687965393,
      "mean_token_accuracy": 0.8523392677307129,
      "num_tokens": 1451740776.0,
      "step": 1147
    },
    {
      "epoch": 0.9795221843003413,
      "grad_norm": 1.328125,
      "learning_rate": 4.7780435189379585e-08,
      "loss": 0.4141175150871277,
      "mean_token_accuracy": 0.8511350154876709,
      "num_tokens": 1452960268.0,
      "step": 1148
    },
    {
      "epoch": 0.9803754266211604,
      "grad_norm": 1.078125,
      "learning_rate": 4.403582434857834e-08,
      "loss": 0.4386860728263855,
      "mean_token_accuracy": 0.8460760116577148,
      "num_tokens": 1454269179.0,
      "step": 1149
    },
    {
      "epoch": 0.9812286689419796,
      "grad_norm": 1.171875,
      "learning_rate": 4.044383493886139e-08,
      "loss": 0.4319445490837097,
      "mean_token_accuracy": 0.8482182621955872,
      "num_tokens": 1455531601.0,
      "step": 1150
    },
    {
      "epoch": 0.9820819112627986,
      "grad_norm": 1.2734375,
      "learning_rate": 3.700449443144516e-08,
      "loss": 0.4054577946662903,
      "mean_token_accuracy": 0.8547571301460266,
      "num_tokens": 1456876829.0,
      "step": 1151
    },
    {
      "epoch": 0.9829351535836177,
      "grad_norm": 1.0859375,
      "learning_rate": 3.371782913009547e-08,
      "loss": 0.4111299216747284,
      "mean_token_accuracy": 0.8559738397598267,
      "num_tokens": 1458093459.0,
      "step": 1152
    },
    {
      "epoch": 0.9837883959044369,
      "grad_norm": 1.0859375,
      "learning_rate": 3.0583864170936616e-08,
      "loss": 0.3922392427921295,
      "mean_token_accuracy": 0.8582087159156799,
      "num_tokens": 1459398721.0,
      "step": 1153
    },
    {
      "epoch": 0.984641638225256,
      "grad_norm": 1.2578125,
      "learning_rate": 2.7602623522253736e-08,
      "loss": 0.4188610315322876,
      "mean_token_accuracy": 0.8473808169364929,
      "num_tokens": 1460689844.0,
      "step": 1154
    },
    {
      "epoch": 0.985494880546075,
      "grad_norm": 1.1328125,
      "learning_rate": 2.4774129984308505e-08,
      "loss": 0.4122602939605713,
      "mean_token_accuracy": 0.8576104640960693,
      "num_tokens": 1462039746.0,
      "step": 1155
    },
    {
      "epoch": 0.9863481228668942,
      "grad_norm": 1.265625,
      "learning_rate": 2.209840518916373e-08,
      "loss": 0.42072227597236633,
      "mean_token_accuracy": 0.8588931560516357,
      "num_tokens": 1463395029.0,
      "step": 1156
    },
    {
      "epoch": 0.9872013651877133,
      "grad_norm": 1.109375,
      "learning_rate": 1.957546960052792e-08,
      "loss": 0.42343375086784363,
      "mean_token_accuracy": 0.8495550751686096,
      "num_tokens": 1464682917.0,
      "step": 1157
    },
    {
      "epoch": 0.9880546075085325,
      "grad_norm": 1.0625,
      "learning_rate": 1.72053425135843e-08,
      "loss": 0.4181426167488098,
      "mean_token_accuracy": 0.8508977293968201,
      "num_tokens": 1465974507.0,
      "step": 1158
    },
    {
      "epoch": 0.9889078498293515,
      "grad_norm": 1.0546875,
      "learning_rate": 1.4988042054855377e-08,
      "loss": 0.41166216135025024,
      "mean_token_accuracy": 0.852851152420044,
      "num_tokens": 1467220868.0,
      "step": 1159
    },
    {
      "epoch": 0.9897610921501706,
      "grad_norm": 1.234375,
      "learning_rate": 1.2923585182054165e-08,
      "loss": 0.4186364412307739,
      "mean_token_accuracy": 0.8530111312866211,
      "num_tokens": 1468552334.0,
      "step": 1160
    },
    {
      "epoch": 0.9906143344709898,
      "grad_norm": 1.1171875,
      "learning_rate": 1.1011987683964276e-08,
      "loss": 0.4342220425605774,
      "mean_token_accuracy": 0.8466411828994751,
      "num_tokens": 1469811599.0,
      "step": 1161
    },
    {
      "epoch": 0.9914675767918089,
      "grad_norm": 1.34375,
      "learning_rate": 9.25326418031114e-09,
      "loss": 0.42412692308425903,
      "mean_token_accuracy": 0.8488216400146484,
      "num_tokens": 1471108515.0,
      "step": 1162
    },
    {
      "epoch": 0.992320819112628,
      "grad_norm": 1.109375,
      "learning_rate": 7.647428121655421e-09,
      "loss": 0.41608482599258423,
      "mean_token_accuracy": 0.8521026372909546,
      "num_tokens": 1472327714.0,
      "step": 1163
    },
    {
      "epoch": 0.9931740614334471,
      "grad_norm": 1.1796875,
      "learning_rate": 6.194491789290879e-09,
      "loss": 0.4466387629508972,
      "mean_token_accuracy": 0.8440028429031372,
      "num_tokens": 1473620662.0,
      "step": 1164
    },
    {
      "epoch": 0.9940273037542662,
      "grad_norm": 1.078125,
      "learning_rate": 4.89446629514001e-09,
      "loss": 0.3968116343021393,
      "mean_token_accuracy": 0.858989417552948,
      "num_tokens": 1474768927.0,
      "step": 1165
    },
    {
      "epoch": 0.9948805460750854,
      "grad_norm": 1.0546875,
      "learning_rate": 3.747361581685205e-09,
      "loss": 0.3952745795249939,
      "mean_token_accuracy": 0.8567055463790894,
      "num_tokens": 1475965330.0,
      "step": 1166
    },
    {
      "epoch": 0.9957337883959044,
      "grad_norm": 1.1015625,
      "learning_rate": 2.753186421877718e-09,
      "loss": 0.42812812328338623,
      "mean_token_accuracy": 0.8515510559082031,
      "num_tokens": 1477312760.0,
      "step": 1167
    },
    {
      "epoch": 0.9965870307167235,
      "grad_norm": 1.0546875,
      "learning_rate": 1.911948419082155e-09,
      "loss": 0.40689677000045776,
      "mean_token_accuracy": 0.8570227026939392,
      "num_tokens": 1478595377.0,
      "step": 1168
    },
    {
      "epoch": 0.9974402730375427,
      "grad_norm": 0.94140625,
      "learning_rate": 1.2236540070142966e-09,
      "loss": 0.36870282888412476,
      "mean_token_accuracy": 0.8673776984214783,
      "num_tokens": 1479953218.0,
      "step": 1169
    },
    {
      "epoch": 0.9982935153583617,
      "grad_norm": 1.0234375,
      "learning_rate": 6.883084496833725e-10,
      "loss": 0.4329805374145508,
      "mean_token_accuracy": 0.8500186800956726,
      "num_tokens": 1481137898.0,
      "step": 1170
    },
    {
      "epoch": 0.9991467576791809,
      "grad_norm": 0.9765625,
      "learning_rate": 3.059158413654117e-10,
      "loss": 0.38551872968673706,
      "mean_token_accuracy": 0.8619588017463684,
      "num_tokens": 1482365430.0,
      "step": 1171
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.265625,
      "learning_rate": 7.647910656771728e-11,
      "loss": 0.4308728575706482,
      "mean_token_accuracy": 0.8495464324951172,
      "num_tokens": 1483667981.0,
      "step": 1172
    }
  ],
  "logging_steps": 1,
  "max_steps": 1172,
  "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": 3.234555188618461e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}