{
  "best_metric": 0.02887474000453949,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold2/checkpoint-1900",
  "epoch": 5.0,
  "eval_steps": 50,
  "global_step": 3225,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015503875968992248,
      "grad_norm": 1.8097485303878784,
      "learning_rate": 3.0959752321981426e-06,
      "loss": 1.678,
      "step": 10
    },
    {
      "epoch": 0.031007751937984496,
      "grad_norm": 1.9774603843688965,
      "learning_rate": 6.191950464396285e-06,
      "loss": 1.7145,
      "step": 20
    },
    {
      "epoch": 0.046511627906976744,
      "grad_norm": 2.036001443862915,
      "learning_rate": 9.287925696594429e-06,
      "loss": 1.5924,
      "step": 30
    },
    {
      "epoch": 0.06201550387596899,
      "grad_norm": 3.265268087387085,
      "learning_rate": 1.238390092879257e-05,
      "loss": 1.2457,
      "step": 40
    },
    {
      "epoch": 0.07751937984496124,
      "grad_norm": 1.1102505922317505,
      "learning_rate": 1.5479876160990712e-05,
      "loss": 0.7519,
      "step": 50
    },
    {
      "epoch": 0.07751937984496124,
      "eval_loss": 0.638915479183197,
      "eval_runtime": 177.4395,
      "eval_samples_per_second": 6.464,
      "eval_steps_per_second": 6.464,
      "step": 50
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 0.6308065056800842,
      "learning_rate": 1.8575851393188857e-05,
      "loss": 0.6351,
      "step": 60
    },
    {
      "epoch": 0.10852713178294573,
      "grad_norm": 0.8012827634811401,
      "learning_rate": 2.1671826625387e-05,
      "loss": 0.4782,
      "step": 70
    },
    {
      "epoch": 0.12403100775193798,
      "grad_norm": 0.6199682354927063,
      "learning_rate": 2.476780185758514e-05,
      "loss": 0.2779,
      "step": 80
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 0.5094736218452454,
      "learning_rate": 2.7863777089783283e-05,
      "loss": 0.1896,
      "step": 90
    },
    {
      "epoch": 0.15503875968992248,
      "grad_norm": 0.5959056615829468,
      "learning_rate": 3.0959752321981425e-05,
      "loss": 0.1485,
      "step": 100
    },
    {
      "epoch": 0.15503875968992248,
      "eval_loss": 0.11244650185108185,
      "eval_runtime": 177.3837,
      "eval_samples_per_second": 6.466,
      "eval_steps_per_second": 6.466,
      "step": 100
    },
    {
      "epoch": 0.17054263565891473,
      "grad_norm": 0.919588029384613,
      "learning_rate": 3.4055727554179566e-05,
      "loss": 0.1271,
      "step": 110
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 0.9113105535507202,
      "learning_rate": 3.7151702786377715e-05,
      "loss": 0.101,
      "step": 120
    },
    {
      "epoch": 0.20155038759689922,
      "grad_norm": 0.74349045753479,
      "learning_rate": 4.024767801857585e-05,
      "loss": 0.1009,
      "step": 130
    },
    {
      "epoch": 0.21705426356589147,
      "grad_norm": 0.6830365061759949,
      "learning_rate": 4.3343653250774e-05,
      "loss": 0.096,
      "step": 140
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 0.5473061203956604,
      "learning_rate": 4.6439628482972134e-05,
      "loss": 0.074,
      "step": 150
    },
    {
      "epoch": 0.23255813953488372,
      "eval_loss": 0.06496970355510712,
      "eval_runtime": 177.3867,
      "eval_samples_per_second": 6.466,
      "eval_steps_per_second": 6.466,
      "step": 150
    },
    {
      "epoch": 0.24806201550387597,
      "grad_norm": 0.45949482917785645,
      "learning_rate": 4.953560371517028e-05,
      "loss": 0.076,
      "step": 160
    },
    {
      "epoch": 0.26356589147286824,
      "grad_norm": 0.6822295188903809,
      "learning_rate": 5.2631578947368424e-05,
      "loss": 0.0689,
      "step": 170
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.5395657420158386,
      "learning_rate": 5.5727554179566566e-05,
      "loss": 0.0634,
      "step": 180
    },
    {
      "epoch": 0.29457364341085274,
      "grad_norm": 1.2032052278518677,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.0687,
      "step": 190
    },
    {
      "epoch": 0.31007751937984496,
      "grad_norm": 0.8278073072433472,
      "learning_rate": 6.191950464396285e-05,
      "loss": 0.0655,
      "step": 200
    },
    {
      "epoch": 0.31007751937984496,
      "eval_loss": 0.06194576248526573,
      "eval_runtime": 177.5104,
      "eval_samples_per_second": 6.462,
      "eval_steps_per_second": 6.462,
      "step": 200
    },
    {
      "epoch": 0.32558139534883723,
      "grad_norm": 0.48666349053382874,
      "learning_rate": 6.501547987616098e-05,
      "loss": 0.0591,
      "step": 210
    },
    {
      "epoch": 0.34108527131782945,
      "grad_norm": 0.6362942457199097,
      "learning_rate": 6.811145510835913e-05,
      "loss": 0.0557,
      "step": 220
    },
    {
      "epoch": 0.35658914728682173,
      "grad_norm": 0.44466376304626465,
      "learning_rate": 7.120743034055728e-05,
      "loss": 0.0531,
      "step": 230
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.473111093044281,
      "learning_rate": 7.430340557275543e-05,
      "loss": 0.0704,
      "step": 240
    },
    {
      "epoch": 0.3875968992248062,
      "grad_norm": 0.6141083836555481,
      "learning_rate": 7.739938080495357e-05,
      "loss": 0.0598,
      "step": 250
    },
    {
      "epoch": 0.3875968992248062,
      "eval_loss": 0.05116499215364456,
      "eval_runtime": 177.5255,
      "eval_samples_per_second": 6.461,
      "eval_steps_per_second": 6.461,
      "step": 250
    },
    {
      "epoch": 0.40310077519379844,
      "grad_norm": 0.34960028529167175,
      "learning_rate": 8.04953560371517e-05,
      "loss": 0.0502,
      "step": 260
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 0.6346738338470459,
      "learning_rate": 8.359133126934985e-05,
      "loss": 0.0469,
      "step": 270
    },
    {
      "epoch": 0.43410852713178294,
      "grad_norm": 0.34611088037490845,
      "learning_rate": 8.6687306501548e-05,
      "loss": 0.0769,
      "step": 280
    },
    {
      "epoch": 0.4496124031007752,
      "grad_norm": 0.746246337890625,
      "learning_rate": 8.978328173374613e-05,
      "loss": 0.0547,
      "step": 290
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.3229033350944519,
      "learning_rate": 9.287925696594427e-05,
      "loss": 0.0414,
      "step": 300
    },
    {
      "epoch": 0.46511627906976744,
      "eval_loss": 0.04494742304086685,
      "eval_runtime": 177.4386,
      "eval_samples_per_second": 6.464,
      "eval_steps_per_second": 6.464,
      "step": 300
    },
    {
      "epoch": 0.4806201550387597,
      "grad_norm": 0.7018037438392639,
      "learning_rate": 9.597523219814242e-05,
      "loss": 0.0469,
      "step": 310
    },
    {
      "epoch": 0.49612403100775193,
      "grad_norm": 0.3599799871444702,
      "learning_rate": 9.907120743034056e-05,
      "loss": 0.0377,
      "step": 320
    },
    {
      "epoch": 0.5116279069767442,
      "grad_norm": 0.3153678774833679,
      "learning_rate": 9.999856438185238e-05,
      "loss": 0.0441,
      "step": 330
    },
    {
      "epoch": 0.5271317829457365,
      "grad_norm": 0.2962203621864319,
      "learning_rate": 9.999153298122152e-05,
      "loss": 0.0522,
      "step": 340
    },
    {
      "epoch": 0.5426356589147286,
      "grad_norm": 0.1656709760427475,
      "learning_rate": 9.997864293614054e-05,
      "loss": 0.0427,
      "step": 350
    },
    {
      "epoch": 0.5426356589147286,
      "eval_loss": 0.04140669107437134,
      "eval_runtime": 177.5437,
      "eval_samples_per_second": 6.46,
      "eval_steps_per_second": 6.46,
      "step": 350
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.2917194068431854,
      "learning_rate": 9.995989575722902e-05,
      "loss": 0.047,
      "step": 360
    },
    {
      "epoch": 0.5736434108527132,
      "grad_norm": 0.3477674424648285,
      "learning_rate": 9.993529364152018e-05,
      "loss": 0.0501,
      "step": 370
    },
    {
      "epoch": 0.5891472868217055,
      "grad_norm": 0.48425808548927307,
      "learning_rate": 9.990483947220319e-05,
      "loss": 0.0593,
      "step": 380
    },
    {
      "epoch": 0.6046511627906976,
      "grad_norm": 0.21694518625736237,
      "learning_rate": 9.986853681828546e-05,
      "loss": 0.0513,
      "step": 390
    },
    {
      "epoch": 0.6201550387596899,
      "grad_norm": 0.5227199792861938,
      "learning_rate": 9.982638993417425e-05,
      "loss": 0.0471,
      "step": 400
    },
    {
      "epoch": 0.6201550387596899,
      "eval_loss": 0.038663338869810104,
      "eval_runtime": 177.5346,
      "eval_samples_per_second": 6.461,
      "eval_steps_per_second": 6.461,
      "step": 400
    },
    {
      "epoch": 0.6356589147286822,
      "grad_norm": 0.25696733593940735,
      "learning_rate": 9.977840375917817e-05,
      "loss": 0.0475,
      "step": 410
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.22936882078647614,
      "learning_rate": 9.972458391692827e-05,
      "loss": 0.0405,
      "step": 420
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.31321045756340027,
      "learning_rate": 9.966493671471904e-05,
      "loss": 0.0449,
      "step": 430
    },
    {
      "epoch": 0.6821705426356589,
      "grad_norm": 0.3151184916496277,
      "learning_rate": 9.959946914276922e-05,
      "loss": 0.056,
      "step": 440
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 0.50681471824646,
      "learning_rate": 9.952818887340257e-05,
      "loss": 0.0433,
      "step": 450
    },
    {
      "epoch": 0.6976744186046512,
      "eval_loss": 0.03619127348065376,
      "eval_runtime": 177.5548,
      "eval_samples_per_second": 6.46,
      "eval_steps_per_second": 6.46,
      "step": 450
    },
    {
      "epoch": 0.7131782945736435,
      "grad_norm": 0.31854957342147827,
      "learning_rate": 9.945110426014878e-05,
      "loss": 0.0494,
      "step": 460
    },
    {
      "epoch": 0.7286821705426356,
      "grad_norm": 0.28269049525260925,
      "learning_rate": 9.936822433676444e-05,
      "loss": 0.0387,
      "step": 470
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.24601627886295319,
      "learning_rate": 9.927955881617444e-05,
      "loss": 0.0479,
      "step": 480
    },
    {
      "epoch": 0.7596899224806202,
      "grad_norm": 0.19982942938804626,
      "learning_rate": 9.918511808933358e-05,
      "loss": 0.0378,
      "step": 490
    },
    {
      "epoch": 0.7751937984496124,
      "grad_norm": 0.20261038839817047,
      "learning_rate": 9.908491322400885e-05,
      "loss": 0.0432,
      "step": 500
    },
    {
      "epoch": 0.7751937984496124,
      "eval_loss": 0.03534514829516411,
      "eval_runtime": 177.4652,
      "eval_samples_per_second": 6.463,
      "eval_steps_per_second": 6.463,
      "step": 500
    },
    {
      "epoch": 0.7906976744186046,
      "grad_norm": 0.1283186674118042,
      "learning_rate": 9.897895596348247e-05,
      "loss": 0.0395,
      "step": 510
    },
    {
      "epoch": 0.8062015503875969,
      "grad_norm": 0.2949117124080658,
      "learning_rate": 9.886725872517552e-05,
      "loss": 0.0326,
      "step": 520
    },
    {
      "epoch": 0.8217054263565892,
      "grad_norm": 0.42478370666503906,
      "learning_rate": 9.874983459919277e-05,
      "loss": 0.0442,
      "step": 530
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.270944207906723,
      "learning_rate": 9.862669734678867e-05,
      "loss": 0.038,
      "step": 540
    },
    {
      "epoch": 0.8527131782945736,
      "grad_norm": 0.19820506870746613,
      "learning_rate": 9.849786139875452e-05,
      "loss": 0.0445,
      "step": 550
    },
    {
      "epoch": 0.8527131782945736,
      "eval_loss": 0.03533301129937172,
      "eval_runtime": 177.5388,
      "eval_samples_per_second": 6.461,
      "eval_steps_per_second": 6.461,
      "step": 550
    },
    {
      "epoch": 0.8682170542635659,
      "grad_norm": 0.33879902958869934,
      "learning_rate": 9.836334185372738e-05,
      "loss": 0.0464,
      "step": 560
    },
    {
      "epoch": 0.8837209302325582,
      "grad_norm": 0.330963134765625,
      "learning_rate": 9.822315447642056e-05,
      "loss": 0.0368,
      "step": 570
    },
    {
      "epoch": 0.8992248062015504,
      "grad_norm": 0.22305184602737427,
      "learning_rate": 9.807731569577615e-05,
      "loss": 0.0296,
      "step": 580
    },
    {
      "epoch": 0.9147286821705426,
      "grad_norm": 0.28611594438552856,
      "learning_rate": 9.792584260303964e-05,
      "loss": 0.0364,
      "step": 590
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.24084076285362244,
      "learning_rate": 9.776875294975698e-05,
      "loss": 0.0529,
      "step": 600
    },
    {
      "epoch": 0.9302325581395349,
      "eval_loss": 0.03533976525068283,
      "eval_runtime": 177.6164,
      "eval_samples_per_second": 6.458,
      "eval_steps_per_second": 6.458,
      "step": 600
    },
    {
      "epoch": 0.9457364341085271,
      "grad_norm": 0.28942936658859253,
      "learning_rate": 9.760606514569416e-05,
      "loss": 0.0381,
      "step": 610
    },
    {
      "epoch": 0.9612403100775194,
      "grad_norm": 0.2056342214345932,
      "learning_rate": 9.743779825667984e-05,
      "loss": 0.0337,
      "step": 620
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 0.14191323518753052,
      "learning_rate": 9.726397200237085e-05,
      "loss": 0.0368,
      "step": 630
    },
    {
      "epoch": 0.9922480620155039,
      "grad_norm": 0.33825528621673584,
      "learning_rate": 9.708460675394126e-05,
      "loss": 0.0465,
      "step": 640
    },
    {
      "epoch": 1.0077519379844961,
      "grad_norm": 0.12557874619960785,
      "learning_rate": 9.689972353169499e-05,
      "loss": 0.0313,
      "step": 650
    },
    {
      "epoch": 1.0077519379844961,
      "eval_loss": 0.03178540617227554,
      "eval_runtime": 177.627,
      "eval_samples_per_second": 6.457,
      "eval_steps_per_second": 6.457,
      "step": 650
    },
    {
      "epoch": 1.0232558139534884,
      "grad_norm": 0.35606396198272705,
      "learning_rate": 9.67093440026024e-05,
      "loss": 0.0303,
      "step": 660
    },
    {
      "epoch": 1.0387596899224807,
      "grad_norm": 0.183203786611557,
      "learning_rate": 9.651349047776112e-05,
      "loss": 0.0386,
      "step": 670
    },
    {
      "epoch": 1.054263565891473,
      "grad_norm": 0.23599869012832642,
      "learning_rate": 9.631218590978126e-05,
      "loss": 0.0222,
      "step": 680
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 0.13795703649520874,
      "learning_rate": 9.610545389009562e-05,
      "loss": 0.0307,
      "step": 690
    },
    {
      "epoch": 1.0852713178294573,
      "grad_norm": 0.2756401002407074,
      "learning_rate": 9.589331864619486e-05,
      "loss": 0.0301,
      "step": 700
    },
    {
      "epoch": 1.0852713178294573,
      "eval_loss": 0.03220707178115845,
      "eval_runtime": 177.6556,
      "eval_samples_per_second": 6.456,
      "eval_steps_per_second": 6.456,
      "step": 700
    },
    {
      "epoch": 1.1007751937984496,
      "grad_norm": 0.22982142865657806,
      "learning_rate": 9.567580503878833e-05,
      "loss": 0.025,
      "step": 710
    },
    {
      "epoch": 1.1162790697674418,
      "grad_norm": 0.21597911417484283,
      "learning_rate": 9.545293855889043e-05,
      "loss": 0.0342,
      "step": 720
    },
    {
      "epoch": 1.1317829457364341,
      "grad_norm": 0.16282838582992554,
      "learning_rate": 9.522474532483337e-05,
      "loss": 0.0275,
      "step": 730
    },
    {
      "epoch": 1.1472868217054264,
      "grad_norm": 0.32020196318626404,
      "learning_rate": 9.499125207920622e-05,
      "loss": 0.0402,
      "step": 740
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.12987101078033447,
      "learning_rate": 9.475248618572096e-05,
      "loss": 0.0289,
      "step": 750
    },
    {
      "epoch": 1.1627906976744187,
      "eval_loss": 0.03383904695510864,
      "eval_runtime": 177.667,
      "eval_samples_per_second": 6.456,
      "eval_steps_per_second": 6.456,
      "step": 750
    },
    {
      "epoch": 1.178294573643411,
      "grad_norm": 0.37133800983428955,
      "learning_rate": 9.45084756260055e-05,
      "loss": 0.035,
      "step": 760
    },
    {
      "epoch": 1.193798449612403,
      "grad_norm": 0.07524913549423218,
      "learning_rate": 9.425924899632458e-05,
      "loss": 0.0291,
      "step": 770
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 0.14250116050243378,
      "learning_rate": 9.400483550422845e-05,
      "loss": 0.0221,
      "step": 780
    },
    {
      "epoch": 1.2248062015503876,
      "grad_norm": 0.33380234241485596,
      "learning_rate": 9.374526496512992e-05,
      "loss": 0.0303,
      "step": 790
    },
    {
      "epoch": 1.2403100775193798,
      "grad_norm": 0.6035688519477844,
      "learning_rate": 9.348056779881025e-05,
      "loss": 0.0267,
      "step": 800
    },
    {
      "epoch": 1.2403100775193798,
      "eval_loss": 0.03135337680578232,
      "eval_runtime": 177.6172,
      "eval_samples_per_second": 6.458,
      "eval_steps_per_second": 6.458,
      "step": 800
    },
    {
      "epoch": 1.255813953488372,
      "grad_norm": 0.25495055317878723,
      "learning_rate": 9.321077502585417e-05,
      "loss": 0.0267,
      "step": 810
    },
    {
      "epoch": 1.2713178294573644,
      "grad_norm": 0.352885365486145,
      "learning_rate": 9.293591826401451e-05,
      "loss": 0.0351,
      "step": 820
    },
    {
      "epoch": 1.2868217054263567,
      "grad_norm": 0.25123533606529236,
      "learning_rate": 9.265602972450673e-05,
      "loss": 0.0328,
      "step": 830
    },
    {
      "epoch": 1.302325581395349,
      "grad_norm": 0.47959160804748535,
      "learning_rate": 9.237114220823413e-05,
      "loss": 0.0326,
      "step": 840
    },
    {
      "epoch": 1.3178294573643412,
      "grad_norm": 0.13225722312927246,
      "learning_rate": 9.208128910194377e-05,
      "loss": 0.0314,
      "step": 850
    },
    {
      "epoch": 1.3178294573643412,
      "eval_loss": 0.03171215206384659,
      "eval_runtime": 177.6409,
      "eval_samples_per_second": 6.457,
      "eval_steps_per_second": 6.457,
      "step": 850
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.27190810441970825,
      "learning_rate": 9.17865043743138e-05,
      "loss": 0.031,
      "step": 860
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 0.10395022481679916,
      "learning_rate": 9.148682257197245e-05,
      "loss": 0.0238,
      "step": 870
    },
    {
      "epoch": 1.3643410852713178,
      "grad_norm": 0.33697041869163513,
      "learning_rate": 9.118227881544967e-05,
      "loss": 0.0397,
      "step": 880
    },
    {
      "epoch": 1.37984496124031,
      "grad_norm": 0.2214643359184265,
      "learning_rate": 9.087290879506104e-05,
      "loss": 0.0281,
      "step": 890
    },
    {
      "epoch": 1.3953488372093024,
      "grad_norm": 0.11761048436164856,
      "learning_rate": 9.055874876672519e-05,
      "loss": 0.0382,
      "step": 900
    },
    {
      "epoch": 1.3953488372093024,
      "eval_loss": 0.03273608908057213,
      "eval_runtime": 177.7052,
      "eval_samples_per_second": 6.455,
      "eval_steps_per_second": 6.455,
      "step": 900
    },
    {
      "epoch": 1.4108527131782946,
      "grad_norm": 0.17550455033779144,
      "learning_rate": 9.023983554771492e-05,
      "loss": 0.0263,
      "step": 910
    },
    {
      "epoch": 1.4263565891472867,
      "grad_norm": 0.21578268706798553,
      "learning_rate": 8.991620651234242e-05,
      "loss": 0.0261,
      "step": 920
    },
    {
      "epoch": 1.441860465116279,
      "grad_norm": 0.21228443086147308,
      "learning_rate": 8.958789958757928e-05,
      "loss": 0.029,
      "step": 930
    },
    {
      "epoch": 1.4573643410852712,
      "grad_norm": 0.2734227180480957,
      "learning_rate": 8.925495324861177e-05,
      "loss": 0.0286,
      "step": 940
    },
    {
      "epoch": 1.4728682170542635,
      "grad_norm": 0.4174731373786926,
      "learning_rate": 8.89174065143318e-05,
      "loss": 0.0354,
      "step": 950
    },
    {
      "epoch": 1.4728682170542635,
      "eval_loss": 0.031981565058231354,
      "eval_runtime": 177.7226,
      "eval_samples_per_second": 6.454,
      "eval_steps_per_second": 6.454,
      "step": 950
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 0.3433147370815277,
      "learning_rate": 8.857529894276412e-05,
      "loss": 0.0352,
      "step": 960
    },
    {
      "epoch": 1.503875968992248,
      "grad_norm": 0.21089136600494385,
      "learning_rate": 8.822867062643059e-05,
      "loss": 0.0298,
      "step": 970
    },
    {
      "epoch": 1.5193798449612403,
      "grad_norm": 0.1780281960964203,
      "learning_rate": 8.78775621876514e-05,
      "loss": 0.0319,
      "step": 980
    },
    {
      "epoch": 1.5348837209302326,
      "grad_norm": 0.3001708686351776,
      "learning_rate": 8.752201477378459e-05,
      "loss": 0.0217,
      "step": 990
    },
    {
      "epoch": 1.550387596899225,
      "grad_norm": 0.15130403637886047,
      "learning_rate": 8.716207005240382e-05,
      "loss": 0.0265,
      "step": 1000
    },
    {
      "epoch": 1.550387596899225,
      "eval_loss": 0.03209313005208969,
      "eval_runtime": 177.6849,
      "eval_samples_per_second": 6.455,
      "eval_steps_per_second": 6.455,
      "step": 1000
    },
    {
      "epoch": 1.5658914728682172,
      "grad_norm": 0.28896772861480713,
      "learning_rate": 8.679777020641525e-05,
      "loss": 0.0345,
      "step": 1010
    },
    {
      "epoch": 1.5813953488372094,
      "grad_norm": 0.4542350471019745,
      "learning_rate": 8.6429157929114e-05,
      "loss": 0.033,
      "step": 1020
    },
    {
      "epoch": 1.5968992248062015,
      "grad_norm": 0.26365530490875244,
      "learning_rate": 8.60562764191808e-05,
      "loss": 0.0277,
      "step": 1030
    },
    {
      "epoch": 1.6124031007751938,
      "grad_norm": 0.17537464201450348,
      "learning_rate": 8.567916937561944e-05,
      "loss": 0.0307,
      "step": 1040
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 0.39150458574295044,
      "learning_rate": 8.529788099263554e-05,
      "loss": 0.0301,
      "step": 1050
    },
    {
      "epoch": 1.627906976744186,
      "eval_loss": 0.03326583281159401,
      "eval_runtime": 177.6473,
      "eval_samples_per_second": 6.457,
      "eval_steps_per_second": 6.457,
      "step": 1050
    },
    {
      "epoch": 1.6434108527131783,
      "grad_norm": 0.2252993881702423,
      "learning_rate": 8.49124559544573e-05,
      "loss": 0.0279,
      "step": 1060
    },
    {
      "epoch": 1.6589147286821704,
      "grad_norm": 0.3254898488521576,
      "learning_rate": 8.452293943009889e-05,
      "loss": 0.0269,
      "step": 1070
    },
    {
      "epoch": 1.6744186046511627,
      "grad_norm": 0.2006569355726242,
      "learning_rate": 8.412937706806692e-05,
      "loss": 0.0361,
      "step": 1080
    },
    {
      "epoch": 1.689922480620155,
      "grad_norm": 0.2423427700996399,
      "learning_rate": 8.373181499101077e-05,
      "loss": 0.0301,
      "step": 1090
    },
    {
      "epoch": 1.7054263565891472,
      "grad_norm": 0.36214545369148254,
      "learning_rate": 8.333029979031737e-05,
      "loss": 0.0262,
      "step": 1100
    },
    {
      "epoch": 1.7054263565891472,
      "eval_loss": 0.031156621873378754,
      "eval_runtime": 177.8049,
      "eval_samples_per_second": 6.451,
      "eval_steps_per_second": 6.451,
      "step": 1100
    },
    {
      "epoch": 1.7209302325581395,
      "grad_norm": 0.17312179505825043,
      "learning_rate": 8.292487852065104e-05,
      "loss": 0.0392,
      "step": 1110
    },
    {
      "epoch": 1.7364341085271318,
      "grad_norm": 0.275713711977005,
      "learning_rate": 8.251559869443897e-05,
      "loss": 0.0322,
      "step": 1120
    },
    {
      "epoch": 1.751937984496124,
      "grad_norm": 0.33000582456588745,
      "learning_rate": 8.210250827630313e-05,
      "loss": 0.029,
      "step": 1130
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 0.29563483595848083,
      "learning_rate": 8.168565567743924e-05,
      "loss": 0.0263,
      "step": 1140
    },
    {
      "epoch": 1.7829457364341086,
      "grad_norm": 0.31916943192481995,
      "learning_rate": 8.126508974994321e-05,
      "loss": 0.0273,
      "step": 1150
    },
    {
      "epoch": 1.7829457364341086,
      "eval_loss": 0.030605774372816086,
      "eval_runtime": 177.725,
      "eval_samples_per_second": 6.454,
      "eval_steps_per_second": 6.454,
      "step": 1150
    },
    {
      "epoch": 1.7984496124031009,
      "grad_norm": 0.1885177344083786,
      "learning_rate": 8.084085978108611e-05,
      "loss": 0.0169,
      "step": 1160
    },
    {
      "epoch": 1.8139534883720931,
      "grad_norm": 0.15678612887859344,
      "learning_rate": 8.041301548753804e-05,
      "loss": 0.0284,
      "step": 1170
    },
    {
      "epoch": 1.8294573643410854,
      "grad_norm": 0.24891909956932068,
      "learning_rate": 7.998160700954165e-05,
      "loss": 0.0244,
      "step": 1180
    },
    {
      "epoch": 1.8449612403100775,
      "grad_norm": 0.27802011370658875,
      "learning_rate": 7.954668490503621e-05,
      "loss": 0.027,
      "step": 1190
    },
    {
      "epoch": 1.8604651162790697,
      "grad_norm": 0.2340599000453949,
      "learning_rate": 7.910830014373237e-05,
      "loss": 0.0283,
      "step": 1200
    },
    {
      "epoch": 1.8604651162790697,
      "eval_loss": 0.029726749286055565,
      "eval_runtime": 177.8483,
      "eval_samples_per_second": 6.449,
      "eval_steps_per_second": 6.449,
      "step": 1200
    },
    {
      "epoch": 1.875968992248062,
      "grad_norm": 0.13624481856822968,
      "learning_rate": 7.866650410113907e-05,
      "loss": 0.0243,
      "step": 1210
    },
    {
      "epoch": 1.8914728682170543,
      "grad_norm": 0.30038079619407654,
      "learning_rate": 7.822134855254263e-05,
      "loss": 0.0296,
      "step": 1220
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 0.23166057467460632,
      "learning_rate": 7.777288566693905e-05,
      "loss": 0.0273,
      "step": 1230
    },
    {
      "epoch": 1.9224806201550386,
      "grad_norm": 0.26034536957740784,
      "learning_rate": 7.732116800092018e-05,
      "loss": 0.029,
      "step": 1240
    },
    {
      "epoch": 1.937984496124031,
      "grad_norm": 0.2269134223461151,
      "learning_rate": 7.686624849251453e-05,
      "loss": 0.0381,
      "step": 1250
    },
    {
      "epoch": 1.937984496124031,
      "eval_loss": 0.029867732897400856,
      "eval_runtime": 177.8147,
      "eval_samples_per_second": 6.451,
      "eval_steps_per_second": 6.451,
      "step": 1250
    },
    {
      "epoch": 1.9534883720930232,
      "grad_norm": 0.30028459429740906,
      "learning_rate": 7.640818045498324e-05,
      "loss": 0.0378,
      "step": 1260
    },
    {
      "epoch": 1.9689922480620154,
      "grad_norm": 0.10794702172279358,
      "learning_rate": 7.59470175705722e-05,
      "loss": 0.0303,
      "step": 1270
    },
    {
      "epoch": 1.9844961240310077,
      "grad_norm": 0.27119511365890503,
      "learning_rate": 7.548281388422088e-05,
      "loss": 0.0361,
      "step": 1280
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.22147968411445618,
      "learning_rate": 7.50156237972286e-05,
      "loss": 0.0263,
      "step": 1290
    },
    {
      "epoch": 2.0155038759689923,
      "grad_norm": 0.12922464311122894,
      "learning_rate": 7.454550206087921e-05,
      "loss": 0.0207,
      "step": 1300
    },
    {
      "epoch": 2.0155038759689923,
      "eval_loss": 0.029372563585639,
      "eval_runtime": 177.7077,
      "eval_samples_per_second": 6.454,
      "eval_steps_per_second": 6.454,
      "step": 1300
    },
    {
      "epoch": 2.0310077519379846,
      "grad_norm": 0.20252908766269684,
      "learning_rate": 7.407250377002451e-05,
      "loss": 0.0184,
      "step": 1310
    },
    {
      "epoch": 2.046511627906977,
      "grad_norm": 0.20628713071346283,
      "learning_rate": 7.359668435662757e-05,
      "loss": 0.0216,
      "step": 1320
    },
    {
      "epoch": 2.062015503875969,
      "grad_norm": 0.16820959746837616,
      "learning_rate": 7.311809958326659e-05,
      "loss": 0.0153,
      "step": 1330
    },
    {
      "epoch": 2.0775193798449614,
      "grad_norm": 0.28085780143737793,
      "learning_rate": 7.263680553659983e-05,
      "loss": 0.0172,
      "step": 1340
    },
    {
      "epoch": 2.0930232558139537,
      "grad_norm": 0.4147201180458069,
      "learning_rate": 7.21528586207927e-05,
      "loss": 0.0163,
      "step": 1350
    },
    {
      "epoch": 2.0930232558139537,
      "eval_loss": 0.03293036296963692,
      "eval_runtime": 177.2709,
      "eval_samples_per_second": 6.47,
      "eval_steps_per_second": 6.47,
      "step": 1350
    },
    {
      "epoch": 2.108527131782946,
      "grad_norm": 0.2516096830368042,
      "learning_rate": 7.166631555090759e-05,
      "loss": 0.025,
      "step": 1360
    },
    {
      "epoch": 2.124031007751938,
      "grad_norm": 0.26919105648994446,
      "learning_rate": 7.117723334625731e-05,
      "loss": 0.024,
      "step": 1370
    },
    {
      "epoch": 2.13953488372093,
      "grad_norm": 0.228290855884552,
      "learning_rate": 7.068566932372278e-05,
      "loss": 0.0183,
      "step": 1380
    },
    {
      "epoch": 2.1550387596899223,
      "grad_norm": 0.28446343541145325,
      "learning_rate": 7.019168109103599e-05,
      "loss": 0.0162,
      "step": 1390
    },
    {
      "epoch": 2.1705426356589146,
      "grad_norm": 0.19133096933364868,
      "learning_rate": 6.96953265400287e-05,
      "loss": 0.0236,
      "step": 1400
    },
    {
      "epoch": 2.1705426356589146,
      "eval_loss": 0.031053410843014717,
      "eval_runtime": 176.3297,
      "eval_samples_per_second": 6.505,
      "eval_steps_per_second": 6.505,
      "step": 1400
    },
    {
      "epoch": 2.186046511627907,
      "grad_norm": 0.14185263216495514,
      "learning_rate": 6.9196663839848e-05,
      "loss": 0.0142,
      "step": 1410
    },
    {
      "epoch": 2.201550387596899,
      "grad_norm": 0.2277035117149353,
      "learning_rate": 6.869575143013934e-05,
      "loss": 0.0166,
      "step": 1420
    },
    {
      "epoch": 2.2170542635658914,
      "grad_norm": 0.24037382006645203,
      "learning_rate": 6.819264801419771e-05,
      "loss": 0.0274,
      "step": 1430
    },
    {
      "epoch": 2.2325581395348837,
      "grad_norm": 0.10074973106384277,
      "learning_rate": 6.768741255208813e-05,
      "loss": 0.0154,
      "step": 1440
    },
    {
      "epoch": 2.248062015503876,
      "grad_norm": 0.24921472370624542,
      "learning_rate": 6.718010425373599e-05,
      "loss": 0.0191,
      "step": 1450
    },
    {
      "epoch": 2.248062015503876,
      "eval_loss": 0.030957119539380074,
      "eval_runtime": 175.6681,
      "eval_samples_per_second": 6.529,
      "eval_steps_per_second": 6.529,
      "step": 1450
    },
    {
      "epoch": 2.2635658914728682,
      "grad_norm": 0.22070886194705963,
      "learning_rate": 6.667078257198796e-05,
      "loss": 0.0184,
      "step": 1460
    },
    {
      "epoch": 2.2790697674418605,
      "grad_norm": 0.39727696776390076,
      "learning_rate": 6.615950719564466e-05,
      "loss": 0.0142,
      "step": 1470
    },
    {
      "epoch": 2.294573643410853,
      "grad_norm": 0.31478437781333923,
      "learning_rate": 6.56463380424655e-05,
      "loss": 0.0252,
      "step": 1480
    },
    {
      "epoch": 2.310077519379845,
      "grad_norm": 0.1986107975244522,
      "learning_rate": 6.51313352521468e-05,
      "loss": 0.0151,
      "step": 1490
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 0.3045531213283539,
      "learning_rate": 6.46145591792738e-05,
      "loss": 0.0243,
      "step": 1500
    },
    {
      "epoch": 2.3255813953488373,
      "eval_loss": 0.03084620088338852,
      "eval_runtime": 175.2668,
      "eval_samples_per_second": 6.544,
      "eval_steps_per_second": 6.544,
      "step": 1500
    },
    {
      "epoch": 2.3410852713178296,
      "grad_norm": 0.08667182922363281,
      "learning_rate": 6.409607038624759e-05,
      "loss": 0.0142,
      "step": 1510
    },
    {
      "epoch": 2.356589147286822,
      "grad_norm": 0.17333078384399414,
      "learning_rate": 6.357592963618761e-05,
      "loss": 0.0241,
      "step": 1520
    },
    {
      "epoch": 2.3720930232558137,
      "grad_norm": 0.20276910066604614,
      "learning_rate": 6.305419788581067e-05,
      "loss": 0.0185,
      "step": 1530
    },
    {
      "epoch": 2.387596899224806,
      "grad_norm": 0.287349134683609,
      "learning_rate": 6.253093627828724e-05,
      "loss": 0.0176,
      "step": 1540
    },
    {
      "epoch": 2.4031007751937983,
      "grad_norm": 0.32019197940826416,
      "learning_rate": 6.200620613607597e-05,
      "loss": 0.0165,
      "step": 1550
    },
    {
      "epoch": 2.4031007751937983,
      "eval_loss": 0.03274994716048241,
      "eval_runtime": 174.5607,
      "eval_samples_per_second": 6.571,
      "eval_steps_per_second": 6.571,
      "step": 1550
    },
    {
      "epoch": 2.4186046511627906,
      "grad_norm": 0.34726911783218384,
      "learning_rate": 6.148006895373706e-05,
      "loss": 0.0196,
      "step": 1560
    },
    {
      "epoch": 2.434108527131783,
      "grad_norm": 0.15877018868923187,
      "learning_rate": 6.0952586390725644e-05,
      "loss": 0.0171,
      "step": 1570
    },
    {
      "epoch": 2.449612403100775,
      "grad_norm": 0.21150444447994232,
      "learning_rate": 6.042382026416563e-05,
      "loss": 0.0164,
      "step": 1580
    },
    {
      "epoch": 2.4651162790697674,
      "grad_norm": 0.19103744626045227,
      "learning_rate": 5.989383254160529e-05,
      "loss": 0.0171,
      "step": 1590
    },
    {
      "epoch": 2.4806201550387597,
      "grad_norm": 0.44117987155914307,
      "learning_rate": 5.936268533375506e-05,
      "loss": 0.0224,
      "step": 1600
    },
    {
      "epoch": 2.4806201550387597,
      "eval_loss": 0.03294115141034126,
      "eval_runtime": 174.4995,
      "eval_samples_per_second": 6.573,
      "eval_steps_per_second": 6.573,
      "step": 1600
    },
    {
      "epoch": 2.496124031007752,
      "grad_norm": 0.3606320321559906,
      "learning_rate": 5.883044088720865e-05,
      "loss": 0.0214,
      "step": 1610
    },
    {
      "epoch": 2.511627906976744,
      "grad_norm": 0.13814851641654968,
      "learning_rate": 5.829716157714814e-05,
      "loss": 0.0177,
      "step": 1620
    },
    {
      "epoch": 2.5271317829457365,
      "grad_norm": 0.23484356701374054,
      "learning_rate": 5.77629099000341e-05,
      "loss": 0.0263,
      "step": 1630
    },
    {
      "epoch": 2.5426356589147288,
      "grad_norm": 0.21404676139354706,
      "learning_rate": 5.7227748466281486e-05,
      "loss": 0.0187,
      "step": 1640
    },
    {
      "epoch": 2.558139534883721,
      "grad_norm": 0.3276318609714508,
      "learning_rate": 5.669173999292208e-05,
      "loss": 0.0289,
      "step": 1650
    },
    {
      "epoch": 2.558139534883721,
      "eval_loss": 0.03190125152468681,
      "eval_runtime": 174.4148,
      "eval_samples_per_second": 6.576,
      "eval_steps_per_second": 6.576,
      "step": 1650
    },
    {
      "epoch": 2.5736434108527133,
      "grad_norm": 0.41611889004707336,
      "learning_rate": 5.615494729625458e-05,
      "loss": 0.0237,
      "step": 1660
    },
    {
      "epoch": 2.5891472868217056,
      "grad_norm": 0.372302383184433,
      "learning_rate": 5.561743328448296e-05,
      "loss": 0.0246,
      "step": 1670
    },
    {
      "epoch": 2.604651162790698,
      "grad_norm": 0.1654350608587265,
      "learning_rate": 5.5079260950344035e-05,
      "loss": 0.019,
      "step": 1680
    },
    {
      "epoch": 2.62015503875969,
      "grad_norm": 0.09135215729475021,
      "learning_rate": 5.454049336372531e-05,
      "loss": 0.0165,
      "step": 1690
    },
    {
      "epoch": 2.6356589147286824,
      "grad_norm": 0.24403300881385803,
      "learning_rate": 5.4001193664273454e-05,
      "loss": 0.014,
      "step": 1700
    },
    {
      "epoch": 2.6356589147286824,
      "eval_loss": 0.03161098435521126,
      "eval_runtime": 174.2648,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 1700
    },
    {
      "epoch": 2.6511627906976747,
      "grad_norm": 0.19514554738998413,
      "learning_rate": 5.346142505399495e-05,
      "loss": 0.0198,
      "step": 1710
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.13203318417072296,
      "learning_rate": 5.292125078984925e-05,
      "loss": 0.0181,
      "step": 1720
    },
    {
      "epoch": 2.682170542635659,
      "grad_norm": 0.09618645161390305,
      "learning_rate": 5.2380734176335425e-05,
      "loss": 0.0171,
      "step": 1730
    },
    {
      "epoch": 2.697674418604651,
      "grad_norm": 0.17840640246868134,
      "learning_rate": 5.183993855807343e-05,
      "loss": 0.0186,
      "step": 1740
    },
    {
      "epoch": 2.7131782945736433,
      "grad_norm": 0.12565214931964874,
      "learning_rate": 5.1298927312380586e-05,
      "loss": 0.0182,
      "step": 1750
    },
    {
      "epoch": 2.7131782945736433,
      "eval_loss": 0.03337998315691948,
      "eval_runtime": 174.2061,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 1750
    },
    {
      "epoch": 2.7286821705426356,
      "grad_norm": 0.1932082623243332,
      "learning_rate": 5.075776384184411e-05,
      "loss": 0.0275,
      "step": 1760
    },
    {
      "epoch": 2.744186046511628,
      "grad_norm": 0.14624637365341187,
      "learning_rate": 5.021651156689094e-05,
      "loss": 0.0284,
      "step": 1770
    },
    {
      "epoch": 2.75968992248062,
      "grad_norm": 0.20881560444831848,
      "learning_rate": 4.967523391835521e-05,
      "loss": 0.0217,
      "step": 1780
    },
    {
      "epoch": 2.7751937984496124,
      "grad_norm": 0.2833938002586365,
      "learning_rate": 4.9133994330044644e-05,
      "loss": 0.0234,
      "step": 1790
    },
    {
      "epoch": 2.7906976744186047,
      "grad_norm": 0.23201899230480194,
      "learning_rate": 4.85928562313066e-05,
      "loss": 0.0175,
      "step": 1800
    },
    {
      "epoch": 2.7906976744186047,
      "eval_loss": 0.029777053743600845,
      "eval_runtime": 174.1005,
      "eval_samples_per_second": 6.588,
      "eval_steps_per_second": 6.588,
      "step": 1800
    },
    {
      "epoch": 2.806201550387597,
      "grad_norm": 0.2119215726852417,
      "learning_rate": 4.8051883039594616e-05,
      "loss": 0.0151,
      "step": 1810
    },
    {
      "epoch": 2.8217054263565893,
      "grad_norm": 0.30587702989578247,
      "learning_rate": 4.751113815303624e-05,
      "loss": 0.0187,
      "step": 1820
    },
    {
      "epoch": 2.8372093023255816,
      "grad_norm": 0.18018727004528046,
      "learning_rate": 4.697068494300343e-05,
      "loss": 0.0151,
      "step": 1830
    },
    {
      "epoch": 2.8527131782945734,
      "grad_norm": 0.21248644590377808,
      "learning_rate": 4.6430586746685724e-05,
      "loss": 0.0222,
      "step": 1840
    },
    {
      "epoch": 2.8682170542635657,
      "grad_norm": 0.18737761676311493,
      "learning_rate": 4.589090685966758e-05,
      "loss": 0.0218,
      "step": 1850
    },
    {
      "epoch": 2.8682170542635657,
      "eval_loss": 0.029682571068406105,
      "eval_runtime": 174.0789,
      "eval_samples_per_second": 6.589,
      "eval_steps_per_second": 6.589,
      "step": 1850
    },
    {
      "epoch": 2.883720930232558,
      "grad_norm": 0.3078942894935608,
      "learning_rate": 4.535170852851073e-05,
      "loss": 0.0265,
      "step": 1860
    },
    {
      "epoch": 2.89922480620155,
      "grad_norm": 0.2036602944135666,
      "learning_rate": 4.481305494334201e-05,
      "loss": 0.0167,
      "step": 1870
    },
    {
      "epoch": 2.9147286821705425,
      "grad_norm": 0.08146923035383224,
      "learning_rate": 4.427500923044801e-05,
      "loss": 0.0199,
      "step": 1880
    },
    {
      "epoch": 2.9302325581395348,
      "grad_norm": 0.12505419552326202,
      "learning_rate": 4.373763444487705e-05,
      "loss": 0.0179,
      "step": 1890
    },
    {
      "epoch": 2.945736434108527,
      "grad_norm": 0.18068896234035492,
      "learning_rate": 4.3200993563049725e-05,
      "loss": 0.018,
      "step": 1900
    },
    {
      "epoch": 2.945736434108527,
      "eval_loss": 0.02887474000453949,
      "eval_runtime": 173.8738,
      "eval_samples_per_second": 6.597,
      "eval_steps_per_second": 6.597,
      "step": 1900
    },
    {
      "epoch": 2.9612403100775193,
      "grad_norm": 0.19300121068954468,
      "learning_rate": 4.266514947537839e-05,
      "loss": 0.0157,
      "step": 1910
    },
    {
      "epoch": 2.9767441860465116,
      "grad_norm": 0.1891685128211975,
      "learning_rate": 4.2130164978896916e-05,
      "loss": 0.0188,
      "step": 1920
    },
    {
      "epoch": 2.992248062015504,
      "grad_norm": 0.027009131386876106,
      "learning_rate": 4.159610276990137e-05,
      "loss": 0.0118,
      "step": 1930
    },
    {
      "epoch": 3.007751937984496,
      "grad_norm": 0.11215299367904663,
      "learning_rate": 4.106302543660235e-05,
      "loss": 0.0164,
      "step": 1940
    },
    {
      "epoch": 3.0232558139534884,
      "grad_norm": 0.07163859158754349,
      "learning_rate": 4.053099545179028e-05,
      "loss": 0.01,
      "step": 1950
    },
    {
      "epoch": 3.0232558139534884,
      "eval_loss": 0.030872860923409462,
      "eval_runtime": 173.945,
      "eval_samples_per_second": 6.594,
      "eval_steps_per_second": 6.594,
      "step": 1950
    },
    {
      "epoch": 3.0387596899224807,
      "grad_norm": 0.12158221006393433,
      "learning_rate": 4.0000075165513845e-05,
      "loss": 0.014,
      "step": 1960
    },
    {
      "epoch": 3.054263565891473,
      "grad_norm": 0.17903171479701996,
      "learning_rate": 3.9470326797773216e-05,
      "loss": 0.0085,
      "step": 1970
    },
    {
      "epoch": 3.0697674418604652,
      "grad_norm": 0.23486894369125366,
      "learning_rate": 3.8941812431228166e-05,
      "loss": 0.012,
      "step": 1980
    },
    {
      "epoch": 3.0852713178294575,
      "grad_norm": 0.19141437113285065,
      "learning_rate": 3.8414594003922515e-05,
      "loss": 0.0141,
      "step": 1990
    },
    {
      "epoch": 3.10077519379845,
      "grad_norm": 0.047276828438043594,
      "learning_rate": 3.788873330202544e-05,
      "loss": 0.0109,
      "step": 2000
    },
    {
      "epoch": 3.10077519379845,
      "eval_loss": 0.033768460154533386,
      "eval_runtime": 173.8687,
      "eval_samples_per_second": 6.597,
      "eval_steps_per_second": 6.597,
      "step": 2000
    },
    {
      "epoch": 3.116279069767442,
      "grad_norm": 0.3079964518547058,
      "learning_rate": 3.736429195259051e-05,
      "loss": 0.0069,
      "step": 2010
    },
    {
      "epoch": 3.1317829457364343,
      "grad_norm": 0.207097128033638,
      "learning_rate": 3.684133141633358e-05,
      "loss": 0.009,
      "step": 2020
    },
    {
      "epoch": 3.147286821705426,
      "grad_norm": 0.16703131794929504,
      "learning_rate": 3.6319912980429846e-05,
      "loss": 0.0098,
      "step": 2030
    },
    {
      "epoch": 3.1627906976744184,
      "grad_norm": 0.1511267125606537,
      "learning_rate": 3.580009775133168e-05,
      "loss": 0.0076,
      "step": 2040
    },
    {
      "epoch": 3.1782945736434107,
      "grad_norm": 0.184535413980484,
      "learning_rate": 3.528194664760714e-05,
      "loss": 0.0076,
      "step": 2050
    },
    {
      "epoch": 3.1782945736434107,
      "eval_loss": 0.03470822051167488,
      "eval_runtime": 173.7828,
      "eval_samples_per_second": 6.6,
      "eval_steps_per_second": 6.6,
      "step": 2050
    },
    {
      "epoch": 3.193798449612403,
      "grad_norm": 0.19592364132404327,
      "learning_rate": 3.476552039280096e-05,
      "loss": 0.0089,
      "step": 2060
    },
    {
      "epoch": 3.2093023255813953,
      "grad_norm": 0.1496165543794632,
      "learning_rate": 3.42508795083181e-05,
      "loss": 0.0092,
      "step": 2070
    },
    {
      "epoch": 3.2248062015503876,
      "grad_norm": 0.19381049275398254,
      "learning_rate": 3.373808430633106e-05,
      "loss": 0.0086,
      "step": 2080
    },
    {
      "epoch": 3.24031007751938,
      "grad_norm": 0.30487295985221863,
      "learning_rate": 3.32271948827118e-05,
      "loss": 0.0099,
      "step": 2090
    },
    {
      "epoch": 3.255813953488372,
      "grad_norm": 0.21724191308021545,
      "learning_rate": 3.2718271109988863e-05,
      "loss": 0.0087,
      "step": 2100
    },
    {
      "epoch": 3.255813953488372,
      "eval_loss": 0.03577126935124397,
      "eval_runtime": 173.7017,
      "eval_samples_per_second": 6.603,
      "eval_steps_per_second": 6.603,
      "step": 2100
    },
    {
      "epoch": 3.2713178294573644,
      "grad_norm": 0.8536518812179565,
      "learning_rate": 3.2211372630330835e-05,
      "loss": 0.0094,
      "step": 2110
    },
    {
      "epoch": 3.2868217054263567,
      "grad_norm": 0.23373013734817505,
      "learning_rate": 3.170655884855661e-05,
      "loss": 0.0121,
      "step": 2120
    },
    {
      "epoch": 3.302325581395349,
      "grad_norm": 0.09981165081262589,
      "learning_rate": 3.120388892517368e-05,
      "loss": 0.0193,
      "step": 2130
    },
    {
      "epoch": 3.317829457364341,
      "grad_norm": 0.2072608470916748,
      "learning_rate": 3.070342176944494e-05,
      "loss": 0.0135,
      "step": 2140
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.07687430083751678,
      "learning_rate": 3.0205216032484805e-05,
      "loss": 0.0092,
      "step": 2150
    },
    {
      "epoch": 3.3333333333333335,
      "eval_loss": 0.032335054129362106,
      "eval_runtime": 173.7494,
      "eval_samples_per_second": 6.601,
      "eval_steps_per_second": 6.601,
      "step": 2150
    },
    {
      "epoch": 3.3488372093023258,
      "grad_norm": 0.2054612785577774,
      "learning_rate": 2.970933010038599e-05,
      "loss": 0.013,
      "step": 2160
    },
    {
      "epoch": 3.3643410852713176,
      "grad_norm": 0.17977559566497803,
      "learning_rate": 2.921582208737681e-05,
      "loss": 0.0088,
      "step": 2170
    },
    {
      "epoch": 3.37984496124031,
      "grad_norm": 0.23377564549446106,
      "learning_rate": 2.872474982901081e-05,
      "loss": 0.0136,
      "step": 2180
    },
    {
      "epoch": 3.395348837209302,
      "grad_norm": 0.16202200949192047,
      "learning_rate": 2.8236170875388744e-05,
      "loss": 0.01,
      "step": 2190
    },
    {
      "epoch": 3.4108527131782944,
      "grad_norm": 0.2739998698234558,
      "learning_rate": 2.775014248441422e-05,
      "loss": 0.0078,
      "step": 2200
    },
    {
      "epoch": 3.4108527131782944,
      "eval_loss": 0.03312241658568382,
      "eval_runtime": 173.7097,
      "eval_samples_per_second": 6.603,
      "eval_steps_per_second": 6.603,
      "step": 2200
    },
    {
      "epoch": 3.4263565891472867,
      "grad_norm": 0.10180757939815521,
      "learning_rate": 2.726672161508341e-05,
      "loss": 0.0098,
      "step": 2210
    },
    {
      "epoch": 3.441860465116279,
      "grad_norm": 0.27830150723457336,
      "learning_rate": 2.678596492080984e-05,
      "loss": 0.0077,
      "step": 2220
    },
    {
      "epoch": 3.4573643410852712,
      "grad_norm": 0.25859320163726807,
      "learning_rate": 2.630792874278516e-05,
      "loss": 0.0099,
      "step": 2230
    },
    {
      "epoch": 3.4728682170542635,
      "grad_norm": 0.4165274500846863,
      "learning_rate": 2.583266910337624e-05,
      "loss": 0.0095,
      "step": 2240
    },
    {
      "epoch": 3.488372093023256,
      "grad_norm": 0.5020641684532166,
      "learning_rate": 2.5360241699559816e-05,
      "loss": 0.0109,
      "step": 2250
    },
    {
      "epoch": 3.488372093023256,
      "eval_loss": 0.035617418587207794,
      "eval_runtime": 173.6138,
      "eval_samples_per_second": 6.607,
      "eval_steps_per_second": 6.607,
      "step": 2250
    },
    {
      "epoch": 3.503875968992248,
      "grad_norm": 0.19337324798107147,
      "learning_rate": 2.4890701896395146e-05,
      "loss": 0.0104,
      "step": 2260
    },
    {
      "epoch": 3.5193798449612403,
      "grad_norm": 0.1879683881998062,
      "learning_rate": 2.4424104720535735e-05,
      "loss": 0.0101,
      "step": 2270
    },
    {
      "epoch": 3.5348837209302326,
      "grad_norm": 0.063876673579216,
      "learning_rate": 2.3960504853780462e-05,
      "loss": 0.0115,
      "step": 2280
    },
    {
      "epoch": 3.550387596899225,
      "grad_norm": 0.11843185126781464,
      "learning_rate": 2.349995662666547e-05,
      "loss": 0.0085,
      "step": 2290
    },
    {
      "epoch": 3.565891472868217,
      "grad_norm": 0.21422308683395386,
      "learning_rate": 2.3042514012096843e-05,
      "loss": 0.0137,
      "step": 2300
    },
    {
      "epoch": 3.565891472868217,
      "eval_loss": 0.035960037261247635,
      "eval_runtime": 173.6653,
      "eval_samples_per_second": 6.605,
      "eval_steps_per_second": 6.605,
      "step": 2300
    },
    {
      "epoch": 3.5813953488372094,
      "grad_norm": 0.354839563369751,
      "learning_rate": 2.2588230619025407e-05,
      "loss": 0.0159,
      "step": 2310
    },
    {
      "epoch": 3.5968992248062017,
      "grad_norm": 0.2568114995956421,
      "learning_rate": 2.213715968616425e-05,
      "loss": 0.0109,
      "step": 2320
    },
    {
      "epoch": 3.612403100775194,
      "grad_norm": 0.06852281093597412,
      "learning_rate": 2.16893540757494e-05,
      "loss": 0.0091,
      "step": 2330
    },
    {
      "epoch": 3.6279069767441863,
      "grad_norm": 0.194693461060524,
      "learning_rate": 2.1244866267344866e-05,
      "loss": 0.0084,
      "step": 2340
    },
    {
      "epoch": 3.6434108527131785,
      "grad_norm": 0.3488699495792389,
      "learning_rate": 2.080374835169235e-05,
      "loss": 0.013,
      "step": 2350
    },
    {
      "epoch": 3.6434108527131785,
      "eval_loss": 0.03496725112199783,
      "eval_runtime": 173.6559,
      "eval_samples_per_second": 6.605,
      "eval_steps_per_second": 6.605,
      "step": 2350
    },
    {
      "epoch": 3.6589147286821704,
      "grad_norm": 0.16774657368659973,
      "learning_rate": 2.0366052024606612e-05,
      "loss": 0.0082,
      "step": 2360
    },
    {
      "epoch": 3.6744186046511627,
      "grad_norm": 0.2563646733760834,
      "learning_rate": 1.9931828580917107e-05,
      "loss": 0.0114,
      "step": 2370
    },
    {
      "epoch": 3.689922480620155,
      "grad_norm": 0.254089891910553,
      "learning_rate": 1.9501128908456523e-05,
      "loss": 0.0087,
      "step": 2380
    },
    {
      "epoch": 3.705426356589147,
      "grad_norm": 0.2673187255859375,
      "learning_rate": 1.9074003482097258e-05,
      "loss": 0.0093,
      "step": 2390
    },
    {
      "epoch": 3.7209302325581395,
      "grad_norm": 0.47327151894569397,
      "learning_rate": 1.8650502357835925e-05,
      "loss": 0.0133,
      "step": 2400
    },
    {
      "epoch": 3.7209302325581395,
      "eval_loss": 0.03533458709716797,
      "eval_runtime": 173.6028,
      "eval_samples_per_second": 6.607,
      "eval_steps_per_second": 6.607,
      "step": 2400
    },
    {
      "epoch": 3.7364341085271318,
      "grad_norm": 0.4366307556629181,
      "learning_rate": 1.82306751669274e-05,
      "loss": 0.0085,
      "step": 2410
    },
    {
      "epoch": 3.751937984496124,
      "grad_norm": 0.39723125100135803,
      "learning_rate": 1.7814571110068135e-05,
      "loss": 0.0123,
      "step": 2420
    },
    {
      "epoch": 3.7674418604651163,
      "grad_norm": 0.16450275480747223,
      "learning_rate": 1.740223895163039e-05,
      "loss": 0.0091,
      "step": 2430
    },
    {
      "epoch": 3.7829457364341086,
      "grad_norm": 0.06844169646501541,
      "learning_rate": 1.6993727013947336e-05,
      "loss": 0.0109,
      "step": 2440
    },
    {
      "epoch": 3.798449612403101,
      "grad_norm": 0.29116564989089966,
      "learning_rate": 1.6589083171649977e-05,
      "loss": 0.0068,
      "step": 2450
    },
    {
      "epoch": 3.798449612403101,
      "eval_loss": 0.035682931542396545,
      "eval_runtime": 173.5613,
      "eval_samples_per_second": 6.609,
      "eval_steps_per_second": 6.609,
      "step": 2450
    },
    {
      "epoch": 3.813953488372093,
      "grad_norm": 0.1012692078948021,
      "learning_rate": 1.6188354846056698e-05,
      "loss": 0.008,
      "step": 2460
    },
    {
      "epoch": 3.8294573643410854,
      "grad_norm": 0.1340951770544052,
      "learning_rate": 1.579158899961575e-05,
      "loss": 0.0054,
      "step": 2470
    },
    {
      "epoch": 3.8449612403100772,
      "grad_norm": 0.22337476909160614,
      "learning_rate": 1.5398832130401637e-05,
      "loss": 0.0147,
      "step": 2480
    },
    {
      "epoch": 3.8604651162790695,
      "grad_norm": 0.18346606194972992,
      "learning_rate": 1.5010130266665807e-05,
      "loss": 0.0108,
      "step": 2490
    },
    {
      "epoch": 3.875968992248062,
      "grad_norm": 0.2582853436470032,
      "learning_rate": 1.4625528961442591e-05,
      "loss": 0.012,
      "step": 2500
    },
    {
      "epoch": 3.875968992248062,
      "eval_loss": 0.034770578145980835,
      "eval_runtime": 173.5234,
      "eval_samples_per_second": 6.61,
      "eval_steps_per_second": 6.61,
      "step": 2500
    },
    {
      "epoch": 3.891472868217054,
      "grad_norm": 0.22837190330028534,
      "learning_rate": 1.4245073287210608e-05,
      "loss": 0.0098,
      "step": 2510
    },
    {
      "epoch": 3.9069767441860463,
      "grad_norm": 0.49539613723754883,
      "learning_rate": 1.3868807830610653e-05,
      "loss": 0.016,
      "step": 2520
    },
    {
      "epoch": 3.9224806201550386,
      "grad_norm": 0.17993128299713135,
      "learning_rate": 1.3496776687220514e-05,
      "loss": 0.0129,
      "step": 2530
    },
    {
      "epoch": 3.937984496124031,
      "grad_norm": 0.25489673018455505,
      "learning_rate": 1.3129023456387151e-05,
      "loss": 0.0068,
      "step": 2540
    },
    {
      "epoch": 3.953488372093023,
      "grad_norm": 0.06681143492460251,
      "learning_rate": 1.2765591236117324e-05,
      "loss": 0.0088,
      "step": 2550
    },
    {
      "epoch": 3.953488372093023,
      "eval_loss": 0.034369926899671555,
      "eval_runtime": 173.574,
      "eval_samples_per_second": 6.608,
      "eval_steps_per_second": 6.608,
      "step": 2550
    },
    {
      "epoch": 3.9689922480620154,
      "grad_norm": 0.14447127282619476,
      "learning_rate": 1.240652261802669e-05,
      "loss": 0.0105,
      "step": 2560
    },
    {
      "epoch": 3.9844961240310077,
      "grad_norm": 0.2608015239238739,
      "learning_rate": 1.205185968234847e-05,
      "loss": 0.0136,
      "step": 2570
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.17687876522541046,
      "learning_rate": 1.1701643993001916e-05,
      "loss": 0.0093,
      "step": 2580
    },
    {
      "epoch": 4.015503875968992,
      "grad_norm": 0.1070009246468544,
      "learning_rate": 1.1355916592721316e-05,
      "loss": 0.0045,
      "step": 2590
    },
    {
      "epoch": 4.0310077519379846,
      "grad_norm": 0.11302992701530457,
      "learning_rate": 1.1014717998246099e-05,
      "loss": 0.0066,
      "step": 2600
    },
    {
      "epoch": 4.0310077519379846,
      "eval_loss": 0.03459889069199562,
      "eval_runtime": 173.5445,
      "eval_samples_per_second": 6.609,
      "eval_steps_per_second": 6.609,
      "step": 2600
    },
    {
      "epoch": 4.046511627906977,
      "grad_norm": 0.107364721596241,
      "learning_rate": 1.0678088195572517e-05,
      "loss": 0.0053,
      "step": 2610
    },
    {
      "epoch": 4.062015503875969,
      "grad_norm": 0.2821330428123474,
      "learning_rate": 1.0346066635267676e-05,
      "loss": 0.0069,
      "step": 2620
    },
    {
      "epoch": 4.077519379844961,
      "grad_norm": 0.17317578196525574,
      "learning_rate": 1.001869222784611e-05,
      "loss": 0.0064,
      "step": 2630
    },
    {
      "epoch": 4.093023255813954,
      "grad_norm": 0.08701343834400177,
      "learning_rate": 9.696003339209819e-06,
      "loss": 0.007,
      "step": 2640
    },
    {
      "epoch": 4.108527131782946,
      "grad_norm": 0.23731686174869537,
      "learning_rate": 9.378037786152055e-06,
      "loss": 0.0052,
      "step": 2650
    },
    {
      "epoch": 4.108527131782946,
      "eval_loss": 0.03613542765378952,
      "eval_runtime": 173.8232,
      "eval_samples_per_second": 6.599,
      "eval_steps_per_second": 6.599,
      "step": 2650
    },
    {
      "epoch": 4.124031007751938,
      "grad_norm": 0.13787880539894104,
      "learning_rate": 9.064832831925441e-06,
      "loss": 0.0054,
      "step": 2660
    },
    {
      "epoch": 4.1395348837209305,
      "grad_norm": 0.1381884068250656,
      "learning_rate": 8.756425181875028e-06,
      "loss": 0.0035,
      "step": 2670
    },
    {
      "epoch": 4.155038759689923,
      "grad_norm": 0.16911053657531738,
      "learning_rate": 8.452850979136617e-06,
      "loss": 0.0045,
      "step": 2680
    },
    {
      "epoch": 4.170542635658915,
      "grad_norm": 0.04475264623761177,
      "learning_rate": 8.15414580040117e-06,
      "loss": 0.0032,
      "step": 2690
    },
    {
      "epoch": 4.186046511627907,
      "grad_norm": 0.09773126989603043,
      "learning_rate": 7.860344651745366e-06,
      "loss": 0.008,
      "step": 2700
    },
    {
      "epoch": 4.186046511627907,
      "eval_loss": 0.037369657307863235,
      "eval_runtime": 173.9253,
      "eval_samples_per_second": 6.595,
      "eval_steps_per_second": 6.595,
      "step": 2700
    },
    {
      "epoch": 4.2015503875969,
      "grad_norm": 0.1751956194639206,
      "learning_rate": 7.571481964529226e-06,
      "loss": 0.0057,
      "step": 2710
    },
    {
      "epoch": 4.217054263565892,
      "grad_norm": 0.06625787168741226,
      "learning_rate": 7.287591591360926e-06,
      "loss": 0.0028,
      "step": 2720
    },
    {
      "epoch": 4.232558139534884,
      "grad_norm": 0.1029408872127533,
      "learning_rate": 7.008706802129606e-06,
      "loss": 0.0069,
      "step": 2730
    },
    {
      "epoch": 4.248062015503876,
      "grad_norm": 0.01715020090341568,
      "learning_rate": 6.734860280106292e-06,
      "loss": 0.0021,
      "step": 2740
    },
    {
      "epoch": 4.263565891472869,
      "grad_norm": 0.5773080587387085,
      "learning_rate": 6.466084118113735e-06,
      "loss": 0.0062,
      "step": 2750
    },
    {
      "epoch": 4.263565891472869,
      "eval_loss": 0.0383269228041172,
      "eval_runtime": 173.7893,
      "eval_samples_per_second": 6.6,
      "eval_steps_per_second": 6.6,
      "step": 2750
    },
    {
      "epoch": 4.27906976744186,
      "grad_norm": 0.09289313852787018,
      "learning_rate": 6.202409814765325e-06,
      "loss": 0.0043,
      "step": 2760
    },
    {
      "epoch": 4.294573643410852,
      "grad_norm": 0.16086533665657043,
      "learning_rate": 5.9438682707736725e-06,
      "loss": 0.0064,
      "step": 2770
    },
    {
      "epoch": 4.310077519379845,
      "grad_norm": 0.175432026386261,
      "learning_rate": 5.690489785329301e-06,
      "loss": 0.0052,
      "step": 2780
    },
    {
      "epoch": 4.325581395348837,
      "grad_norm": 0.03287180885672569,
      "learning_rate": 5.442304052549752e-06,
      "loss": 0.0037,
      "step": 2790
    },
    {
      "epoch": 4.341085271317829,
      "grad_norm": 0.09991113096475601,
      "learning_rate": 5.199340157999733e-06,
      "loss": 0.005,
      "step": 2800
    },
    {
      "epoch": 4.341085271317829,
      "eval_loss": 0.038646504282951355,
      "eval_runtime": 173.5908,
      "eval_samples_per_second": 6.607,
      "eval_steps_per_second": 6.607,
      "step": 2800
    },
    {
      "epoch": 4.3565891472868215,
      "grad_norm": 0.07446504384279251,
      "learning_rate": 4.961626575282396e-06,
      "loss": 0.0028,
      "step": 2810
    },
    {
      "epoch": 4.372093023255814,
      "grad_norm": 0.1563124656677246,
      "learning_rate": 4.7291911627025235e-06,
      "loss": 0.0045,
      "step": 2820
    },
    {
      "epoch": 4.387596899224806,
      "grad_norm": 0.09954715520143509,
      "learning_rate": 4.502061160001725e-06,
      "loss": 0.0038,
      "step": 2830
    },
    {
      "epoch": 4.403100775193798,
      "grad_norm": 0.03728143870830536,
      "learning_rate": 4.280263185166078e-06,
      "loss": 0.007,
      "step": 2840
    },
    {
      "epoch": 4.4186046511627906,
      "grad_norm": 0.14879867434501648,
      "learning_rate": 4.063823231306757e-06,
      "loss": 0.004,
      "step": 2850
    },
    {
      "epoch": 4.4186046511627906,
      "eval_loss": 0.03950056806206703,
      "eval_runtime": 173.7546,
      "eval_samples_per_second": 6.601,
      "eval_steps_per_second": 6.601,
      "step": 2850
    },
    {
      "epoch": 4.434108527131783,
      "grad_norm": 0.04011745750904083,
      "learning_rate": 3.8527666636137885e-06,
      "loss": 0.0039,
      "step": 2860
    },
    {
      "epoch": 4.449612403100775,
      "grad_norm": 0.16904973983764648,
      "learning_rate": 3.6471182163834825e-06,
      "loss": 0.0075,
      "step": 2870
    },
    {
      "epoch": 4.465116279069767,
      "grad_norm": 0.05882781371474266,
      "learning_rate": 3.4469019901197054e-06,
      "loss": 0.0027,
      "step": 2880
    },
    {
      "epoch": 4.48062015503876,
      "grad_norm": 0.36288192868232727,
      "learning_rate": 3.252141448709495e-06,
      "loss": 0.0052,
      "step": 2890
    },
    {
      "epoch": 4.496124031007752,
      "grad_norm": 0.25636863708496094,
      "learning_rate": 3.0628594166732693e-06,
      "loss": 0.0075,
      "step": 2900
    },
    {
      "epoch": 4.496124031007752,
      "eval_loss": 0.040012069046497345,
      "eval_runtime": 173.5684,
      "eval_samples_per_second": 6.608,
      "eval_steps_per_second": 6.608,
      "step": 2900
    },
    {
      "epoch": 4.511627906976744,
      "grad_norm": 0.07002587616443634,
      "learning_rate": 2.8790780764899384e-06,
      "loss": 0.0053,
      "step": 2910
    },
    {
      "epoch": 4.5271317829457365,
      "grad_norm": 0.18481026589870453,
      "learning_rate": 2.700818965997315e-06,
      "loss": 0.0064,
      "step": 2920
    },
    {
      "epoch": 4.542635658914729,
      "grad_norm": 0.1123792976140976,
      "learning_rate": 2.528102975867991e-06,
      "loss": 0.0033,
      "step": 2930
    },
    {
      "epoch": 4.558139534883721,
      "grad_norm": 0.039597250521183014,
      "learning_rate": 2.3609503471611284e-06,
      "loss": 0.0041,
      "step": 2940
    },
    {
      "epoch": 4.573643410852713,
      "grad_norm": 0.05195772275328636,
      "learning_rate": 2.1993806689503738e-06,
      "loss": 0.003,
      "step": 2950
    },
    {
      "epoch": 4.573643410852713,
      "eval_loss": 0.040207840502262115,
      "eval_runtime": 173.5139,
      "eval_samples_per_second": 6.61,
      "eval_steps_per_second": 6.61,
      "step": 2950
    },
    {
      "epoch": 4.589147286821706,
      "grad_norm": 0.06371759623289108,
      "learning_rate": 2.0434128760281167e-06,
      "loss": 0.0021,
      "step": 2960
    },
    {
      "epoch": 4.604651162790698,
      "grad_norm": 0.0519726499915123,
      "learning_rate": 1.893065246686504e-06,
      "loss": 0.0029,
      "step": 2970
    },
    {
      "epoch": 4.62015503875969,
      "grad_norm": 0.05901438370347023,
      "learning_rate": 1.74835540057533e-06,
      "loss": 0.0067,
      "step": 2980
    },
    {
      "epoch": 4.635658914728682,
      "grad_norm": 0.16322703659534454,
      "learning_rate": 1.6093002966371617e-06,
      "loss": 0.0044,
      "step": 2990
    },
    {
      "epoch": 4.651162790697675,
      "grad_norm": 0.25651177763938904,
      "learning_rate": 1.4759162311198783e-06,
      "loss": 0.0066,
      "step": 3000
    },
    {
      "epoch": 4.651162790697675,
      "eval_loss": 0.0405086986720562,
      "eval_runtime": 173.4642,
      "eval_samples_per_second": 6.612,
      "eval_steps_per_second": 6.612,
      "step": 3000
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.009503856301307678,
      "learning_rate": 1.348218835666859e-06,
      "loss": 0.0032,
      "step": 3010
    },
    {
      "epoch": 4.682170542635659,
      "grad_norm": 0.12524276971817017,
      "learning_rate": 1.2262230754850445e-06,
      "loss": 0.0032,
      "step": 3020
    },
    {
      "epoch": 4.6976744186046515,
      "grad_norm": 0.06170937791466713,
      "learning_rate": 1.109943247591172e-06,
      "loss": 0.0045,
      "step": 3030
    },
    {
      "epoch": 4.713178294573644,
      "grad_norm": 0.12521466612815857,
      "learning_rate": 9.993929791362323e-07,
      "loss": 0.0052,
      "step": 3040
    },
    {
      "epoch": 4.728682170542635,
      "grad_norm": 0.17083865404129028,
      "learning_rate": 8.945852258084863e-07,
      "loss": 0.005,
      "step": 3050
    },
    {
      "epoch": 4.728682170542635,
      "eval_loss": 0.04061457887291908,
      "eval_runtime": 173.5187,
      "eval_samples_per_second": 6.61,
      "eval_steps_per_second": 6.61,
      "step": 3050
    },
    {
      "epoch": 4.7441860465116275,
      "grad_norm": 0.35159188508987427,
      "learning_rate": 7.955322703151358e-07,
      "loss": 0.0046,
      "step": 3060
    },
    {
      "epoch": 4.75968992248062,
      "grad_norm": 0.2449195384979248,
      "learning_rate": 7.022457209428901e-07,
      "loss": 0.0044,
      "step": 3070
    },
    {
      "epoch": 4.775193798449612,
      "grad_norm": 0.2619343101978302,
      "learning_rate": 6.147365101975666e-07,
      "loss": 0.0035,
      "step": 3080
    },
    {
      "epoch": 4.790697674418604,
      "grad_norm": 0.09077528119087219,
      "learning_rate": 5.33014893522854e-07,
      "loss": 0.0032,
      "step": 3090
    },
    {
      "epoch": 4.8062015503875966,
      "grad_norm": 0.14284414052963257,
      "learning_rate": 4.57090448098485e-07,
      "loss": 0.0067,
      "step": 3100
    },
    {
      "epoch": 4.8062015503875966,
      "eval_loss": 0.04071947932243347,
      "eval_runtime": 173.7232,
      "eval_samples_per_second": 6.602,
      "eval_steps_per_second": 6.602,
      "step": 3100
    },
    {
      "epoch": 4.821705426356589,
      "grad_norm": 0.10482044517993927,
      "learning_rate": 3.8697207171781714e-07,
      "loss": 0.0037,
      "step": 3110
    },
    {
      "epoch": 4.837209302325581,
      "grad_norm": 0.03253592178225517,
      "learning_rate": 3.2266798174512837e-07,
      "loss": 0.0045,
      "step": 3120
    },
    {
      "epoch": 4.852713178294573,
      "grad_norm": 0.03622818365693092,
      "learning_rate": 2.6418571415255387e-07,
      "loss": 0.0093,
      "step": 3130
    },
    {
      "epoch": 4.868217054263566,
      "grad_norm": 0.09220151603221893,
      "learning_rate": 2.1153212263695378e-07,
      "loss": 0.0046,
      "step": 3140
    },
    {
      "epoch": 4.883720930232558,
      "grad_norm": 0.1942395716905594,
      "learning_rate": 1.6471337781669982e-07,
      "loss": 0.0067,
      "step": 3150
    },
    {
      "epoch": 4.883720930232558,
      "eval_loss": 0.040707945823669434,
      "eval_runtime": 173.7396,
      "eval_samples_per_second": 6.602,
      "eval_steps_per_second": 6.602,
      "step": 3150
    },
    {
      "epoch": 4.89922480620155,
      "grad_norm": 0.09173876792192459,
      "learning_rate": 1.237349665085097e-07,
      "loss": 0.0049,
      "step": 3160
    },
    {
      "epoch": 4.9147286821705425,
      "grad_norm": 0.045584529638290405,
      "learning_rate": 8.86016910844667e-08,
      "loss": 0.003,
      "step": 3170
    },
    {
      "epoch": 4.930232558139535,
      "grad_norm": 0.46576347947120667,
      "learning_rate": 5.9317668909192323e-08,
      "loss": 0.0061,
      "step": 3180
    },
    {
      "epoch": 4.945736434108527,
      "grad_norm": 0.059987977147102356,
      "learning_rate": 3.588633185730994e-08,
      "loss": 0.0062,
      "step": 3190
    },
    {
      "epoch": 4.961240310077519,
      "grad_norm": 0.2531404197216034,
      "learning_rate": 1.8310425911294283e-08,
      "loss": 0.006,
      "step": 3200
    },
    {
      "epoch": 4.961240310077519,
      "eval_loss": 0.04073048010468483,
      "eval_runtime": 173.54,
      "eval_samples_per_second": 6.609,
      "eval_steps_per_second": 6.609,
      "step": 3200
    },
    {
      "epoch": 4.976744186046512,
      "grad_norm": 0.24996767938137054,
      "learning_rate": 6.592010839612251e-09,
      "loss": 0.0076,
      "step": 3210
    },
    {
      "epoch": 4.992248062015504,
      "grad_norm": 0.020084669813513756,
      "learning_rate": 7.324599553770739e-10,
      "loss": 0.0025,
      "step": 3220
    },
    {
      "epoch": 5.0,
      "step": 3225,
      "total_flos": 8.718478050646426e+17,
      "train_loss": 0.049027663960821866,
      "train_runtime": 36714.2196,
      "train_samples_per_second": 1.405,
      "train_steps_per_second": 0.088
    }
  ],
  "logging_steps": 10,
  "max_steps": 3225,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.718478050646426e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}