{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 34595,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07226477814713109,
      "grad_norm": 5.687336444854736,
      "learning_rate": 2.9566411331117216e-05,
      "loss": 1.6256,
      "step": 500
    },
    {
      "epoch": 0.14452955629426217,
      "grad_norm": 13.803632736206055,
      "learning_rate": 2.9132822662234428e-05,
      "loss": 1.1029,
      "step": 1000
    },
    {
      "epoch": 0.21679433444139326,
      "grad_norm": 10.664984703063965,
      "learning_rate": 2.8699233993351643e-05,
      "loss": 0.9223,
      "step": 1500
    },
    {
      "epoch": 0.28905911258852435,
      "grad_norm": 10.714991569519043,
      "learning_rate": 2.8265645324468854e-05,
      "loss": 0.9753,
      "step": 2000
    },
    {
      "epoch": 0.36132389073565546,
      "grad_norm": 20.49856185913086,
      "learning_rate": 2.783205665558607e-05,
      "loss": 0.9205,
      "step": 2500
    },
    {
      "epoch": 0.4335886688827865,
      "grad_norm": 32.648468017578125,
      "learning_rate": 2.739846798670328e-05,
      "loss": 0.8662,
      "step": 3000
    },
    {
      "epoch": 0.5058534470299176,
      "grad_norm": 1.8131040334701538,
      "learning_rate": 2.6964879317820496e-05,
      "loss": 0.8642,
      "step": 3500
    },
    {
      "epoch": 0.5781182251770487,
      "grad_norm": 108.92765808105469,
      "learning_rate": 2.6531290648937708e-05,
      "loss": 0.8668,
      "step": 4000
    },
    {
      "epoch": 0.6503830033241798,
      "grad_norm": 70.46207427978516,
      "learning_rate": 2.6097701980054923e-05,
      "loss": 0.8507,
      "step": 4500
    },
    {
      "epoch": 0.7226477814713109,
      "grad_norm": 41.309329986572266,
      "learning_rate": 2.5664113311172135e-05,
      "loss": 0.8232,
      "step": 5000
    },
    {
      "epoch": 0.794912559618442,
      "grad_norm": 66.08831787109375,
      "learning_rate": 2.523052464228935e-05,
      "loss": 0.8212,
      "step": 5500
    },
    {
      "epoch": 0.867177337765573,
      "grad_norm": 8.820863723754883,
      "learning_rate": 2.4796935973406565e-05,
      "loss": 0.7775,
      "step": 6000
    },
    {
      "epoch": 0.9394421159127041,
      "grad_norm": 3.9847397804260254,
      "learning_rate": 2.4363347304523777e-05,
      "loss": 0.7593,
      "step": 6500
    },
    {
      "epoch": 1.0117068940598353,
      "grad_norm": 2.6096384525299072,
      "learning_rate": 2.392975863564099e-05,
      "loss": 0.7395,
      "step": 7000
    },
    {
      "epoch": 1.0839716722069663,
      "grad_norm": 0.2627348303794861,
      "learning_rate": 2.34961699667582e-05,
      "loss": 0.5237,
      "step": 7500
    },
    {
      "epoch": 1.1562364503540974,
      "grad_norm": 13.379576683044434,
      "learning_rate": 2.3062581297875416e-05,
      "loss": 0.5469,
      "step": 8000
    },
    {
      "epoch": 1.2285012285012284,
      "grad_norm": 1.7146592140197754,
      "learning_rate": 2.2628992628992627e-05,
      "loss": 0.555,
      "step": 8500
    },
    {
      "epoch": 1.3007660066483595,
      "grad_norm": 40.48280334472656,
      "learning_rate": 2.2195403960109842e-05,
      "loss": 0.503,
      "step": 9000
    },
    {
      "epoch": 1.3730307847954908,
      "grad_norm": 42.039703369140625,
      "learning_rate": 2.1761815291227054e-05,
      "loss": 0.4805,
      "step": 9500
    },
    {
      "epoch": 1.4452955629426218,
      "grad_norm": 88.76701354980469,
      "learning_rate": 2.132822662234427e-05,
      "loss": 0.449,
      "step": 10000
    },
    {
      "epoch": 1.517560341089753,
      "grad_norm": 2.241396188735962,
      "learning_rate": 2.089463795346148e-05,
      "loss": 0.546,
      "step": 10500
    },
    {
      "epoch": 1.589825119236884,
      "grad_norm": 40.65458297729492,
      "learning_rate": 2.0461049284578696e-05,
      "loss": 0.5435,
      "step": 11000
    },
    {
      "epoch": 1.662089897384015,
      "grad_norm": 8.819799423217773,
      "learning_rate": 2.002746061569591e-05,
      "loss": 0.4942,
      "step": 11500
    },
    {
      "epoch": 1.734354675531146,
      "grad_norm": 2.431903839111328,
      "learning_rate": 1.9593871946813123e-05,
      "loss": 0.5563,
      "step": 12000
    },
    {
      "epoch": 1.8066194536782771,
      "grad_norm": 45.27000427246094,
      "learning_rate": 1.9160283277930338e-05,
      "loss": 0.4843,
      "step": 12500
    },
    {
      "epoch": 1.8788842318254084,
      "grad_norm": 0.5083319544792175,
      "learning_rate": 1.872669460904755e-05,
      "loss": 0.5188,
      "step": 13000
    },
    {
      "epoch": 1.9511490099725393,
      "grad_norm": 146.18194580078125,
      "learning_rate": 1.8293105940164765e-05,
      "loss": 0.5161,
      "step": 13500
    },
    {
      "epoch": 2.0234137881196705,
      "grad_norm": 6.673525333404541,
      "learning_rate": 1.7859517271281977e-05,
      "loss": 0.4454,
      "step": 14000
    },
    {
      "epoch": 2.0956785662668014,
      "grad_norm": 2.377631187438965,
      "learning_rate": 1.7425928602399192e-05,
      "loss": 0.2991,
      "step": 14500
    },
    {
      "epoch": 2.1679433444139327,
      "grad_norm": 2.739825963973999,
      "learning_rate": 1.6992339933516404e-05,
      "loss": 0.3334,
      "step": 15000
    },
    {
      "epoch": 2.240208122561064,
      "grad_norm": 2.2864904403686523,
      "learning_rate": 1.655875126463362e-05,
      "loss": 0.3069,
      "step": 15500
    },
    {
      "epoch": 2.312472900708195,
      "grad_norm": 2.948648691177368,
      "learning_rate": 1.6125162595750834e-05,
      "loss": 0.3202,
      "step": 16000
    },
    {
      "epoch": 2.384737678855326,
      "grad_norm": 0.33637821674346924,
      "learning_rate": 1.5691573926868046e-05,
      "loss": 0.3559,
      "step": 16500
    },
    {
      "epoch": 2.457002457002457,
      "grad_norm": 61.67799377441406,
      "learning_rate": 1.5257985257985259e-05,
      "loss": 0.3195,
      "step": 17000
    },
    {
      "epoch": 2.529267235149588,
      "grad_norm": 0.03277577832341194,
      "learning_rate": 1.482439658910247e-05,
      "loss": 0.3241,
      "step": 17500
    },
    {
      "epoch": 2.601532013296719,
      "grad_norm": 0.016185609623789787,
      "learning_rate": 1.4390807920219684e-05,
      "loss": 0.3099,
      "step": 18000
    },
    {
      "epoch": 2.6737967914438503,
      "grad_norm": 27.654953002929688,
      "learning_rate": 1.39572192513369e-05,
      "loss": 0.286,
      "step": 18500
    },
    {
      "epoch": 2.7460615695909816,
      "grad_norm": 0.4255935251712799,
      "learning_rate": 1.3523630582454113e-05,
      "loss": 0.3362,
      "step": 19000
    },
    {
      "epoch": 2.8183263477381124,
      "grad_norm": 17.48617935180664,
      "learning_rate": 1.3090041913571326e-05,
      "loss": 0.346,
      "step": 19500
    },
    {
      "epoch": 2.8905911258852437,
      "grad_norm": 184.4835968017578,
      "learning_rate": 1.265645324468854e-05,
      "loss": 0.3076,
      "step": 20000
    },
    {
      "epoch": 2.9628559040323745,
      "grad_norm": 4.913192272186279,
      "learning_rate": 1.2222864575805753e-05,
      "loss": 0.2971,
      "step": 20500
    },
    {
      "epoch": 3.035120682179506,
      "grad_norm": 19.735275268554688,
      "learning_rate": 1.1789275906922966e-05,
      "loss": 0.2479,
      "step": 21000
    },
    {
      "epoch": 3.1073854603266367,
      "grad_norm": 0.6235244274139404,
      "learning_rate": 1.135568723804018e-05,
      "loss": 0.1986,
      "step": 21500
    },
    {
      "epoch": 3.179650238473768,
      "grad_norm": 0.0023228218778967857,
      "learning_rate": 1.0922098569157393e-05,
      "loss": 0.2097,
      "step": 22000
    },
    {
      "epoch": 3.2519150166208988,
      "grad_norm": 0.006465112790465355,
      "learning_rate": 1.0488509900274607e-05,
      "loss": 0.1967,
      "step": 22500
    },
    {
      "epoch": 3.32417979476803,
      "grad_norm": 0.03349950537085533,
      "learning_rate": 1.0054921231391818e-05,
      "loss": 0.1744,
      "step": 23000
    },
    {
      "epoch": 3.3964445729151613,
      "grad_norm": 0.0003899051807820797,
      "learning_rate": 9.621332562509032e-06,
      "loss": 0.1687,
      "step": 23500
    },
    {
      "epoch": 3.468709351062292,
      "grad_norm": 0.004888344090431929,
      "learning_rate": 9.187743893626247e-06,
      "loss": 0.19,
      "step": 24000
    },
    {
      "epoch": 3.5409741292094234,
      "grad_norm": 0.2528671324253082,
      "learning_rate": 8.75415522474346e-06,
      "loss": 0.1723,
      "step": 24500
    },
    {
      "epoch": 3.6132389073565543,
      "grad_norm": 32.603153228759766,
      "learning_rate": 8.320566555860674e-06,
      "loss": 0.1692,
      "step": 25000
    },
    {
      "epoch": 3.6855036855036856,
      "grad_norm": 0.0015944232000038028,
      "learning_rate": 7.886977886977887e-06,
      "loss": 0.1671,
      "step": 25500
    },
    {
      "epoch": 3.757768463650817,
      "grad_norm": 0.02464761771261692,
      "learning_rate": 7.453389218095101e-06,
      "loss": 0.1688,
      "step": 26000
    },
    {
      "epoch": 3.8300332417979477,
      "grad_norm": 0.4081017076969147,
      "learning_rate": 7.019800549212314e-06,
      "loss": 0.1813,
      "step": 26500
    },
    {
      "epoch": 3.9022980199450785,
      "grad_norm": 0.32034188508987427,
      "learning_rate": 6.586211880329528e-06,
      "loss": 0.1801,
      "step": 27000
    },
    {
      "epoch": 3.97456279809221,
      "grad_norm": 3.1471328735351562,
      "learning_rate": 6.152623211446741e-06,
      "loss": 0.1667,
      "step": 27500
    },
    {
      "epoch": 4.046827576239341,
      "grad_norm": 0.0016349386423826218,
      "learning_rate": 5.7190345425639545e-06,
      "loss": 0.1641,
      "step": 28000
    },
    {
      "epoch": 4.119092354386472,
      "grad_norm": 1.3927046060562134,
      "learning_rate": 5.285445873681168e-06,
      "loss": 0.1069,
      "step": 28500
    },
    {
      "epoch": 4.191357132533603,
      "grad_norm": 0.0006769502069801092,
      "learning_rate": 4.851857204798381e-06,
      "loss": 0.0815,
      "step": 29000
    },
    {
      "epoch": 4.263621910680734,
      "grad_norm": 0.0053017823956906796,
      "learning_rate": 4.418268535915595e-06,
      "loss": 0.1065,
      "step": 29500
    },
    {
      "epoch": 4.335886688827865,
      "grad_norm": 0.0034885455388575792,
      "learning_rate": 3.984679867032808e-06,
      "loss": 0.085,
      "step": 30000
    },
    {
      "epoch": 4.408151466974997,
      "grad_norm": 1.5222100019454956,
      "learning_rate": 3.5510911981500216e-06,
      "loss": 0.1032,
      "step": 30500
    },
    {
      "epoch": 4.480416245122128,
      "grad_norm": 0.0017290544928982854,
      "learning_rate": 3.117502529267235e-06,
      "loss": 0.0941,
      "step": 31000
    },
    {
      "epoch": 4.552681023269258,
      "grad_norm": 0.0026269140653312206,
      "learning_rate": 2.683913860384449e-06,
      "loss": 0.0981,
      "step": 31500
    },
    {
      "epoch": 4.62494580141639,
      "grad_norm": 5.1134233474731445,
      "learning_rate": 2.250325191501662e-06,
      "loss": 0.0995,
      "step": 32000
    },
    {
      "epoch": 4.697210579563521,
      "grad_norm": 0.0005454424535855651,
      "learning_rate": 1.8167365226188756e-06,
      "loss": 0.0927,
      "step": 32500
    },
    {
      "epoch": 4.769475357710652,
      "grad_norm": 0.0004920898354612291,
      "learning_rate": 1.383147853736089e-06,
      "loss": 0.0756,
      "step": 33000
    },
    {
      "epoch": 4.841740135857783,
      "grad_norm": 0.0010318453423678875,
      "learning_rate": 9.495591848533026e-07,
      "loss": 0.0837,
      "step": 33500
    },
    {
      "epoch": 4.914004914004914,
      "grad_norm": 45.28452682495117,
      "learning_rate": 5.15970515970516e-07,
      "loss": 0.0928,
      "step": 34000
    },
    {
      "epoch": 4.986269692152045,
      "grad_norm": 0.00040028223884291947,
      "learning_rate": 8.238184708772944e-08,
      "loss": 0.0767,
      "step": 34500
    },
    {
      "epoch": 5.0,
      "step": 34595,
      "total_flos": 3.6156938713344e+16,
      "train_loss": 0.40735532518307727,
      "train_runtime": 4631.1638,
      "train_samples_per_second": 29.879,
      "train_steps_per_second": 7.47
    }
  ],
  "logging_steps": 500,
  "max_steps": 34595,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 3.6156938713344e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}