{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998461775111521,
  "eval_steps": 500,
  "global_step": 3250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015382248884786957,
      "grad_norm": 1.2864877128592522,
      "learning_rate": 1.0204081632653063e-05,
      "loss": 1.211,
      "step": 50
    },
    {
      "epoch": 0.030764497769573913,
      "grad_norm": 1.227813083166002,
      "learning_rate": 1.999998013186146e-05,
      "loss": 1.0177,
      "step": 100
    },
    {
      "epoch": 0.046146746654360866,
      "grad_norm": 1.16006238153043,
      "learning_rate": 1.9986572140101312e-05,
      "loss": 0.9909,
      "step": 150
    },
    {
      "epoch": 0.06152899553914783,
      "grad_norm": 1.2120743368126807,
      "learning_rate": 1.9948367447837503e-05,
      "loss": 0.9853,
      "step": 200
    },
    {
      "epoch": 0.07691124442393478,
      "grad_norm": 1.0434928463049529,
      "learning_rate": 1.9885460917481137e-05,
      "loss": 0.9766,
      "step": 250
    },
    {
      "epoch": 0.09229349330872173,
      "grad_norm": 1.0799628915416184,
      "learning_rate": 1.9798008746210894e-05,
      "loss": 0.9819,
      "step": 300
    },
    {
      "epoch": 0.10767574219350869,
      "grad_norm": 1.0390775292248948,
      "learning_rate": 1.968622807813479e-05,
      "loss": 0.9678,
      "step": 350
    },
    {
      "epoch": 0.12305799107829565,
      "grad_norm": 1.3974223368156584,
      "learning_rate": 1.955039646512041e-05,
      "loss": 0.9782,
      "step": 400
    },
    {
      "epoch": 0.1384402399630826,
      "grad_norm": 1.1107339192286112,
      "learning_rate": 1.9390851177632496e-05,
      "loss": 0.9602,
      "step": 450
    },
    {
      "epoch": 0.15382248884786956,
      "grad_norm": 1.0579993349228267,
      "learning_rate": 1.9207988367288956e-05,
      "loss": 0.9571,
      "step": 500
    },
    {
      "epoch": 0.1692047377326565,
      "grad_norm": 1.1144151233888806,
      "learning_rate": 1.9002262083214728e-05,
      "loss": 0.9669,
      "step": 550
    },
    {
      "epoch": 0.18458698661744347,
      "grad_norm": 1.187591950721483,
      "learning_rate": 1.8774183144635894e-05,
      "loss": 0.958,
      "step": 600
    },
    {
      "epoch": 0.19996923550223042,
      "grad_norm": 1.3767826488829,
      "learning_rate": 1.8524317872513434e-05,
      "loss": 0.9481,
      "step": 650
    },
    {
      "epoch": 0.21535148438701737,
      "grad_norm": 1.0853939205331895,
      "learning_rate": 1.8253286683365855e-05,
      "loss": 0.9434,
      "step": 700
    },
    {
      "epoch": 0.23073373327180433,
      "grad_norm": 1.0358359930096885,
      "learning_rate": 1.7961762548772413e-05,
      "loss": 0.9524,
      "step": 750
    },
    {
      "epoch": 0.2461159821565913,
      "grad_norm": 1.0885430781645802,
      "learning_rate": 1.7650469324381843e-05,
      "loss": 0.9443,
      "step": 800
    },
    {
      "epoch": 0.26149823104137826,
      "grad_norm": 1.139432989706956,
      "learning_rate": 1.732017995257575e-05,
      "loss": 0.9398,
      "step": 850
    },
    {
      "epoch": 0.2768804799261652,
      "grad_norm": 1.1616765089635157,
      "learning_rate": 1.697171454324943e-05,
      "loss": 0.9318,
      "step": 900
    },
    {
      "epoch": 0.29226272881095217,
      "grad_norm": 1.102418985915195,
      "learning_rate": 1.660593833747557e-05,
      "loss": 0.9348,
      "step": 950
    },
    {
      "epoch": 0.3076449776957391,
      "grad_norm": 1.0377834162761999,
      "learning_rate": 1.6223759559107028e-05,
      "loss": 0.9206,
      "step": 1000
    },
    {
      "epoch": 0.3230272265805261,
      "grad_norm": 1.103549274843009,
      "learning_rate": 1.5826127159653203e-05,
      "loss": 0.9353,
      "step": 1050
    },
    {
      "epoch": 0.338409475465313,
      "grad_norm": 1.0090406165137933,
      "learning_rate": 1.5414028462029454e-05,
      "loss": 0.9363,
      "step": 1100
    },
    {
      "epoch": 0.3537917243501,
      "grad_norm": 1.1329344585144494,
      "learning_rate": 1.4988486709030166e-05,
      "loss": 0.9347,
      "step": 1150
    },
    {
      "epoch": 0.36917397323488693,
      "grad_norm": 1.0307615452725163,
      "learning_rate": 1.4550558522612584e-05,
      "loss": 0.9231,
      "step": 1200
    },
    {
      "epoch": 0.3845562221196739,
      "grad_norm": 1.1133951229282608,
      "learning_rate": 1.410133128030009e-05,
      "loss": 0.9281,
      "step": 1250
    },
    {
      "epoch": 0.39993847100446084,
      "grad_norm": 1.0490189419573543,
      "learning_rate": 1.3641920415219224e-05,
      "loss": 0.9229,
      "step": 1300
    },
    {
      "epoch": 0.4153207198892478,
      "grad_norm": 1.0225887710538208,
      "learning_rate": 1.3173466646474575e-05,
      "loss": 0.9115,
      "step": 1350
    },
    {
      "epoch": 0.43070296877403474,
      "grad_norm": 1.0818630597333305,
      "learning_rate": 1.2697133146738452e-05,
      "loss": 0.9124,
      "step": 1400
    },
    {
      "epoch": 0.4460852176588217,
      "grad_norm": 1.1222485167556653,
      "learning_rate": 1.2214102654088248e-05,
      "loss": 0.9166,
      "step": 1450
    },
    {
      "epoch": 0.46146746654360865,
      "grad_norm": 0.9980409625749838,
      "learning_rate": 1.1725574535262863e-05,
      "loss": 0.9047,
      "step": 1500
    },
    {
      "epoch": 0.4768497154283956,
      "grad_norm": 1.0397039742249545,
      "learning_rate": 1.1232761807630032e-05,
      "loss": 0.9079,
      "step": 1550
    },
    {
      "epoch": 0.4922319643131826,
      "grad_norm": 1.0715968646982938,
      "learning_rate": 1.0736888127259132e-05,
      "loss": 0.9061,
      "step": 1600
    },
    {
      "epoch": 0.5076142131979695,
      "grad_norm": 1.0724763367894545,
      "learning_rate": 1.023918475057803e-05,
      "loss": 0.8947,
      "step": 1650
    },
    {
      "epoch": 0.5229964620827565,
      "grad_norm": 1.03882946183293,
      "learning_rate": 9.740887477158194e-06,
      "loss": 0.9053,
      "step": 1700
    },
    {
      "epoch": 0.5383787109675434,
      "grad_norm": 1.0204487030578802,
      "learning_rate": 9.24323358121919e-06,
      "loss": 0.9016,
      "step": 1750
    },
    {
      "epoch": 0.5537609598523304,
      "grad_norm": 1.0085542891249055,
      "learning_rate": 8.747458739471603e-06,
      "loss": 0.8988,
      "step": 1800
    },
    {
      "epoch": 0.5691432087371173,
      "grad_norm": 1.1293770623412562,
      "learning_rate": 8.254793962926579e-06,
      "loss": 0.8967,
      "step": 1850
    },
    {
      "epoch": 0.5845254576219043,
      "grad_norm": 1.1860769177336596,
      "learning_rate": 7.76646254029032e-06,
      "loss": 0.8967,
      "step": 1900
    },
    {
      "epoch": 0.5999077065066912,
      "grad_norm": 1.062132771926321,
      "learning_rate": 7.283677000533106e-06,
      "loss": 0.8936,
      "step": 1950
    },
    {
      "epoch": 0.6152899553914782,
      "grad_norm": 1.1679358844534626,
      "learning_rate": 6.807636102174764e-06,
      "loss": 0.8928,
      "step": 2000
    },
    {
      "epoch": 0.6306722042762651,
      "grad_norm": 1.0610336394985995,
      "learning_rate": 6.339521856762254e-06,
      "loss": 0.8845,
      "step": 2050
    },
    {
      "epoch": 0.6460544531610521,
      "grad_norm": 1.084745782348564,
      "learning_rate": 5.880496593930021e-06,
      "loss": 0.8869,
      "step": 2100
    },
    {
      "epoch": 0.6614367020458392,
      "grad_norm": 0.9948936563241075,
      "learning_rate": 5.431700075330654e-06,
      "loss": 0.8817,
      "step": 2150
    },
    {
      "epoch": 0.676818950930626,
      "grad_norm": 1.0955833872820746,
      "learning_rate": 4.9942466646019e-06,
      "loss": 0.8857,
      "step": 2200
    },
    {
      "epoch": 0.6922011998154131,
      "grad_norm": 1.0328397346172227,
      "learning_rate": 4.569222560397126e-06,
      "loss": 0.8842,
      "step": 2250
    },
    {
      "epoch": 0.7075834487002,
      "grad_norm": 1.0439140807601641,
      "learning_rate": 4.157683099349444e-06,
      "loss": 0.8797,
      "step": 2300
    },
    {
      "epoch": 0.722965697584987,
      "grad_norm": 1.0659188075287298,
      "learning_rate": 3.760650135666476e-06,
      "loss": 0.8765,
      "step": 2350
    },
    {
      "epoch": 0.7383479464697739,
      "grad_norm": 1.0691950735737183,
      "learning_rate": 3.3791095038620515e-06,
      "loss": 0.8715,
      "step": 2400
    },
    {
      "epoch": 0.7537301953545609,
      "grad_norm": 1.0882575411106976,
      "learning_rate": 3.0140085709249666e-06,
      "loss": 0.8754,
      "step": 2450
    },
    {
      "epoch": 0.7691124442393478,
      "grad_norm": 1.1112353348166593,
      "learning_rate": 2.6662538840027984e-06,
      "loss": 0.873,
      "step": 2500
    },
    {
      "epoch": 0.7844946931241348,
      "grad_norm": 1.1312531360561675,
      "learning_rate": 2.3367089194415226e-06,
      "loss": 0.8776,
      "step": 2550
    },
    {
      "epoch": 0.7998769420089217,
      "grad_norm": 1.118822941453432,
      "learning_rate": 2.026191938770128e-06,
      "loss": 0.8769,
      "step": 2600
    },
    {
      "epoch": 0.8152591908937087,
      "grad_norm": 1.0617562102838363,
      "learning_rate": 1.7354739569538081e-06,
      "loss": 0.8611,
      "step": 2650
    },
    {
      "epoch": 0.8306414397784956,
      "grad_norm": 1.1190891753348724,
      "learning_rate": 1.4652768279605734e-06,
      "loss": 0.8778,
      "step": 2700
    },
    {
      "epoch": 0.8460236886632826,
      "grad_norm": 1.1025665252752193,
      "learning_rate": 1.2162714523948104e-06,
      "loss": 0.8582,
      "step": 2750
    },
    {
      "epoch": 0.8614059375480695,
      "grad_norm": 1.0621744666965152,
      "learning_rate": 9.89076111648284e-07,
      "loss": 0.8643,
      "step": 2800
    },
    {
      "epoch": 0.8767881864328565,
      "grad_norm": 0.9763313630276178,
      "learning_rate": 7.842549327048366e-07,
      "loss": 0.8667,
      "step": 2850
    },
    {
      "epoch": 0.8921704353176434,
      "grad_norm": 1.0393762329975094,
      "learning_rate": 6.023164874107202e-07,
      "loss": 0.8677,
      "step": 2900
    },
    {
      "epoch": 0.9075526842024304,
      "grad_norm": 1.1026189621389317,
      "learning_rate": 4.437125296885847e-07,
      "loss": 0.8649,
      "step": 2950
    },
    {
      "epoch": 0.9229349330872173,
      "grad_norm": 1.073850367410434,
      "learning_rate": 3.088368738305747e-07,
      "loss": 0.8771,
      "step": 3000
    },
    {
      "epoch": 0.9383171819720043,
      "grad_norm": 1.0327578416174346,
      "learning_rate": 1.9802441665580208e-07,
      "loss": 0.8678,
      "step": 3050
    },
    {
      "epoch": 0.9536994308567912,
      "grad_norm": 1.035131432577654,
      "learning_rate": 1.1155030596013638e-07,
      "loss": 0.8662,
      "step": 3100
    },
    {
      "epoch": 0.9690816797415782,
      "grad_norm": 1.0361099744000193,
      "learning_rate": 4.962925732307278e-08,
      "loss": 0.8588,
      "step": 3150
    },
    {
      "epoch": 0.9844639286263652,
      "grad_norm": 1.087919608094763,
      "learning_rate": 1.2415020968049585e-08,
      "loss": 0.8733,
      "step": 3200
    },
    {
      "epoch": 0.9998461775111521,
      "grad_norm": 1.0473814351391442,
      "learning_rate": 0.0,
      "loss": 0.8751,
      "step": 3250
    },
    {
      "epoch": 0.9998461775111521,
      "step": 3250,
      "total_flos": 1.7053270040117248e+16,
      "train_loss": 0.9156783118614784,
      "train_runtime": 46310.2792,
      "train_samples_per_second": 0.562,
      "train_steps_per_second": 0.07
    }
  ],
  "logging_steps": 50,
  "max_steps": 3250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 1.7053270040117248e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}