{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12734195771253845,
  "eval_steps": 2,
  "global_step": 18720,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 3.921875,
      "learning_rate": 4.896848787123102e-07,
      "loss": 1.9119,
      "step": 144
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.5625,
      "learning_rate": 9.793697574246203e-07,
      "loss": 1.9115,
      "step": 288
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.96875,
      "learning_rate": 1.4690546361369305e-06,
      "loss": 1.9003,
      "step": 432
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.359375,
      "learning_rate": 1.9587395148492407e-06,
      "loss": 1.9034,
      "step": 576
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.390625,
      "learning_rate": 2.4484243935615507e-06,
      "loss": 1.9145,
      "step": 720
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.5,
      "learning_rate": 2.938109272273861e-06,
      "loss": 1.8969,
      "step": 864
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.234375,
      "learning_rate": 3.427794150986171e-06,
      "loss": 1.8644,
      "step": 1008
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.328125,
      "learning_rate": 3.917479029698481e-06,
      "loss": 1.8823,
      "step": 1152
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.421875,
      "learning_rate": 4.4071639084107905e-06,
      "loss": 1.8635,
      "step": 1296
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.15625,
      "learning_rate": 4.896848787123101e-06,
      "loss": 1.8545,
      "step": 1440
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.890625,
      "learning_rate": 5.386533665835411e-06,
      "loss": 1.871,
      "step": 1584
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.046875,
      "learning_rate": 5.876218544547722e-06,
      "loss": 1.8609,
      "step": 1728
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.40625,
      "learning_rate": 6.365903423260031e-06,
      "loss": 1.8788,
      "step": 1872
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.140625,
      "learning_rate": 6.855588301972342e-06,
      "loss": 1.8672,
      "step": 2016
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.03125,
      "learning_rate": 7.345273180684652e-06,
      "loss": 1.8723,
      "step": 2160
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.078125,
      "learning_rate": 7.834958059396963e-06,
      "loss": 1.879,
      "step": 2304
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.359375,
      "learning_rate": 8.324642938109274e-06,
      "loss": 1.8622,
      "step": 2448
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.03125,
      "learning_rate": 8.814327816821581e-06,
      "loss": 1.8533,
      "step": 2592
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.125,
      "learning_rate": 9.304012695533892e-06,
      "loss": 1.8503,
      "step": 2736
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.1875,
      "learning_rate": 9.793697574246203e-06,
      "loss": 1.8314,
      "step": 2880
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.96875,
      "learning_rate": 1.0283382452958513e-05,
      "loss": 1.835,
      "step": 3024
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.546875,
      "learning_rate": 1.0773067331670823e-05,
      "loss": 1.8526,
      "step": 3168
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.96875,
      "learning_rate": 1.1262752210383133e-05,
      "loss": 1.8434,
      "step": 3312
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.03125,
      "learning_rate": 1.1752437089095444e-05,
      "loss": 1.8381,
      "step": 3456
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.984375,
      "learning_rate": 1.2242121967807753e-05,
      "loss": 1.848,
      "step": 3600
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.828125,
      "learning_rate": 1.2731806846520062e-05,
      "loss": 1.8476,
      "step": 3744
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.984375,
      "learning_rate": 1.3221491725232373e-05,
      "loss": 1.8307,
      "step": 3888
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.953125,
      "learning_rate": 1.3711176603944684e-05,
      "loss": 1.8328,
      "step": 4032
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.875,
      "learning_rate": 1.4200861482656995e-05,
      "loss": 1.833,
      "step": 4176
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.765625,
      "learning_rate": 1.4690546361369304e-05,
      "loss": 1.8559,
      "step": 4320
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.765625,
      "learning_rate": 1.4999994886948122e-05,
      "loss": 1.8258,
      "step": 4464
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.765625,
      "learning_rate": 1.499992935844023e-05,
      "loss": 1.8256,
      "step": 4608
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.15625,
      "learning_rate": 1.4999788341738499e-05,
      "loss": 1.8315,
      "step": 4752
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.796875,
      "learning_rate": 1.4999571838262296e-05,
      "loss": 1.8583,
      "step": 4896
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.71875,
      "learning_rate": 1.4999279850190762e-05,
      "loss": 1.8287,
      "step": 5040
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.953125,
      "learning_rate": 1.4998912380462815e-05,
      "loss": 1.8134,
      "step": 5184
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.59375,
      "learning_rate": 1.4998469432777108e-05,
      "loss": 1.8111,
      "step": 5328
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.015625,
      "learning_rate": 1.4997951011591987e-05,
      "loss": 1.8199,
      "step": 5472
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.6875,
      "learning_rate": 1.4997357122125465e-05,
      "loss": 1.8188,
      "step": 5616
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.625,
      "learning_rate": 1.4996687770355145e-05,
      "loss": 1.8023,
      "step": 5760
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.015625,
      "learning_rate": 1.4995942963018183e-05,
      "loss": 1.8131,
      "step": 5904
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.71875,
      "learning_rate": 1.49951227076112e-05,
      "loss": 1.8056,
      "step": 6048
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.578125,
      "learning_rate": 1.4994227012390224e-05,
      "loss": 1.8034,
      "step": 6192
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.8125,
      "learning_rate": 1.4993255886370593e-05,
      "loss": 1.8221,
      "step": 6336
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.0,
      "learning_rate": 1.4992209339326872e-05,
      "loss": 1.7937,
      "step": 6480
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.453125,
      "learning_rate": 1.4991087381792756e-05,
      "loss": 1.7936,
      "step": 6624
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.59375,
      "learning_rate": 1.4989890025060953e-05,
      "loss": 1.804,
      "step": 6768
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.671875,
      "learning_rate": 1.4988617281183087e-05,
      "loss": 1.779,
      "step": 6912
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.671875,
      "learning_rate": 1.4987269162969559e-05,
      "loss": 1.7907,
      "step": 7056
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.65625,
      "learning_rate": 1.4985845683989435e-05,
      "loss": 1.7945,
      "step": 7200
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6875,
      "learning_rate": 1.4984346858570297e-05,
      "loss": 1.7887,
      "step": 7344
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.9375,
      "learning_rate": 1.4982772701798105e-05,
      "loss": 1.7657,
      "step": 7488
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.703125,
      "learning_rate": 1.4981123229517039e-05,
      "loss": 1.7914,
      "step": 7632
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.546875,
      "learning_rate": 1.4979398458329347e-05,
      "loss": 1.7716,
      "step": 7776
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.46875,
      "learning_rate": 1.4977598405595177e-05,
      "loss": 1.7979,
      "step": 7920
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.59375,
      "learning_rate": 1.4975723089432394e-05,
      "loss": 1.7751,
      "step": 8064
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.5625,
      "learning_rate": 1.4973772528716404e-05,
      "loss": 1.7868,
      "step": 8208
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.78125,
      "learning_rate": 1.4971746743079972e-05,
      "loss": 1.7974,
      "step": 8352
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.59375,
      "learning_rate": 1.4969645752913006e-05,
      "loss": 1.7895,
      "step": 8496
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.578125,
      "learning_rate": 1.4967469579362364e-05,
      "loss": 1.7919,
      "step": 8640
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.484375,
      "learning_rate": 1.4965218244331647e-05,
      "loss": 1.7856,
      "step": 8784
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.578125,
      "learning_rate": 1.4962891770480958e-05,
      "loss": 1.7777,
      "step": 8928
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.5,
      "learning_rate": 1.4960490181226693e-05,
      "loss": 1.7796,
      "step": 9072
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.421875,
      "learning_rate": 1.49580135007413e-05,
      "loss": 1.7769,
      "step": 9216
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.59375,
      "learning_rate": 1.495546175395303e-05,
      "loss": 1.7843,
      "step": 9360
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.640625,
      "learning_rate": 1.495283496654569e-05,
      "loss": 1.7763,
      "step": 9504
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.765625,
      "learning_rate": 1.4950133164958392e-05,
      "loss": 1.7709,
      "step": 9648
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.4375,
      "learning_rate": 1.4947356376385275e-05,
      "loss": 1.781,
      "step": 9792
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.5625,
      "learning_rate": 1.4944504628775233e-05,
      "loss": 1.7831,
      "step": 9936
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.71875,
      "learning_rate": 1.4941577950831642e-05,
      "loss": 1.7835,
      "step": 10080
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.640625,
      "learning_rate": 1.4938576372012062e-05,
      "loss": 1.7648,
      "step": 10224
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.703125,
      "learning_rate": 1.4935499922527949e-05,
      "loss": 1.774,
      "step": 10368
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.640625,
      "learning_rate": 1.4932348633344338e-05,
      "loss": 1.7989,
      "step": 10512
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.5,
      "learning_rate": 1.492912253617955e-05,
      "loss": 1.7624,
      "step": 10656
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.4375,
      "learning_rate": 1.4925821663504856e-05,
      "loss": 1.7584,
      "step": 10800
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.640625,
      "learning_rate": 1.4922446048544157e-05,
      "loss": 1.7719,
      "step": 10944
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.703125,
      "learning_rate": 1.4918995725273649e-05,
      "loss": 1.7416,
      "step": 11088
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.59375,
      "learning_rate": 1.4915470728421481e-05,
      "loss": 1.7587,
      "step": 11232
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.609375,
      "learning_rate": 1.4911871093467414e-05,
      "loss": 1.7562,
      "step": 11376
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.90625,
      "learning_rate": 1.4908196856642441e-05,
      "loss": 1.7579,
      "step": 11520
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.625,
      "learning_rate": 1.4904448054928444e-05,
      "loss": 1.7734,
      "step": 11664
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.734375,
      "learning_rate": 1.4900624726057816e-05,
      "loss": 1.7781,
      "step": 11808
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.640625,
      "learning_rate": 1.489672690851308e-05,
      "loss": 1.7576,
      "step": 11952
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.65625,
      "learning_rate": 1.4892754641526498e-05,
      "loss": 1.7561,
      "step": 12096
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.375,
      "learning_rate": 1.4888707965079682e-05,
      "loss": 1.7602,
      "step": 12240
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.5625,
      "learning_rate": 1.4884586919903186e-05,
      "loss": 1.7676,
      "step": 12384
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.515625,
      "learning_rate": 1.4880391547476106e-05,
      "loss": 1.7546,
      "step": 12528
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.234375,
      "learning_rate": 1.4876121890025647e-05,
      "loss": 1.7484,
      "step": 12672
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.765625,
      "learning_rate": 1.4871777990526713e-05,
      "loss": 1.7425,
      "step": 12816
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.703125,
      "learning_rate": 1.4867359892701466e-05,
      "loss": 1.7495,
      "step": 12960
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.578125,
      "learning_rate": 1.4862867641018887e-05,
      "loss": 1.7478,
      "step": 13104
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.453125,
      "learning_rate": 1.4858301280694332e-05,
      "loss": 1.7648,
      "step": 13248
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.546875,
      "learning_rate": 1.4853660857689073e-05,
      "loss": 1.7779,
      "step": 13392
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.0,
      "learning_rate": 1.4848946418709838e-05,
      "loss": 1.746,
      "step": 13536
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.1875,
      "learning_rate": 1.4844158011208336e-05,
      "loss": 1.7482,
      "step": 13680
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.4375,
      "learning_rate": 1.4839295683380785e-05,
      "loss": 1.7614,
      "step": 13824
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.625,
      "learning_rate": 1.4834359484167429e-05,
      "loss": 1.7394,
      "step": 13968
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.65625,
      "learning_rate": 1.4829349463252034e-05,
      "loss": 1.7529,
      "step": 14112
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.578125,
      "learning_rate": 1.48242656710614e-05,
      "loss": 1.741,
      "step": 14256
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.484375,
      "learning_rate": 1.4819108158764847e-05,
      "loss": 1.7496,
      "step": 14400
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.40625,
      "learning_rate": 1.4813876978273708e-05,
      "loss": 1.7537,
      "step": 14544
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.46875,
      "learning_rate": 1.4808572182240786e-05,
      "loss": 1.7435,
      "step": 14688
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.703125,
      "learning_rate": 1.4803193824059852e-05,
      "loss": 1.7665,
      "step": 14832
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.546875,
      "learning_rate": 1.4797741957865087e-05,
      "loss": 1.7547,
      "step": 14976
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.59375,
      "learning_rate": 1.4792216638530545e-05,
      "loss": 1.7356,
      "step": 15120
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.703125,
      "learning_rate": 1.4786617921669603e-05,
      "loss": 1.7397,
      "step": 15264
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.5625,
      "learning_rate": 1.478094586363439e-05,
      "loss": 1.7239,
      "step": 15408
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.96875,
      "learning_rate": 1.4775200521515237e-05,
      "loss": 1.7428,
      "step": 15552
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.765625,
      "learning_rate": 1.4769381953140084e-05,
      "loss": 1.745,
      "step": 15696
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.453125,
      "learning_rate": 1.4763490217073917e-05,
      "loss": 1.7382,
      "step": 15840
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.5,
      "learning_rate": 1.475752537261816e-05,
      "loss": 1.7524,
      "step": 15984
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.5,
      "learning_rate": 1.475148747981009e-05,
      "loss": 1.7288,
      "step": 16128
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.484375,
      "learning_rate": 1.4745376599422232e-05,
      "loss": 1.7336,
      "step": 16272
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.78125,
      "learning_rate": 1.4739192792961736e-05,
      "loss": 1.7313,
      "step": 16416
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.53125,
      "learning_rate": 1.4732936122669777e-05,
      "loss": 1.7537,
      "step": 16560
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.671875,
      "learning_rate": 1.4726606651520911e-05,
      "loss": 1.7428,
      "step": 16704
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.90625,
      "learning_rate": 1.4720204443222452e-05,
      "loss": 1.7299,
      "step": 16848
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.4375,
      "learning_rate": 1.4713729562213825e-05,
      "loss": 1.7276,
      "step": 16992
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.5,
      "learning_rate": 1.4707182073665921e-05,
      "loss": 1.7312,
      "step": 17136
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.71875,
      "learning_rate": 1.470056204348044e-05,
      "loss": 1.7477,
      "step": 17280
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.453125,
      "learning_rate": 1.4693869538289221e-05,
      "loss": 1.7296,
      "step": 17424
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.890625,
      "learning_rate": 1.4687104625453587e-05,
      "loss": 1.7353,
      "step": 17568
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.359375,
      "learning_rate": 1.4680267373063651e-05,
      "loss": 1.7265,
      "step": 17712
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.421875,
      "learning_rate": 1.467335784993764e-05,
      "loss": 1.7406,
      "step": 17856
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.828125,
      "learning_rate": 1.4666376125621198e-05,
      "loss": 1.7313,
      "step": 18000
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.46875,
      "learning_rate": 1.4659322270386687e-05,
      "loss": 1.7614,
      "step": 18144
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.609375,
      "learning_rate": 1.4652196355232481e-05,
      "loss": 1.7313,
      "step": 18288
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.5,
      "learning_rate": 1.4644998451882254e-05,
      "loss": 1.7307,
      "step": 18432
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.546875,
      "learning_rate": 1.463772863278425e-05,
      "loss": 1.7291,
      "step": 18576
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.359375,
      "learning_rate": 1.4630386971110556e-05,
      "loss": 1.7315,
      "step": 18720
    }
  ],
  "logging_steps": 144,
  "max_steps": 147005,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1440,
  "total_flos": 7.842747163103724e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}