{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 231,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030303030303030304,
      "grad_norm": 6.1662449684862635,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.9939,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 5.926832666052827,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.9561,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 5.805313217907784,
      "learning_rate": 2.5e-06,
      "loss": 0.938,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.55509698585396,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.961,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.486440189273938,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9264,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.408895022676021,
      "learning_rate": 5e-06,
      "loss": 0.9174,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.121455001597856,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9116,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.8019241428078905,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8283,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.6187885027183055,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9159,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.0643169752929555,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8038,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.783031909713882,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8098,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.438760320298039,
      "learning_rate": 1e-05,
      "loss": 0.821,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.628773406330169,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8084,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 2.101437067106728,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.7885,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 2.561542710313462,
      "learning_rate": 1.25e-05,
      "loss": 0.7777,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 2.2785501289872334,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7545,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.968640882509224,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7642,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.6503703412450261,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7628,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.4610036540858915,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7217,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.594258697669721,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7391,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.2982657030797504,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7558,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.0275238545989651,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7515,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.2175982602177642,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7078,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.978638646811626,
      "learning_rate": 2e-05,
      "loss": 0.6884,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 1.683011429828035,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.702,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 1.0931954590818262,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.6852,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.8715902949932376,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.7329,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.990814722476233,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7158,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.6789750119065185,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.6423,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.7651415539681655,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.673,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.8222544887382023,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6799,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.751118418600838,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.7229,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7888195361031916,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6565,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.7632501204658517,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6436,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.6086752426127852,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6151,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.8066193590823785,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6314,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 1.1030417035982745,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6214,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.6965305674287905,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6208,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.661397189829578,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.581,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6336256906697842,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6057,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.6465036633378368,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.5725,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.5712468871654879,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6014,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.7172775233709918,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.5815,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.6107948933451833,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.5866,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.7023917048774757,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.5991,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.5671287437450974,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.5615,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.5765875377611958,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6012,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6274147021363536,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6328,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.5368149245236333,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.5613,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.6337972161485507,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.5832,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6171486003526793,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.5838,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.5603325633045169,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.5339,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.5389820315910553,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.5773,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.5429018155097364,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.5466,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5724521489491811,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.5509,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.620067318704003,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.5793,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.6136157221307668,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.5671,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.6450739638427054,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.5506,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.6083411255518985,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.5837,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5897387017387672,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.5425,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.5663018675161458,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.5429,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.6428642915447594,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.5603,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.5712158761459469,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.5785,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.5551909227299838,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.5344,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5537830037919519,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.5489,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.49418511392667863,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.5436,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7298615347623922,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.4442,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5705959817676246,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.4398,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.9452028606450429,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.482,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.6414340153162598,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.4454,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.6095992156500246,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4068,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.633492423374324,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.4467,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.5766748185628832,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.4379,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.6493188235932105,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.4285,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5875271817089605,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.4248,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.6836831324261866,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4556,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5994128423825578,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4522,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.6154956352319594,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.4199,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.5877709032730231,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.4124,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.5892784501997475,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4331,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.6242910018158795,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4319,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.6228180471801781,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4165,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.5402621814858859,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4055,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5595621903761638,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4422,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.6413413305825129,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4294,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5415817515005147,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.4149,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.5544319394198701,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.4247,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.6144118839198031,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.4365,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.5446802572812841,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4383,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.5689161159445739,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4379,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.5492503315743634,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4418,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5392114628318508,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.4252,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.5478969216527346,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4251,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.48484475366752094,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4216,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5377139356990187,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.4056,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.46532979406304303,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4347,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5292797105904653,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4152,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.5152230681655751,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4037,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5075010344446873,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.406,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.7004565074576138,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.324,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.5594911072199484,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.349,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.7928415574075284,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3383,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.8766946961105866,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3126,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.6189076333597993,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3193,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.6607874395540838,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.3028,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.6301967705525754,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.2999,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.598461288995727,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3099,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.5451454783797907,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3221,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.5784296850807539,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.2909,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.5312856048236173,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.2906,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.5030414595167957,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.2975,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.5221182249057237,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.2992,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.5724656407639066,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.2977,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.518540802412599,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.3121,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.5223811212675639,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.2908,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.5974743640725056,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.3186,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5036916149914951,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.2943,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.508386422586809,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.3015,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.584351819466433,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.2856,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.4872141748235865,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.2896,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.5408074680307653,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3172,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.5105203754242232,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.3007,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.5445370598010902,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.3172,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.4895255068077135,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.2998,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.5786414162994448,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.3042,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.5001034460131794,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.3009,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.5119024212807315,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.2884,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.4628325795263138,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.2932,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.533493962833977,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3013,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.5274924073071741,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.2918,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.46727460788527936,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3052,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4822524135582183,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.2834,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.661746107519592,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.2278,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.6397085148245383,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.2271,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.4908022751367689,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.2043,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.8334404424838376,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.2056,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.837920195325812,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.1999,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.5506264180485309,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.2111,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.572426050315769,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2014,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.6621673962587367,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.1869,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.5128932478999959,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.2076,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.5946416969442491,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.221,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.5215721795171269,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2033,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.5683997512163645,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.2071,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.5866881210425684,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.204,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.562003979966829,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.2138,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.49985419766069167,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.1877,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.5014148363668974,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.183,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.5038545610493653,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.2034,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.49647239050566155,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.195,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.4584385620351258,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.2027,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.47546922157242544,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.2083,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.49689345224555154,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.2043,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.46878208729115467,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.1953,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.4698553684319248,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.2068,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.44468673922047613,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.2101,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.44993912665904695,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.1864,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.4665441935350063,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.2085,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.4580176973643987,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.204,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.43758849275205464,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.1988,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.45654609148343556,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.206,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.43753429677640177,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2146,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.44647965540707885,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.2103,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.4691773289667456,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.2126,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4416591731641226,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.2061,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.643025128316058,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.1596,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.5619564574227774,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.1539,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.4136786730819999,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1339,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.44959067786516094,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.1464,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.585073867487514,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.1501,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.7086760613447944,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.1554,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.5109939737394803,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.1274,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.44830266505851224,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.1387,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.4078668406206426,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1375,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.41607544830431914,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.1388,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.47450388401997157,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.1506,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.4910892637500035,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.1488,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.4276902125306567,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.1554,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.4201862611484017,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.1317,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.41780450402606534,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.1312,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.4195823719446442,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1586,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.44835308557346876,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.1259,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.426595882824439,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1239,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.42662528620444984,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.12,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.40749760545660685,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.1283,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.41199813491947007,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.1356,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.42019815995415866,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1409,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.41208887217912016,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.1528,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.3990765743451002,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.141,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.38171306063601207,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.1463,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.40413940289376943,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.1492,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.38922281904737593,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.1485,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.4247437386167156,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.1296,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.40240779274143895,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.1404,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.3860929921312101,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1322,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.3934361089659041,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.1356,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.3834350604462262,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.1176,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3935084222726297,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.1237,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.4814095419788397,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1119,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.41053514554468046,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1144,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.47115063572925003,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.1179,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.4554455561826325,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1149,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.35049204479060087,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1201,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.3678326252161325,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1067,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.35538764885156826,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.1131,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.41808922461912157,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1104,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.3879261022919951,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1019,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.4398724480741241,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1195,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.4025673824605803,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1181,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.4694756044822564,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1367,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.3732292067600177,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.1128,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.4035079173294576,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.1135,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.3883097344703561,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.1125,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.37718913958054917,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.1026,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.3579556256555221,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1224,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.3481175246578372,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1117,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.3333995616124938,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.1043,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.3610788904119615,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.1126,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.3728688773006187,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.1227,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.33934861266082333,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1108,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.33995427566719505,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.1117,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.35251038578696026,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1156,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.3382323920986913,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1117,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.3403288442239302,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1117,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.36123560169204555,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1126,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.3688237953252601,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.1048,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.3399828127266381,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.1129,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.3521762533636998,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.1113,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.35513065923202436,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.1137,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.35544645398521824,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1105,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.35675586952978544,
      "learning_rate": 0.0,
      "loss": 0.1124,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 2.93440642585985e+17,
      "train_loss": 0.36575089378919434,
      "train_runtime": 6464.1361,
      "train_samples_per_second": 3.422,
      "train_steps_per_second": 0.036
    }
  ],
  "logging_steps": 1,
  "max_steps": 231,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 2.93440642585985e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}