{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9911678115799804,
  "eval_steps": 500,
  "global_step": 101,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009813542688910697,
      "grad_norm": 2.8672487474187687,
      "learning_rate": 2.5e-06,
      "loss": 2.1544,
      "step": 1
    },
    {
      "epoch": 0.019627085377821395,
      "grad_norm": 2.754588790109863,
      "learning_rate": 5e-06,
      "loss": 2.142,
      "step": 2
    },
    {
      "epoch": 0.029440628066732092,
      "grad_norm": 2.590669275853557,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.9074,
      "step": 3
    },
    {
      "epoch": 0.03925417075564279,
      "grad_norm": 1.912791537410904,
      "learning_rate": 1e-05,
      "loss": 1.733,
      "step": 4
    },
    {
      "epoch": 0.04906771344455348,
      "grad_norm": 1.2960446152840717,
      "learning_rate": 9.997377845227577e-06,
      "loss": 1.4053,
      "step": 5
    },
    {
      "epoch": 0.058881256133464184,
      "grad_norm": 1.5094470155263513,
      "learning_rate": 9.98951413118856e-06,
      "loss": 1.24,
      "step": 6
    },
    {
      "epoch": 0.06869479882237488,
      "grad_norm": 1.4729523990902567,
      "learning_rate": 9.97641710583307e-06,
      "loss": 1.1846,
      "step": 7
    },
    {
      "epoch": 0.07850834151128558,
      "grad_norm": 0.6192450033938127,
      "learning_rate": 9.958100506132127e-06,
      "loss": 1.2066,
      "step": 8
    },
    {
      "epoch": 0.08832188420019627,
      "grad_norm": 0.5248188615161075,
      "learning_rate": 9.934583543669454e-06,
      "loss": 0.9077,
      "step": 9
    },
    {
      "epoch": 0.09813542688910697,
      "grad_norm": 0.4652794249979759,
      "learning_rate": 9.905890884491196e-06,
      "loss": 0.8559,
      "step": 10
    },
    {
      "epoch": 0.10794896957801767,
      "grad_norm": 0.524548716007966,
      "learning_rate": 9.872052623234632e-06,
      "loss": 0.9704,
      "step": 11
    },
    {
      "epoch": 0.11776251226692837,
      "grad_norm": 0.5314194335904315,
      "learning_rate": 9.833104251563058e-06,
      "loss": 0.9223,
      "step": 12
    },
    {
      "epoch": 0.12757605495583907,
      "grad_norm": 0.3047635309199024,
      "learning_rate": 9.789086620939936e-06,
      "loss": 0.9302,
      "step": 13
    },
    {
      "epoch": 0.13738959764474976,
      "grad_norm": 0.3819529820066455,
      "learning_rate": 9.740045899781353e-06,
      "loss": 0.8211,
      "step": 14
    },
    {
      "epoch": 0.14720314033366044,
      "grad_norm": 0.39909585000271475,
      "learning_rate": 9.68603352503172e-06,
      "loss": 0.8913,
      "step": 15
    },
    {
      "epoch": 0.15701668302257116,
      "grad_norm": 0.3479302056108582,
      "learning_rate": 9.627106148213521e-06,
      "loss": 0.9723,
      "step": 16
    },
    {
      "epoch": 0.16683022571148184,
      "grad_norm": 0.3766959942077643,
      "learning_rate": 9.563325576007702e-06,
      "loss": 0.8522,
      "step": 17
    },
    {
      "epoch": 0.17664376840039253,
      "grad_norm": 0.2857198795401394,
      "learning_rate": 9.494758705426978e-06,
      "loss": 0.9633,
      "step": 18
    },
    {
      "epoch": 0.18645731108930325,
      "grad_norm": 0.2993174163090425,
      "learning_rate": 9.421477453650118e-06,
      "loss": 0.8605,
      "step": 19
    },
    {
      "epoch": 0.19627085377821393,
      "grad_norm": 0.3107401663608036,
      "learning_rate": 9.343558682590757e-06,
      "loss": 0.8296,
      "step": 20
    },
    {
      "epoch": 0.20608439646712462,
      "grad_norm": 0.3270368688125023,
      "learning_rate": 9.261084118279846e-06,
      "loss": 0.8632,
      "step": 21
    },
    {
      "epoch": 0.21589793915603533,
      "grad_norm": 0.32890526306764684,
      "learning_rate": 9.174140265146356e-06,
      "loss": 0.954,
      "step": 22
    },
    {
      "epoch": 0.22571148184494602,
      "grad_norm": 0.2208062222434778,
      "learning_rate": 9.082818315286054e-06,
      "loss": 0.7973,
      "step": 23
    },
    {
      "epoch": 0.23552502453385674,
      "grad_norm": 0.23994712657664863,
      "learning_rate": 8.987214052813605e-06,
      "loss": 0.8003,
      "step": 24
    },
    {
      "epoch": 0.24533856722276742,
      "grad_norm": 0.3579410121713744,
      "learning_rate": 8.887427753398249e-06,
      "loss": 0.9163,
      "step": 25
    },
    {
      "epoch": 0.25515210991167814,
      "grad_norm": 0.25858791865785286,
      "learning_rate": 8.783564079088478e-06,
      "loss": 0.9203,
      "step": 26
    },
    {
      "epoch": 0.2649656526005888,
      "grad_norm": 0.2221965052804783,
      "learning_rate": 8.675731968536004e-06,
      "loss": 0.7986,
      "step": 27
    },
    {
      "epoch": 0.2747791952894995,
      "grad_norm": 0.25869875995203495,
      "learning_rate": 8.564044522734147e-06,
      "loss": 0.8584,
      "step": 28
    },
    {
      "epoch": 0.2845927379784102,
      "grad_norm": 0.25639549077125273,
      "learning_rate": 8.448618886390523e-06,
      "loss": 0.9153,
      "step": 29
    },
    {
      "epoch": 0.2944062806673209,
      "grad_norm": 0.28969420637722326,
      "learning_rate": 8.329576125058406e-06,
      "loss": 1.0017,
      "step": 30
    },
    {
      "epoch": 0.3042198233562316,
      "grad_norm": 0.27142079291709187,
      "learning_rate": 8.207041098155701e-06,
      "loss": 0.8834,
      "step": 31
    },
    {
      "epoch": 0.3140333660451423,
      "grad_norm": 0.24855994809239515,
      "learning_rate": 8.081142328004638e-06,
      "loss": 0.8804,
      "step": 32
    },
    {
      "epoch": 0.323846908734053,
      "grad_norm": 0.2932484533076321,
      "learning_rate": 7.952011865029614e-06,
      "loss": 0.8055,
      "step": 33
    },
    {
      "epoch": 0.3336604514229637,
      "grad_norm": 0.2899949936434053,
      "learning_rate": 7.819785149254534e-06,
      "loss": 0.7479,
      "step": 34
    },
    {
      "epoch": 0.3434739941118744,
      "grad_norm": 0.23167648356664933,
      "learning_rate": 7.68460086824492e-06,
      "loss": 0.8886,
      "step": 35
    },
    {
      "epoch": 0.35328753680078506,
      "grad_norm": 0.2640358818005348,
      "learning_rate": 7.546600811643816e-06,
      "loss": 0.9551,
      "step": 36
    },
    {
      "epoch": 0.3631010794896958,
      "grad_norm": 0.2595883854771292,
      "learning_rate": 7.405929722454026e-06,
      "loss": 0.9107,
      "step": 37
    },
    {
      "epoch": 0.3729146221786065,
      "grad_norm": 0.2471525297780812,
      "learning_rate": 7.262735145222696e-06,
      "loss": 0.8207,
      "step": 38
    },
    {
      "epoch": 0.38272816486751715,
      "grad_norm": 0.2476329062250312,
      "learning_rate": 7.117167271287453e-06,
      "loss": 0.854,
      "step": 39
    },
    {
      "epoch": 0.39254170755642787,
      "grad_norm": 0.26094540269164535,
      "learning_rate": 6.969378781246436e-06,
      "loss": 0.876,
      "step": 40
    },
    {
      "epoch": 0.4023552502453386,
      "grad_norm": 0.26106316938836766,
      "learning_rate": 6.819524684817439e-06,
      "loss": 0.8629,
      "step": 41
    },
    {
      "epoch": 0.41216879293424924,
      "grad_norm": 0.23060264618437654,
      "learning_rate": 6.667762158254104e-06,
      "loss": 0.8538,
      "step": 42
    },
    {
      "epoch": 0.42198233562315995,
      "grad_norm": 0.24185186221890845,
      "learning_rate": 6.514250379489754e-06,
      "loss": 0.9191,
      "step": 43
    },
    {
      "epoch": 0.43179587831207067,
      "grad_norm": 0.2393529024626251,
      "learning_rate": 6.3591503611817155e-06,
      "loss": 0.8808,
      "step": 44
    },
    {
      "epoch": 0.44160942100098133,
      "grad_norm": 0.2605127572935315,
      "learning_rate": 6.202624781831269e-06,
      "loss": 0.8904,
      "step": 45
    },
    {
      "epoch": 0.45142296368989204,
      "grad_norm": 0.2480445894592462,
      "learning_rate": 6.044837815156377e-06,
      "loss": 0.8717,
      "step": 46
    },
    {
      "epoch": 0.46123650637880276,
      "grad_norm": 0.2598210274633706,
      "learning_rate": 5.885954957896115e-06,
      "loss": 0.8659,
      "step": 47
    },
    {
      "epoch": 0.47105004906771347,
      "grad_norm": 0.25886899490619003,
      "learning_rate": 5.726142856227453e-06,
      "loss": 0.9661,
      "step": 48
    },
    {
      "epoch": 0.48086359175662413,
      "grad_norm": 0.27445286917757566,
      "learning_rate": 5.5655691309764225e-06,
      "loss": 0.7741,
      "step": 49
    },
    {
      "epoch": 0.49067713444553485,
      "grad_norm": 0.2998354383728768,
      "learning_rate": 5.404402201807022e-06,
      "loss": 0.7016,
      "step": 50
    },
    {
      "epoch": 0.5004906771344455,
      "grad_norm": 0.30281526320811003,
      "learning_rate": 5.242811110572243e-06,
      "loss": 0.8998,
      "step": 51
    },
    {
      "epoch": 0.5103042198233563,
      "grad_norm": 0.23291295344605198,
      "learning_rate": 5.080965344012509e-06,
      "loss": 0.8599,
      "step": 52
    },
    {
      "epoch": 0.5201177625122669,
      "grad_norm": 0.2564082609308688,
      "learning_rate": 4.919034655987493e-06,
      "loss": 0.8412,
      "step": 53
    },
    {
      "epoch": 0.5299313052011776,
      "grad_norm": 0.27404998782840256,
      "learning_rate": 4.757188889427761e-06,
      "loss": 0.8401,
      "step": 54
    },
    {
      "epoch": 0.5397448478900884,
      "grad_norm": 0.25818007029770623,
      "learning_rate": 4.59559779819298e-06,
      "loss": 0.9063,
      "step": 55
    },
    {
      "epoch": 0.549558390578999,
      "grad_norm": 0.26623848076266293,
      "learning_rate": 4.434430869023579e-06,
      "loss": 0.7651,
      "step": 56
    },
    {
      "epoch": 0.5593719332679097,
      "grad_norm": 0.2658949551049741,
      "learning_rate": 4.27385714377255e-06,
      "loss": 0.9671,
      "step": 57
    },
    {
      "epoch": 0.5691854759568205,
      "grad_norm": 0.2954557630383791,
      "learning_rate": 4.1140450421038865e-06,
      "loss": 0.8364,
      "step": 58
    },
    {
      "epoch": 0.5789990186457311,
      "grad_norm": 0.24989225651779404,
      "learning_rate": 3.955162184843625e-06,
      "loss": 0.8642,
      "step": 59
    },
    {
      "epoch": 0.5888125613346418,
      "grad_norm": 0.2540070747708594,
      "learning_rate": 3.7973752181687336e-06,
      "loss": 0.9487,
      "step": 60
    },
    {
      "epoch": 0.5986261040235525,
      "grad_norm": 0.2875487613543649,
      "learning_rate": 3.6408496388182857e-06,
      "loss": 0.8931,
      "step": 61
    },
    {
      "epoch": 0.6084396467124632,
      "grad_norm": 0.25257746440863926,
      "learning_rate": 3.4857496205102475e-06,
      "loss": 0.7283,
      "step": 62
    },
    {
      "epoch": 0.6182531894013739,
      "grad_norm": 0.2612423385739647,
      "learning_rate": 3.3322378417458985e-06,
      "loss": 0.8973,
      "step": 63
    },
    {
      "epoch": 0.6280667320902846,
      "grad_norm": 0.2863739428894319,
      "learning_rate": 3.180475315182563e-06,
      "loss": 0.8244,
      "step": 64
    },
    {
      "epoch": 0.6378802747791953,
      "grad_norm": 0.25895617346448274,
      "learning_rate": 3.0306212187535653e-06,
      "loss": 0.8444,
      "step": 65
    },
    {
      "epoch": 0.647693817468106,
      "grad_norm": 0.25861219708555827,
      "learning_rate": 2.882832728712551e-06,
      "loss": 0.9331,
      "step": 66
    },
    {
      "epoch": 0.6575073601570167,
      "grad_norm": 0.259266449527201,
      "learning_rate": 2.7372648547773063e-06,
      "loss": 0.9111,
      "step": 67
    },
    {
      "epoch": 0.6673209028459274,
      "grad_norm": 0.2536619784605648,
      "learning_rate": 2.594070277545975e-06,
      "loss": 0.8717,
      "step": 68
    },
    {
      "epoch": 0.677134445534838,
      "grad_norm": 0.26066935129753177,
      "learning_rate": 2.4533991883561868e-06,
      "loss": 0.7797,
      "step": 69
    },
    {
      "epoch": 0.6869479882237488,
      "grad_norm": 0.2517815528639677,
      "learning_rate": 2.315399131755081e-06,
      "loss": 0.8631,
      "step": 70
    },
    {
      "epoch": 0.6967615309126595,
      "grad_norm": 0.25996941562937426,
      "learning_rate": 2.1802148507454675e-06,
      "loss": 0.8719,
      "step": 71
    },
    {
      "epoch": 0.7065750736015701,
      "grad_norm": 0.24588266413337945,
      "learning_rate": 2.0479881349703885e-06,
      "loss": 0.9193,
      "step": 72
    },
    {
      "epoch": 0.7163886162904809,
      "grad_norm": 0.2745377438843979,
      "learning_rate": 1.9188576719953635e-06,
      "loss": 0.861,
      "step": 73
    },
    {
      "epoch": 0.7262021589793916,
      "grad_norm": 0.22640992501852827,
      "learning_rate": 1.7929589018443016e-06,
      "loss": 0.7663,
      "step": 74
    },
    {
      "epoch": 0.7360157016683022,
      "grad_norm": 0.24363259853967661,
      "learning_rate": 1.6704238749415958e-06,
      "loss": 0.7472,
      "step": 75
    },
    {
      "epoch": 0.745829244357213,
      "grad_norm": 0.2777670891337088,
      "learning_rate": 1.5513811136094786e-06,
      "loss": 0.9389,
      "step": 76
    },
    {
      "epoch": 0.7556427870461236,
      "grad_norm": 0.26198217510926863,
      "learning_rate": 1.4359554772658551e-06,
      "loss": 0.8868,
      "step": 77
    },
    {
      "epoch": 0.7654563297350343,
      "grad_norm": 0.25420697034959644,
      "learning_rate": 1.3242680314639995e-06,
      "loss": 0.7417,
      "step": 78
    },
    {
      "epoch": 0.7752698724239451,
      "grad_norm": 0.27383288657704047,
      "learning_rate": 1.2164359209115235e-06,
      "loss": 0.7767,
      "step": 79
    },
    {
      "epoch": 0.7850834151128557,
      "grad_norm": 0.26155693194049967,
      "learning_rate": 1.1125722466017547e-06,
      "loss": 0.8705,
      "step": 80
    },
    {
      "epoch": 0.7948969578017664,
      "grad_norm": 0.2892989585599706,
      "learning_rate": 1.012785947186397e-06,
      "loss": 0.8472,
      "step": 81
    },
    {
      "epoch": 0.8047105004906772,
      "grad_norm": 0.257453864175166,
      "learning_rate": 9.171816847139447e-07,
      "loss": 0.9102,
      "step": 82
    },
    {
      "epoch": 0.8145240431795878,
      "grad_norm": 0.25933961134109923,
      "learning_rate": 8.258597348536452e-07,
      "loss": 0.8377,
      "step": 83
    },
    {
      "epoch": 0.8243375858684985,
      "grad_norm": 0.2504962150041825,
      "learning_rate": 7.389158817201541e-07,
      "loss": 0.8331,
      "step": 84
    },
    {
      "epoch": 0.8341511285574092,
      "grad_norm": 0.2420056571684986,
      "learning_rate": 6.564413174092443e-07,
      "loss": 0.8771,
      "step": 85
    },
    {
      "epoch": 0.8439646712463199,
      "grad_norm": 0.23509290212940573,
      "learning_rate": 5.785225463498828e-07,
      "loss": 0.8197,
      "step": 86
    },
    {
      "epoch": 0.8537782139352306,
      "grad_norm": 0.26997192628106204,
      "learning_rate": 5.05241294573024e-07,
      "loss": 0.8369,
      "step": 87
    },
    {
      "epoch": 0.8635917566241413,
      "grad_norm": 0.2820942983849838,
      "learning_rate": 4.3667442399229985e-07,
      "loss": 0.7722,
      "step": 88
    },
    {
      "epoch": 0.873405299313052,
      "grad_norm": 0.2516407661322018,
      "learning_rate": 3.728938517864794e-07,
      "loss": 0.8593,
      "step": 89
    },
    {
      "epoch": 0.8832188420019627,
      "grad_norm": 0.26231652652086895,
      "learning_rate": 3.1396647496828245e-07,
      "loss": 0.8056,
      "step": 90
    },
    {
      "epoch": 0.8930323846908734,
      "grad_norm": 0.27028297041866356,
      "learning_rate": 2.599541002186479e-07,
      "loss": 0.7567,
      "step": 91
    },
    {
      "epoch": 0.9028459273797841,
      "grad_norm": 0.23972498640175965,
      "learning_rate": 2.109133790600648e-07,
      "loss": 0.8646,
      "step": 92
    },
    {
      "epoch": 0.9126594700686947,
      "grad_norm": 0.25770792802506376,
      "learning_rate": 1.6689574843694433e-07,
      "loss": 0.8596,
      "step": 93
    },
    {
      "epoch": 0.9224730127576055,
      "grad_norm": 0.25190515551071735,
      "learning_rate": 1.2794737676536993e-07,
      "loss": 0.711,
      "step": 94
    },
    {
      "epoch": 0.9322865554465162,
      "grad_norm": 0.24771560518251803,
      "learning_rate": 9.410911550880474e-08,
      "loss": 0.8539,
      "step": 95
    },
    {
      "epoch": 0.9421000981354269,
      "grad_norm": 0.2534003058388717,
      "learning_rate": 6.54164563305465e-08,
      "loss": 0.8507,
      "step": 96
    },
    {
      "epoch": 0.9519136408243376,
      "grad_norm": 0.27958663472352846,
      "learning_rate": 4.189949386787462e-08,
      "loss": 0.7761,
      "step": 97
    },
    {
      "epoch": 0.9617271835132483,
      "grad_norm": 0.28280544357692566,
      "learning_rate": 2.358289416693027e-08,
      "loss": 0.8915,
      "step": 98
    },
    {
      "epoch": 0.971540726202159,
      "grad_norm": 0.251854611552598,
      "learning_rate": 1.0485868811441757e-08,
      "loss": 0.9458,
      "step": 99
    },
    {
      "epoch": 0.9813542688910697,
      "grad_norm": 0.2651175828136124,
      "learning_rate": 2.6221547724253337e-09,
      "loss": 0.7782,
      "step": 100
    },
    {
      "epoch": 0.9911678115799804,
      "grad_norm": 0.262045684118293,
      "learning_rate": 0.0,
      "loss": 0.7791,
      "step": 101
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 101,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 412567592239104.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}