{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 361,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002770083102493075,
      "grad_norm": 15.603601455688477,
      "learning_rate": 7.5e-06,
      "loss": 1.4608,
      "step": 1
    },
    {
      "epoch": 0.00554016620498615,
      "grad_norm": 14.760490417480469,
      "learning_rate": 1.5e-05,
      "loss": 1.428,
      "step": 2
    },
    {
      "epoch": 0.008310249307479225,
      "grad_norm": 19.810142517089844,
      "learning_rate": 2.25e-05,
      "loss": 1.2831,
      "step": 3
    },
    {
      "epoch": 0.0110803324099723,
      "grad_norm": 6.9140729904174805,
      "learning_rate": 3e-05,
      "loss": 1.0688,
      "step": 4
    },
    {
      "epoch": 0.013850415512465374,
      "grad_norm": 6.4214253425598145,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.2067,
      "step": 5
    },
    {
      "epoch": 0.01662049861495845,
      "grad_norm": 11.87337589263916,
      "learning_rate": 4.5e-05,
      "loss": 1.4816,
      "step": 6
    },
    {
      "epoch": 0.019390581717451522,
      "grad_norm": 6.669366359710693,
      "learning_rate": 5.25e-05,
      "loss": 1.1287,
      "step": 7
    },
    {
      "epoch": 0.0221606648199446,
      "grad_norm": 8.616178512573242,
      "learning_rate": 6e-05,
      "loss": 1.2203,
      "step": 8
    },
    {
      "epoch": 0.024930747922437674,
      "grad_norm": 9.265680313110352,
      "learning_rate": 5.999881193903433e-05,
      "loss": 1.1897,
      "step": 9
    },
    {
      "epoch": 0.027700831024930747,
      "grad_norm": 5.8777852058410645,
      "learning_rate": 5.999524785023657e-05,
      "loss": 1.0788,
      "step": 10
    },
    {
      "epoch": 0.030470914127423823,
      "grad_norm": 5.538949012756348,
      "learning_rate": 5.998930801589704e-05,
      "loss": 1.0736,
      "step": 11
    },
    {
      "epoch": 0.0332409972299169,
      "grad_norm": 5.686643123626709,
      "learning_rate": 5.9980992906474764e-05,
      "loss": 1.1573,
      "step": 12
    },
    {
      "epoch": 0.036011080332409975,
      "grad_norm": 6.211599349975586,
      "learning_rate": 5.9970303180560196e-05,
      "loss": 1.1618,
      "step": 13
    },
    {
      "epoch": 0.038781163434903045,
      "grad_norm": 5.079687595367432,
      "learning_rate": 5.995723968482309e-05,
      "loss": 0.9948,
      "step": 14
    },
    {
      "epoch": 0.04155124653739612,
      "grad_norm": 5.150979042053223,
      "learning_rate": 5.994180345394539e-05,
      "loss": 1.0328,
      "step": 15
    },
    {
      "epoch": 0.0443213296398892,
      "grad_norm": 5.891695499420166,
      "learning_rate": 5.9923995710539324e-05,
      "loss": 1.1854,
      "step": 16
    },
    {
      "epoch": 0.04709141274238227,
      "grad_norm": 5.33378791809082,
      "learning_rate": 5.990381786505055e-05,
      "loss": 1.0424,
      "step": 17
    },
    {
      "epoch": 0.04986149584487535,
      "grad_norm": 4.58151912689209,
      "learning_rate": 5.988127151564644e-05,
      "loss": 1.0392,
      "step": 18
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 5.48449182510376,
      "learning_rate": 5.9856358448089506e-05,
      "loss": 1.147,
      "step": 19
    },
    {
      "epoch": 0.055401662049861494,
      "grad_norm": 5.061849117279053,
      "learning_rate": 5.9829080635595944e-05,
      "loss": 1.1072,
      "step": 20
    },
    {
      "epoch": 0.05817174515235457,
      "grad_norm": 4.828271389007568,
      "learning_rate": 5.979944023867938e-05,
      "loss": 1.1132,
      "step": 21
    },
    {
      "epoch": 0.060941828254847646,
      "grad_norm": 5.035257816314697,
      "learning_rate": 5.976743960497973e-05,
      "loss": 1.0272,
      "step": 22
    },
    {
      "epoch": 0.06371191135734072,
      "grad_norm": 5.55271577835083,
      "learning_rate": 5.973308126907723e-05,
      "loss": 1.1674,
      "step": 23
    },
    {
      "epoch": 0.0664819944598338,
      "grad_norm": 4.926880836486816,
      "learning_rate": 5.969636795229172e-05,
      "loss": 1.0765,
      "step": 24
    },
    {
      "epoch": 0.06925207756232687,
      "grad_norm": 5.437366962432861,
      "learning_rate": 5.965730256246713e-05,
      "loss": 1.1344,
      "step": 25
    },
    {
      "epoch": 0.07202216066481995,
      "grad_norm": 4.4522786140441895,
      "learning_rate": 5.9615888193741094e-05,
      "loss": 0.9976,
      "step": 26
    },
    {
      "epoch": 0.07479224376731301,
      "grad_norm": 5.204329967498779,
      "learning_rate": 5.9572128126299946e-05,
      "loss": 1.1298,
      "step": 27
    },
    {
      "epoch": 0.07756232686980609,
      "grad_norm": 5.264033317565918,
      "learning_rate": 5.9526025826118885e-05,
      "loss": 1.0428,
      "step": 28
    },
    {
      "epoch": 0.08033240997229917,
      "grad_norm": 5.672830581665039,
      "learning_rate": 5.947758494468746e-05,
      "loss": 1.1636,
      "step": 29
    },
    {
      "epoch": 0.08310249307479224,
      "grad_norm": 5.798141002655029,
      "learning_rate": 5.942680931872036e-05,
      "loss": 1.1112,
      "step": 30
    },
    {
      "epoch": 0.08587257617728532,
      "grad_norm": 4.760410308837891,
      "learning_rate": 5.937370296985354e-05,
      "loss": 1.0444,
      "step": 31
    },
    {
      "epoch": 0.0886426592797784,
      "grad_norm": 5.4676313400268555,
      "learning_rate": 5.931827010432566e-05,
      "loss": 1.0957,
      "step": 32
    },
    {
      "epoch": 0.09141274238227147,
      "grad_norm": 4.983479976654053,
      "learning_rate": 5.9260515112644995e-05,
      "loss": 1.0734,
      "step": 33
    },
    {
      "epoch": 0.09418282548476455,
      "grad_norm": 5.242843151092529,
      "learning_rate": 5.9200442569241606e-05,
      "loss": 1.0879,
      "step": 34
    },
    {
      "epoch": 0.09695290858725762,
      "grad_norm": 5.12119722366333,
      "learning_rate": 5.9138057232105084e-05,
      "loss": 1.0841,
      "step": 35
    },
    {
      "epoch": 0.0997229916897507,
      "grad_norm": 5.087031841278076,
      "learning_rate": 5.9073364042407705e-05,
      "loss": 1.0343,
      "step": 36
    },
    {
      "epoch": 0.10249307479224377,
      "grad_norm": 5.226377487182617,
      "learning_rate": 5.900636812411301e-05,
      "loss": 1.0775,
      "step": 37
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 5.388728141784668,
      "learning_rate": 5.893707478357005e-05,
      "loss": 1.0439,
      "step": 38
    },
    {
      "epoch": 0.10803324099722991,
      "grad_norm": 5.6382951736450195,
      "learning_rate": 5.886548950909301e-05,
      "loss": 1.096,
      "step": 39
    },
    {
      "epoch": 0.11080332409972299,
      "grad_norm": 5.046523094177246,
      "learning_rate": 5.879161797052658e-05,
      "loss": 0.9928,
      "step": 40
    },
    {
      "epoch": 0.11357340720221606,
      "grad_norm": 5.98545503616333,
      "learning_rate": 5.8715466018796865e-05,
      "loss": 1.1639,
      "step": 41
    },
    {
      "epoch": 0.11634349030470914,
      "grad_norm": 5.064982891082764,
      "learning_rate": 5.863703968544795e-05,
      "loss": 0.9878,
      "step": 42
    },
    {
      "epoch": 0.11911357340720222,
      "grad_norm": 5.36715841293335,
      "learning_rate": 5.8556345182164176e-05,
      "loss": 1.056,
      "step": 43
    },
    {
      "epoch": 0.12188365650969529,
      "grad_norm": 4.7610368728637695,
      "learning_rate": 5.8473388900278206e-05,
      "loss": 0.9625,
      "step": 44
    },
    {
      "epoch": 0.12465373961218837,
      "grad_norm": 5.5188889503479,
      "learning_rate": 5.83881774102647e-05,
      "loss": 1.0828,
      "step": 45
    },
    {
      "epoch": 0.12742382271468145,
      "grad_norm": 5.703125476837158,
      "learning_rate": 5.8300717461220027e-05,
      "loss": 1.0632,
      "step": 46
    },
    {
      "epoch": 0.13019390581717452,
      "grad_norm": 5.264132499694824,
      "learning_rate": 5.821101598032759e-05,
      "loss": 1.0057,
      "step": 47
    },
    {
      "epoch": 0.1329639889196676,
      "grad_norm": 6.05692720413208,
      "learning_rate": 5.811908007230929e-05,
      "loss": 1.0565,
      "step": 48
    },
    {
      "epoch": 0.13573407202216067,
      "grad_norm": 4.773650169372559,
      "learning_rate": 5.802491701886268e-05,
      "loss": 0.9764,
      "step": 49
    },
    {
      "epoch": 0.13850415512465375,
      "grad_norm": 5.461302757263184,
      "learning_rate": 5.792853427808431e-05,
      "loss": 1.0002,
      "step": 50
    },
    {
      "epoch": 0.14127423822714683,
      "grad_norm": 5.83071231842041,
      "learning_rate": 5.7829939483878996e-05,
      "loss": 1.0353,
      "step": 51
    },
    {
      "epoch": 0.1440443213296399,
      "grad_norm": 5.715704917907715,
      "learning_rate": 5.772914044535516e-05,
      "loss": 0.985,
      "step": 52
    },
    {
      "epoch": 0.14681440443213298,
      "grad_norm": 5.309004306793213,
      "learning_rate": 5.762614514620634e-05,
      "loss": 1.0106,
      "step": 53
    },
    {
      "epoch": 0.14958448753462603,
      "grad_norm": 5.693915843963623,
      "learning_rate": 5.752096174407884e-05,
      "loss": 1.0049,
      "step": 54
    },
    {
      "epoch": 0.1523545706371191,
      "grad_norm": 5.644995212554932,
      "learning_rate": 5.741359856992561e-05,
      "loss": 1.0411,
      "step": 55
    },
    {
      "epoch": 0.15512465373961218,
      "grad_norm": 5.765170574188232,
      "learning_rate": 5.73040641273464e-05,
      "loss": 1.0039,
      "step": 56
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 5.643161296844482,
      "learning_rate": 5.719236709191428e-05,
      "loss": 1.0173,
      "step": 57
    },
    {
      "epoch": 0.16066481994459833,
      "grad_norm": 5.821686267852783,
      "learning_rate": 5.707851631048841e-05,
      "loss": 0.9725,
      "step": 58
    },
    {
      "epoch": 0.1634349030470914,
      "grad_norm": 6.034667015075684,
      "learning_rate": 5.6962520800513414e-05,
      "loss": 1.0381,
      "step": 59
    },
    {
      "epoch": 0.16620498614958448,
      "grad_norm": 5.336913585662842,
      "learning_rate": 5.6844389749305136e-05,
      "loss": 0.9266,
      "step": 60
    },
    {
      "epoch": 0.16897506925207756,
      "grad_norm": 5.65543270111084,
      "learning_rate": 5.672413251332297e-05,
      "loss": 0.9768,
      "step": 61
    },
    {
      "epoch": 0.17174515235457063,
      "grad_norm": 5.3824615478515625,
      "learning_rate": 5.6601758617428766e-05,
      "loss": 0.9276,
      "step": 62
    },
    {
      "epoch": 0.1745152354570637,
      "grad_norm": 5.195915222167969,
      "learning_rate": 5.647727775413245e-05,
      "loss": 0.9332,
      "step": 63
    },
    {
      "epoch": 0.1772853185595568,
      "grad_norm": 5.240732669830322,
      "learning_rate": 5.6350699782824346e-05,
      "loss": 0.9012,
      "step": 64
    },
    {
      "epoch": 0.18005540166204986,
      "grad_norm": 5.112541675567627,
      "learning_rate": 5.622203472899423e-05,
      "loss": 0.9211,
      "step": 65
    },
    {
      "epoch": 0.18282548476454294,
      "grad_norm": 6.605805397033691,
      "learning_rate": 5.609129278343731e-05,
      "loss": 0.9847,
      "step": 66
    },
    {
      "epoch": 0.18559556786703602,
      "grad_norm": 5.1862568855285645,
      "learning_rate": 5.595848430144705e-05,
      "loss": 0.9064,
      "step": 67
    },
    {
      "epoch": 0.1883656509695291,
      "grad_norm": 6.03863525390625,
      "learning_rate": 5.582361980199504e-05,
      "loss": 0.9405,
      "step": 68
    },
    {
      "epoch": 0.19113573407202217,
      "grad_norm": 6.4853901863098145,
      "learning_rate": 5.568670996689773e-05,
      "loss": 0.9676,
      "step": 69
    },
    {
      "epoch": 0.19390581717451524,
      "grad_norm": 5.415260314941406,
      "learning_rate": 5.554776563997056e-05,
      "loss": 0.8873,
      "step": 70
    },
    {
      "epoch": 0.19667590027700832,
      "grad_norm": 5.11320686340332,
      "learning_rate": 5.540679782616892e-05,
      "loss": 0.8752,
      "step": 71
    },
    {
      "epoch": 0.1994459833795014,
      "grad_norm": 4.8276214599609375,
      "learning_rate": 5.52638176907166e-05,
      "loss": 0.8441,
      "step": 72
    },
    {
      "epoch": 0.20221606648199447,
      "grad_norm": 5.574111461639404,
      "learning_rate": 5.5118836558221475e-05,
      "loss": 0.907,
      "step": 73
    },
    {
      "epoch": 0.20498614958448755,
      "grad_norm": 5.575868606567383,
      "learning_rate": 5.497186591177849e-05,
      "loss": 0.8691,
      "step": 74
    },
    {
      "epoch": 0.2077562326869806,
      "grad_norm": 5.388404846191406,
      "learning_rate": 5.4822917392060184e-05,
      "loss": 0.8935,
      "step": 75
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 6.2548747062683105,
      "learning_rate": 5.46720027963947e-05,
      "loss": 0.8814,
      "step": 76
    },
    {
      "epoch": 0.21329639889196675,
      "grad_norm": 5.358329772949219,
      "learning_rate": 5.4519134077831395e-05,
      "loss": 0.9041,
      "step": 77
    },
    {
      "epoch": 0.21606648199445982,
      "grad_norm": 6.0121564865112305,
      "learning_rate": 5.4364323344194095e-05,
      "loss": 0.8568,
      "step": 78
    },
    {
      "epoch": 0.2188365650969529,
      "grad_norm": 6.2294020652771,
      "learning_rate": 5.420758285712211e-05,
      "loss": 0.9205,
      "step": 79
    },
    {
      "epoch": 0.22160664819944598,
      "grad_norm": 5.889683723449707,
      "learning_rate": 5.404892503109906e-05,
      "loss": 0.8264,
      "step": 80
    },
    {
      "epoch": 0.22437673130193905,
      "grad_norm": 5.401793956756592,
      "learning_rate": 5.388836243246963e-05,
      "loss": 0.8574,
      "step": 81
    },
    {
      "epoch": 0.22714681440443213,
      "grad_norm": 6.194601058959961,
      "learning_rate": 5.372590777844421e-05,
      "loss": 0.8742,
      "step": 82
    },
    {
      "epoch": 0.2299168975069252,
      "grad_norm": 5.504939079284668,
      "learning_rate": 5.356157393609167e-05,
      "loss": 0.8261,
      "step": 83
    },
    {
      "epoch": 0.23268698060941828,
      "grad_norm": 5.818756103515625,
      "learning_rate": 5.339537392132025e-05,
      "loss": 0.8364,
      "step": 84
    },
    {
      "epoch": 0.23545706371191136,
      "grad_norm": 5.8196024894714355,
      "learning_rate": 5.322732089784661e-05,
      "loss": 0.8577,
      "step": 85
    },
    {
      "epoch": 0.23822714681440443,
      "grad_norm": 6.102773189544678,
      "learning_rate": 5.305742817615325e-05,
      "loss": 0.8708,
      "step": 86
    },
    {
      "epoch": 0.2409972299168975,
      "grad_norm": 6.225861549377441,
      "learning_rate": 5.288570921243423e-05,
      "loss": 0.8654,
      "step": 87
    },
    {
      "epoch": 0.24376731301939059,
      "grad_norm": 6.0107340812683105,
      "learning_rate": 5.2712177607529405e-05,
      "loss": 0.8401,
      "step": 88
    },
    {
      "epoch": 0.24653739612188366,
      "grad_norm": 6.3431715965271,
      "learning_rate": 5.2536847105847185e-05,
      "loss": 0.855,
      "step": 89
    },
    {
      "epoch": 0.24930747922437674,
      "grad_norm": 5.196923732757568,
      "learning_rate": 5.235973159427591e-05,
      "loss": 0.756,
      "step": 90
    },
    {
      "epoch": 0.2520775623268698,
      "grad_norm": 6.157230377197266,
      "learning_rate": 5.218084510108397e-05,
      "loss": 0.856,
      "step": 91
    },
    {
      "epoch": 0.2548476454293629,
      "grad_norm": 5.813492774963379,
      "learning_rate": 5.200020179480868e-05,
      "loss": 0.7829,
      "step": 92
    },
    {
      "epoch": 0.25761772853185594,
      "grad_norm": 6.468319416046143,
      "learning_rate": 5.181781598313409e-05,
      "loss": 0.8327,
      "step": 93
    },
    {
      "epoch": 0.26038781163434904,
      "grad_norm": 6.1590423583984375,
      "learning_rate": 5.16337021117578e-05,
      "loss": 0.8035,
      "step": 94
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 6.378724575042725,
      "learning_rate": 5.14478747632467e-05,
      "loss": 0.8134,
      "step": 95
    },
    {
      "epoch": 0.2659279778393352,
      "grad_norm": 5.47311544418335,
      "learning_rate": 5.126034865588208e-05,
      "loss": 0.7149,
      "step": 96
    },
    {
      "epoch": 0.26869806094182824,
      "grad_norm": 5.790994644165039,
      "learning_rate": 5.107113864249381e-05,
      "loss": 0.7714,
      "step": 97
    },
    {
      "epoch": 0.27146814404432135,
      "grad_norm": 5.880769729614258,
      "learning_rate": 5.088025970928399e-05,
      "loss": 0.7519,
      "step": 98
    },
    {
      "epoch": 0.2742382271468144,
      "grad_norm": 5.894231796264648,
      "learning_rate": 5.068772697463992e-05,
      "loss": 0.8124,
      "step": 99
    },
    {
      "epoch": 0.2770083102493075,
      "grad_norm": 6.011329174041748,
      "learning_rate": 5.0493555687936704e-05,
      "loss": 0.7536,
      "step": 100
    },
    {
      "epoch": 0.27977839335180055,
      "grad_norm": 5.813382148742676,
      "learning_rate": 5.029776122832945e-05,
      "loss": 0.7483,
      "step": 101
    },
    {
      "epoch": 0.28254847645429365,
      "grad_norm": 6.445542335510254,
      "learning_rate": 5.0100359103535134e-05,
      "loss": 0.7376,
      "step": 102
    },
    {
      "epoch": 0.2853185595567867,
      "grad_norm": 5.571536540985107,
      "learning_rate": 4.9901364948604366e-05,
      "loss": 0.7538,
      "step": 103
    },
    {
      "epoch": 0.2880886426592798,
      "grad_norm": 5.566730499267578,
      "learning_rate": 4.970079452468298e-05,
      "loss": 0.7102,
      "step": 104
    },
    {
      "epoch": 0.29085872576177285,
      "grad_norm": 6.336560249328613,
      "learning_rate": 4.949866371776376e-05,
      "loss": 0.7469,
      "step": 105
    },
    {
      "epoch": 0.29362880886426596,
      "grad_norm": 5.440810203552246,
      "learning_rate": 4.929498853742816e-05,
      "loss": 0.7252,
      "step": 106
    },
    {
      "epoch": 0.296398891966759,
      "grad_norm": 6.162292957305908,
      "learning_rate": 4.908978511557827e-05,
      "loss": 0.7569,
      "step": 107
    },
    {
      "epoch": 0.29916897506925205,
      "grad_norm": 6.059966087341309,
      "learning_rate": 4.8883069705159104e-05,
      "loss": 0.7136,
      "step": 108
    },
    {
      "epoch": 0.30193905817174516,
      "grad_norm": 6.249420642852783,
      "learning_rate": 4.867485867887136e-05,
      "loss": 0.7408,
      "step": 109
    },
    {
      "epoch": 0.3047091412742382,
      "grad_norm": 5.173600673675537,
      "learning_rate": 4.846516852787457e-05,
      "loss": 0.7106,
      "step": 110
    },
    {
      "epoch": 0.3074792243767313,
      "grad_norm": 5.633162975311279,
      "learning_rate": 4.8254015860480934e-05,
      "loss": 0.6951,
      "step": 111
    },
    {
      "epoch": 0.31024930747922436,
      "grad_norm": 5.940577507019043,
      "learning_rate": 4.804141740083993e-05,
      "loss": 0.7022,
      "step": 112
    },
    {
      "epoch": 0.31301939058171746,
      "grad_norm": 5.466473579406738,
      "learning_rate": 4.7827389987613635e-05,
      "loss": 0.6694,
      "step": 113
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 6.02973747253418,
      "learning_rate": 4.761195057264307e-05,
      "loss": 0.6799,
      "step": 114
    },
    {
      "epoch": 0.3185595567867036,
      "grad_norm": 5.916580677032471,
      "learning_rate": 4.739511621960551e-05,
      "loss": 0.7019,
      "step": 115
    },
    {
      "epoch": 0.32132963988919666,
      "grad_norm": 5.600930213928223,
      "learning_rate": 4.717690410266304e-05,
      "loss": 0.6818,
      "step": 116
    },
    {
      "epoch": 0.32409972299168976,
      "grad_norm": 5.327589988708496,
      "learning_rate": 4.69573315051022e-05,
      "loss": 0.6811,
      "step": 117
    },
    {
      "epoch": 0.3268698060941828,
      "grad_norm": 5.755538463592529,
      "learning_rate": 4.673641581796515e-05,
      "loss": 0.6705,
      "step": 118
    },
    {
      "epoch": 0.3296398891966759,
      "grad_norm": 6.217751979827881,
      "learning_rate": 4.6514174538672187e-05,
      "loss": 0.6856,
      "step": 119
    },
    {
      "epoch": 0.33240997229916897,
      "grad_norm": 5.41038179397583,
      "learning_rate": 4.629062526963592e-05,
      "loss": 0.6509,
      "step": 120
    },
    {
      "epoch": 0.33518005540166207,
      "grad_norm": 5.7826247215271,
      "learning_rate": 4.606578571686703e-05,
      "loss": 0.662,
      "step": 121
    },
    {
      "epoch": 0.3379501385041551,
      "grad_norm": 6.125760555267334,
      "learning_rate": 4.583967368857194e-05,
      "loss": 0.6518,
      "step": 122
    },
    {
      "epoch": 0.3407202216066482,
      "grad_norm": 5.408969879150391,
      "learning_rate": 4.561230709374229e-05,
      "loss": 0.6391,
      "step": 123
    },
    {
      "epoch": 0.34349030470914127,
      "grad_norm": 5.880269527435303,
      "learning_rate": 4.5383703940736494e-05,
      "loss": 0.6388,
      "step": 124
    },
    {
      "epoch": 0.3462603878116344,
      "grad_norm": 6.065842628479004,
      "learning_rate": 4.51538823358534e-05,
      "loss": 0.653,
      "step": 125
    },
    {
      "epoch": 0.3490304709141274,
      "grad_norm": 6.418548107147217,
      "learning_rate": 4.49228604818982e-05,
      "loss": 0.6607,
      "step": 126
    },
    {
      "epoch": 0.3518005540166205,
      "grad_norm": 5.7509074211120605,
      "learning_rate": 4.469065667674069e-05,
      "loss": 0.6289,
      "step": 127
    },
    {
      "epoch": 0.3545706371191136,
      "grad_norm": 5.758950710296631,
      "learning_rate": 4.445728931186599e-05,
      "loss": 0.6332,
      "step": 128
    },
    {
      "epoch": 0.3573407202216066,
      "grad_norm": 5.907932758331299,
      "learning_rate": 4.42227768709179e-05,
      "loss": 0.634,
      "step": 129
    },
    {
      "epoch": 0.3601108033240997,
      "grad_norm": 5.904737949371338,
      "learning_rate": 4.398713792823489e-05,
      "loss": 0.5948,
      "step": 130
    },
    {
      "epoch": 0.3628808864265928,
      "grad_norm": 6.62101411819458,
      "learning_rate": 4.375039114737895e-05,
      "loss": 0.6332,
      "step": 131
    },
    {
      "epoch": 0.3656509695290859,
      "grad_norm": 6.03896951675415,
      "learning_rate": 4.3512555279657344e-05,
      "loss": 0.6153,
      "step": 132
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 5.788692474365234,
      "learning_rate": 4.3273649162637454e-05,
      "loss": 0.6223,
      "step": 133
    },
    {
      "epoch": 0.37119113573407203,
      "grad_norm": 5.691258907318115,
      "learning_rate": 4.3033691718654767e-05,
      "loss": 0.6048,
      "step": 134
    },
    {
      "epoch": 0.3739612188365651,
      "grad_norm": 5.318130016326904,
      "learning_rate": 4.2792701953314085e-05,
      "loss": 0.6023,
      "step": 135
    },
    {
      "epoch": 0.3767313019390582,
      "grad_norm": 6.409132480621338,
      "learning_rate": 4.2550698953984335e-05,
      "loss": 0.5975,
      "step": 136
    },
    {
      "epoch": 0.37950138504155123,
      "grad_norm": 5.915623664855957,
      "learning_rate": 4.230770188828665e-05,
      "loss": 0.6048,
      "step": 137
    },
    {
      "epoch": 0.38227146814404434,
      "grad_norm": 6.316535472869873,
      "learning_rate": 4.2063730002576244e-05,
      "loss": 0.5944,
      "step": 138
    },
    {
      "epoch": 0.3850415512465374,
      "grad_norm": 6.343788146972656,
      "learning_rate": 4.181880262041808e-05,
      "loss": 0.6115,
      "step": 139
    },
    {
      "epoch": 0.3878116343490305,
      "grad_norm": 6.030927658081055,
      "learning_rate": 4.15729391410563e-05,
      "loss": 0.5709,
      "step": 140
    },
    {
      "epoch": 0.39058171745152354,
      "grad_norm": 5.820158004760742,
      "learning_rate": 4.132615903787774e-05,
      "loss": 0.5882,
      "step": 141
    },
    {
      "epoch": 0.39335180055401664,
      "grad_norm": 5.914207935333252,
      "learning_rate": 4.1078481856869586e-05,
      "loss": 0.5579,
      "step": 142
    },
    {
      "epoch": 0.3961218836565097,
      "grad_norm": 5.729140758514404,
      "learning_rate": 4.0829927215071234e-05,
      "loss": 0.569,
      "step": 143
    },
    {
      "epoch": 0.3988919667590028,
      "grad_norm": 6.345619201660156,
      "learning_rate": 4.0580514799020525e-05,
      "loss": 0.5574,
      "step": 144
    },
    {
      "epoch": 0.40166204986149584,
      "grad_norm": 5.319207668304443,
      "learning_rate": 4.033026436319452e-05,
      "loss": 0.5584,
      "step": 145
    },
    {
      "epoch": 0.40443213296398894,
      "grad_norm": 6.186460018157959,
      "learning_rate": 4.0079195728444845e-05,
      "loss": 0.5476,
      "step": 146
    },
    {
      "epoch": 0.407202216066482,
      "grad_norm": 6.213712692260742,
      "learning_rate": 3.982732878042782e-05,
      "loss": 0.5403,
      "step": 147
    },
    {
      "epoch": 0.4099722991689751,
      "grad_norm": 5.564090251922607,
      "learning_rate": 3.95746834680294e-05,
      "loss": 0.5391,
      "step": 148
    },
    {
      "epoch": 0.41274238227146814,
      "grad_norm": 6.060794830322266,
      "learning_rate": 3.932127980178519e-05,
      "loss": 0.5332,
      "step": 149
    },
    {
      "epoch": 0.4155124653739612,
      "grad_norm": 5.621760368347168,
      "learning_rate": 3.906713785229546e-05,
      "loss": 0.5426,
      "step": 150
    },
    {
      "epoch": 0.4182825484764543,
      "grad_norm": 5.4746012687683105,
      "learning_rate": 3.8812277748635554e-05,
      "loss": 0.5444,
      "step": 151
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 4.968982696533203,
      "learning_rate": 3.855671967676153e-05,
      "loss": 0.525,
      "step": 152
    },
    {
      "epoch": 0.42382271468144045,
      "grad_norm": 5.877557277679443,
      "learning_rate": 3.830048387791136e-05,
      "loss": 0.5485,
      "step": 153
    },
    {
      "epoch": 0.4265927977839335,
      "grad_norm": 5.59464168548584,
      "learning_rate": 3.804359064700175e-05,
      "loss": 0.5274,
      "step": 154
    },
    {
      "epoch": 0.4293628808864266,
      "grad_norm": 5.718829154968262,
      "learning_rate": 3.778606033102071e-05,
      "loss": 0.5321,
      "step": 155
    },
    {
      "epoch": 0.43213296398891965,
      "grad_norm": 5.996484279632568,
      "learning_rate": 3.752791332741596e-05,
      "loss": 0.5083,
      "step": 156
    },
    {
      "epoch": 0.43490304709141275,
      "grad_norm": 5.6276373863220215,
      "learning_rate": 3.7269170082479376e-05,
      "loss": 0.5466,
      "step": 157
    },
    {
      "epoch": 0.4376731301939058,
      "grad_norm": 5.630644798278809,
      "learning_rate": 3.7009851089727615e-05,
      "loss": 0.5168,
      "step": 158
    },
    {
      "epoch": 0.4404432132963989,
      "grad_norm": 5.379649639129639,
      "learning_rate": 3.6749976888278864e-05,
      "loss": 0.529,
      "step": 159
    },
    {
      "epoch": 0.44321329639889195,
      "grad_norm": 5.533950328826904,
      "learning_rate": 3.6489568061226106e-05,
      "loss": 0.4908,
      "step": 160
    },
    {
      "epoch": 0.44598337950138506,
      "grad_norm": 6.063986301422119,
      "learning_rate": 3.6228645234006835e-05,
      "loss": 0.5209,
      "step": 161
    },
    {
      "epoch": 0.4487534626038781,
      "grad_norm": 6.18126916885376,
      "learning_rate": 3.596722907276946e-05,
      "loss": 0.4909,
      "step": 162
    },
    {
      "epoch": 0.4515235457063712,
      "grad_norm": 5.693446636199951,
      "learning_rate": 3.570534028273645e-05,
      "loss": 0.5257,
      "step": 163
    },
    {
      "epoch": 0.45429362880886426,
      "grad_norm": 4.973173141479492,
      "learning_rate": 3.5442999606564376e-05,
      "loss": 0.4936,
      "step": 164
    },
    {
      "epoch": 0.45706371191135736,
      "grad_norm": 5.8764729499816895,
      "learning_rate": 3.518022782270106e-05,
      "loss": 0.4793,
      "step": 165
    },
    {
      "epoch": 0.4598337950138504,
      "grad_norm": 4.693322658538818,
      "learning_rate": 3.491704574373978e-05,
      "loss": 0.4836,
      "step": 166
    },
    {
      "epoch": 0.4626038781163435,
      "grad_norm": 5.623289108276367,
      "learning_rate": 3.465347421477087e-05,
      "loss": 0.4814,
      "step": 167
    },
    {
      "epoch": 0.46537396121883656,
      "grad_norm": 5.063323497772217,
      "learning_rate": 3.438953411173066e-05,
      "loss": 0.4801,
      "step": 168
    },
    {
      "epoch": 0.46814404432132967,
      "grad_norm": 5.001865863800049,
      "learning_rate": 3.412524633974808e-05,
      "loss": 0.5035,
      "step": 169
    },
    {
      "epoch": 0.4709141274238227,
      "grad_norm": 4.878357410430908,
      "learning_rate": 3.386063183148883e-05,
      "loss": 0.4814,
      "step": 170
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 6.276041507720947,
      "learning_rate": 3.359571154549747e-05,
      "loss": 0.4853,
      "step": 171
    },
    {
      "epoch": 0.47645429362880887,
      "grad_norm": 5.440047740936279,
      "learning_rate": 3.3330506464537356e-05,
      "loss": 0.4837,
      "step": 172
    },
    {
      "epoch": 0.4792243767313019,
      "grad_norm": 4.922585964202881,
      "learning_rate": 3.306503759392882e-05,
      "loss": 0.4957,
      "step": 173
    },
    {
      "epoch": 0.481994459833795,
      "grad_norm": 4.886894226074219,
      "learning_rate": 3.279932595988538e-05,
      "loss": 0.4831,
      "step": 174
    },
    {
      "epoch": 0.48476454293628807,
      "grad_norm": 4.792039394378662,
      "learning_rate": 3.253339260784838e-05,
      "loss": 0.5031,
      "step": 175
    },
    {
      "epoch": 0.48753462603878117,
      "grad_norm": 4.9557647705078125,
      "learning_rate": 3.226725860082018e-05,
      "loss": 0.4812,
      "step": 176
    },
    {
      "epoch": 0.4903047091412742,
      "grad_norm": 5.699478626251221,
      "learning_rate": 3.200094501769581e-05,
      "loss": 0.4898,
      "step": 177
    },
    {
      "epoch": 0.4930747922437673,
      "grad_norm": 4.814623832702637,
      "learning_rate": 3.173447295159344e-05,
      "loss": 0.4469,
      "step": 178
    },
    {
      "epoch": 0.49584487534626037,
      "grad_norm": 5.314681053161621,
      "learning_rate": 3.1467863508183735e-05,
      "loss": 0.4889,
      "step": 179
    },
    {
      "epoch": 0.4986149584487535,
      "grad_norm": 5.161649703979492,
      "learning_rate": 3.1201137804018246e-05,
      "loss": 0.4543,
      "step": 180
    },
    {
      "epoch": 0.5013850415512465,
      "grad_norm": 5.343588829040527,
      "learning_rate": 3.093431696485679e-05,
      "loss": 0.4557,
      "step": 181
    },
    {
      "epoch": 0.5041551246537396,
      "grad_norm": 4.8219499588012695,
      "learning_rate": 3.0667422123994304e-05,
      "loss": 0.4607,
      "step": 182
    },
    {
      "epoch": 0.5069252077562327,
      "grad_norm": 5.002239227294922,
      "learning_rate": 3.040047442058694e-05,
      "loss": 0.4693,
      "step": 183
    },
    {
      "epoch": 0.5096952908587258,
      "grad_norm": 4.721946716308594,
      "learning_rate": 3.0133494997977792e-05,
      "loss": 0.4445,
      "step": 184
    },
    {
      "epoch": 0.5124653739612188,
      "grad_norm": 4.8258538246154785,
      "learning_rate": 2.9866505002022223e-05,
      "loss": 0.4712,
      "step": 185
    },
    {
      "epoch": 0.5152354570637119,
      "grad_norm": 5.217929840087891,
      "learning_rate": 2.9599525579413066e-05,
      "loss": 0.4282,
      "step": 186
    },
    {
      "epoch": 0.518005540166205,
      "grad_norm": 5.506514072418213,
      "learning_rate": 2.9332577876005707e-05,
      "loss": 0.4733,
      "step": 187
    },
    {
      "epoch": 0.5207756232686981,
      "grad_norm": 4.479320049285889,
      "learning_rate": 2.9065683035143213e-05,
      "loss": 0.4892,
      "step": 188
    },
    {
      "epoch": 0.5235457063711911,
      "grad_norm": 4.9687089920043945,
      "learning_rate": 2.8798862195981763e-05,
      "loss": 0.4549,
      "step": 189
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 5.03638219833374,
      "learning_rate": 2.8532136491816266e-05,
      "loss": 0.4461,
      "step": 190
    },
    {
      "epoch": 0.5290858725761773,
      "grad_norm": 5.558093547821045,
      "learning_rate": 2.8265527048406564e-05,
      "loss": 0.45,
      "step": 191
    },
    {
      "epoch": 0.5318559556786704,
      "grad_norm": 4.820427417755127,
      "learning_rate": 2.7999054982304192e-05,
      "loss": 0.4567,
      "step": 192
    },
    {
      "epoch": 0.5346260387811634,
      "grad_norm": 4.981771945953369,
      "learning_rate": 2.7732741399179813e-05,
      "loss": 0.4754,
      "step": 193
    },
    {
      "epoch": 0.5373961218836565,
      "grad_norm": 4.8197245597839355,
      "learning_rate": 2.746660739215162e-05,
      "loss": 0.4623,
      "step": 194
    },
    {
      "epoch": 0.5401662049861495,
      "grad_norm": 4.506078243255615,
      "learning_rate": 2.7200674040114627e-05,
      "loss": 0.4579,
      "step": 195
    },
    {
      "epoch": 0.5429362880886427,
      "grad_norm": 5.280223369598389,
      "learning_rate": 2.693496240607119e-05,
      "loss": 0.4455,
      "step": 196
    },
    {
      "epoch": 0.5457063711911357,
      "grad_norm": 4.943070411682129,
      "learning_rate": 2.666949353546265e-05,
      "loss": 0.4516,
      "step": 197
    },
    {
      "epoch": 0.5484764542936288,
      "grad_norm": 4.689980983734131,
      "learning_rate": 2.6404288454502546e-05,
      "loss": 0.4353,
      "step": 198
    },
    {
      "epoch": 0.5512465373961218,
      "grad_norm": 4.386908531188965,
      "learning_rate": 2.6139368168511176e-05,
      "loss": 0.47,
      "step": 199
    },
    {
      "epoch": 0.554016620498615,
      "grad_norm": 4.444925785064697,
      "learning_rate": 2.5874753660251916e-05,
      "loss": 0.4538,
      "step": 200
    },
    {
      "epoch": 0.556786703601108,
      "grad_norm": 4.536672592163086,
      "learning_rate": 2.561046588826934e-05,
      "loss": 0.4416,
      "step": 201
    },
    {
      "epoch": 0.5595567867036011,
      "grad_norm": 4.070997714996338,
      "learning_rate": 2.5346525785229134e-05,
      "loss": 0.4422,
      "step": 202
    },
    {
      "epoch": 0.5623268698060941,
      "grad_norm": 4.6580281257629395,
      "learning_rate": 2.5082954256260227e-05,
      "loss": 0.4526,
      "step": 203
    },
    {
      "epoch": 0.5650969529085873,
      "grad_norm": 4.6260247230529785,
      "learning_rate": 2.481977217729894e-05,
      "loss": 0.4232,
      "step": 204
    },
    {
      "epoch": 0.5678670360110804,
      "grad_norm": 4.411331653594971,
      "learning_rate": 2.4557000393435636e-05,
      "loss": 0.4599,
      "step": 205
    },
    {
      "epoch": 0.5706371191135734,
      "grad_norm": 4.103204250335693,
      "learning_rate": 2.429465971726356e-05,
      "loss": 0.4349,
      "step": 206
    },
    {
      "epoch": 0.5734072022160664,
      "grad_norm": 4.877063274383545,
      "learning_rate": 2.403277092723055e-05,
      "loss": 0.4521,
      "step": 207
    },
    {
      "epoch": 0.5761772853185596,
      "grad_norm": 4.473416328430176,
      "learning_rate": 2.3771354765993177e-05,
      "loss": 0.4398,
      "step": 208
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 4.630215644836426,
      "learning_rate": 2.351043193877391e-05,
      "loss": 0.4708,
      "step": 209
    },
    {
      "epoch": 0.5817174515235457,
      "grad_norm": 4.25671911239624,
      "learning_rate": 2.3250023111721137e-05,
      "loss": 0.4167,
      "step": 210
    },
    {
      "epoch": 0.5844875346260388,
      "grad_norm": 4.418769836425781,
      "learning_rate": 2.2990148910272383e-05,
      "loss": 0.4562,
      "step": 211
    },
    {
      "epoch": 0.5872576177285319,
      "grad_norm": 4.783448219299316,
      "learning_rate": 2.273082991752063e-05,
      "loss": 0.4371,
      "step": 212
    },
    {
      "epoch": 0.590027700831025,
      "grad_norm": 4.563108921051025,
      "learning_rate": 2.2472086672584048e-05,
      "loss": 0.4335,
      "step": 213
    },
    {
      "epoch": 0.592797783933518,
      "grad_norm": 4.505840301513672,
      "learning_rate": 2.22139396689793e-05,
      "loss": 0.4421,
      "step": 214
    },
    {
      "epoch": 0.5955678670360111,
      "grad_norm": 3.867152452468872,
      "learning_rate": 2.1956409352998253e-05,
      "loss": 0.479,
      "step": 215
    },
    {
      "epoch": 0.5983379501385041,
      "grad_norm": 4.328498363494873,
      "learning_rate": 2.169951612208865e-05,
      "loss": 0.4359,
      "step": 216
    },
    {
      "epoch": 0.6011080332409973,
      "grad_norm": 4.354103088378906,
      "learning_rate": 2.144328032323848e-05,
      "loss": 0.477,
      "step": 217
    },
    {
      "epoch": 0.6038781163434903,
      "grad_norm": 4.329044342041016,
      "learning_rate": 2.1187722251364455e-05,
      "loss": 0.4553,
      "step": 218
    },
    {
      "epoch": 0.6066481994459834,
      "grad_norm": 4.295444488525391,
      "learning_rate": 2.093286214770454e-05,
      "loss": 0.4762,
      "step": 219
    },
    {
      "epoch": 0.6094182825484764,
      "grad_norm": 3.934124231338501,
      "learning_rate": 2.0678720198214814e-05,
      "loss": 0.4415,
      "step": 220
    },
    {
      "epoch": 0.6121883656509696,
      "grad_norm": 4.6317973136901855,
      "learning_rate": 2.04253165319706e-05,
      "loss": 0.4382,
      "step": 221
    },
    {
      "epoch": 0.6149584487534626,
      "grad_norm": 4.10235071182251,
      "learning_rate": 2.0172671219572175e-05,
      "loss": 0.438,
      "step": 222
    },
    {
      "epoch": 0.6177285318559557,
      "grad_norm": 4.445311546325684,
      "learning_rate": 1.9920804271555153e-05,
      "loss": 0.4442,
      "step": 223
    },
    {
      "epoch": 0.6204986149584487,
      "grad_norm": 4.412539958953857,
      "learning_rate": 1.9669735636805483e-05,
      "loss": 0.4343,
      "step": 224
    },
    {
      "epoch": 0.6232686980609419,
      "grad_norm": 4.816469669342041,
      "learning_rate": 1.941948520097948e-05,
      "loss": 0.4482,
      "step": 225
    },
    {
      "epoch": 0.6260387811634349,
      "grad_norm": 4.047571182250977,
      "learning_rate": 1.9170072784928764e-05,
      "loss": 0.4466,
      "step": 226
    },
    {
      "epoch": 0.628808864265928,
      "grad_norm": 4.237831115722656,
      "learning_rate": 1.8921518143130422e-05,
      "loss": 0.4543,
      "step": 227
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 3.3804972171783447,
      "learning_rate": 1.867384096212226e-05,
      "loss": 0.4177,
      "step": 228
    },
    {
      "epoch": 0.6343490304709142,
      "grad_norm": 4.166910648345947,
      "learning_rate": 1.842706085894371e-05,
      "loss": 0.4292,
      "step": 229
    },
    {
      "epoch": 0.6371191135734072,
      "grad_norm": 4.192974090576172,
      "learning_rate": 1.818119737958192e-05,
      "loss": 0.4366,
      "step": 230
    },
    {
      "epoch": 0.6398891966759003,
      "grad_norm": 4.138956069946289,
      "learning_rate": 1.7936269997423754e-05,
      "loss": 0.4494,
      "step": 231
    },
    {
      "epoch": 0.6426592797783933,
      "grad_norm": 3.8270256519317627,
      "learning_rate": 1.7692298111713357e-05,
      "loss": 0.4416,
      "step": 232
    },
    {
      "epoch": 0.6454293628808865,
      "grad_norm": 4.01252555847168,
      "learning_rate": 1.744930104601566e-05,
      "loss": 0.4533,
      "step": 233
    },
    {
      "epoch": 0.6481994459833795,
      "grad_norm": 4.926167964935303,
      "learning_rate": 1.7207298046685913e-05,
      "loss": 0.446,
      "step": 234
    },
    {
      "epoch": 0.6509695290858726,
      "grad_norm": 4.479677200317383,
      "learning_rate": 1.696630828134525e-05,
      "loss": 0.4389,
      "step": 235
    },
    {
      "epoch": 0.6537396121883656,
      "grad_norm": 4.288498878479004,
      "learning_rate": 1.672635083736255e-05,
      "loss": 0.4411,
      "step": 236
    },
    {
      "epoch": 0.6565096952908587,
      "grad_norm": 3.79034423828125,
      "learning_rate": 1.6487444720342657e-05,
      "loss": 0.4244,
      "step": 237
    },
    {
      "epoch": 0.6592797783933518,
      "grad_norm": 4.163748264312744,
      "learning_rate": 1.6249608852621057e-05,
      "loss": 0.4365,
      "step": 238
    },
    {
      "epoch": 0.6620498614958449,
      "grad_norm": 4.326589584350586,
      "learning_rate": 1.601286207176511e-05,
      "loss": 0.4338,
      "step": 239
    },
    {
      "epoch": 0.6648199445983379,
      "grad_norm": 4.408019065856934,
      "learning_rate": 1.5777223129082104e-05,
      "loss": 0.4274,
      "step": 240
    },
    {
      "epoch": 0.667590027700831,
      "grad_norm": 4.198231220245361,
      "learning_rate": 1.554271068813402e-05,
      "loss": 0.4363,
      "step": 241
    },
    {
      "epoch": 0.6703601108033241,
      "grad_norm": 4.108282566070557,
      "learning_rate": 1.5309343323259316e-05,
      "loss": 0.4254,
      "step": 242
    },
    {
      "epoch": 0.6731301939058172,
      "grad_norm": 4.349982738494873,
      "learning_rate": 1.5077139518101804e-05,
      "loss": 0.4222,
      "step": 243
    },
    {
      "epoch": 0.6759002770083102,
      "grad_norm": 3.9607222080230713,
      "learning_rate": 1.4846117664146597e-05,
      "loss": 0.426,
      "step": 244
    },
    {
      "epoch": 0.6786703601108033,
      "grad_norm": 4.317805290222168,
      "learning_rate": 1.461629605926351e-05,
      "loss": 0.4356,
      "step": 245
    },
    {
      "epoch": 0.6814404432132964,
      "grad_norm": 4.254571437835693,
      "learning_rate": 1.4387692906257705e-05,
      "loss": 0.4234,
      "step": 246
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 4.319187641143799,
      "learning_rate": 1.4160326311428072e-05,
      "loss": 0.419,
      "step": 247
    },
    {
      "epoch": 0.6869806094182825,
      "grad_norm": 4.034308910369873,
      "learning_rate": 1.3934214283132973e-05,
      "loss": 0.4349,
      "step": 248
    },
    {
      "epoch": 0.6897506925207756,
      "grad_norm": 4.316988468170166,
      "learning_rate": 1.3709374730364091e-05,
      "loss": 0.4257,
      "step": 249
    },
    {
      "epoch": 0.6925207756232687,
      "grad_norm": 3.927248001098633,
      "learning_rate": 1.3485825461327812e-05,
      "loss": 0.4034,
      "step": 250
    },
    {
      "epoch": 0.6952908587257618,
      "grad_norm": 4.411158561706543,
      "learning_rate": 1.3263584182034849e-05,
      "loss": 0.4296,
      "step": 251
    },
    {
      "epoch": 0.6980609418282548,
      "grad_norm": 4.725271224975586,
      "learning_rate": 1.3042668494897804e-05,
      "loss": 0.4591,
      "step": 252
    },
    {
      "epoch": 0.7008310249307479,
      "grad_norm": 4.10283899307251,
      "learning_rate": 1.2823095897336969e-05,
      "loss": 0.4464,
      "step": 253
    },
    {
      "epoch": 0.703601108033241,
      "grad_norm": 3.92057728767395,
      "learning_rate": 1.2604883780394499e-05,
      "loss": 0.431,
      "step": 254
    },
    {
      "epoch": 0.7063711911357341,
      "grad_norm": 4.471419334411621,
      "learning_rate": 1.2388049427356943e-05,
      "loss": 0.4082,
      "step": 255
    },
    {
      "epoch": 0.7091412742382271,
      "grad_norm": 4.321521282196045,
      "learning_rate": 1.2172610012386376e-05,
      "loss": 0.4472,
      "step": 256
    },
    {
      "epoch": 0.7119113573407202,
      "grad_norm": 4.131201267242432,
      "learning_rate": 1.1958582599160073e-05,
      "loss": 0.427,
      "step": 257
    },
    {
      "epoch": 0.7146814404432132,
      "grad_norm": 3.820596933364868,
      "learning_rate": 1.1745984139519071e-05,
      "loss": 0.4138,
      "step": 258
    },
    {
      "epoch": 0.7174515235457064,
      "grad_norm": 4.444502353668213,
      "learning_rate": 1.153483147212544e-05,
      "loss": 0.4109,
      "step": 259
    },
    {
      "epoch": 0.7202216066481995,
      "grad_norm": 4.224623203277588,
      "learning_rate": 1.1325141321128649e-05,
      "loss": 0.4202,
      "step": 260
    },
    {
      "epoch": 0.7229916897506925,
      "grad_norm": 3.9233286380767822,
      "learning_rate": 1.11169302948409e-05,
      "loss": 0.4377,
      "step": 261
    },
    {
      "epoch": 0.7257617728531855,
      "grad_norm": 4.564990997314453,
      "learning_rate": 1.0910214884421735e-05,
      "loss": 0.4184,
      "step": 262
    },
    {
      "epoch": 0.7285318559556787,
      "grad_norm": 3.8863961696624756,
      "learning_rate": 1.0705011462571842e-05,
      "loss": 0.4326,
      "step": 263
    },
    {
      "epoch": 0.7313019390581718,
      "grad_norm": 3.9038331508636475,
      "learning_rate": 1.050133628223623e-05,
      "loss": 0.4407,
      "step": 264
    },
    {
      "epoch": 0.7340720221606648,
      "grad_norm": 4.037172317504883,
      "learning_rate": 1.0299205475317022e-05,
      "loss": 0.4184,
      "step": 265
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 4.162524700164795,
      "learning_rate": 1.0098635051395648e-05,
      "loss": 0.437,
      "step": 266
    },
    {
      "epoch": 0.739612188365651,
      "grad_norm": 3.8530848026275635,
      "learning_rate": 9.899640896464874e-06,
      "loss": 0.4172,
      "step": 267
    },
    {
      "epoch": 0.7423822714681441,
      "grad_norm": 4.2646284103393555,
      "learning_rate": 9.70223877167055e-06,
      "loss": 0.4331,
      "step": 268
    },
    {
      "epoch": 0.7451523545706371,
      "grad_norm": 4.318920135498047,
      "learning_rate": 9.5064443120633e-06,
      "loss": 0.4275,
      "step": 269
    },
    {
      "epoch": 0.7479224376731302,
      "grad_norm": 3.900696277618408,
      "learning_rate": 9.312273025360084e-06,
      "loss": 0.4575,
      "step": 270
    },
    {
      "epoch": 0.7506925207756233,
      "grad_norm": 4.387091159820557,
      "learning_rate": 9.119740290716017e-06,
      "loss": 0.4565,
      "step": 271
    },
    {
      "epoch": 0.7534626038781164,
      "grad_norm": 4.237148761749268,
      "learning_rate": 8.928861357506193e-06,
      "loss": 0.4242,
      "step": 272
    },
    {
      "epoch": 0.7562326869806094,
      "grad_norm": 4.315036773681641,
      "learning_rate": 8.739651344117926e-06,
      "loss": 0.4222,
      "step": 273
    },
    {
      "epoch": 0.7590027700831025,
      "grad_norm": 4.285457134246826,
      "learning_rate": 8.552125236753309e-06,
      "loss": 0.4275,
      "step": 274
    },
    {
      "epoch": 0.7617728531855956,
      "grad_norm": 3.956871747970581,
      "learning_rate": 8.366297888242205e-06,
      "loss": 0.4026,
      "step": 275
    },
    {
      "epoch": 0.7645429362880887,
      "grad_norm": 4.065526008605957,
      "learning_rate": 8.18218401686591e-06,
      "loss": 0.4079,
      "step": 276
    },
    {
      "epoch": 0.7673130193905817,
      "grad_norm": 3.7003588676452637,
      "learning_rate": 7.999798205191323e-06,
      "loss": 0.4166,
      "step": 277
    },
    {
      "epoch": 0.7700831024930748,
      "grad_norm": 4.1025800704956055,
      "learning_rate": 7.819154898916042e-06,
      "loss": 0.4222,
      "step": 278
    },
    {
      "epoch": 0.7728531855955678,
      "grad_norm": 4.2863593101501465,
      "learning_rate": 7.640268405724095e-06,
      "loss": 0.4072,
      "step": 279
    },
    {
      "epoch": 0.775623268698061,
      "grad_norm": 3.834885358810425,
      "learning_rate": 7.463152894152826e-06,
      "loss": 0.461,
      "step": 280
    },
    {
      "epoch": 0.778393351800554,
      "grad_norm": 3.8217790126800537,
      "learning_rate": 7.287822392470604e-06,
      "loss": 0.4323,
      "step": 281
    },
    {
      "epoch": 0.7811634349030471,
      "grad_norm": 3.9972996711730957,
      "learning_rate": 7.114290787565773e-06,
      "loss": 0.4443,
      "step": 282
    },
    {
      "epoch": 0.7839335180055401,
      "grad_norm": 4.411800861358643,
      "learning_rate": 6.942571823846755e-06,
      "loss": 0.4323,
      "step": 283
    },
    {
      "epoch": 0.7867036011080333,
      "grad_norm": 4.5063300132751465,
      "learning_rate": 6.772679102153395e-06,
      "loss": 0.4228,
      "step": 284
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 4.130795955657959,
      "learning_rate": 6.6046260786797585e-06,
      "loss": 0.4197,
      "step": 285
    },
    {
      "epoch": 0.7922437673130194,
      "grad_norm": 4.331823348999023,
      "learning_rate": 6.438426063908331e-06,
      "loss": 0.4269,
      "step": 286
    },
    {
      "epoch": 0.7950138504155124,
      "grad_norm": 3.887230396270752,
      "learning_rate": 6.2740922215558e-06,
      "loss": 0.4402,
      "step": 287
    },
    {
      "epoch": 0.7977839335180056,
      "grad_norm": 4.196553707122803,
      "learning_rate": 6.11163756753037e-06,
      "loss": 0.4417,
      "step": 288
    },
    {
      "epoch": 0.8005540166204986,
      "grad_norm": 4.163293361663818,
      "learning_rate": 5.951074968900941e-06,
      "loss": 0.4216,
      "step": 289
    },
    {
      "epoch": 0.8033240997229917,
      "grad_norm": 4.441506385803223,
      "learning_rate": 5.792417142877902e-06,
      "loss": 0.415,
      "step": 290
    },
    {
      "epoch": 0.8060941828254847,
      "grad_norm": 4.561277866363525,
      "learning_rate": 5.635676655805913e-06,
      "loss": 0.4089,
      "step": 291
    },
    {
      "epoch": 0.8088642659279779,
      "grad_norm": 4.779412746429443,
      "learning_rate": 5.480865922168612e-06,
      "loss": 0.4076,
      "step": 292
    },
    {
      "epoch": 0.8116343490304709,
      "grad_norm": 4.437448024749756,
      "learning_rate": 5.3279972036053005e-06,
      "loss": 0.3955,
      "step": 293
    },
    {
      "epoch": 0.814404432132964,
      "grad_norm": 4.1041035652160645,
      "learning_rate": 5.177082607939821e-06,
      "loss": 0.4441,
      "step": 294
    },
    {
      "epoch": 0.817174515235457,
      "grad_norm": 4.502457141876221,
      "learning_rate": 5.028134088221509e-06,
      "loss": 0.4097,
      "step": 295
    },
    {
      "epoch": 0.8199445983379502,
      "grad_norm": 4.109027862548828,
      "learning_rate": 4.881163441778526e-06,
      "loss": 0.4292,
      "step": 296
    },
    {
      "epoch": 0.8227146814404432,
      "grad_norm": 4.413381099700928,
      "learning_rate": 4.736182309283401e-06,
      "loss": 0.4361,
      "step": 297
    },
    {
      "epoch": 0.8254847645429363,
      "grad_norm": 4.096319198608398,
      "learning_rate": 4.593202173831094e-06,
      "loss": 0.4059,
      "step": 298
    },
    {
      "epoch": 0.8282548476454293,
      "grad_norm": 4.19406270980835,
      "learning_rate": 4.452234360029446e-06,
      "loss": 0.4164,
      "step": 299
    },
    {
      "epoch": 0.8310249307479224,
      "grad_norm": 3.696166753768921,
      "learning_rate": 4.3132900331022695e-06,
      "loss": 0.4374,
      "step": 300
    },
    {
      "epoch": 0.8337950138504155,
      "grad_norm": 4.229091167449951,
      "learning_rate": 4.176380198004969e-06,
      "loss": 0.4347,
      "step": 301
    },
    {
      "epoch": 0.8365650969529086,
      "grad_norm": 4.072672367095947,
      "learning_rate": 4.041515698552953e-06,
      "loss": 0.4268,
      "step": 302
    },
    {
      "epoch": 0.8393351800554016,
      "grad_norm": 4.1049981117248535,
      "learning_rate": 3.908707216562705e-06,
      "loss": 0.4186,
      "step": 303
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 4.255549907684326,
      "learning_rate": 3.7779652710057797e-06,
      "loss": 0.4232,
      "step": 304
    },
    {
      "epoch": 0.8448753462603878,
      "grad_norm": 4.021961688995361,
      "learning_rate": 3.649300217175661e-06,
      "loss": 0.408,
      "step": 305
    },
    {
      "epoch": 0.8476454293628809,
      "grad_norm": 4.191747665405273,
      "learning_rate": 3.522722245867548e-06,
      "loss": 0.4251,
      "step": 306
    },
    {
      "epoch": 0.850415512465374,
      "grad_norm": 4.252630233764648,
      "learning_rate": 3.398241382571237e-06,
      "loss": 0.395,
      "step": 307
    },
    {
      "epoch": 0.853185595567867,
      "grad_norm": 3.839334726333618,
      "learning_rate": 3.2758674866770278e-06,
      "loss": 0.4218,
      "step": 308
    },
    {
      "epoch": 0.8559556786703602,
      "grad_norm": 4.2264838218688965,
      "learning_rate": 3.155610250694866e-06,
      "loss": 0.4468,
      "step": 309
    },
    {
      "epoch": 0.8587257617728532,
      "grad_norm": 3.8925042152404785,
      "learning_rate": 3.03747919948659e-06,
      "loss": 0.4315,
      "step": 310
    },
    {
      "epoch": 0.8614958448753463,
      "grad_norm": 4.53446102142334,
      "learning_rate": 2.9214836895115947e-06,
      "loss": 0.4353,
      "step": 311
    },
    {
      "epoch": 0.8642659279778393,
      "grad_norm": 4.308258533477783,
      "learning_rate": 2.8076329080857277e-06,
      "loss": 0.4441,
      "step": 312
    },
    {
      "epoch": 0.8670360110803325,
      "grad_norm": 4.659024715423584,
      "learning_rate": 2.6959358726535964e-06,
      "loss": 0.4165,
      "step": 313
    },
    {
      "epoch": 0.8698060941828255,
      "grad_norm": 4.282739639282227,
      "learning_rate": 2.5864014300743964e-06,
      "loss": 0.3965,
      "step": 314
    },
    {
      "epoch": 0.8725761772853186,
      "grad_norm": 4.223807334899902,
      "learning_rate": 2.479038255921171e-06,
      "loss": 0.405,
      "step": 315
    },
    {
      "epoch": 0.8753462603878116,
      "grad_norm": 4.042827129364014,
      "learning_rate": 2.373854853793671e-06,
      "loss": 0.4223,
      "step": 316
    },
    {
      "epoch": 0.8781163434903048,
      "grad_norm": 3.7920212745666504,
      "learning_rate": 2.2708595546448442e-06,
      "loss": 0.4104,
      "step": 317
    },
    {
      "epoch": 0.8808864265927978,
      "grad_norm": 4.02117395401001,
      "learning_rate": 2.170060516121012e-06,
      "loss": 0.401,
      "step": 318
    },
    {
      "epoch": 0.8836565096952909,
      "grad_norm": 4.442043304443359,
      "learning_rate": 2.071465721915694e-06,
      "loss": 0.4345,
      "step": 319
    },
    {
      "epoch": 0.8864265927977839,
      "grad_norm": 4.755099773406982,
      "learning_rate": 1.9750829811373307e-06,
      "loss": 0.3976,
      "step": 320
    },
    {
      "epoch": 0.889196675900277,
      "grad_norm": 4.009551525115967,
      "learning_rate": 1.8809199276907196e-06,
      "loss": 0.4175,
      "step": 321
    },
    {
      "epoch": 0.8919667590027701,
      "grad_norm": 3.991119623184204,
      "learning_rate": 1.7889840196724083e-06,
      "loss": 0.4082,
      "step": 322
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 3.973052740097046,
      "learning_rate": 1.6992825387799816e-06,
      "loss": 0.4091,
      "step": 323
    },
    {
      "epoch": 0.8975069252077562,
      "grad_norm": 4.348862171173096,
      "learning_rate": 1.6118225897352967e-06,
      "loss": 0.4279,
      "step": 324
    },
    {
      "epoch": 0.9002770083102493,
      "grad_norm": 4.407948017120361,
      "learning_rate": 1.5266110997218009e-06,
      "loss": 0.4195,
      "step": 325
    },
    {
      "epoch": 0.9030470914127424,
      "grad_norm": 4.2243428230285645,
      "learning_rate": 1.4436548178358188e-06,
      "loss": 0.4193,
      "step": 326
    },
    {
      "epoch": 0.9058171745152355,
      "grad_norm": 4.009667873382568,
      "learning_rate": 1.3629603145520554e-06,
      "loss": 0.3961,
      "step": 327
    },
    {
      "epoch": 0.9085872576177285,
      "grad_norm": 4.0680060386657715,
      "learning_rate": 1.2845339812031397e-06,
      "loss": 0.4235,
      "step": 328
    },
    {
      "epoch": 0.9113573407202216,
      "grad_norm": 3.899540901184082,
      "learning_rate": 1.208382029473426e-06,
      "loss": 0.4234,
      "step": 329
    },
    {
      "epoch": 0.9141274238227147,
      "grad_norm": 4.105435848236084,
      "learning_rate": 1.1345104909069947e-06,
      "loss": 0.4522,
      "step": 330
    },
    {
      "epoch": 0.9168975069252078,
      "grad_norm": 4.221428871154785,
      "learning_rate": 1.062925216429953e-06,
      "loss": 0.4117,
      "step": 331
    },
    {
      "epoch": 0.9196675900277008,
      "grad_norm": 4.16056489944458,
      "learning_rate": 9.936318758869877e-07,
      "loss": 0.4211,
      "step": 332
    },
    {
      "epoch": 0.9224376731301939,
      "grad_norm": 4.145576000213623,
      "learning_rate": 9.26635957592299e-07,
      "loss": 0.4086,
      "step": 333
    },
    {
      "epoch": 0.925207756232687,
      "grad_norm": 4.158425807952881,
      "learning_rate": 8.619427678949165e-07,
      "loss": 0.4123,
      "step": 334
    },
    {
      "epoch": 0.9279778393351801,
      "grad_norm": 3.9155080318450928,
      "learning_rate": 7.995574307583975e-07,
      "loss": 0.4271,
      "step": 335
    },
    {
      "epoch": 0.9307479224376731,
      "grad_norm": 4.067850589752197,
      "learning_rate": 7.394848873550064e-07,
      "loss": 0.4156,
      "step": 336
    },
    {
      "epoch": 0.9335180055401662,
      "grad_norm": 4.133598804473877,
      "learning_rate": 6.817298956743356e-07,
      "loss": 0.4449,
      "step": 337
    },
    {
      "epoch": 0.9362880886426593,
      "grad_norm": 4.627798080444336,
      "learning_rate": 6.262970301464632e-07,
      "loss": 0.4143,
      "step": 338
    },
    {
      "epoch": 0.9390581717451524,
      "grad_norm": 4.110208034515381,
      "learning_rate": 5.731906812796395e-07,
      "loss": 0.4238,
      "step": 339
    },
    {
      "epoch": 0.9418282548476454,
      "grad_norm": 4.190328598022461,
      "learning_rate": 5.22415055312545e-07,
      "loss": 0.3903,
      "step": 340
    },
    {
      "epoch": 0.9445983379501385,
      "grad_norm": 4.284666538238525,
      "learning_rate": 4.7397417388111766e-07,
      "loss": 0.3984,
      "step": 341
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 4.148164749145508,
      "learning_rate": 4.278718737000564e-07,
      "loss": 0.4286,
      "step": 342
    },
    {
      "epoch": 0.9501385041551247,
      "grad_norm": 4.629441261291504,
      "learning_rate": 3.8411180625890886e-07,
      "loss": 0.409,
      "step": 343
    },
    {
      "epoch": 0.9529085872576177,
      "grad_norm": 4.130425453186035,
      "learning_rate": 3.4269743753287175e-07,
      "loss": 0.423,
      "step": 344
    },
    {
      "epoch": 0.9556786703601108,
      "grad_norm": 3.745643377304077,
      "learning_rate": 3.03632047708281e-07,
      "loss": 0.4262,
      "step": 345
    },
    {
      "epoch": 0.9584487534626038,
      "grad_norm": 4.220255374908447,
      "learning_rate": 2.669187309227794e-07,
      "loss": 0.4183,
      "step": 346
    },
    {
      "epoch": 0.961218836565097,
      "grad_norm": 3.9583799839019775,
      "learning_rate": 2.3256039502027592e-07,
      "loss": 0.4221,
      "step": 347
    },
    {
      "epoch": 0.96398891966759,
      "grad_norm": 3.985778331756592,
      "learning_rate": 2.005597613206167e-07,
      "loss": 0.4088,
      "step": 348
    },
    {
      "epoch": 0.9667590027700831,
      "grad_norm": 4.007012844085693,
      "learning_rate": 1.7091936440405742e-07,
      "loss": 0.4379,
      "step": 349
    },
    {
      "epoch": 0.9695290858725761,
      "grad_norm": 4.563694000244141,
      "learning_rate": 1.436415519104972e-07,
      "loss": 0.4149,
      "step": 350
    },
    {
      "epoch": 0.9722991689750693,
      "grad_norm": 4.10459041595459,
      "learning_rate": 1.1872848435356076e-07,
      "loss": 0.4297,
      "step": 351
    },
    {
      "epoch": 0.9750692520775623,
      "grad_norm": 4.2543487548828125,
      "learning_rate": 9.618213494945183e-08,
      "loss": 0.3925,
      "step": 352
    },
    {
      "epoch": 0.9778393351800554,
      "grad_norm": 4.755935192108154,
      "learning_rate": 7.600428946067828e-08,
      "loss": 0.4242,
      "step": 353
    },
    {
      "epoch": 0.9806094182825484,
      "grad_norm": 4.033304214477539,
      "learning_rate": 5.819654605461522e-08,
      "loss": 0.4195,
      "step": 354
    },
    {
      "epoch": 0.9833795013850416,
      "grad_norm": 3.9951982498168945,
      "learning_rate": 4.276031517691292e-08,
      "loss": 0.4187,
      "step": 355
    },
    {
      "epoch": 0.9861495844875346,
      "grad_norm": 4.105679988861084,
      "learning_rate": 2.9696819439802846e-08,
      "loss": 0.4328,
      "step": 356
    },
    {
      "epoch": 0.9889196675900277,
      "grad_norm": 4.221299648284912,
      "learning_rate": 1.900709352524177e-08,
      "loss": 0.4281,
      "step": 357
    },
    {
      "epoch": 0.9916897506925207,
      "grad_norm": 4.06874418258667,
      "learning_rate": 1.0691984102964014e-08,
      "loss": 0.4126,
      "step": 358
    },
    {
      "epoch": 0.9944598337950139,
      "grad_norm": 4.161386489868164,
      "learning_rate": 4.7521497634317454e-09,
      "loss": 0.39,
      "step": 359
    },
    {
      "epoch": 0.997229916897507,
      "grad_norm": 4.167057991027832,
      "learning_rate": 1.1880609656733653e-09,
      "loss": 0.4371,
      "step": 360
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.841407537460327,
      "learning_rate": 0.0,
      "loss": 0.4198,
      "step": 361
    }
  ],
  "logging_steps": 1,
  "max_steps": 361,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.230029212895805e+17,
  "train_batch_size": 9,
  "trial_name": null,
  "trial_params": null
}