{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 116,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017241379310344827,
      "grad_norm": 1.34375,
      "learning_rate": 0.0,
      "loss": 2.5553,
      "step": 1
    },
    {
      "epoch": 0.034482758620689655,
      "grad_norm": 1.359375,
      "learning_rate": 1.25e-06,
      "loss": 2.5516,
      "step": 2
    },
    {
      "epoch": 0.05172413793103448,
      "grad_norm": 1.25,
      "learning_rate": 2.5e-06,
      "loss": 2.5156,
      "step": 3
    },
    {
      "epoch": 0.06896551724137931,
      "grad_norm": 1.3125,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 2.5199,
      "step": 4
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 1.1640625,
      "learning_rate": 5e-06,
      "loss": 2.5059,
      "step": 5
    },
    {
      "epoch": 0.10344827586206896,
      "grad_norm": 1.2109375,
      "learning_rate": 6.25e-06,
      "loss": 2.4909,
      "step": 6
    },
    {
      "epoch": 0.1206896551724138,
      "grad_norm": 1.03125,
      "learning_rate": 7.500000000000001e-06,
      "loss": 2.5399,
      "step": 7
    },
    {
      "epoch": 0.13793103448275862,
      "grad_norm": 0.91796875,
      "learning_rate": 8.750000000000001e-06,
      "loss": 2.5355,
      "step": 8
    },
    {
      "epoch": 0.15517241379310345,
      "grad_norm": 0.828125,
      "learning_rate": 1e-05,
      "loss": 2.5359,
      "step": 9
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.70703125,
      "learning_rate": 9.99910461334869e-06,
      "loss": 2.5257,
      "step": 10
    },
    {
      "epoch": 0.1896551724137931,
      "grad_norm": 0.6875,
      "learning_rate": 9.996418774081658e-06,
      "loss": 2.5138,
      "step": 11
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 0.6953125,
      "learning_rate": 9.991943444144758e-06,
      "loss": 2.5139,
      "step": 12
    },
    {
      "epoch": 0.22413793103448276,
      "grad_norm": 0.87109375,
      "learning_rate": 9.985680226398261e-06,
      "loss": 2.4983,
      "step": 13
    },
    {
      "epoch": 0.2413793103448276,
      "grad_norm": 0.69921875,
      "learning_rate": 9.977631364042796e-06,
      "loss": 2.5036,
      "step": 14
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.703125,
      "learning_rate": 9.967799739815925e-06,
      "loss": 2.5121,
      "step": 15
    },
    {
      "epoch": 0.27586206896551724,
      "grad_norm": 0.703125,
      "learning_rate": 9.956188874959686e-06,
      "loss": 2.539,
      "step": 16
    },
    {
      "epoch": 0.29310344827586204,
      "grad_norm": 0.70703125,
      "learning_rate": 9.942802927959444e-06,
      "loss": 2.5107,
      "step": 17
    },
    {
      "epoch": 0.3103448275862069,
      "grad_norm": 0.67578125,
      "learning_rate": 9.927646693054498e-06,
      "loss": 2.4907,
      "step": 18
    },
    {
      "epoch": 0.3275862068965517,
      "grad_norm": 0.66015625,
      "learning_rate": 9.910725598521014e-06,
      "loss": 2.4836,
      "step": 19
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.66796875,
      "learning_rate": 9.892045704727864e-06,
      "loss": 2.494,
      "step": 20
    },
    {
      "epoch": 0.3620689655172414,
      "grad_norm": 0.61328125,
      "learning_rate": 9.871613701966067e-06,
      "loss": 2.4733,
      "step": 21
    },
    {
      "epoch": 0.3793103448275862,
      "grad_norm": 0.66796875,
      "learning_rate": 9.849436908052636e-06,
      "loss": 2.4141,
      "step": 22
    },
    {
      "epoch": 0.39655172413793105,
      "grad_norm": 0.61328125,
      "learning_rate": 9.825523265709667e-06,
      "loss": 2.521,
      "step": 23
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 0.640625,
      "learning_rate": 9.799881339719615e-06,
      "loss": 2.4647,
      "step": 24
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 0.62109375,
      "learning_rate": 9.772520313857777e-06,
      "loss": 2.4892,
      "step": 25
    },
    {
      "epoch": 0.4482758620689655,
      "grad_norm": 0.6015625,
      "learning_rate": 9.743449987603082e-06,
      "loss": 2.4734,
      "step": 26
    },
    {
      "epoch": 0.46551724137931033,
      "grad_norm": 0.609375,
      "learning_rate": 9.712680772628365e-06,
      "loss": 2.5229,
      "step": 27
    },
    {
      "epoch": 0.4827586206896552,
      "grad_norm": 0.6171875,
      "learning_rate": 9.680223689071364e-06,
      "loss": 2.5502,
      "step": 28
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6171875,
      "learning_rate": 9.646090361587828e-06,
      "loss": 2.5008,
      "step": 29
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.63671875,
      "learning_rate": 9.610293015188067e-06,
      "loss": 2.5319,
      "step": 30
    },
    {
      "epoch": 0.5344827586206896,
      "grad_norm": 0.61328125,
      "learning_rate": 9.572844470858537e-06,
      "loss": 2.5004,
      "step": 31
    },
    {
      "epoch": 0.5517241379310345,
      "grad_norm": 0.6171875,
      "learning_rate": 9.533758140969913e-06,
      "loss": 2.4611,
      "step": 32
    },
    {
      "epoch": 0.5689655172413793,
      "grad_norm": 0.625,
      "learning_rate": 9.493048024473413e-06,
      "loss": 2.4883,
      "step": 33
    },
    {
      "epoch": 0.5862068965517241,
      "grad_norm": 0.59765625,
      "learning_rate": 9.450728701886985e-06,
      "loss": 2.5014,
      "step": 34
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 0.59765625,
      "learning_rate": 9.406815330073244e-06,
      "loss": 2.4959,
      "step": 35
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 0.60546875,
      "learning_rate": 9.36132363681097e-06,
      "loss": 2.4901,
      "step": 36
    },
    {
      "epoch": 0.6379310344827587,
      "grad_norm": 0.59765625,
      "learning_rate": 9.314269915162115e-06,
      "loss": 2.54,
      "step": 37
    },
    {
      "epoch": 0.6551724137931034,
      "grad_norm": 0.609375,
      "learning_rate": 9.265671017636384e-06,
      "loss": 2.5077,
      "step": 38
    },
    {
      "epoch": 0.6724137931034483,
      "grad_norm": 0.60546875,
      "learning_rate": 9.215544350155423e-06,
      "loss": 2.5093,
      "step": 39
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.6015625,
      "learning_rate": 9.163907865818806e-06,
      "loss": 2.5134,
      "step": 40
    },
    {
      "epoch": 0.7068965517241379,
      "grad_norm": 0.609375,
      "learning_rate": 9.110780058474052e-06,
      "loss": 2.5651,
      "step": 41
    },
    {
      "epoch": 0.7241379310344828,
      "grad_norm": 0.59375,
      "learning_rate": 9.056179956092961e-06,
      "loss": 2.4615,
      "step": 42
    },
    {
      "epoch": 0.7413793103448276,
      "grad_norm": 0.59375,
      "learning_rate": 9.000127113956673e-06,
      "loss": 2.4763,
      "step": 43
    },
    {
      "epoch": 0.7586206896551724,
      "grad_norm": 0.58203125,
      "learning_rate": 8.94264160765183e-06,
      "loss": 2.4732,
      "step": 44
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.58203125,
      "learning_rate": 8.883744025880429e-06,
      "loss": 2.49,
      "step": 45
    },
    {
      "epoch": 0.7931034482758621,
      "grad_norm": 0.5859375,
      "learning_rate": 8.823455463085873e-06,
      "loss": 2.4944,
      "step": 46
    },
    {
      "epoch": 0.8103448275862069,
      "grad_norm": 0.58203125,
      "learning_rate": 8.761797511897907e-06,
      "loss": 2.5177,
      "step": 47
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 0.6015625,
      "learning_rate": 8.698792255399104e-06,
      "loss": 2.5083,
      "step": 48
    },
    {
      "epoch": 0.8448275862068966,
      "grad_norm": 0.58984375,
      "learning_rate": 8.634462259215719e-06,
      "loss": 2.5096,
      "step": 49
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.60546875,
      "learning_rate": 8.568830563435695e-06,
      "loss": 2.5267,
      "step": 50
    },
    {
      "epoch": 0.8793103448275862,
      "grad_norm": 0.6015625,
      "learning_rate": 8.501920674356755e-06,
      "loss": 2.5041,
      "step": 51
    },
    {
      "epoch": 0.896551724137931,
      "grad_norm": 0.59765625,
      "learning_rate": 8.433756556067506e-06,
      "loss": 2.5081,
      "step": 52
    },
    {
      "epoch": 0.9137931034482759,
      "grad_norm": 0.578125,
      "learning_rate": 8.364362621864595e-06,
      "loss": 2.4329,
      "step": 53
    },
    {
      "epoch": 0.9310344827586207,
      "grad_norm": 0.58984375,
      "learning_rate": 8.29376372550897e-06,
      "loss": 2.4784,
      "step": 54
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 0.60546875,
      "learning_rate": 8.221985152324385e-06,
      "loss": 2.5528,
      "step": 55
    },
    {
      "epoch": 0.9655172413793104,
      "grad_norm": 0.59375,
      "learning_rate": 8.149052610141357e-06,
      "loss": 2.4803,
      "step": 56
    },
    {
      "epoch": 0.9827586206896551,
      "grad_norm": 0.578125,
      "learning_rate": 8.07499222008977e-06,
      "loss": 2.5519,
      "step": 57
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.625,
      "learning_rate": 7.999830507243478e-06,
      "loss": 3.0347,
      "step": 58
    },
    {
      "epoch": 1.0172413793103448,
      "grad_norm": 0.5703125,
      "learning_rate": 7.923594391120237e-06,
      "loss": 2.504,
      "step": 59
    },
    {
      "epoch": 1.0344827586206897,
      "grad_norm": 0.5859375,
      "learning_rate": 7.846311176040331e-06,
      "loss": 2.4902,
      "step": 60
    },
    {
      "epoch": 1.0517241379310345,
      "grad_norm": 0.578125,
      "learning_rate": 7.768008541347423e-06,
      "loss": 2.4702,
      "step": 61
    },
    {
      "epoch": 1.0689655172413792,
      "grad_norm": 0.578125,
      "learning_rate": 7.688714531495061e-06,
      "loss": 2.467,
      "step": 62
    },
    {
      "epoch": 1.0862068965517242,
      "grad_norm": 0.6015625,
      "learning_rate": 7.608457546002423e-06,
      "loss": 2.4747,
      "step": 63
    },
    {
      "epoch": 1.103448275862069,
      "grad_norm": 0.58203125,
      "learning_rate": 7.527266329282905e-06,
      "loss": 2.4916,
      "step": 64
    },
    {
      "epoch": 1.1206896551724137,
      "grad_norm": 0.5859375,
      "learning_rate": 7.445169960349167e-06,
      "loss": 2.4334,
      "step": 65
    },
    {
      "epoch": 1.1379310344827587,
      "grad_norm": 0.58984375,
      "learning_rate": 7.362197842398355e-06,
      "loss": 2.4852,
      "step": 66
    },
    {
      "epoch": 1.1551724137931034,
      "grad_norm": 0.58984375,
      "learning_rate": 7.278379692281209e-06,
      "loss": 2.5411,
      "step": 67
    },
    {
      "epoch": 1.1724137931034484,
      "grad_norm": 0.57421875,
      "learning_rate": 7.193745529858827e-06,
      "loss": 2.4498,
      "step": 68
    },
    {
      "epoch": 1.1896551724137931,
      "grad_norm": 0.6015625,
      "learning_rate": 7.10832566725092e-06,
      "loss": 2.505,
      "step": 69
    },
    {
      "epoch": 1.206896551724138,
      "grad_norm": 0.59765625,
      "learning_rate": 7.022150697979385e-06,
      "loss": 2.5124,
      "step": 70
    },
    {
      "epoch": 1.2241379310344827,
      "grad_norm": 0.56640625,
      "learning_rate": 6.9352514860110876e-06,
      "loss": 2.4896,
      "step": 71
    },
    {
      "epoch": 1.2413793103448276,
      "grad_norm": 0.58203125,
      "learning_rate": 6.847659154703785e-06,
      "loss": 2.4544,
      "step": 72
    },
    {
      "epoch": 1.2586206896551724,
      "grad_norm": 0.58984375,
      "learning_rate": 6.759405075659165e-06,
      "loss": 2.4906,
      "step": 73
    },
    {
      "epoch": 1.2758620689655173,
      "grad_norm": 0.5703125,
      "learning_rate": 6.6705208574869504e-06,
      "loss": 2.5192,
      "step": 74
    },
    {
      "epoch": 1.293103448275862,
      "grad_norm": 0.57421875,
      "learning_rate": 6.58103833448412e-06,
      "loss": 2.4903,
      "step": 75
    },
    {
      "epoch": 1.3103448275862069,
      "grad_norm": 0.59765625,
      "learning_rate": 6.490989555233328e-06,
      "loss": 2.4906,
      "step": 76
    },
    {
      "epoch": 1.3275862068965516,
      "grad_norm": 0.578125,
      "learning_rate": 6.4004067711245366e-06,
      "loss": 2.53,
      "step": 77
    },
    {
      "epoch": 1.3448275862068966,
      "grad_norm": 0.56640625,
      "learning_rate": 6.309322424804034e-06,
      "loss": 2.5156,
      "step": 78
    },
    {
      "epoch": 1.3620689655172413,
      "grad_norm": 0.5859375,
      "learning_rate": 6.2177691385549595e-06,
      "loss": 2.4715,
      "step": 79
    },
    {
      "epoch": 1.3793103448275863,
      "grad_norm": 0.59375,
      "learning_rate": 6.125779702613471e-06,
      "loss": 2.5168,
      "step": 80
    },
    {
      "epoch": 1.396551724137931,
      "grad_norm": 0.58984375,
      "learning_rate": 6.033387063424765e-06,
      "loss": 2.4368,
      "step": 81
    },
    {
      "epoch": 1.4137931034482758,
      "grad_norm": 0.578125,
      "learning_rate": 5.94062431184317e-06,
      "loss": 2.5114,
      "step": 82
    },
    {
      "epoch": 1.4310344827586206,
      "grad_norm": 0.59375,
      "learning_rate": 5.8475246712804845e-06,
      "loss": 2.5196,
      "step": 83
    },
    {
      "epoch": 1.4482758620689655,
      "grad_norm": 0.59375,
      "learning_rate": 5.7541214858068705e-06,
      "loss": 2.5102,
      "step": 84
    },
    {
      "epoch": 1.4655172413793103,
      "grad_norm": 0.578125,
      "learning_rate": 5.660448208208513e-06,
      "loss": 2.5055,
      "step": 85
    },
    {
      "epoch": 1.4827586206896552,
      "grad_norm": 0.58984375,
      "learning_rate": 5.566538388006351e-06,
      "loss": 2.4588,
      "step": 86
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.5859375,
      "learning_rate": 5.472425659440157e-06,
      "loss": 2.5023,
      "step": 87
    },
    {
      "epoch": 1.5172413793103448,
      "grad_norm": 0.58203125,
      "learning_rate": 5.378143729422285e-06,
      "loss": 2.5304,
      "step": 88
    },
    {
      "epoch": 1.5344827586206895,
      "grad_norm": 0.58203125,
      "learning_rate": 5.2837263654653715e-06,
      "loss": 2.4919,
      "step": 89
    },
    {
      "epoch": 1.5517241379310345,
      "grad_norm": 0.57421875,
      "learning_rate": 5.189207383588353e-06,
      "loss": 2.4797,
      "step": 90
    },
    {
      "epoch": 1.5689655172413794,
      "grad_norm": 0.58203125,
      "learning_rate": 5.094620636205096e-06,
      "loss": 2.4814,
      "step": 91
    },
    {
      "epoch": 1.5862068965517242,
      "grad_norm": 0.58984375,
      "learning_rate": 5e-06,
      "loss": 2.5611,
      "step": 92
    },
    {
      "epoch": 1.603448275862069,
      "grad_norm": 0.59765625,
      "learning_rate": 4.905379363794907e-06,
      "loss": 2.5033,
      "step": 93
    },
    {
      "epoch": 1.6206896551724137,
      "grad_norm": 0.59765625,
      "learning_rate": 4.81079261641165e-06,
      "loss": 2.4482,
      "step": 94
    },
    {
      "epoch": 1.6379310344827587,
      "grad_norm": 0.5859375,
      "learning_rate": 4.71627363453463e-06,
      "loss": 2.4975,
      "step": 95
    },
    {
      "epoch": 1.6551724137931034,
      "grad_norm": 0.58984375,
      "learning_rate": 4.6218562705777185e-06,
      "loss": 2.5358,
      "step": 96
    },
    {
      "epoch": 1.6724137931034484,
      "grad_norm": 0.58984375,
      "learning_rate": 4.527574340559844e-06,
      "loss": 2.4869,
      "step": 97
    },
    {
      "epoch": 1.6896551724137931,
      "grad_norm": 0.56640625,
      "learning_rate": 4.4334616119936516e-06,
      "loss": 2.4609,
      "step": 98
    },
    {
      "epoch": 1.706896551724138,
      "grad_norm": 0.5703125,
      "learning_rate": 4.33955179179149e-06,
      "loss": 2.5028,
      "step": 99
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 0.5703125,
      "learning_rate": 4.245878514193131e-06,
      "loss": 2.4754,
      "step": 100
    },
    {
      "epoch": 1.7413793103448276,
      "grad_norm": 0.59375,
      "learning_rate": 4.152475328719517e-06,
      "loss": 2.4999,
      "step": 101
    },
    {
      "epoch": 1.7586206896551724,
      "grad_norm": 0.5625,
      "learning_rate": 4.059375688156833e-06,
      "loss": 2.4888,
      "step": 102
    },
    {
      "epoch": 1.7758620689655173,
      "grad_norm": 0.5859375,
      "learning_rate": 3.966612936575235e-06,
      "loss": 2.4616,
      "step": 103
    },
    {
      "epoch": 1.793103448275862,
      "grad_norm": 0.578125,
      "learning_rate": 3.87422029738653e-06,
      "loss": 2.5035,
      "step": 104
    },
    {
      "epoch": 1.8103448275862069,
      "grad_norm": 0.578125,
      "learning_rate": 3.782230861445041e-06,
      "loss": 2.4758,
      "step": 105
    },
    {
      "epoch": 1.8275862068965516,
      "grad_norm": 0.5859375,
      "learning_rate": 3.6906775751959667e-06,
      "loss": 2.4702,
      "step": 106
    },
    {
      "epoch": 1.8448275862068966,
      "grad_norm": 0.58203125,
      "learning_rate": 3.5995932288754655e-06,
      "loss": 2.4774,
      "step": 107
    },
    {
      "epoch": 1.8620689655172413,
      "grad_norm": 0.5703125,
      "learning_rate": 3.509010444766674e-06,
      "loss": 2.4498,
      "step": 108
    },
    {
      "epoch": 1.8793103448275863,
      "grad_norm": 0.57421875,
      "learning_rate": 3.4189616655158803e-06,
      "loss": 2.4637,
      "step": 109
    },
    {
      "epoch": 1.896551724137931,
      "grad_norm": 0.59375,
      "learning_rate": 3.3294791425130512e-06,
      "loss": 2.4424,
      "step": 110
    },
    {
      "epoch": 1.9137931034482758,
      "grad_norm": 0.578125,
      "learning_rate": 3.240594924340835e-06,
      "loss": 2.5001,
      "step": 111
    },
    {
      "epoch": 1.9310344827586206,
      "grad_norm": 0.5703125,
      "learning_rate": 3.1523408452962156e-06,
      "loss": 2.4744,
      "step": 112
    },
    {
      "epoch": 1.9482758620689655,
      "grad_norm": 0.578125,
      "learning_rate": 3.0647485139889145e-06,
      "loss": 2.4345,
      "step": 113
    },
    {
      "epoch": 1.9655172413793105,
      "grad_norm": 0.58203125,
      "learning_rate": 2.9778493020206155e-06,
      "loss": 2.4341,
      "step": 114
    },
    {
      "epoch": 1.9827586206896552,
      "grad_norm": 0.578125,
      "learning_rate": 2.89167433274908e-06,
      "loss": 2.4849,
      "step": 115
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.5,
      "learning_rate": 2.806254470141174e-06,
      "loss": 2.672,
      "step": 116
    }
  ],
  "logging_steps": 1,
  "max_steps": 174,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 58,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.175788450912666e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}