{
  "best_metric": 1.257092833518982,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 2.345029239766082,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011695906432748537,
      "grad_norm": 0.7432779669761658,
      "learning_rate": 1e-05,
      "loss": 5.6697,
      "step": 1
    },
    {
      "epoch": 0.011695906432748537,
      "eval_loss": 1.4917855262756348,
      "eval_runtime": 14.065,
      "eval_samples_per_second": 10.238,
      "eval_steps_per_second": 2.56,
      "step": 1
    },
    {
      "epoch": 0.023391812865497075,
      "grad_norm": 0.7208809852600098,
      "learning_rate": 2e-05,
      "loss": 5.5813,
      "step": 2
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 0.7191916108131409,
      "learning_rate": 3e-05,
      "loss": 5.5128,
      "step": 3
    },
    {
      "epoch": 0.04678362573099415,
      "grad_norm": 0.6861053109169006,
      "learning_rate": 4e-05,
      "loss": 5.8694,
      "step": 4
    },
    {
      "epoch": 0.05847953216374269,
      "grad_norm": 0.7745478749275208,
      "learning_rate": 5e-05,
      "loss": 5.3542,
      "step": 5
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 0.6765620112419128,
      "learning_rate": 6e-05,
      "loss": 5.4969,
      "step": 6
    },
    {
      "epoch": 0.08187134502923976,
      "grad_norm": 0.6339160203933716,
      "learning_rate": 7e-05,
      "loss": 6.0554,
      "step": 7
    },
    {
      "epoch": 0.0935672514619883,
      "grad_norm": 0.787830114364624,
      "learning_rate": 8e-05,
      "loss": 6.1915,
      "step": 8
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 1.064342737197876,
      "learning_rate": 9e-05,
      "loss": 6.4802,
      "step": 9
    },
    {
      "epoch": 0.11695906432748537,
      "grad_norm": 0.8624706864356995,
      "learning_rate": 0.0001,
      "loss": 5.7992,
      "step": 10
    },
    {
      "epoch": 0.1286549707602339,
      "grad_norm": 0.6974483132362366,
      "learning_rate": 9.999316524962345e-05,
      "loss": 5.3927,
      "step": 11
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 0.8129259347915649,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.5676,
      "step": 12
    },
    {
      "epoch": 0.15204678362573099,
      "grad_norm": 0.7599148154258728,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.6346,
      "step": 13
    },
    {
      "epoch": 0.16374269005847952,
      "grad_norm": 0.8230239152908325,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.4399,
      "step": 14
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 0.6070770025253296,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.6014,
      "step": 15
    },
    {
      "epoch": 0.1871345029239766,
      "grad_norm": 0.6046319007873535,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.5709,
      "step": 16
    },
    {
      "epoch": 0.19883040935672514,
      "grad_norm": 0.5819116234779358,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.7219,
      "step": 17
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.6522576212882996,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.4101,
      "step": 18
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 1.0478955507278442,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.244,
      "step": 19
    },
    {
      "epoch": 0.23391812865497075,
      "grad_norm": 0.602959930896759,
      "learning_rate": 9.931806517013612e-05,
      "loss": 4.7152,
      "step": 20
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 0.6481666564941406,
      "learning_rate": 9.917525374361912e-05,
      "loss": 4.919,
      "step": 21
    },
    {
      "epoch": 0.2573099415204678,
      "grad_norm": 0.5746079683303833,
      "learning_rate": 9.901899829374047e-05,
      "loss": 4.9853,
      "step": 22
    },
    {
      "epoch": 0.26900584795321636,
      "grad_norm": 0.5986230969429016,
      "learning_rate": 9.884934153917997e-05,
      "loss": 4.7646,
      "step": 23
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 0.51701420545578,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.0542,
      "step": 24
    },
    {
      "epoch": 0.29239766081871343,
      "grad_norm": 0.4578527510166168,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.2084,
      "step": 25
    },
    {
      "epoch": 0.30409356725146197,
      "grad_norm": 0.5963104367256165,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.1664,
      "step": 26
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 0.4963514804840088,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.229,
      "step": 27
    },
    {
      "epoch": 0.32748538011695905,
      "grad_norm": 0.7828831076622009,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.2419,
      "step": 28
    },
    {
      "epoch": 0.3391812865497076,
      "grad_norm": 0.5033621788024902,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.4469,
      "step": 29
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 0.47355058789253235,
      "learning_rate": 9.729086208503174e-05,
      "loss": 4.9648,
      "step": 30
    },
    {
      "epoch": 0.36257309941520466,
      "grad_norm": 0.515572190284729,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.4122,
      "step": 31
    },
    {
      "epoch": 0.3742690058479532,
      "grad_norm": 0.5490233302116394,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.3553,
      "step": 32
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 0.5334340333938599,
      "learning_rate": 9.642770192448536e-05,
      "loss": 4.8582,
      "step": 33
    },
    {
      "epoch": 0.39766081871345027,
      "grad_norm": 0.6100642681121826,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.327,
      "step": 34
    },
    {
      "epoch": 0.4093567251461988,
      "grad_norm": 0.5442097187042236,
      "learning_rate": 9.578866633275288e-05,
      "loss": 5.2261,
      "step": 35
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.5721125602722168,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.1833,
      "step": 36
    },
    {
      "epoch": 0.4327485380116959,
      "grad_norm": 0.6107903718948364,
      "learning_rate": 9.509956150664796e-05,
      "loss": 5.1851,
      "step": 37
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.6423254013061523,
      "learning_rate": 9.473646649103818e-05,
      "loss": 4.7939,
      "step": 38
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 0.9591628909111023,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.3958,
      "step": 39
    },
    {
      "epoch": 0.4678362573099415,
      "grad_norm": 0.5825883746147156,
      "learning_rate": 9.397368756032445e-05,
      "loss": 4.7548,
      "step": 40
    },
    {
      "epoch": 0.47953216374269003,
      "grad_norm": 0.622710108757019,
      "learning_rate": 9.357421218136386e-05,
      "loss": 5.1341,
      "step": 41
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 0.6731835007667542,
      "learning_rate": 9.316282404787871e-05,
      "loss": 5.2531,
      "step": 42
    },
    {
      "epoch": 0.5029239766081871,
      "grad_norm": 0.5775891542434692,
      "learning_rate": 9.273963562927695e-05,
      "loss": 4.906,
      "step": 43
    },
    {
      "epoch": 0.5146198830409356,
      "grad_norm": 0.5359857082366943,
      "learning_rate": 9.230476262104677e-05,
      "loss": 5.0036,
      "step": 44
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.5207698345184326,
      "learning_rate": 9.185832391312644e-05,
      "loss": 4.7729,
      "step": 45
    },
    {
      "epoch": 0.5380116959064327,
      "grad_norm": 0.5065916776657104,
      "learning_rate": 9.140044155740101e-05,
      "loss": 5.2115,
      "step": 46
    },
    {
      "epoch": 0.5497076023391813,
      "grad_norm": 0.4974070191383362,
      "learning_rate": 9.093124073433463e-05,
      "loss": 5.0765,
      "step": 47
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.8208938837051392,
      "learning_rate": 9.045084971874738e-05,
      "loss": 5.286,
      "step": 48
    },
    {
      "epoch": 0.5730994152046783,
      "grad_norm": 0.567413866519928,
      "learning_rate": 8.995939984474624e-05,
      "loss": 5.2909,
      "step": 49
    },
    {
      "epoch": 0.5847953216374269,
      "grad_norm": 0.5367448329925537,
      "learning_rate": 8.945702546981969e-05,
      "loss": 5.1381,
      "step": 50
    },
    {
      "epoch": 0.5847953216374269,
      "eval_loss": 1.323288083076477,
      "eval_runtime": 14.2826,
      "eval_samples_per_second": 10.082,
      "eval_steps_per_second": 2.521,
      "step": 50
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 0.6580852270126343,
      "learning_rate": 8.894386393810563e-05,
      "loss": 4.9489,
      "step": 51
    },
    {
      "epoch": 0.6081871345029239,
      "grad_norm": 0.5204082727432251,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.0322,
      "step": 52
    },
    {
      "epoch": 0.6198830409356725,
      "grad_norm": 0.5475398898124695,
      "learning_rate": 8.788574348801675e-05,
      "loss": 5.074,
      "step": 53
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.5349409580230713,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.1456,
      "step": 54
    },
    {
      "epoch": 0.6432748538011696,
      "grad_norm": 0.6114522814750671,
      "learning_rate": 8.678619553365659e-05,
      "loss": 4.7651,
      "step": 55
    },
    {
      "epoch": 0.6549707602339181,
      "grad_norm": 0.5422197580337524,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.2278,
      "step": 56
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.7123456001281738,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.3771,
      "step": 57
    },
    {
      "epoch": 0.6783625730994152,
      "grad_norm": 0.7903186082839966,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.0217,
      "step": 58
    },
    {
      "epoch": 0.6900584795321637,
      "grad_norm": 0.6229669451713562,
      "learning_rate": 8.44676704559283e-05,
      "loss": 4.7591,
      "step": 59
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.637251079082489,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.1638,
      "step": 60
    },
    {
      "epoch": 0.7134502923976608,
      "grad_norm": 0.7046337127685547,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.4028,
      "step": 61
    },
    {
      "epoch": 0.7251461988304093,
      "grad_norm": 0.6663981676101685,
      "learning_rate": 8.262928807620843e-05,
      "loss": 4.5089,
      "step": 62
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.6950377225875854,
      "learning_rate": 8.199842702516583e-05,
      "loss": 4.7752,
      "step": 63
    },
    {
      "epoch": 0.7485380116959064,
      "grad_norm": 0.5766004323959351,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.2349,
      "step": 64
    },
    {
      "epoch": 0.7602339181286549,
      "grad_norm": 0.5690869092941284,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.4043,
      "step": 65
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 0.5272654891014099,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.848,
      "step": 66
    },
    {
      "epoch": 0.783625730994152,
      "grad_norm": 0.5626436471939087,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.0628,
      "step": 67
    },
    {
      "epoch": 0.7953216374269005,
      "grad_norm": 0.4988678991794586,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.1741,
      "step": 68
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 0.6162065863609314,
      "learning_rate": 7.803575286758364e-05,
      "loss": 4.7162,
      "step": 69
    },
    {
      "epoch": 0.8187134502923976,
      "grad_norm": 0.5828086733818054,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.4411,
      "step": 70
    },
    {
      "epoch": 0.8304093567251462,
      "grad_norm": 0.6132132411003113,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.245,
      "step": 71
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.5989595055580139,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.5671,
      "step": 72
    },
    {
      "epoch": 0.8538011695906432,
      "grad_norm": 0.6122727394104004,
      "learning_rate": 7.52382768867422e-05,
      "loss": 4.652,
      "step": 73
    },
    {
      "epoch": 0.8654970760233918,
      "grad_norm": 0.5506373643875122,
      "learning_rate": 7.452117519152542e-05,
      "loss": 4.8493,
      "step": 74
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.5534459948539734,
      "learning_rate": 7.379736965185368e-05,
      "loss": 4.7269,
      "step": 75
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.5943086743354797,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.0294,
      "step": 76
    },
    {
      "epoch": 0.9005847953216374,
      "grad_norm": 0.6175689101219177,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.3012,
      "step": 77
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.5720868110656738,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.0087,
      "step": 78
    },
    {
      "epoch": 0.9239766081871345,
      "grad_norm": 1.8394052982330322,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.3475,
      "step": 79
    },
    {
      "epoch": 0.935672514619883,
      "grad_norm": 0.6584748029708862,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.4526,
      "step": 80
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 0.6851667761802673,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.0865,
      "step": 81
    },
    {
      "epoch": 0.9590643274853801,
      "grad_norm": 1.1946464776992798,
      "learning_rate": 6.855986244591104e-05,
      "loss": 4.347,
      "step": 82
    },
    {
      "epoch": 0.9707602339181286,
      "grad_norm": 0.8016842007637024,
      "learning_rate": 6.778969234612584e-05,
      "loss": 4.6189,
      "step": 83
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.7933606505393982,
      "learning_rate": 6.701465872208216e-05,
      "loss": 4.5684,
      "step": 84
    },
    {
      "epoch": 0.9941520467836257,
      "grad_norm": 0.5581173300743103,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.4069,
      "step": 85
    },
    {
      "epoch": 1.0087719298245614,
      "grad_norm": 0.5724794268608093,
      "learning_rate": 6.545084971874738e-05,
      "loss": 4.5882,
      "step": 86
    },
    {
      "epoch": 1.0204678362573099,
      "grad_norm": 0.573429524898529,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.1707,
      "step": 87
    },
    {
      "epoch": 1.0321637426900585,
      "grad_norm": 0.5287179350852966,
      "learning_rate": 6.387014543809223e-05,
      "loss": 4.8823,
      "step": 88
    },
    {
      "epoch": 1.043859649122807,
      "grad_norm": 0.5025233626365662,
      "learning_rate": 6.307399704769099e-05,
      "loss": 4.5625,
      "step": 89
    },
    {
      "epoch": 1.0555555555555556,
      "grad_norm": 0.514818549156189,
      "learning_rate": 6.227427435703997e-05,
      "loss": 4.73,
      "step": 90
    },
    {
      "epoch": 1.067251461988304,
      "grad_norm": 0.5340942144393921,
      "learning_rate": 6.147119600233758e-05,
      "loss": 4.8756,
      "step": 91
    },
    {
      "epoch": 1.0789473684210527,
      "grad_norm": 0.6041892170906067,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.2854,
      "step": 92
    },
    {
      "epoch": 1.090643274853801,
      "grad_norm": 0.7493144869804382,
      "learning_rate": 5.985585137257401e-05,
      "loss": 5.3384,
      "step": 93
    },
    {
      "epoch": 1.1023391812865497,
      "grad_norm": 0.5995897054672241,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.5511,
      "step": 94
    },
    {
      "epoch": 1.1140350877192982,
      "grad_norm": 0.618107259273529,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 4.8538,
      "step": 95
    },
    {
      "epoch": 1.1257309941520468,
      "grad_norm": 0.9848514199256897,
      "learning_rate": 5.74131823855921e-05,
      "loss": 4.8317,
      "step": 96
    },
    {
      "epoch": 1.1374269005847952,
      "grad_norm": 0.6017423868179321,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 4.3356,
      "step": 97
    },
    {
      "epoch": 1.1491228070175439,
      "grad_norm": 0.6649016737937927,
      "learning_rate": 5.577423184847932e-05,
      "loss": 4.9057,
      "step": 98
    },
    {
      "epoch": 1.1608187134502923,
      "grad_norm": 0.682602047920227,
      "learning_rate": 5.495227651252315e-05,
      "loss": 5.3497,
      "step": 99
    },
    {
      "epoch": 1.172514619883041,
      "grad_norm": 0.8660184741020203,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 5.1194,
      "step": 100
    },
    {
      "epoch": 1.172514619883041,
      "eval_loss": 1.2805209159851074,
      "eval_runtime": 14.2582,
      "eval_samples_per_second": 10.099,
      "eval_steps_per_second": 2.525,
      "step": 100
    },
    {
      "epoch": 1.1842105263157894,
      "grad_norm": 1.3263887166976929,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.1145,
      "step": 101
    },
    {
      "epoch": 1.195906432748538,
      "grad_norm": 0.7568484544754028,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.7292,
      "step": 102
    },
    {
      "epoch": 1.2076023391812865,
      "grad_norm": 0.7868157029151917,
      "learning_rate": 5.165316846586541e-05,
      "loss": 4.9724,
      "step": 103
    },
    {
      "epoch": 1.219298245614035,
      "grad_norm": 0.8283912539482117,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.5678,
      "step": 104
    },
    {
      "epoch": 1.2309941520467835,
      "grad_norm": 0.7498999238014221,
      "learning_rate": 5e-05,
      "loss": 5.1414,
      "step": 105
    },
    {
      "epoch": 1.2426900584795322,
      "grad_norm": 0.7905023097991943,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.3368,
      "step": 106
    },
    {
      "epoch": 1.2543859649122808,
      "grad_norm": 0.6527309417724609,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.9133,
      "step": 107
    },
    {
      "epoch": 1.2660818713450293,
      "grad_norm": 0.6032450199127197,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 4.579,
      "step": 108
    },
    {
      "epoch": 1.2777777777777777,
      "grad_norm": 0.6000065207481384,
      "learning_rate": 4.669547078371504e-05,
      "loss": 4.4797,
      "step": 109
    },
    {
      "epoch": 1.2894736842105263,
      "grad_norm": 0.6161346435546875,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.2997,
      "step": 110
    },
    {
      "epoch": 1.301169590643275,
      "grad_norm": 0.6743441224098206,
      "learning_rate": 4.504772348747687e-05,
      "loss": 4.7612,
      "step": 111
    },
    {
      "epoch": 1.3128654970760234,
      "grad_norm": 0.6210641860961914,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 4.8589,
      "step": 112
    },
    {
      "epoch": 1.3245614035087718,
      "grad_norm": 0.7048709988594055,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.5511,
      "step": 113
    },
    {
      "epoch": 1.3362573099415205,
      "grad_norm": 0.690811812877655,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 4.4486,
      "step": 114
    },
    {
      "epoch": 1.3479532163742691,
      "grad_norm": 0.7149151563644409,
      "learning_rate": 4.17702704859633e-05,
      "loss": 4.888,
      "step": 115
    },
    {
      "epoch": 1.3596491228070176,
      "grad_norm": 0.6492061614990234,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.1674,
      "step": 116
    },
    {
      "epoch": 1.371345029239766,
      "grad_norm": 0.6934666633605957,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 4.7724,
      "step": 117
    },
    {
      "epoch": 1.3830409356725146,
      "grad_norm": 0.6952803134918213,
      "learning_rate": 3.933501846281267e-05,
      "loss": 4.8122,
      "step": 118
    },
    {
      "epoch": 1.3947368421052633,
      "grad_norm": 0.7041831016540527,
      "learning_rate": 3.852880399766243e-05,
      "loss": 4.9574,
      "step": 119
    },
    {
      "epoch": 1.4064327485380117,
      "grad_norm": 0.7406746745109558,
      "learning_rate": 3.772572564296005e-05,
      "loss": 4.6055,
      "step": 120
    },
    {
      "epoch": 1.4181286549707601,
      "grad_norm": 0.7634146809577942,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 4.9763,
      "step": 121
    },
    {
      "epoch": 1.4298245614035088,
      "grad_norm": 0.8683531284332275,
      "learning_rate": 3.612985456190778e-05,
      "loss": 4.6887,
      "step": 122
    },
    {
      "epoch": 1.4415204678362574,
      "grad_norm": 0.7926579713821411,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.0066,
      "step": 123
    },
    {
      "epoch": 1.4532163742690059,
      "grad_norm": 0.8141331672668457,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 4.7194,
      "step": 124
    },
    {
      "epoch": 1.4649122807017543,
      "grad_norm": 0.7802368998527527,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 3.9383,
      "step": 125
    },
    {
      "epoch": 1.476608187134503,
      "grad_norm": 0.8326878547668457,
      "learning_rate": 3.298534127791785e-05,
      "loss": 4.421,
      "step": 126
    },
    {
      "epoch": 1.4883040935672516,
      "grad_norm": 0.8888916969299316,
      "learning_rate": 3.221030765387417e-05,
      "loss": 4.3634,
      "step": 127
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.7237643599510193,
      "learning_rate": 3.144013755408895e-05,
      "loss": 4.7557,
      "step": 128
    },
    {
      "epoch": 1.5116959064327484,
      "grad_norm": 0.6778761744499207,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 4.5118,
      "step": 129
    },
    {
      "epoch": 1.523391812865497,
      "grad_norm": 0.7118635773658752,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.1603,
      "step": 130
    },
    {
      "epoch": 1.5350877192982457,
      "grad_norm": 0.685716450214386,
      "learning_rate": 2.916090697523549e-05,
      "loss": 4.7323,
      "step": 131
    },
    {
      "epoch": 1.5467836257309941,
      "grad_norm": 0.8787785172462463,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 4.9417,
      "step": 132
    },
    {
      "epoch": 1.5584795321637426,
      "grad_norm": 0.7456165552139282,
      "learning_rate": 2.766955965735968e-05,
      "loss": 4.9567,
      "step": 133
    },
    {
      "epoch": 1.5701754385964912,
      "grad_norm": 0.6835184097290039,
      "learning_rate": 2.693294185106562e-05,
      "loss": 4.5495,
      "step": 134
    },
    {
      "epoch": 1.5818713450292399,
      "grad_norm": 0.6832260489463806,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 4.9867,
      "step": 135
    },
    {
      "epoch": 1.5935672514619883,
      "grad_norm": 0.7048373818397522,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.1378,
      "step": 136
    },
    {
      "epoch": 1.6052631578947367,
      "grad_norm": 0.7293301224708557,
      "learning_rate": 2.476172311325783e-05,
      "loss": 4.7399,
      "step": 137
    },
    {
      "epoch": 1.6169590643274854,
      "grad_norm": 0.6833894848823547,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.0154,
      "step": 138
    },
    {
      "epoch": 1.628654970760234,
      "grad_norm": 0.7147215008735657,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 4.7814,
      "step": 139
    },
    {
      "epoch": 1.6403508771929824,
      "grad_norm": 0.7362138032913208,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 4.9131,
      "step": 140
    },
    {
      "epoch": 1.6520467836257309,
      "grad_norm": 0.7377339005470276,
      "learning_rate": 2.196424713241637e-05,
      "loss": 4.7527,
      "step": 141
    },
    {
      "epoch": 1.6637426900584795,
      "grad_norm": 0.7660468220710754,
      "learning_rate": 2.128356686585282e-05,
      "loss": 4.462,
      "step": 142
    },
    {
      "epoch": 1.6754385964912282,
      "grad_norm": 0.7579042315483093,
      "learning_rate": 2.061073738537635e-05,
      "loss": 4.8809,
      "step": 143
    },
    {
      "epoch": 1.6871345029239766,
      "grad_norm": 0.7728883028030396,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 4.9308,
      "step": 144
    },
    {
      "epoch": 1.698830409356725,
      "grad_norm": 0.7957561612129211,
      "learning_rate": 1.928936436551661e-05,
      "loss": 4.9666,
      "step": 145
    },
    {
      "epoch": 1.7105263157894737,
      "grad_norm": 0.8299893140792847,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 4.5053,
      "step": 146
    },
    {
      "epoch": 1.7222222222222223,
      "grad_norm": 0.8418483734130859,
      "learning_rate": 1.800157297483417e-05,
      "loss": 4.1139,
      "step": 147
    },
    {
      "epoch": 1.7339181286549707,
      "grad_norm": 0.8716121315956116,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.5639,
      "step": 148
    },
    {
      "epoch": 1.7456140350877192,
      "grad_norm": 0.7277250289916992,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 4.8218,
      "step": 149
    },
    {
      "epoch": 1.7573099415204678,
      "grad_norm": 0.7173898220062256,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 4.4455,
      "step": 150
    },
    {
      "epoch": 1.7573099415204678,
      "eval_loss": 1.2639923095703125,
      "eval_runtime": 14.2483,
      "eval_samples_per_second": 10.106,
      "eval_steps_per_second": 2.527,
      "step": 150
    },
    {
      "epoch": 1.7690058479532165,
      "grad_norm": 0.899341881275177,
      "learning_rate": 1.553232954407171e-05,
      "loss": 4.2755,
      "step": 151
    },
    {
      "epoch": 1.780701754385965,
      "grad_norm": 0.6698784828186035,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.3786,
      "step": 152
    },
    {
      "epoch": 1.7923976608187133,
      "grad_norm": 0.7427481412887573,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.1591,
      "step": 153
    },
    {
      "epoch": 1.804093567251462,
      "grad_norm": 0.6991140842437744,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 4.6719,
      "step": 154
    },
    {
      "epoch": 1.8157894736842106,
      "grad_norm": 0.6866679787635803,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.7145,
      "step": 155
    },
    {
      "epoch": 1.827485380116959,
      "grad_norm": 0.6441006064414978,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 4.4043,
      "step": 156
    },
    {
      "epoch": 1.8391812865497075,
      "grad_norm": 0.6817840933799744,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 4.8823,
      "step": 157
    },
    {
      "epoch": 1.8508771929824561,
      "grad_norm": 0.7140611410140991,
      "learning_rate": 1.157994445715706e-05,
      "loss": 4.728,
      "step": 158
    },
    {
      "epoch": 1.8625730994152048,
      "grad_norm": 0.7639045119285583,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 4.7192,
      "step": 159
    },
    {
      "epoch": 1.8742690058479532,
      "grad_norm": 0.7013947367668152,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 4.5624,
      "step": 160
    },
    {
      "epoch": 1.8859649122807016,
      "grad_norm": 0.7424225211143494,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.2048,
      "step": 161
    },
    {
      "epoch": 1.8976608187134503,
      "grad_norm": 0.7256579995155334,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.6092,
      "step": 162
    },
    {
      "epoch": 1.909356725146199,
      "grad_norm": 0.8256704807281494,
      "learning_rate": 9.068759265665384e-06,
      "loss": 4.4324,
      "step": 163
    },
    {
      "epoch": 1.9210526315789473,
      "grad_norm": 0.8200169205665588,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.0588,
      "step": 164
    },
    {
      "epoch": 1.9327485380116958,
      "grad_norm": 0.7896498441696167,
      "learning_rate": 8.141676086873572e-06,
      "loss": 4.2785,
      "step": 165
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.8187960982322693,
      "learning_rate": 7.695237378953223e-06,
      "loss": 4.645,
      "step": 166
    },
    {
      "epoch": 1.956140350877193,
      "grad_norm": 0.8252620697021484,
      "learning_rate": 7.260364370723044e-06,
      "loss": 4.629,
      "step": 167
    },
    {
      "epoch": 1.9678362573099415,
      "grad_norm": 0.9667232036590576,
      "learning_rate": 6.837175952121306e-06,
      "loss": 4.2799,
      "step": 168
    },
    {
      "epoch": 1.97953216374269,
      "grad_norm": 0.954495906829834,
      "learning_rate": 6.425787818636131e-06,
      "loss": 4.5271,
      "step": 169
    },
    {
      "epoch": 1.9912280701754386,
      "grad_norm": 0.7367401719093323,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.8202,
      "step": 170
    },
    {
      "epoch": 2.0058479532163744,
      "grad_norm": 0.8099879622459412,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 4.936,
      "step": 171
    },
    {
      "epoch": 2.017543859649123,
      "grad_norm": 0.5857845544815063,
      "learning_rate": 5.263533508961827e-06,
      "loss": 4.2838,
      "step": 172
    },
    {
      "epoch": 2.0292397660818713,
      "grad_norm": 0.6351048946380615,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.6657,
      "step": 173
    },
    {
      "epoch": 2.0409356725146197,
      "grad_norm": 0.6121008396148682,
      "learning_rate": 4.549673247541875e-06,
      "loss": 4.483,
      "step": 174
    },
    {
      "epoch": 2.0526315789473686,
      "grad_norm": 0.6245721578598022,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.714,
      "step": 175
    },
    {
      "epoch": 2.064327485380117,
      "grad_norm": 0.6177881360054016,
      "learning_rate": 3.885512251130763e-06,
      "loss": 4.5448,
      "step": 176
    },
    {
      "epoch": 2.0760233918128654,
      "grad_norm": 0.6620082259178162,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.0884,
      "step": 177
    },
    {
      "epoch": 2.087719298245614,
      "grad_norm": 0.6682341694831848,
      "learning_rate": 3.271776770026963e-06,
      "loss": 4.7471,
      "step": 178
    },
    {
      "epoch": 2.0994152046783627,
      "grad_norm": 0.7520459294319153,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.2967,
      "step": 179
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 0.7431067228317261,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.3128,
      "step": 180
    },
    {
      "epoch": 2.1228070175438596,
      "grad_norm": 0.742266833782196,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.0478,
      "step": 181
    },
    {
      "epoch": 2.134502923976608,
      "grad_norm": 0.6587474942207336,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 4.351,
      "step": 182
    },
    {
      "epoch": 2.146198830409357,
      "grad_norm": 0.7246972918510437,
      "learning_rate": 1.962316193157593e-06,
      "loss": 4.4303,
      "step": 183
    },
    {
      "epoch": 2.1578947368421053,
      "grad_norm": 0.7119033932685852,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 4.1313,
      "step": 184
    },
    {
      "epoch": 2.1695906432748537,
      "grad_norm": 0.8276904821395874,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 4.4922,
      "step": 185
    },
    {
      "epoch": 2.181286549707602,
      "grad_norm": 0.7426488995552063,
      "learning_rate": 1.333670137599713e-06,
      "loss": 4.4075,
      "step": 186
    },
    {
      "epoch": 2.192982456140351,
      "grad_norm": 0.8933841586112976,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 4.6855,
      "step": 187
    },
    {
      "epoch": 2.2046783625730995,
      "grad_norm": 0.8134186863899231,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.0226,
      "step": 188
    },
    {
      "epoch": 2.216374269005848,
      "grad_norm": 0.8100214600563049,
      "learning_rate": 8.247462563808817e-07,
      "loss": 4.7842,
      "step": 189
    },
    {
      "epoch": 2.2280701754385963,
      "grad_norm": 3.2747926712036133,
      "learning_rate": 6.819348298638839e-07,
      "loss": 4.2482,
      "step": 190
    },
    {
      "epoch": 2.239766081871345,
      "grad_norm": 0.852902889251709,
      "learning_rate": 5.526064699265753e-07,
      "loss": 4.1164,
      "step": 191
    },
    {
      "epoch": 2.2514619883040936,
      "grad_norm": 0.7660697102546692,
      "learning_rate": 4.367965336512403e-07,
      "loss": 4.6685,
      "step": 192
    },
    {
      "epoch": 2.263157894736842,
      "grad_norm": 0.5799196362495422,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 4.5511,
      "step": 193
    },
    {
      "epoch": 2.2748538011695905,
      "grad_norm": 0.7791092395782471,
      "learning_rate": 2.458548727494292e-07,
      "loss": 4.755,
      "step": 194
    },
    {
      "epoch": 2.2865497076023393,
      "grad_norm": 0.6427508592605591,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 4.4126,
      "step": 195
    },
    {
      "epoch": 2.2982456140350878,
      "grad_norm": 0.6345090866088867,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 4.191,
      "step": 196
    },
    {
      "epoch": 2.309941520467836,
      "grad_norm": 0.6280609965324402,
      "learning_rate": 6.150154258476315e-08,
      "loss": 4.7681,
      "step": 197
    },
    {
      "epoch": 2.3216374269005846,
      "grad_norm": 0.6592674255371094,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 4.8152,
      "step": 198
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6667617559432983,
      "learning_rate": 6.834750376549792e-09,
      "loss": 4.7783,
      "step": 199
    },
    {
      "epoch": 2.345029239766082,
      "grad_norm": 0.6581698656082153,
      "learning_rate": 0.0,
      "loss": 4.4587,
      "step": 200
    },
    {
      "epoch": 2.345029239766082,
      "eval_loss": 1.257092833518982,
      "eval_runtime": 14.2489,
      "eval_samples_per_second": 10.106,
      "eval_steps_per_second": 2.527,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.021563448262656e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}