{
  "best_metric": 1.3243097066879272,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.4158004158004158,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002079002079002079,
      "grad_norm": 28.071395874023438,
      "learning_rate": 1e-05,
      "loss": 6.7852,
      "step": 1
    },
    {
      "epoch": 0.002079002079002079,
      "eval_loss": 1.588921070098877,
      "eval_runtime": 16.5922,
      "eval_samples_per_second": 48.879,
      "eval_steps_per_second": 12.235,
      "step": 1
    },
    {
      "epoch": 0.004158004158004158,
      "grad_norm": 24.824087142944336,
      "learning_rate": 2e-05,
      "loss": 6.8426,
      "step": 2
    },
    {
      "epoch": 0.006237006237006237,
      "grad_norm": 26.048978805541992,
      "learning_rate": 3e-05,
      "loss": 7.2334,
      "step": 3
    },
    {
      "epoch": 0.008316008316008316,
      "grad_norm": 26.67098617553711,
      "learning_rate": 4e-05,
      "loss": 6.8714,
      "step": 4
    },
    {
      "epoch": 0.010395010395010396,
      "grad_norm": 30.142372131347656,
      "learning_rate": 5e-05,
      "loss": 6.7315,
      "step": 5
    },
    {
      "epoch": 0.012474012474012475,
      "grad_norm": 25.800376892089844,
      "learning_rate": 6e-05,
      "loss": 6.4853,
      "step": 6
    },
    {
      "epoch": 0.014553014553014554,
      "grad_norm": 25.5001220703125,
      "learning_rate": 7e-05,
      "loss": 6.8934,
      "step": 7
    },
    {
      "epoch": 0.016632016632016633,
      "grad_norm": 24.654605865478516,
      "learning_rate": 8e-05,
      "loss": 6.5447,
      "step": 8
    },
    {
      "epoch": 0.018711018711018712,
      "grad_norm": 25.30592155456543,
      "learning_rate": 9e-05,
      "loss": 6.2901,
      "step": 9
    },
    {
      "epoch": 0.02079002079002079,
      "grad_norm": 24.669618606567383,
      "learning_rate": 0.0001,
      "loss": 6.5575,
      "step": 10
    },
    {
      "epoch": 0.02286902286902287,
      "grad_norm": 27.357210159301758,
      "learning_rate": 9.999316524962345e-05,
      "loss": 6.2223,
      "step": 11
    },
    {
      "epoch": 0.02494802494802495,
      "grad_norm": 26.247827529907227,
      "learning_rate": 9.997266286704631e-05,
      "loss": 6.1598,
      "step": 12
    },
    {
      "epoch": 0.02702702702702703,
      "grad_norm": 25.70038604736328,
      "learning_rate": 9.993849845741524e-05,
      "loss": 6.1561,
      "step": 13
    },
    {
      "epoch": 0.029106029106029108,
      "grad_norm": 24.153085708618164,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.6821,
      "step": 14
    },
    {
      "epoch": 0.031185031185031187,
      "grad_norm": 24.06270408630371,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.6772,
      "step": 15
    },
    {
      "epoch": 0.033264033264033266,
      "grad_norm": 23.677181243896484,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.6262,
      "step": 16
    },
    {
      "epoch": 0.035343035343035345,
      "grad_norm": 25.706100463867188,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.5083,
      "step": 17
    },
    {
      "epoch": 0.037422037422037424,
      "grad_norm": 24.57635498046875,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.528,
      "step": 18
    },
    {
      "epoch": 0.0395010395010395,
      "grad_norm": 24.061233520507812,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.6109,
      "step": 19
    },
    {
      "epoch": 0.04158004158004158,
      "grad_norm": 24.62213134765625,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.1977,
      "step": 20
    },
    {
      "epoch": 0.04365904365904366,
      "grad_norm": 27.57269287109375,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.6462,
      "step": 21
    },
    {
      "epoch": 0.04573804573804574,
      "grad_norm": 26.805295944213867,
      "learning_rate": 9.901899829374047e-05,
      "loss": 4.9515,
      "step": 22
    },
    {
      "epoch": 0.04781704781704782,
      "grad_norm": 28.417388916015625,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.6827,
      "step": 23
    },
    {
      "epoch": 0.0498960498960499,
      "grad_norm": 31.16328239440918,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.4636,
      "step": 24
    },
    {
      "epoch": 0.05197505197505198,
      "grad_norm": 32.09971618652344,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.5845,
      "step": 25
    },
    {
      "epoch": 0.05405405405405406,
      "grad_norm": 29.880390167236328,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.2932,
      "step": 26
    },
    {
      "epoch": 0.056133056133056136,
      "grad_norm": 37.16545486450195,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.1862,
      "step": 27
    },
    {
      "epoch": 0.058212058212058215,
      "grad_norm": 34.66633605957031,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.4445,
      "step": 28
    },
    {
      "epoch": 0.060291060291060294,
      "grad_norm": 29.252239227294922,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.3222,
      "step": 29
    },
    {
      "epoch": 0.062370062370062374,
      "grad_norm": 30.50757598876953,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.458,
      "step": 30
    },
    {
      "epoch": 0.06444906444906445,
      "grad_norm": 33.433143615722656,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.2476,
      "step": 31
    },
    {
      "epoch": 0.06652806652806653,
      "grad_norm": 29.555431365966797,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.0169,
      "step": 32
    },
    {
      "epoch": 0.06860706860706861,
      "grad_norm": 28.021678924560547,
      "learning_rate": 9.642770192448536e-05,
      "loss": 5.7114,
      "step": 33
    },
    {
      "epoch": 0.07068607068607069,
      "grad_norm": 28.683151245117188,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.6333,
      "step": 34
    },
    {
      "epoch": 0.07276507276507277,
      "grad_norm": 24.848201751708984,
      "learning_rate": 9.578866633275288e-05,
      "loss": 5.0257,
      "step": 35
    },
    {
      "epoch": 0.07484407484407485,
      "grad_norm": 29.15994644165039,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.6732,
      "step": 36
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 37.718177795410156,
      "learning_rate": 9.509956150664796e-05,
      "loss": 5.3217,
      "step": 37
    },
    {
      "epoch": 0.079002079002079,
      "grad_norm": 25.195199966430664,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.4996,
      "step": 38
    },
    {
      "epoch": 0.08108108108108109,
      "grad_norm": 22.9675350189209,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.5896,
      "step": 39
    },
    {
      "epoch": 0.08316008316008316,
      "grad_norm": 22.624597549438477,
      "learning_rate": 9.397368756032445e-05,
      "loss": 5.7527,
      "step": 40
    },
    {
      "epoch": 0.08523908523908524,
      "grad_norm": 22.4384708404541,
      "learning_rate": 9.357421218136386e-05,
      "loss": 5.3034,
      "step": 41
    },
    {
      "epoch": 0.08731808731808732,
      "grad_norm": 19.613401412963867,
      "learning_rate": 9.316282404787871e-05,
      "loss": 5.4214,
      "step": 42
    },
    {
      "epoch": 0.0893970893970894,
      "grad_norm": 18.841463088989258,
      "learning_rate": 9.273963562927695e-05,
      "loss": 5.5083,
      "step": 43
    },
    {
      "epoch": 0.09147609147609148,
      "grad_norm": 17.60077476501465,
      "learning_rate": 9.230476262104677e-05,
      "loss": 5.8899,
      "step": 44
    },
    {
      "epoch": 0.09355509355509356,
      "grad_norm": 16.823047637939453,
      "learning_rate": 9.185832391312644e-05,
      "loss": 5.7908,
      "step": 45
    },
    {
      "epoch": 0.09563409563409564,
      "grad_norm": 16.220130920410156,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.2148,
      "step": 46
    },
    {
      "epoch": 0.09771309771309772,
      "grad_norm": 16.249300003051758,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.2011,
      "step": 47
    },
    {
      "epoch": 0.0997920997920998,
      "grad_norm": 18.441320419311523,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.2117,
      "step": 48
    },
    {
      "epoch": 0.10187110187110188,
      "grad_norm": 16.127399444580078,
      "learning_rate": 8.995939984474624e-05,
      "loss": 6.4037,
      "step": 49
    },
    {
      "epoch": 0.10395010395010396,
      "grad_norm": 25.184375762939453,
      "learning_rate": 8.945702546981969e-05,
      "loss": 5.4199,
      "step": 50
    },
    {
      "epoch": 0.10395010395010396,
      "eval_loss": 1.4382290840148926,
      "eval_runtime": 16.5896,
      "eval_samples_per_second": 48.886,
      "eval_steps_per_second": 12.237,
      "step": 50
    },
    {
      "epoch": 0.10602910602910603,
      "grad_norm": 21.945032119750977,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.6972,
      "step": 51
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 22.294170379638672,
      "learning_rate": 8.842005554284296e-05,
      "loss": 6.2051,
      "step": 52
    },
    {
      "epoch": 0.1101871101871102,
      "grad_norm": 19.184036254882812,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.3798,
      "step": 53
    },
    {
      "epoch": 0.11226611226611227,
      "grad_norm": 19.589115142822266,
      "learning_rate": 8.73410738492077e-05,
      "loss": 6.4645,
      "step": 54
    },
    {
      "epoch": 0.11434511434511435,
      "grad_norm": 18.968124389648438,
      "learning_rate": 8.678619553365659e-05,
      "loss": 6.2279,
      "step": 55
    },
    {
      "epoch": 0.11642411642411643,
      "grad_norm": 15.437923431396484,
      "learning_rate": 8.622126023955446e-05,
      "loss": 6.3658,
      "step": 56
    },
    {
      "epoch": 0.11850311850311851,
      "grad_norm": 14.735605239868164,
      "learning_rate": 8.564642241456986e-05,
      "loss": 6.1889,
      "step": 57
    },
    {
      "epoch": 0.12058212058212059,
      "grad_norm": 19.146366119384766,
      "learning_rate": 8.506183921362443e-05,
      "loss": 6.1798,
      "step": 58
    },
    {
      "epoch": 0.12266112266112267,
      "grad_norm": 17.00002670288086,
      "learning_rate": 8.44676704559283e-05,
      "loss": 6.229,
      "step": 59
    },
    {
      "epoch": 0.12474012474012475,
      "grad_norm": 15.448655128479004,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.9698,
      "step": 60
    },
    {
      "epoch": 0.12681912681912683,
      "grad_norm": 14.371026992797852,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.8222,
      "step": 61
    },
    {
      "epoch": 0.1288981288981289,
      "grad_norm": 13.9545259475708,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.5434,
      "step": 62
    },
    {
      "epoch": 0.13097713097713098,
      "grad_norm": 15.803035736083984,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.8329,
      "step": 63
    },
    {
      "epoch": 0.13305613305613306,
      "grad_norm": 15.736432075500488,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.5458,
      "step": 64
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 12.779462814331055,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.3702,
      "step": 65
    },
    {
      "epoch": 0.13721413721413722,
      "grad_norm": 14.004919052124023,
      "learning_rate": 8.005405736415126e-05,
      "loss": 5.2535,
      "step": 66
    },
    {
      "epoch": 0.1392931392931393,
      "grad_norm": 13.838112831115723,
      "learning_rate": 7.938926261462366e-05,
      "loss": 4.862,
      "step": 67
    },
    {
      "epoch": 0.14137214137214138,
      "grad_norm": 14.023115158081055,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.2739,
      "step": 68
    },
    {
      "epoch": 0.14345114345114346,
      "grad_norm": 14.582015991210938,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.3618,
      "step": 69
    },
    {
      "epoch": 0.14553014553014554,
      "grad_norm": 15.975716590881348,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.2481,
      "step": 70
    },
    {
      "epoch": 0.14760914760914762,
      "grad_norm": 13.392427444458008,
      "learning_rate": 7.66515864363997e-05,
      "loss": 4.8628,
      "step": 71
    },
    {
      "epoch": 0.1496881496881497,
      "grad_norm": 14.639575004577637,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.299,
      "step": 72
    },
    {
      "epoch": 0.15176715176715178,
      "grad_norm": 16.03403663635254,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.1636,
      "step": 73
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 14.60657787322998,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.438,
      "step": 74
    },
    {
      "epoch": 0.15592515592515593,
      "grad_norm": 14.212603569030762,
      "learning_rate": 7.379736965185368e-05,
      "loss": 4.9637,
      "step": 75
    },
    {
      "epoch": 0.158004158004158,
      "grad_norm": 17.060789108276367,
      "learning_rate": 7.30670581489344e-05,
      "loss": 4.5159,
      "step": 76
    },
    {
      "epoch": 0.1600831600831601,
      "grad_norm": 14.055794715881348,
      "learning_rate": 7.233044034264034e-05,
      "loss": 4.9214,
      "step": 77
    },
    {
      "epoch": 0.16216216216216217,
      "grad_norm": 14.713263511657715,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.0005,
      "step": 78
    },
    {
      "epoch": 0.16424116424116425,
      "grad_norm": 16.1826114654541,
      "learning_rate": 7.083909302476453e-05,
      "loss": 4.7481,
      "step": 79
    },
    {
      "epoch": 0.16632016632016633,
      "grad_norm": 19.140520095825195,
      "learning_rate": 7.008477123264848e-05,
      "loss": 4.8512,
      "step": 80
    },
    {
      "epoch": 0.1683991683991684,
      "grad_norm": 15.271452903747559,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.2141,
      "step": 81
    },
    {
      "epoch": 0.1704781704781705,
      "grad_norm": 16.169696807861328,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.0205,
      "step": 82
    },
    {
      "epoch": 0.17255717255717257,
      "grad_norm": 16.73891258239746,
      "learning_rate": 6.778969234612584e-05,
      "loss": 4.7573,
      "step": 83
    },
    {
      "epoch": 0.17463617463617465,
      "grad_norm": 18.32731819152832,
      "learning_rate": 6.701465872208216e-05,
      "loss": 5.2433,
      "step": 84
    },
    {
      "epoch": 0.17671517671517672,
      "grad_norm": 17.923723220825195,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.2455,
      "step": 85
    },
    {
      "epoch": 0.1787941787941788,
      "grad_norm": 16.739099502563477,
      "learning_rate": 6.545084971874738e-05,
      "loss": 5.4368,
      "step": 86
    },
    {
      "epoch": 0.18087318087318088,
      "grad_norm": 15.544157981872559,
      "learning_rate": 6.466250186922325e-05,
      "loss": 4.8306,
      "step": 87
    },
    {
      "epoch": 0.18295218295218296,
      "grad_norm": 18.448301315307617,
      "learning_rate": 6.387014543809223e-05,
      "loss": 4.9866,
      "step": 88
    },
    {
      "epoch": 0.18503118503118504,
      "grad_norm": 19.25309944152832,
      "learning_rate": 6.307399704769099e-05,
      "loss": 5.5389,
      "step": 89
    },
    {
      "epoch": 0.18711018711018712,
      "grad_norm": 18.53395652770996,
      "learning_rate": 6.227427435703997e-05,
      "loss": 5.4347,
      "step": 90
    },
    {
      "epoch": 0.1891891891891892,
      "grad_norm": 16.402286529541016,
      "learning_rate": 6.147119600233758e-05,
      "loss": 5.6028,
      "step": 91
    },
    {
      "epoch": 0.19126819126819128,
      "grad_norm": 21.19646644592285,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.7664,
      "step": 92
    },
    {
      "epoch": 0.19334719334719336,
      "grad_norm": 17.12851333618164,
      "learning_rate": 5.985585137257401e-05,
      "loss": 5.9171,
      "step": 93
    },
    {
      "epoch": 0.19542619542619544,
      "grad_norm": 18.081279754638672,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.4797,
      "step": 94
    },
    {
      "epoch": 0.19750519750519752,
      "grad_norm": 16.03946304321289,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 5.9026,
      "step": 95
    },
    {
      "epoch": 0.1995841995841996,
      "grad_norm": 16.010799407958984,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.4291,
      "step": 96
    },
    {
      "epoch": 0.20166320166320167,
      "grad_norm": 13.145974159240723,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.0234,
      "step": 97
    },
    {
      "epoch": 0.20374220374220375,
      "grad_norm": 16.17687225341797,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.5085,
      "step": 98
    },
    {
      "epoch": 0.20582120582120583,
      "grad_norm": 14.096548080444336,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.3102,
      "step": 99
    },
    {
      "epoch": 0.2079002079002079,
      "grad_norm": 15.08219051361084,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 5.1916,
      "step": 100
    },
    {
      "epoch": 0.2079002079002079,
      "eval_loss": 1.3682379722595215,
      "eval_runtime": 16.7017,
      "eval_samples_per_second": 48.558,
      "eval_steps_per_second": 12.154,
      "step": 100
    },
    {
      "epoch": 0.20997920997921,
      "grad_norm": 18.226085662841797,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.8869,
      "step": 101
    },
    {
      "epoch": 0.21205821205821207,
      "grad_norm": 17.85428237915039,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.0493,
      "step": 102
    },
    {
      "epoch": 0.21413721413721415,
      "grad_norm": 23.202178955078125,
      "learning_rate": 5.165316846586541e-05,
      "loss": 6.2234,
      "step": 103
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 16.699892044067383,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 6.1155,
      "step": 104
    },
    {
      "epoch": 0.2182952182952183,
      "grad_norm": 57.3818244934082,
      "learning_rate": 5e-05,
      "loss": 5.9773,
      "step": 105
    },
    {
      "epoch": 0.2203742203742204,
      "grad_norm": 15.62403392791748,
      "learning_rate": 4.917330276168208e-05,
      "loss": 6.3167,
      "step": 106
    },
    {
      "epoch": 0.22245322245322247,
      "grad_norm": 20.42095375061035,
      "learning_rate": 4.834683153413459e-05,
      "loss": 5.7811,
      "step": 107
    },
    {
      "epoch": 0.22453222453222454,
      "grad_norm": 15.216282844543457,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 5.8864,
      "step": 108
    },
    {
      "epoch": 0.22661122661122662,
      "grad_norm": 14.746712684631348,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.7187,
      "step": 109
    },
    {
      "epoch": 0.2286902286902287,
      "grad_norm": 12.363672256469727,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.5358,
      "step": 110
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 14.588408470153809,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.806,
      "step": 111
    },
    {
      "epoch": 0.23284823284823286,
      "grad_norm": 14.796876907348633,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.7591,
      "step": 112
    },
    {
      "epoch": 0.23492723492723494,
      "grad_norm": 11.773994445800781,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.2513,
      "step": 113
    },
    {
      "epoch": 0.23700623700623702,
      "grad_norm": 12.673705101013184,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.0809,
      "step": 114
    },
    {
      "epoch": 0.2390852390852391,
      "grad_norm": 10.781575202941895,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.1155,
      "step": 115
    },
    {
      "epoch": 0.24116424116424118,
      "grad_norm": 11.707293510437012,
      "learning_rate": 4.095597328339452e-05,
      "loss": 4.9225,
      "step": 116
    },
    {
      "epoch": 0.24324324324324326,
      "grad_norm": 12.583459854125977,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.206,
      "step": 117
    },
    {
      "epoch": 0.24532224532224534,
      "grad_norm": 11.232467651367188,
      "learning_rate": 3.933501846281267e-05,
      "loss": 4.9289,
      "step": 118
    },
    {
      "epoch": 0.24740124740124741,
      "grad_norm": 10.965503692626953,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.1957,
      "step": 119
    },
    {
      "epoch": 0.2494802494802495,
      "grad_norm": 11.507230758666992,
      "learning_rate": 3.772572564296005e-05,
      "loss": 4.8472,
      "step": 120
    },
    {
      "epoch": 0.2515592515592516,
      "grad_norm": 10.629424095153809,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 5.0717,
      "step": 121
    },
    {
      "epoch": 0.25363825363825365,
      "grad_norm": 10.779979705810547,
      "learning_rate": 3.612985456190778e-05,
      "loss": 4.8315,
      "step": 122
    },
    {
      "epoch": 0.25571725571725573,
      "grad_norm": 11.182655334472656,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.2714,
      "step": 123
    },
    {
      "epoch": 0.2577962577962578,
      "grad_norm": 11.317505836486816,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.015,
      "step": 124
    },
    {
      "epoch": 0.2598752598752599,
      "grad_norm": 9.808978080749512,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.0205,
      "step": 125
    },
    {
      "epoch": 0.26195426195426197,
      "grad_norm": 12.914642333984375,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.2824,
      "step": 126
    },
    {
      "epoch": 0.26403326403326405,
      "grad_norm": 11.557820320129395,
      "learning_rate": 3.221030765387417e-05,
      "loss": 4.9653,
      "step": 127
    },
    {
      "epoch": 0.2661122661122661,
      "grad_norm": 12.997471809387207,
      "learning_rate": 3.144013755408895e-05,
      "loss": 4.5631,
      "step": 128
    },
    {
      "epoch": 0.2681912681912682,
      "grad_norm": 10.460030555725098,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 4.6488,
      "step": 129
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 10.787405967712402,
      "learning_rate": 2.991522876735154e-05,
      "loss": 4.9893,
      "step": 130
    },
    {
      "epoch": 0.27234927234927236,
      "grad_norm": 11.53608512878418,
      "learning_rate": 2.916090697523549e-05,
      "loss": 4.9134,
      "step": 131
    },
    {
      "epoch": 0.27442827442827444,
      "grad_norm": 10.501526832580566,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 4.8084,
      "step": 132
    },
    {
      "epoch": 0.2765072765072765,
      "grad_norm": 10.274712562561035,
      "learning_rate": 2.766955965735968e-05,
      "loss": 4.8796,
      "step": 133
    },
    {
      "epoch": 0.2785862785862786,
      "grad_norm": 10.84492015838623,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.2624,
      "step": 134
    },
    {
      "epoch": 0.2806652806652807,
      "grad_norm": 11.600790023803711,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.1239,
      "step": 135
    },
    {
      "epoch": 0.28274428274428276,
      "grad_norm": 11.047086715698242,
      "learning_rate": 2.547882480847461e-05,
      "loss": 4.993,
      "step": 136
    },
    {
      "epoch": 0.28482328482328484,
      "grad_norm": 11.589991569519043,
      "learning_rate": 2.476172311325783e-05,
      "loss": 4.7531,
      "step": 137
    },
    {
      "epoch": 0.2869022869022869,
      "grad_norm": 10.702897071838379,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.2714,
      "step": 138
    },
    {
      "epoch": 0.288981288981289,
      "grad_norm": 9.661480903625488,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 4.7616,
      "step": 139
    },
    {
      "epoch": 0.2910602910602911,
      "grad_norm": 11.115570068359375,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 4.908,
      "step": 140
    },
    {
      "epoch": 0.29313929313929316,
      "grad_norm": 11.266924858093262,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.0043,
      "step": 141
    },
    {
      "epoch": 0.29521829521829523,
      "grad_norm": 11.180541038513184,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.5747,
      "step": 142
    },
    {
      "epoch": 0.2972972972972973,
      "grad_norm": 9.850950241088867,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.5562,
      "step": 143
    },
    {
      "epoch": 0.2993762993762994,
      "grad_norm": 11.097012519836426,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.363,
      "step": 144
    },
    {
      "epoch": 0.30145530145530147,
      "grad_norm": 13.673072814941406,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.9579,
      "step": 145
    },
    {
      "epoch": 0.30353430353430355,
      "grad_norm": 11.051887512207031,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 6.2673,
      "step": 146
    },
    {
      "epoch": 0.30561330561330563,
      "grad_norm": 11.962884902954102,
      "learning_rate": 1.800157297483417e-05,
      "loss": 5.6779,
      "step": 147
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 11.67485237121582,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 5.7515,
      "step": 148
    },
    {
      "epoch": 0.3097713097713098,
      "grad_norm": 12.355193138122559,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.4577,
      "step": 149
    },
    {
      "epoch": 0.31185031185031187,
      "grad_norm": 14.26075553894043,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.2312,
      "step": 150
    },
    {
      "epoch": 0.31185031185031187,
      "eval_loss": 1.3336721658706665,
      "eval_runtime": 16.6173,
      "eval_samples_per_second": 48.805,
      "eval_steps_per_second": 12.216,
      "step": 150
    },
    {
      "epoch": 0.31392931392931395,
      "grad_norm": 14.017719268798828,
      "learning_rate": 1.553232954407171e-05,
      "loss": 6.0484,
      "step": 151
    },
    {
      "epoch": 0.316008316008316,
      "grad_norm": 14.372090339660645,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.6974,
      "step": 152
    },
    {
      "epoch": 0.3180873180873181,
      "grad_norm": 14.471624374389648,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.9147,
      "step": 153
    },
    {
      "epoch": 0.3201663201663202,
      "grad_norm": 13.52252197265625,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 5.854,
      "step": 154
    },
    {
      "epoch": 0.32224532224532226,
      "grad_norm": 15.499825477600098,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 5.8749,
      "step": 155
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 15.15707015991211,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 6.1993,
      "step": 156
    },
    {
      "epoch": 0.3264033264033264,
      "grad_norm": 11.873882293701172,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.8482,
      "step": 157
    },
    {
      "epoch": 0.3284823284823285,
      "grad_norm": 11.463122367858887,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.5539,
      "step": 158
    },
    {
      "epoch": 0.3305613305613306,
      "grad_norm": 11.364886283874512,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.6778,
      "step": 159
    },
    {
      "epoch": 0.33264033264033266,
      "grad_norm": 10.653420448303223,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.92,
      "step": 160
    },
    {
      "epoch": 0.33471933471933474,
      "grad_norm": 10.782219886779785,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.4004,
      "step": 161
    },
    {
      "epoch": 0.3367983367983368,
      "grad_norm": 13.527227401733398,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.6093,
      "step": 162
    },
    {
      "epoch": 0.3388773388773389,
      "grad_norm": 9.673450469970703,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.1625,
      "step": 163
    },
    {
      "epoch": 0.340956340956341,
      "grad_norm": 9.782866477966309,
      "learning_rate": 8.599558442598998e-06,
      "loss": 4.9882,
      "step": 164
    },
    {
      "epoch": 0.34303534303534305,
      "grad_norm": 10.166117668151855,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.2874,
      "step": 165
    },
    {
      "epoch": 0.34511434511434513,
      "grad_norm": 10.824355125427246,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.2443,
      "step": 166
    },
    {
      "epoch": 0.3471933471933472,
      "grad_norm": 10.134610176086426,
      "learning_rate": 7.260364370723044e-06,
      "loss": 4.8686,
      "step": 167
    },
    {
      "epoch": 0.3492723492723493,
      "grad_norm": 9.993563652038574,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.1396,
      "step": 168
    },
    {
      "epoch": 0.35135135135135137,
      "grad_norm": 10.504877090454102,
      "learning_rate": 6.425787818636131e-06,
      "loss": 4.9754,
      "step": 169
    },
    {
      "epoch": 0.35343035343035345,
      "grad_norm": 9.01434326171875,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.4939,
      "step": 170
    },
    {
      "epoch": 0.35550935550935553,
      "grad_norm": 9.73424243927002,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 4.9358,
      "step": 171
    },
    {
      "epoch": 0.3575883575883576,
      "grad_norm": 10.597433090209961,
      "learning_rate": 5.263533508961827e-06,
      "loss": 4.6608,
      "step": 172
    },
    {
      "epoch": 0.3596673596673597,
      "grad_norm": 9.57521915435791,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.7085,
      "step": 173
    },
    {
      "epoch": 0.36174636174636177,
      "grad_norm": 9.50676155090332,
      "learning_rate": 4.549673247541875e-06,
      "loss": 4.7269,
      "step": 174
    },
    {
      "epoch": 0.36382536382536385,
      "grad_norm": 10.125983238220215,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.6006,
      "step": 175
    },
    {
      "epoch": 0.3659043659043659,
      "grad_norm": 9.474956512451172,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.4297,
      "step": 176
    },
    {
      "epoch": 0.367983367983368,
      "grad_norm": 10.156993865966797,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 4.9677,
      "step": 177
    },
    {
      "epoch": 0.3700623700623701,
      "grad_norm": 11.636570930480957,
      "learning_rate": 3.271776770026963e-06,
      "loss": 4.9507,
      "step": 178
    },
    {
      "epoch": 0.37214137214137216,
      "grad_norm": 10.199498176574707,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 4.803,
      "step": 179
    },
    {
      "epoch": 0.37422037422037424,
      "grad_norm": 9.61894416809082,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 4.7661,
      "step": 180
    },
    {
      "epoch": 0.3762993762993763,
      "grad_norm": 9.834057807922363,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 4.8538,
      "step": 181
    },
    {
      "epoch": 0.3783783783783784,
      "grad_norm": 10.69369125366211,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 4.6799,
      "step": 182
    },
    {
      "epoch": 0.3804573804573805,
      "grad_norm": 9.11548137664795,
      "learning_rate": 1.962316193157593e-06,
      "loss": 4.9974,
      "step": 183
    },
    {
      "epoch": 0.38253638253638256,
      "grad_norm": 9.363030433654785,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 4.8709,
      "step": 184
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 12.58370590209961,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 4.7915,
      "step": 185
    },
    {
      "epoch": 0.3866943866943867,
      "grad_norm": 10.919997215270996,
      "learning_rate": 1.333670137599713e-06,
      "loss": 4.9068,
      "step": 186
    },
    {
      "epoch": 0.3887733887733888,
      "grad_norm": 10.264413833618164,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.1283,
      "step": 187
    },
    {
      "epoch": 0.3908523908523909,
      "grad_norm": 10.425904273986816,
      "learning_rate": 9.810017062595322e-07,
      "loss": 4.5543,
      "step": 188
    },
    {
      "epoch": 0.39293139293139295,
      "grad_norm": 10.88646125793457,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.1774,
      "step": 189
    },
    {
      "epoch": 0.39501039501039503,
      "grad_norm": 10.167569160461426,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.4664,
      "step": 190
    },
    {
      "epoch": 0.3970893970893971,
      "grad_norm": 10.223713874816895,
      "learning_rate": 5.526064699265753e-07,
      "loss": 4.9284,
      "step": 191
    },
    {
      "epoch": 0.3991683991683992,
      "grad_norm": 10.080347061157227,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.4166,
      "step": 192
    },
    {
      "epoch": 0.40124740124740127,
      "grad_norm": 9.010005950927734,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.4865,
      "step": 193
    },
    {
      "epoch": 0.40332640332640335,
      "grad_norm": 9.998534202575684,
      "learning_rate": 2.458548727494292e-07,
      "loss": 5.724,
      "step": 194
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 9.751326560974121,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.5383,
      "step": 195
    },
    {
      "epoch": 0.4074844074844075,
      "grad_norm": 10.707924842834473,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 6.0355,
      "step": 196
    },
    {
      "epoch": 0.4095634095634096,
      "grad_norm": 9.888345718383789,
      "learning_rate": 6.150154258476315e-08,
      "loss": 5.8942,
      "step": 197
    },
    {
      "epoch": 0.41164241164241167,
      "grad_norm": 10.965300559997559,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 5.9991,
      "step": 198
    },
    {
      "epoch": 0.41372141372141374,
      "grad_norm": 10.796441078186035,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.2615,
      "step": 199
    },
    {
      "epoch": 0.4158004158004158,
      "grad_norm": 10.783673286437988,
      "learning_rate": 0.0,
      "loss": 4.837,
      "step": 200
    },
    {
      "epoch": 0.4158004158004158,
      "eval_loss": 1.3243097066879272,
      "eval_runtime": 16.6366,
      "eval_samples_per_second": 48.748,
      "eval_steps_per_second": 12.202,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 2.98048718635008e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}