{
  "best_metric": 1.6453725099563599,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.13003901170351106,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006501950585175553,
      "grad_norm": 3.150578022003174,
      "learning_rate": 1e-05,
      "loss": 2.4779,
      "step": 1
    },
    {
      "epoch": 0.0006501950585175553,
      "eval_loss": 3.207399845123291,
      "eval_runtime": 275.4986,
      "eval_samples_per_second": 9.405,
      "eval_steps_per_second": 2.352,
      "step": 1
    },
    {
      "epoch": 0.0013003901170351106,
      "grad_norm": 3.1236329078674316,
      "learning_rate": 2e-05,
      "loss": 2.5118,
      "step": 2
    },
    {
      "epoch": 0.0019505851755526658,
      "grad_norm": 2.805406093597412,
      "learning_rate": 3e-05,
      "loss": 2.4361,
      "step": 3
    },
    {
      "epoch": 0.002600780234070221,
      "grad_norm": 1.5615613460540771,
      "learning_rate": 4e-05,
      "loss": 2.4698,
      "step": 4
    },
    {
      "epoch": 0.003250975292587776,
      "grad_norm": 2.6339380741119385,
      "learning_rate": 5e-05,
      "loss": 2.3301,
      "step": 5
    },
    {
      "epoch": 0.0039011703511053317,
      "grad_norm": 3.0237369537353516,
      "learning_rate": 6e-05,
      "loss": 2.3439,
      "step": 6
    },
    {
      "epoch": 0.004551365409622887,
      "grad_norm": 2.2433011531829834,
      "learning_rate": 7e-05,
      "loss": 2.2133,
      "step": 7
    },
    {
      "epoch": 0.005201560468140442,
      "grad_norm": 1.3404631614685059,
      "learning_rate": 8e-05,
      "loss": 2.2192,
      "step": 8
    },
    {
      "epoch": 0.005851755526657998,
      "grad_norm": 1.2992517948150635,
      "learning_rate": 9e-05,
      "loss": 2.093,
      "step": 9
    },
    {
      "epoch": 0.006501950585175552,
      "grad_norm": 1.346543312072754,
      "learning_rate": 0.0001,
      "loss": 2.1585,
      "step": 10
    },
    {
      "epoch": 0.007152145643693108,
      "grad_norm": 1.5351017713546753,
      "learning_rate": 9.999316524962345e-05,
      "loss": 2.2105,
      "step": 11
    },
    {
      "epoch": 0.007802340702210663,
      "grad_norm": 1.1878572702407837,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.0165,
      "step": 12
    },
    {
      "epoch": 0.008452535760728219,
      "grad_norm": 1.1663057804107666,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.0755,
      "step": 13
    },
    {
      "epoch": 0.009102730819245773,
      "grad_norm": 1.0004663467407227,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.9999,
      "step": 14
    },
    {
      "epoch": 0.00975292587776333,
      "grad_norm": 1.1744166612625122,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.949,
      "step": 15
    },
    {
      "epoch": 0.010403120936280884,
      "grad_norm": 1.0961233377456665,
      "learning_rate": 9.975414512725057e-05,
      "loss": 1.9241,
      "step": 16
    },
    {
      "epoch": 0.011053315994798439,
      "grad_norm": 1.1497384309768677,
      "learning_rate": 9.966546331768191e-05,
      "loss": 1.9065,
      "step": 17
    },
    {
      "epoch": 0.011703511053315995,
      "grad_norm": 1.2461047172546387,
      "learning_rate": 9.956320346634876e-05,
      "loss": 2.1369,
      "step": 18
    },
    {
      "epoch": 0.01235370611183355,
      "grad_norm": 1.2522481679916382,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.082,
      "step": 19
    },
    {
      "epoch": 0.013003901170351105,
      "grad_norm": 1.3264752626419067,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.1983,
      "step": 20
    },
    {
      "epoch": 0.013654096228868661,
      "grad_norm": 1.3068052530288696,
      "learning_rate": 9.917525374361912e-05,
      "loss": 1.989,
      "step": 21
    },
    {
      "epoch": 0.014304291287386216,
      "grad_norm": 1.293046236038208,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.1082,
      "step": 22
    },
    {
      "epoch": 0.01495448634590377,
      "grad_norm": 1.3728796243667603,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.0111,
      "step": 23
    },
    {
      "epoch": 0.015604681404421327,
      "grad_norm": 1.4067012071609497,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.0141,
      "step": 24
    },
    {
      "epoch": 0.016254876462938883,
      "grad_norm": 1.4745562076568604,
      "learning_rate": 9.847001329696653e-05,
      "loss": 1.9794,
      "step": 25
    },
    {
      "epoch": 0.016905071521456438,
      "grad_norm": 1.5138723850250244,
      "learning_rate": 9.826044551386744e-05,
      "loss": 1.9876,
      "step": 26
    },
    {
      "epoch": 0.017555266579973992,
      "grad_norm": 1.9568480253219604,
      "learning_rate": 9.803768380684242e-05,
      "loss": 1.9259,
      "step": 27
    },
    {
      "epoch": 0.018205461638491547,
      "grad_norm": 1.8425397872924805,
      "learning_rate": 9.780178907671789e-05,
      "loss": 1.9473,
      "step": 28
    },
    {
      "epoch": 0.0188556566970091,
      "grad_norm": 1.86009681224823,
      "learning_rate": 9.755282581475769e-05,
      "loss": 1.7664,
      "step": 29
    },
    {
      "epoch": 0.01950585175552666,
      "grad_norm": 1.9022787809371948,
      "learning_rate": 9.729086208503174e-05,
      "loss": 1.6201,
      "step": 30
    },
    {
      "epoch": 0.020156046814044214,
      "grad_norm": 1.9312118291854858,
      "learning_rate": 9.701596950580806e-05,
      "loss": 1.7167,
      "step": 31
    },
    {
      "epoch": 0.02080624187256177,
      "grad_norm": 2.2026944160461426,
      "learning_rate": 9.672822322997305e-05,
      "loss": 1.9026,
      "step": 32
    },
    {
      "epoch": 0.021456436931079324,
      "grad_norm": 2.2948923110961914,
      "learning_rate": 9.642770192448536e-05,
      "loss": 1.8421,
      "step": 33
    },
    {
      "epoch": 0.022106631989596878,
      "grad_norm": 2.651780366897583,
      "learning_rate": 9.611448774886924e-05,
      "loss": 1.965,
      "step": 34
    },
    {
      "epoch": 0.022756827048114433,
      "grad_norm": 2.7910056114196777,
      "learning_rate": 9.578866633275288e-05,
      "loss": 1.8887,
      "step": 35
    },
    {
      "epoch": 0.02340702210663199,
      "grad_norm": 2.727020740509033,
      "learning_rate": 9.545032675245813e-05,
      "loss": 1.8668,
      "step": 36
    },
    {
      "epoch": 0.024057217165149546,
      "grad_norm": 3.5199644565582275,
      "learning_rate": 9.509956150664796e-05,
      "loss": 1.9967,
      "step": 37
    },
    {
      "epoch": 0.0247074122236671,
      "grad_norm": 3.35101580619812,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.2491,
      "step": 38
    },
    {
      "epoch": 0.025357607282184655,
      "grad_norm": 3.314497232437134,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.091,
      "step": 39
    },
    {
      "epoch": 0.02600780234070221,
      "grad_norm": 3.239468574523926,
      "learning_rate": 9.397368756032445e-05,
      "loss": 1.8457,
      "step": 40
    },
    {
      "epoch": 0.026657997399219768,
      "grad_norm": 3.625401258468628,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.1501,
      "step": 41
    },
    {
      "epoch": 0.027308192457737322,
      "grad_norm": 3.375628709793091,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.0271,
      "step": 42
    },
    {
      "epoch": 0.027958387516254877,
      "grad_norm": 3.509922504425049,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.1569,
      "step": 43
    },
    {
      "epoch": 0.02860858257477243,
      "grad_norm": 3.69523024559021,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.8829,
      "step": 44
    },
    {
      "epoch": 0.029258777633289986,
      "grad_norm": 4.548573970794678,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.1611,
      "step": 45
    },
    {
      "epoch": 0.02990897269180754,
      "grad_norm": 5.596911430358887,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.3799,
      "step": 46
    },
    {
      "epoch": 0.0305591677503251,
      "grad_norm": 5.344971179962158,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.0883,
      "step": 47
    },
    {
      "epoch": 0.031209362808842653,
      "grad_norm": 5.836085319519043,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.3294,
      "step": 48
    },
    {
      "epoch": 0.03185955786736021,
      "grad_norm": 5.933544158935547,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.6573,
      "step": 49
    },
    {
      "epoch": 0.032509752925877766,
      "grad_norm": 6.288773059844971,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.9916,
      "step": 50
    },
    {
      "epoch": 0.032509752925877766,
      "eval_loss": 2.141352415084839,
      "eval_runtime": 277.5504,
      "eval_samples_per_second": 9.335,
      "eval_steps_per_second": 2.335,
      "step": 50
    },
    {
      "epoch": 0.03315994798439532,
      "grad_norm": 1.6872538328170776,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.0142,
      "step": 51
    },
    {
      "epoch": 0.033810143042912875,
      "grad_norm": 1.4419867992401123,
      "learning_rate": 8.842005554284296e-05,
      "loss": 1.9571,
      "step": 52
    },
    {
      "epoch": 0.03446033810143043,
      "grad_norm": 1.0730379819869995,
      "learning_rate": 8.788574348801675e-05,
      "loss": 1.8388,
      "step": 53
    },
    {
      "epoch": 0.035110533159947985,
      "grad_norm": 1.009609580039978,
      "learning_rate": 8.73410738492077e-05,
      "loss": 1.7745,
      "step": 54
    },
    {
      "epoch": 0.03576072821846554,
      "grad_norm": 0.9554789066314697,
      "learning_rate": 8.678619553365659e-05,
      "loss": 1.8567,
      "step": 55
    },
    {
      "epoch": 0.036410923276983094,
      "grad_norm": 0.8234381079673767,
      "learning_rate": 8.622126023955446e-05,
      "loss": 1.7594,
      "step": 56
    },
    {
      "epoch": 0.03706111833550065,
      "grad_norm": 0.9714184403419495,
      "learning_rate": 8.564642241456986e-05,
      "loss": 1.7398,
      "step": 57
    },
    {
      "epoch": 0.0377113133940182,
      "grad_norm": 1.0201642513275146,
      "learning_rate": 8.506183921362443e-05,
      "loss": 1.6566,
      "step": 58
    },
    {
      "epoch": 0.03836150845253576,
      "grad_norm": 0.9000511169433594,
      "learning_rate": 8.44676704559283e-05,
      "loss": 1.762,
      "step": 59
    },
    {
      "epoch": 0.03901170351105332,
      "grad_norm": 0.8679172396659851,
      "learning_rate": 8.386407858128706e-05,
      "loss": 1.6917,
      "step": 60
    },
    {
      "epoch": 0.03966189856957087,
      "grad_norm": 0.9420127272605896,
      "learning_rate": 8.32512286056924e-05,
      "loss": 1.749,
      "step": 61
    },
    {
      "epoch": 0.04031209362808843,
      "grad_norm": 0.9565222859382629,
      "learning_rate": 8.262928807620843e-05,
      "loss": 1.8037,
      "step": 62
    },
    {
      "epoch": 0.04096228868660598,
      "grad_norm": 0.9313763380050659,
      "learning_rate": 8.199842702516583e-05,
      "loss": 1.8003,
      "step": 63
    },
    {
      "epoch": 0.04161248374512354,
      "grad_norm": 0.9148924946784973,
      "learning_rate": 8.135881792367686e-05,
      "loss": 1.7142,
      "step": 64
    },
    {
      "epoch": 0.04226267880364109,
      "grad_norm": 0.9595693945884705,
      "learning_rate": 8.07106356344834e-05,
      "loss": 1.6453,
      "step": 65
    },
    {
      "epoch": 0.04291287386215865,
      "grad_norm": 1.0527570247650146,
      "learning_rate": 8.005405736415126e-05,
      "loss": 1.7549,
      "step": 66
    },
    {
      "epoch": 0.043563068920676205,
      "grad_norm": 1.0396835803985596,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.6481,
      "step": 67
    },
    {
      "epoch": 0.044213263979193757,
      "grad_norm": 1.10210382938385,
      "learning_rate": 7.871643313414718e-05,
      "loss": 1.679,
      "step": 68
    },
    {
      "epoch": 0.044863459037711315,
      "grad_norm": 1.1436645984649658,
      "learning_rate": 7.803575286758364e-05,
      "loss": 1.7465,
      "step": 69
    },
    {
      "epoch": 0.045513654096228866,
      "grad_norm": 1.1897839307785034,
      "learning_rate": 7.734740790612136e-05,
      "loss": 1.7896,
      "step": 70
    },
    {
      "epoch": 0.046163849154746424,
      "grad_norm": 1.1813805103302002,
      "learning_rate": 7.66515864363997e-05,
      "loss": 1.5994,
      "step": 71
    },
    {
      "epoch": 0.04681404421326398,
      "grad_norm": 1.369110345840454,
      "learning_rate": 7.594847868906076e-05,
      "loss": 1.8466,
      "step": 72
    },
    {
      "epoch": 0.04746423927178153,
      "grad_norm": 1.33378267288208,
      "learning_rate": 7.52382768867422e-05,
      "loss": 1.6806,
      "step": 73
    },
    {
      "epoch": 0.04811443433029909,
      "grad_norm": 1.3799965381622314,
      "learning_rate": 7.452117519152542e-05,
      "loss": 1.7023,
      "step": 74
    },
    {
      "epoch": 0.04876462938881664,
      "grad_norm": 1.5445268154144287,
      "learning_rate": 7.379736965185368e-05,
      "loss": 1.8027,
      "step": 75
    },
    {
      "epoch": 0.0494148244473342,
      "grad_norm": 1.459384560585022,
      "learning_rate": 7.30670581489344e-05,
      "loss": 1.636,
      "step": 76
    },
    {
      "epoch": 0.05006501950585176,
      "grad_norm": 1.5591135025024414,
      "learning_rate": 7.233044034264034e-05,
      "loss": 1.6102,
      "step": 77
    },
    {
      "epoch": 0.05071521456436931,
      "grad_norm": 1.6465017795562744,
      "learning_rate": 7.158771761692464e-05,
      "loss": 1.8642,
      "step": 78
    },
    {
      "epoch": 0.05136540962288687,
      "grad_norm": 1.774129033088684,
      "learning_rate": 7.083909302476453e-05,
      "loss": 1.6801,
      "step": 79
    },
    {
      "epoch": 0.05201560468140442,
      "grad_norm": 1.789703607559204,
      "learning_rate": 7.008477123264848e-05,
      "loss": 1.8245,
      "step": 80
    },
    {
      "epoch": 0.05266579973992198,
      "grad_norm": 1.846843957901001,
      "learning_rate": 6.932495846462261e-05,
      "loss": 1.6955,
      "step": 81
    },
    {
      "epoch": 0.053315994798439535,
      "grad_norm": 2.1186845302581787,
      "learning_rate": 6.855986244591104e-05,
      "loss": 1.7972,
      "step": 82
    },
    {
      "epoch": 0.053966189856957086,
      "grad_norm": 2.200887680053711,
      "learning_rate": 6.778969234612584e-05,
      "loss": 1.648,
      "step": 83
    },
    {
      "epoch": 0.054616384915474644,
      "grad_norm": 2.3800294399261475,
      "learning_rate": 6.701465872208216e-05,
      "loss": 1.7689,
      "step": 84
    },
    {
      "epoch": 0.055266579973992196,
      "grad_norm": 2.290611505508423,
      "learning_rate": 6.623497346023418e-05,
      "loss": 1.5584,
      "step": 85
    },
    {
      "epoch": 0.055916775032509754,
      "grad_norm": 2.658085346221924,
      "learning_rate": 6.545084971874738e-05,
      "loss": 1.82,
      "step": 86
    },
    {
      "epoch": 0.056566970091027305,
      "grad_norm": 2.6808366775512695,
      "learning_rate": 6.466250186922325e-05,
      "loss": 1.7131,
      "step": 87
    },
    {
      "epoch": 0.05721716514954486,
      "grad_norm": 2.5249409675598145,
      "learning_rate": 6.387014543809223e-05,
      "loss": 1.4725,
      "step": 88
    },
    {
      "epoch": 0.05786736020806242,
      "grad_norm": 3.1884984970092773,
      "learning_rate": 6.307399704769099e-05,
      "loss": 1.7134,
      "step": 89
    },
    {
      "epoch": 0.05851755526657997,
      "grad_norm": 3.4987826347351074,
      "learning_rate": 6.227427435703997e-05,
      "loss": 1.9913,
      "step": 90
    },
    {
      "epoch": 0.05916775032509753,
      "grad_norm": 4.086397647857666,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.0549,
      "step": 91
    },
    {
      "epoch": 0.05981794538361508,
      "grad_norm": 3.71805477142334,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.8826,
      "step": 92
    },
    {
      "epoch": 0.06046814044213264,
      "grad_norm": 3.697395086288452,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.9195,
      "step": 93
    },
    {
      "epoch": 0.0611183355006502,
      "grad_norm": 4.0528669357299805,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.8876,
      "step": 94
    },
    {
      "epoch": 0.06176853055916775,
      "grad_norm": 4.652343273162842,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.269,
      "step": 95
    },
    {
      "epoch": 0.06241872561768531,
      "grad_norm": 3.74043607711792,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.646,
      "step": 96
    },
    {
      "epoch": 0.06306892067620286,
      "grad_norm": 4.780050277709961,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.992,
      "step": 97
    },
    {
      "epoch": 0.06371911573472042,
      "grad_norm": 5.511683940887451,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.4362,
      "step": 98
    },
    {
      "epoch": 0.06436931079323797,
      "grad_norm": 6.257039546966553,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.5217,
      "step": 99
    },
    {
      "epoch": 0.06501950585175553,
      "grad_norm": 6.353720664978027,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.5057,
      "step": 100
    },
    {
      "epoch": 0.06501950585175553,
      "eval_loss": 1.8925142288208008,
      "eval_runtime": 277.2751,
      "eval_samples_per_second": 9.345,
      "eval_steps_per_second": 2.337,
      "step": 100
    },
    {
      "epoch": 0.06566970091027308,
      "grad_norm": 1.5737056732177734,
      "learning_rate": 5.330452921628497e-05,
      "loss": 1.7769,
      "step": 101
    },
    {
      "epoch": 0.06631989596879063,
      "grad_norm": 1.196608543395996,
      "learning_rate": 5.247918773366112e-05,
      "loss": 1.6637,
      "step": 102
    },
    {
      "epoch": 0.06697009102730819,
      "grad_norm": 1.0653871297836304,
      "learning_rate": 5.165316846586541e-05,
      "loss": 1.6869,
      "step": 103
    },
    {
      "epoch": 0.06762028608582575,
      "grad_norm": 1.0002870559692383,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 1.637,
      "step": 104
    },
    {
      "epoch": 0.06827048114434331,
      "grad_norm": 0.8883082270622253,
      "learning_rate": 5e-05,
      "loss": 1.5519,
      "step": 105
    },
    {
      "epoch": 0.06892067620286085,
      "grad_norm": 0.9232680797576904,
      "learning_rate": 4.917330276168208e-05,
      "loss": 1.6784,
      "step": 106
    },
    {
      "epoch": 0.06957087126137841,
      "grad_norm": 0.9425049424171448,
      "learning_rate": 4.834683153413459e-05,
      "loss": 1.7163,
      "step": 107
    },
    {
      "epoch": 0.07022106631989597,
      "grad_norm": 0.8800442218780518,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 1.5535,
      "step": 108
    },
    {
      "epoch": 0.07087126137841353,
      "grad_norm": 0.9336991906166077,
      "learning_rate": 4.669547078371504e-05,
      "loss": 1.591,
      "step": 109
    },
    {
      "epoch": 0.07152145643693109,
      "grad_norm": 0.951155960559845,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 1.5177,
      "step": 110
    },
    {
      "epoch": 0.07217165149544863,
      "grad_norm": 0.9872428774833679,
      "learning_rate": 4.504772348747687e-05,
      "loss": 1.554,
      "step": 111
    },
    {
      "epoch": 0.07282184655396619,
      "grad_norm": 1.0234076976776123,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 1.6269,
      "step": 112
    },
    {
      "epoch": 0.07347204161248375,
      "grad_norm": 0.9843485951423645,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 1.7401,
      "step": 113
    },
    {
      "epoch": 0.0741222366710013,
      "grad_norm": 1.0014989376068115,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 1.4811,
      "step": 114
    },
    {
      "epoch": 0.07477243172951886,
      "grad_norm": 1.0141053199768066,
      "learning_rate": 4.17702704859633e-05,
      "loss": 1.5014,
      "step": 115
    },
    {
      "epoch": 0.0754226267880364,
      "grad_norm": 1.0666749477386475,
      "learning_rate": 4.095597328339452e-05,
      "loss": 1.7403,
      "step": 116
    },
    {
      "epoch": 0.07607282184655396,
      "grad_norm": 1.0580428838729858,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 1.5787,
      "step": 117
    },
    {
      "epoch": 0.07672301690507152,
      "grad_norm": 1.2409926652908325,
      "learning_rate": 3.933501846281267e-05,
      "loss": 1.6983,
      "step": 118
    },
    {
      "epoch": 0.07737321196358908,
      "grad_norm": 1.1604779958724976,
      "learning_rate": 3.852880399766243e-05,
      "loss": 1.603,
      "step": 119
    },
    {
      "epoch": 0.07802340702210664,
      "grad_norm": 1.2414381504058838,
      "learning_rate": 3.772572564296005e-05,
      "loss": 1.6353,
      "step": 120
    },
    {
      "epoch": 0.07867360208062418,
      "grad_norm": 1.2233747243881226,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 1.5125,
      "step": 121
    },
    {
      "epoch": 0.07932379713914174,
      "grad_norm": 1.249691128730774,
      "learning_rate": 3.612985456190778e-05,
      "loss": 1.6696,
      "step": 122
    },
    {
      "epoch": 0.0799739921976593,
      "grad_norm": 1.3441444635391235,
      "learning_rate": 3.533749813077677e-05,
      "loss": 1.6229,
      "step": 123
    },
    {
      "epoch": 0.08062418725617686,
      "grad_norm": 1.4655382633209229,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 1.7122,
      "step": 124
    },
    {
      "epoch": 0.0812743823146944,
      "grad_norm": 1.497069239616394,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 1.651,
      "step": 125
    },
    {
      "epoch": 0.08192457737321196,
      "grad_norm": 1.6520121097564697,
      "learning_rate": 3.298534127791785e-05,
      "loss": 1.6502,
      "step": 126
    },
    {
      "epoch": 0.08257477243172952,
      "grad_norm": 1.6223881244659424,
      "learning_rate": 3.221030765387417e-05,
      "loss": 1.6955,
      "step": 127
    },
    {
      "epoch": 0.08322496749024708,
      "grad_norm": 1.6531888246536255,
      "learning_rate": 3.144013755408895e-05,
      "loss": 1.6713,
      "step": 128
    },
    {
      "epoch": 0.08387516254876463,
      "grad_norm": 1.7278934717178345,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 1.7045,
      "step": 129
    },
    {
      "epoch": 0.08452535760728218,
      "grad_norm": 2.1566648483276367,
      "learning_rate": 2.991522876735154e-05,
      "loss": 1.7417,
      "step": 130
    },
    {
      "epoch": 0.08517555266579974,
      "grad_norm": 1.9919639825820923,
      "learning_rate": 2.916090697523549e-05,
      "loss": 1.71,
      "step": 131
    },
    {
      "epoch": 0.0858257477243173,
      "grad_norm": 2.004662275314331,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 1.4817,
      "step": 132
    },
    {
      "epoch": 0.08647594278283485,
      "grad_norm": 2.1898128986358643,
      "learning_rate": 2.766955965735968e-05,
      "loss": 1.7235,
      "step": 133
    },
    {
      "epoch": 0.08712613784135241,
      "grad_norm": 2.319091558456421,
      "learning_rate": 2.693294185106562e-05,
      "loss": 1.6722,
      "step": 134
    },
    {
      "epoch": 0.08777633289986995,
      "grad_norm": 2.796513557434082,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 1.8985,
      "step": 135
    },
    {
      "epoch": 0.08842652795838751,
      "grad_norm": 3.0333023071289062,
      "learning_rate": 2.547882480847461e-05,
      "loss": 1.9842,
      "step": 136
    },
    {
      "epoch": 0.08907672301690507,
      "grad_norm": 3.1889796257019043,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.2586,
      "step": 137
    },
    {
      "epoch": 0.08972691807542263,
      "grad_norm": 3.9190168380737305,
      "learning_rate": 2.405152131093926e-05,
      "loss": 1.9954,
      "step": 138
    },
    {
      "epoch": 0.09037711313394019,
      "grad_norm": 3.3880276679992676,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 1.8613,
      "step": 139
    },
    {
      "epoch": 0.09102730819245773,
      "grad_norm": 4.007750511169434,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 1.9913,
      "step": 140
    },
    {
      "epoch": 0.09167750325097529,
      "grad_norm": 3.6146087646484375,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.9652,
      "step": 141
    },
    {
      "epoch": 0.09232769830949285,
      "grad_norm": 3.7341558933258057,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.0953,
      "step": 142
    },
    {
      "epoch": 0.0929778933680104,
      "grad_norm": 3.877579689025879,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.0519,
      "step": 143
    },
    {
      "epoch": 0.09362808842652796,
      "grad_norm": 4.308061599731445,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.9195,
      "step": 144
    },
    {
      "epoch": 0.09427828348504551,
      "grad_norm": 4.26055908203125,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.0899,
      "step": 145
    },
    {
      "epoch": 0.09492847854356307,
      "grad_norm": 5.060686111450195,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.205,
      "step": 146
    },
    {
      "epoch": 0.09557867360208062,
      "grad_norm": 5.254965305328369,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.9174,
      "step": 147
    },
    {
      "epoch": 0.09622886866059818,
      "grad_norm": 4.360957622528076,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.6288,
      "step": 148
    },
    {
      "epoch": 0.09687906371911574,
      "grad_norm": 5.653236389160156,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 1.8962,
      "step": 149
    },
    {
      "epoch": 0.09752925877763328,
      "grad_norm": 8.28332233428955,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.6263,
      "step": 150
    },
    {
      "epoch": 0.09752925877763328,
      "eval_loss": 1.681923270225525,
      "eval_runtime": 277.1289,
      "eval_samples_per_second": 9.349,
      "eval_steps_per_second": 2.338,
      "step": 150
    },
    {
      "epoch": 0.09817945383615084,
      "grad_norm": 1.150153398513794,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.4591,
      "step": 151
    },
    {
      "epoch": 0.0988296488946684,
      "grad_norm": 1.1365052461624146,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.4918,
      "step": 152
    },
    {
      "epoch": 0.09947984395318596,
      "grad_norm": 0.9435325264930725,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.4369,
      "step": 153
    },
    {
      "epoch": 0.10013003901170352,
      "grad_norm": 0.9590620398521423,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 1.5005,
      "step": 154
    },
    {
      "epoch": 0.10078023407022106,
      "grad_norm": 0.8837301731109619,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.4298,
      "step": 155
    },
    {
      "epoch": 0.10143042912873862,
      "grad_norm": 0.8774361610412598,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.37,
      "step": 156
    },
    {
      "epoch": 0.10208062418725618,
      "grad_norm": 0.9308640956878662,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 1.4281,
      "step": 157
    },
    {
      "epoch": 0.10273081924577374,
      "grad_norm": 1.035635232925415,
      "learning_rate": 1.157994445715706e-05,
      "loss": 1.501,
      "step": 158
    },
    {
      "epoch": 0.1033810143042913,
      "grad_norm": 1.0066900253295898,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 1.5392,
      "step": 159
    },
    {
      "epoch": 0.10403120936280884,
      "grad_norm": 1.0178413391113281,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.4993,
      "step": 160
    },
    {
      "epoch": 0.1046814044213264,
      "grad_norm": 1.0421100854873657,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 1.579,
      "step": 161
    },
    {
      "epoch": 0.10533159947984395,
      "grad_norm": 0.9771685004234314,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.5652,
      "step": 162
    },
    {
      "epoch": 0.10598179453836151,
      "grad_norm": 1.0289626121520996,
      "learning_rate": 9.068759265665384e-06,
      "loss": 1.6163,
      "step": 163
    },
    {
      "epoch": 0.10663198959687907,
      "grad_norm": 0.9808756113052368,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.4928,
      "step": 164
    },
    {
      "epoch": 0.10728218465539661,
      "grad_norm": 1.117025375366211,
      "learning_rate": 8.141676086873572e-06,
      "loss": 1.6242,
      "step": 165
    },
    {
      "epoch": 0.10793237971391417,
      "grad_norm": 1.067068338394165,
      "learning_rate": 7.695237378953223e-06,
      "loss": 1.553,
      "step": 166
    },
    {
      "epoch": 0.10858257477243173,
      "grad_norm": 1.0739459991455078,
      "learning_rate": 7.260364370723044e-06,
      "loss": 1.5172,
      "step": 167
    },
    {
      "epoch": 0.10923276983094929,
      "grad_norm": 1.1459896564483643,
      "learning_rate": 6.837175952121306e-06,
      "loss": 1.7501,
      "step": 168
    },
    {
      "epoch": 0.10988296488946683,
      "grad_norm": 1.2288655042648315,
      "learning_rate": 6.425787818636131e-06,
      "loss": 1.5428,
      "step": 169
    },
    {
      "epoch": 0.11053315994798439,
      "grad_norm": 1.1887933015823364,
      "learning_rate": 6.026312439675552e-06,
      "loss": 1.526,
      "step": 170
    },
    {
      "epoch": 0.11118335500650195,
      "grad_norm": 1.3447848558425903,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 1.5976,
      "step": 171
    },
    {
      "epoch": 0.11183355006501951,
      "grad_norm": 1.2542001008987427,
      "learning_rate": 5.263533508961827e-06,
      "loss": 1.5402,
      "step": 172
    },
    {
      "epoch": 0.11248374512353707,
      "grad_norm": 1.3023407459259033,
      "learning_rate": 4.900438493352055e-06,
      "loss": 1.6799,
      "step": 173
    },
    {
      "epoch": 0.11313394018205461,
      "grad_norm": 1.3968709707260132,
      "learning_rate": 4.549673247541875e-06,
      "loss": 1.5237,
      "step": 174
    },
    {
      "epoch": 0.11378413524057217,
      "grad_norm": 1.5408791303634644,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 1.7118,
      "step": 175
    },
    {
      "epoch": 0.11443433029908973,
      "grad_norm": 1.5033438205718994,
      "learning_rate": 3.885512251130763e-06,
      "loss": 1.5824,
      "step": 176
    },
    {
      "epoch": 0.11508452535760728,
      "grad_norm": 1.4552483558654785,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 1.4535,
      "step": 177
    },
    {
      "epoch": 0.11573472041612484,
      "grad_norm": 1.5696234703063965,
      "learning_rate": 3.271776770026963e-06,
      "loss": 1.4988,
      "step": 178
    },
    {
      "epoch": 0.11638491547464239,
      "grad_norm": 2.0923192501068115,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.4789,
      "step": 179
    },
    {
      "epoch": 0.11703511053315994,
      "grad_norm": 1.9070204496383667,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.7235,
      "step": 180
    },
    {
      "epoch": 0.1176853055916775,
      "grad_norm": 1.7953242063522339,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.3923,
      "step": 181
    },
    {
      "epoch": 0.11833550065019506,
      "grad_norm": 2.046569585800171,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.6432,
      "step": 182
    },
    {
      "epoch": 0.11898569570871262,
      "grad_norm": 2.286076784133911,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.5482,
      "step": 183
    },
    {
      "epoch": 0.11963589076723016,
      "grad_norm": 2.4285292625427246,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 1.8515,
      "step": 184
    },
    {
      "epoch": 0.12028608582574772,
      "grad_norm": 2.57780385017395,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.8352,
      "step": 185
    },
    {
      "epoch": 0.12093628088426528,
      "grad_norm": 3.3238699436187744,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.0833,
      "step": 186
    },
    {
      "epoch": 0.12158647594278284,
      "grad_norm": 3.0119619369506836,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 1.7637,
      "step": 187
    },
    {
      "epoch": 0.1222366710013004,
      "grad_norm": 3.2846052646636963,
      "learning_rate": 9.810017062595322e-07,
      "loss": 1.8315,
      "step": 188
    },
    {
      "epoch": 0.12288686605981794,
      "grad_norm": 2.96270751953125,
      "learning_rate": 8.247462563808817e-07,
      "loss": 1.734,
      "step": 189
    },
    {
      "epoch": 0.1235370611183355,
      "grad_norm": 3.2978405952453613,
      "learning_rate": 6.819348298638839e-07,
      "loss": 1.7422,
      "step": 190
    },
    {
      "epoch": 0.12418725617685306,
      "grad_norm": 3.5689170360565186,
      "learning_rate": 5.526064699265753e-07,
      "loss": 1.7226,
      "step": 191
    },
    {
      "epoch": 0.12483745123537061,
      "grad_norm": 3.528951406478882,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.76,
      "step": 192
    },
    {
      "epoch": 0.12548764629388817,
      "grad_norm": 3.643585205078125,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 1.7848,
      "step": 193
    },
    {
      "epoch": 0.12613784135240572,
      "grad_norm": 5.217795372009277,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.1788,
      "step": 194
    },
    {
      "epoch": 0.1267880364109233,
      "grad_norm": 4.407168865203857,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.0547,
      "step": 195
    },
    {
      "epoch": 0.12743823146944083,
      "grad_norm": 3.9154980182647705,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.6681,
      "step": 196
    },
    {
      "epoch": 0.12808842652795838,
      "grad_norm": 3.8873953819274902,
      "learning_rate": 6.150154258476315e-08,
      "loss": 1.4196,
      "step": 197
    },
    {
      "epoch": 0.12873862158647595,
      "grad_norm": 4.719685077667236,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.862,
      "step": 198
    },
    {
      "epoch": 0.1293888166449935,
      "grad_norm": 4.893847465515137,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.0108,
      "step": 199
    },
    {
      "epoch": 0.13003901170351106,
      "grad_norm": 6.191561222076416,
      "learning_rate": 0.0,
      "loss": 2.2662,
      "step": 200
    },
    {
      "epoch": 0.13003901170351106,
      "eval_loss": 1.6453725099563599,
      "eval_runtime": 277.2849,
      "eval_samples_per_second": 9.344,
      "eval_steps_per_second": 2.337,
      "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": 3.358172450390016e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}