{
  "best_metric": 0.7726338505744934,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.46893317702227433,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0023446658851113715,
      "grad_norm": 23.239633560180664,
      "learning_rate": 1e-05,
      "loss": 5.4807,
      "step": 1
    },
    {
      "epoch": 0.0023446658851113715,
      "eval_loss": 1.620888352394104,
      "eval_runtime": 52.8298,
      "eval_samples_per_second": 13.591,
      "eval_steps_per_second": 3.407,
      "step": 1
    },
    {
      "epoch": 0.004689331770222743,
      "grad_norm": 30.096290588378906,
      "learning_rate": 2e-05,
      "loss": 5.8935,
      "step": 2
    },
    {
      "epoch": 0.007033997655334115,
      "grad_norm": 25.60515785217285,
      "learning_rate": 3e-05,
      "loss": 5.8402,
      "step": 3
    },
    {
      "epoch": 0.009378663540445486,
      "grad_norm": 23.096664428710938,
      "learning_rate": 4e-05,
      "loss": 5.3388,
      "step": 4
    },
    {
      "epoch": 0.011723329425556858,
      "grad_norm": 23.348421096801758,
      "learning_rate": 5e-05,
      "loss": 5.397,
      "step": 5
    },
    {
      "epoch": 0.01406799531066823,
      "grad_norm": 16.431598663330078,
      "learning_rate": 6e-05,
      "loss": 4.4782,
      "step": 6
    },
    {
      "epoch": 0.016412661195779603,
      "grad_norm": 16.66216468811035,
      "learning_rate": 7e-05,
      "loss": 4.2424,
      "step": 7
    },
    {
      "epoch": 0.01875732708089097,
      "grad_norm": 13.334493637084961,
      "learning_rate": 8e-05,
      "loss": 4.0755,
      "step": 8
    },
    {
      "epoch": 0.021101992966002344,
      "grad_norm": 11.975020408630371,
      "learning_rate": 9e-05,
      "loss": 3.9145,
      "step": 9
    },
    {
      "epoch": 0.023446658851113716,
      "grad_norm": 11.74121379852295,
      "learning_rate": 0.0001,
      "loss": 4.0372,
      "step": 10
    },
    {
      "epoch": 0.02579132473622509,
      "grad_norm": 12.151472091674805,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.9669,
      "step": 11
    },
    {
      "epoch": 0.02813599062133646,
      "grad_norm": 10.813011169433594,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.9362,
      "step": 12
    },
    {
      "epoch": 0.03048065650644783,
      "grad_norm": 14.49845027923584,
      "learning_rate": 9.993849845741524e-05,
      "loss": 4.221,
      "step": 13
    },
    {
      "epoch": 0.032825322391559206,
      "grad_norm": 9.16484546661377,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.3724,
      "step": 14
    },
    {
      "epoch": 0.035169988276670575,
      "grad_norm": 13.538769721984863,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.8166,
      "step": 15
    },
    {
      "epoch": 0.03751465416178194,
      "grad_norm": 9.84959602355957,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.6727,
      "step": 16
    },
    {
      "epoch": 0.03985932004689332,
      "grad_norm": 10.60831356048584,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.5212,
      "step": 17
    },
    {
      "epoch": 0.04220398593200469,
      "grad_norm": 9.566568374633789,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.7478,
      "step": 18
    },
    {
      "epoch": 0.044548651817116064,
      "grad_norm": 8.169366836547852,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.3065,
      "step": 19
    },
    {
      "epoch": 0.04689331770222743,
      "grad_norm": 7.885672569274902,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.1369,
      "step": 20
    },
    {
      "epoch": 0.0492379835873388,
      "grad_norm": 8.924588203430176,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.5698,
      "step": 21
    },
    {
      "epoch": 0.05158264947245018,
      "grad_norm": 8.621098518371582,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.5412,
      "step": 22
    },
    {
      "epoch": 0.053927315357561546,
      "grad_norm": 8.363326072692871,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.3901,
      "step": 23
    },
    {
      "epoch": 0.05627198124267292,
      "grad_norm": 8.602740287780762,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.0478,
      "step": 24
    },
    {
      "epoch": 0.05861664712778429,
      "grad_norm": 11.961050987243652,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.8124,
      "step": 25
    },
    {
      "epoch": 0.06096131301289566,
      "grad_norm": 8.368843078613281,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.413,
      "step": 26
    },
    {
      "epoch": 0.06330597889800703,
      "grad_norm": 13.979700088500977,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.3493,
      "step": 27
    },
    {
      "epoch": 0.06565064478311841,
      "grad_norm": 10.302687644958496,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.1922,
      "step": 28
    },
    {
      "epoch": 0.06799531066822978,
      "grad_norm": 8.424847602844238,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.0255,
      "step": 29
    },
    {
      "epoch": 0.07033997655334115,
      "grad_norm": 8.677651405334473,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.9156,
      "step": 30
    },
    {
      "epoch": 0.07268464243845252,
      "grad_norm": 8.514183044433594,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.0409,
      "step": 31
    },
    {
      "epoch": 0.07502930832356389,
      "grad_norm": 8.841915130615234,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.0594,
      "step": 32
    },
    {
      "epoch": 0.07737397420867527,
      "grad_norm": 9.110976219177246,
      "learning_rate": 9.642770192448536e-05,
      "loss": 3.2144,
      "step": 33
    },
    {
      "epoch": 0.07971864009378664,
      "grad_norm": 8.615434646606445,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.9209,
      "step": 34
    },
    {
      "epoch": 0.08206330597889801,
      "grad_norm": 8.627106666564941,
      "learning_rate": 9.578866633275288e-05,
      "loss": 3.3399,
      "step": 35
    },
    {
      "epoch": 0.08440797186400938,
      "grad_norm": 8.770901679992676,
      "learning_rate": 9.545032675245813e-05,
      "loss": 3.0987,
      "step": 36
    },
    {
      "epoch": 0.08675263774912075,
      "grad_norm": 9.875587463378906,
      "learning_rate": 9.509956150664796e-05,
      "loss": 3.2772,
      "step": 37
    },
    {
      "epoch": 0.08909730363423213,
      "grad_norm": 9.106793403625488,
      "learning_rate": 9.473646649103818e-05,
      "loss": 3.3099,
      "step": 38
    },
    {
      "epoch": 0.0914419695193435,
      "grad_norm": 7.894020080566406,
      "learning_rate": 9.43611409721806e-05,
      "loss": 3.2431,
      "step": 39
    },
    {
      "epoch": 0.09378663540445487,
      "grad_norm": 8.276036262512207,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.3587,
      "step": 40
    },
    {
      "epoch": 0.09613130128956623,
      "grad_norm": 12.071616172790527,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.2565,
      "step": 41
    },
    {
      "epoch": 0.0984759671746776,
      "grad_norm": 8.251750946044922,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.8408,
      "step": 42
    },
    {
      "epoch": 0.10082063305978899,
      "grad_norm": 7.910401344299316,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.7768,
      "step": 43
    },
    {
      "epoch": 0.10316529894490035,
      "grad_norm": 9.235706329345703,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.1056,
      "step": 44
    },
    {
      "epoch": 0.10550996483001172,
      "grad_norm": 9.143514633178711,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.5952,
      "step": 45
    },
    {
      "epoch": 0.10785463071512309,
      "grad_norm": 8.614946365356445,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.6248,
      "step": 46
    },
    {
      "epoch": 0.11019929660023446,
      "grad_norm": 8.619750022888184,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.4423,
      "step": 47
    },
    {
      "epoch": 0.11254396248534584,
      "grad_norm": 9.125701904296875,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.4764,
      "step": 48
    },
    {
      "epoch": 0.11488862837045721,
      "grad_norm": 9.610570907592773,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.1642,
      "step": 49
    },
    {
      "epoch": 0.11723329425556858,
      "grad_norm": 11.419136047363281,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.8759,
      "step": 50
    },
    {
      "epoch": 0.11723329425556858,
      "eval_loss": 1.0751440525054932,
      "eval_runtime": 53.7599,
      "eval_samples_per_second": 13.356,
      "eval_steps_per_second": 3.348,
      "step": 50
    },
    {
      "epoch": 0.11957796014067995,
      "grad_norm": 26.25602149963379,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.8854,
      "step": 51
    },
    {
      "epoch": 0.12192262602579132,
      "grad_norm": 18.95648956298828,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.8606,
      "step": 52
    },
    {
      "epoch": 0.1242672919109027,
      "grad_norm": 9.83833122253418,
      "learning_rate": 8.788574348801675e-05,
      "loss": 4.3101,
      "step": 53
    },
    {
      "epoch": 0.12661195779601406,
      "grad_norm": 9.714905738830566,
      "learning_rate": 8.73410738492077e-05,
      "loss": 4.5283,
      "step": 54
    },
    {
      "epoch": 0.12895662368112543,
      "grad_norm": 10.457443237304688,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.788,
      "step": 55
    },
    {
      "epoch": 0.13130128956623682,
      "grad_norm": 8.392775535583496,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.9433,
      "step": 56
    },
    {
      "epoch": 0.1336459554513482,
      "grad_norm": 8.235962867736816,
      "learning_rate": 8.564642241456986e-05,
      "loss": 4.0651,
      "step": 57
    },
    {
      "epoch": 0.13599062133645956,
      "grad_norm": 9.83517074584961,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.9383,
      "step": 58
    },
    {
      "epoch": 0.13833528722157093,
      "grad_norm": 7.149994850158691,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.9611,
      "step": 59
    },
    {
      "epoch": 0.1406799531066823,
      "grad_norm": 7.5323333740234375,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.5999,
      "step": 60
    },
    {
      "epoch": 0.14302461899179367,
      "grad_norm": 9.259871482849121,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.634,
      "step": 61
    },
    {
      "epoch": 0.14536928487690504,
      "grad_norm": 8.562212944030762,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.7187,
      "step": 62
    },
    {
      "epoch": 0.1477139507620164,
      "grad_norm": 8.278037071228027,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.6186,
      "step": 63
    },
    {
      "epoch": 0.15005861664712777,
      "grad_norm": 8.255841255187988,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.5898,
      "step": 64
    },
    {
      "epoch": 0.15240328253223914,
      "grad_norm": 8.422991752624512,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.465,
      "step": 65
    },
    {
      "epoch": 0.15474794841735054,
      "grad_norm": 7.391692638397217,
      "learning_rate": 8.005405736415126e-05,
      "loss": 3.7774,
      "step": 66
    },
    {
      "epoch": 0.1570926143024619,
      "grad_norm": 6.739409923553467,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.4421,
      "step": 67
    },
    {
      "epoch": 0.15943728018757328,
      "grad_norm": 8.762239456176758,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.3068,
      "step": 68
    },
    {
      "epoch": 0.16178194607268465,
      "grad_norm": 8.467728614807129,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.8213,
      "step": 69
    },
    {
      "epoch": 0.16412661195779601,
      "grad_norm": 6.069597244262695,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.2578,
      "step": 70
    },
    {
      "epoch": 0.16647127784290738,
      "grad_norm": 8.404447555541992,
      "learning_rate": 7.66515864363997e-05,
      "loss": 3.959,
      "step": 71
    },
    {
      "epoch": 0.16881594372801875,
      "grad_norm": 7.48832368850708,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.9806,
      "step": 72
    },
    {
      "epoch": 0.17116060961313012,
      "grad_norm": 6.894347667694092,
      "learning_rate": 7.52382768867422e-05,
      "loss": 3.0812,
      "step": 73
    },
    {
      "epoch": 0.1735052754982415,
      "grad_norm": 6.712637901306152,
      "learning_rate": 7.452117519152542e-05,
      "loss": 3.208,
      "step": 74
    },
    {
      "epoch": 0.17584994138335286,
      "grad_norm": 6.727443695068359,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.1937,
      "step": 75
    },
    {
      "epoch": 0.17819460726846426,
      "grad_norm": 7.373191833496094,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.103,
      "step": 76
    },
    {
      "epoch": 0.18053927315357562,
      "grad_norm": 6.736911296844482,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.7025,
      "step": 77
    },
    {
      "epoch": 0.182883939038687,
      "grad_norm": 8.696050643920898,
      "learning_rate": 7.158771761692464e-05,
      "loss": 3.0454,
      "step": 78
    },
    {
      "epoch": 0.18522860492379836,
      "grad_norm": 8.053873062133789,
      "learning_rate": 7.083909302476453e-05,
      "loss": 3.0847,
      "step": 79
    },
    {
      "epoch": 0.18757327080890973,
      "grad_norm": 7.666583061218262,
      "learning_rate": 7.008477123264848e-05,
      "loss": 3.2359,
      "step": 80
    },
    {
      "epoch": 0.1899179366940211,
      "grad_norm": 6.970043182373047,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.5973,
      "step": 81
    },
    {
      "epoch": 0.19226260257913247,
      "grad_norm": 7.732367992401123,
      "learning_rate": 6.855986244591104e-05,
      "loss": 3.0107,
      "step": 82
    },
    {
      "epoch": 0.19460726846424384,
      "grad_norm": 7.392098426818848,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.7857,
      "step": 83
    },
    {
      "epoch": 0.1969519343493552,
      "grad_norm": 7.76128625869751,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.9686,
      "step": 84
    },
    {
      "epoch": 0.19929660023446658,
      "grad_norm": 7.578400135040283,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.1077,
      "step": 85
    },
    {
      "epoch": 0.20164126611957797,
      "grad_norm": 8.27650260925293,
      "learning_rate": 6.545084971874738e-05,
      "loss": 3.2833,
      "step": 86
    },
    {
      "epoch": 0.20398593200468934,
      "grad_norm": 9.004765510559082,
      "learning_rate": 6.466250186922325e-05,
      "loss": 3.2852,
      "step": 87
    },
    {
      "epoch": 0.2063305978898007,
      "grad_norm": 8.086798667907715,
      "learning_rate": 6.387014543809223e-05,
      "loss": 3.0326,
      "step": 88
    },
    {
      "epoch": 0.20867526377491208,
      "grad_norm": 8.294204711914062,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.8488,
      "step": 89
    },
    {
      "epoch": 0.21101992966002345,
      "grad_norm": 7.493625164031982,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.7359,
      "step": 90
    },
    {
      "epoch": 0.21336459554513482,
      "grad_norm": 8.053818702697754,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.1962,
      "step": 91
    },
    {
      "epoch": 0.21570926143024619,
      "grad_norm": 7.4961395263671875,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.2751,
      "step": 92
    },
    {
      "epoch": 0.21805392731535755,
      "grad_norm": 7.001191139221191,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.8502,
      "step": 93
    },
    {
      "epoch": 0.22039859320046892,
      "grad_norm": 7.234589576721191,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.8708,
      "step": 94
    },
    {
      "epoch": 0.2227432590855803,
      "grad_norm": 7.190276145935059,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.4968,
      "step": 95
    },
    {
      "epoch": 0.2250879249706917,
      "grad_norm": 7.514248371124268,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.7352,
      "step": 96
    },
    {
      "epoch": 0.22743259085580306,
      "grad_norm": 7.901513576507568,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.9496,
      "step": 97
    },
    {
      "epoch": 0.22977725674091443,
      "grad_norm": 6.377729892730713,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.2929,
      "step": 98
    },
    {
      "epoch": 0.2321219226260258,
      "grad_norm": 7.4685797691345215,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.2377,
      "step": 99
    },
    {
      "epoch": 0.23446658851113716,
      "grad_norm": 7.034140586853027,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.407,
      "step": 100
    },
    {
      "epoch": 0.23446658851113716,
      "eval_loss": 0.9282410144805908,
      "eval_runtime": 53.7567,
      "eval_samples_per_second": 13.356,
      "eval_steps_per_second": 3.348,
      "step": 100
    },
    {
      "epoch": 0.23681125439624853,
      "grad_norm": 13.492389678955078,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.208,
      "step": 101
    },
    {
      "epoch": 0.2391559202813599,
      "grad_norm": 11.557668685913086,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.7059,
      "step": 102
    },
    {
      "epoch": 0.24150058616647127,
      "grad_norm": 7.9621806144714355,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.8156,
      "step": 103
    },
    {
      "epoch": 0.24384525205158264,
      "grad_norm": 6.837264537811279,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.1534,
      "step": 104
    },
    {
      "epoch": 0.246189917936694,
      "grad_norm": 6.286181449890137,
      "learning_rate": 5e-05,
      "loss": 3.4996,
      "step": 105
    },
    {
      "epoch": 0.2485345838218054,
      "grad_norm": 7.520151615142822,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.0659,
      "step": 106
    },
    {
      "epoch": 0.2508792497069168,
      "grad_norm": 7.174684524536133,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.6981,
      "step": 107
    },
    {
      "epoch": 0.2532239155920281,
      "grad_norm": 5.834714889526367,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.3225,
      "step": 108
    },
    {
      "epoch": 0.2555685814771395,
      "grad_norm": 6.822752475738525,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.6656,
      "step": 109
    },
    {
      "epoch": 0.25791324736225085,
      "grad_norm": 7.0578227043151855,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.5365,
      "step": 110
    },
    {
      "epoch": 0.26025791324736225,
      "grad_norm": 6.7616424560546875,
      "learning_rate": 4.504772348747687e-05,
      "loss": 3.4461,
      "step": 111
    },
    {
      "epoch": 0.26260257913247365,
      "grad_norm": 6.884385585784912,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.5808,
      "step": 112
    },
    {
      "epoch": 0.264947245017585,
      "grad_norm": 7.327887058258057,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 4.027,
      "step": 113
    },
    {
      "epoch": 0.2672919109026964,
      "grad_norm": 7.241699695587158,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.6953,
      "step": 114
    },
    {
      "epoch": 0.2696365767878077,
      "grad_norm": 6.595090866088867,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.1933,
      "step": 115
    },
    {
      "epoch": 0.2719812426729191,
      "grad_norm": 6.4801483154296875,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.3897,
      "step": 116
    },
    {
      "epoch": 0.27432590855803046,
      "grad_norm": 6.858499050140381,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.2126,
      "step": 117
    },
    {
      "epoch": 0.27667057444314186,
      "grad_norm": 6.374567985534668,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.1547,
      "step": 118
    },
    {
      "epoch": 0.2790152403282532,
      "grad_norm": 6.215252876281738,
      "learning_rate": 3.852880399766243e-05,
      "loss": 3.3142,
      "step": 119
    },
    {
      "epoch": 0.2813599062133646,
      "grad_norm": 6.304596424102783,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.1328,
      "step": 120
    },
    {
      "epoch": 0.283704572098476,
      "grad_norm": 6.3406524658203125,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.1292,
      "step": 121
    },
    {
      "epoch": 0.28604923798358733,
      "grad_norm": 7.53887414932251,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.6153,
      "step": 122
    },
    {
      "epoch": 0.28839390386869873,
      "grad_norm": 6.54124641418457,
      "learning_rate": 3.533749813077677e-05,
      "loss": 3.32,
      "step": 123
    },
    {
      "epoch": 0.29073856975381007,
      "grad_norm": 6.543540000915527,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.9915,
      "step": 124
    },
    {
      "epoch": 0.29308323563892147,
      "grad_norm": 7.439838886260986,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.8839,
      "step": 125
    },
    {
      "epoch": 0.2954279015240328,
      "grad_norm": 6.743582248687744,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.8826,
      "step": 126
    },
    {
      "epoch": 0.2977725674091442,
      "grad_norm": 6.0309014320373535,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.6459,
      "step": 127
    },
    {
      "epoch": 0.30011723329425555,
      "grad_norm": 7.520136833190918,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.8574,
      "step": 128
    },
    {
      "epoch": 0.30246189917936694,
      "grad_norm": 6.969455242156982,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.3561,
      "step": 129
    },
    {
      "epoch": 0.3048065650644783,
      "grad_norm": 6.428686618804932,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.8129,
      "step": 130
    },
    {
      "epoch": 0.3071512309495897,
      "grad_norm": 6.685698986053467,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.1487,
      "step": 131
    },
    {
      "epoch": 0.3094958968347011,
      "grad_norm": 6.508021354675293,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 3.1696,
      "step": 132
    },
    {
      "epoch": 0.3118405627198124,
      "grad_norm": 6.594387531280518,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.6793,
      "step": 133
    },
    {
      "epoch": 0.3141852286049238,
      "grad_norm": 5.8391337394714355,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.4195,
      "step": 134
    },
    {
      "epoch": 0.31652989449003516,
      "grad_norm": 7.751064300537109,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.9144,
      "step": 135
    },
    {
      "epoch": 0.31887456037514655,
      "grad_norm": 6.841287612915039,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.771,
      "step": 136
    },
    {
      "epoch": 0.3212192262602579,
      "grad_norm": 6.453044414520264,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.8674,
      "step": 137
    },
    {
      "epoch": 0.3235638921453693,
      "grad_norm": 7.089900970458984,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.7596,
      "step": 138
    },
    {
      "epoch": 0.32590855803048063,
      "grad_norm": 7.597554683685303,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 3.0539,
      "step": 139
    },
    {
      "epoch": 0.32825322391559203,
      "grad_norm": 7.377899646759033,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.7922,
      "step": 140
    },
    {
      "epoch": 0.3305978898007034,
      "grad_norm": 7.7280192375183105,
      "learning_rate": 2.196424713241637e-05,
      "loss": 3.266,
      "step": 141
    },
    {
      "epoch": 0.33294255568581477,
      "grad_norm": 7.4540324211120605,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.7233,
      "step": 142
    },
    {
      "epoch": 0.33528722157092616,
      "grad_norm": 7.697388648986816,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.783,
      "step": 143
    },
    {
      "epoch": 0.3376318874560375,
      "grad_norm": 7.493178844451904,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.6292,
      "step": 144
    },
    {
      "epoch": 0.3399765533411489,
      "grad_norm": 6.404391765594482,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.3068,
      "step": 145
    },
    {
      "epoch": 0.34232121922626024,
      "grad_norm": 8.373858451843262,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.8398,
      "step": 146
    },
    {
      "epoch": 0.34466588511137164,
      "grad_norm": 6.972586631774902,
      "learning_rate": 1.800157297483417e-05,
      "loss": 2.2926,
      "step": 147
    },
    {
      "epoch": 0.347010550996483,
      "grad_norm": 7.07899808883667,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.3815,
      "step": 148
    },
    {
      "epoch": 0.3493552168815944,
      "grad_norm": 7.6493144035339355,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.4759,
      "step": 149
    },
    {
      "epoch": 0.3516998827667057,
      "grad_norm": 7.906857490539551,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.5526,
      "step": 150
    },
    {
      "epoch": 0.3516998827667057,
      "eval_loss": 0.8359604477882385,
      "eval_runtime": 53.7652,
      "eval_samples_per_second": 13.354,
      "eval_steps_per_second": 3.348,
      "step": 150
    },
    {
      "epoch": 0.3540445486518171,
      "grad_norm": 13.72651481628418,
      "learning_rate": 1.553232954407171e-05,
      "loss": 4.8476,
      "step": 151
    },
    {
      "epoch": 0.3563892145369285,
      "grad_norm": 11.892254829406738,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.2858,
      "step": 152
    },
    {
      "epoch": 0.35873388042203985,
      "grad_norm": 11.731127738952637,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.0372,
      "step": 153
    },
    {
      "epoch": 0.36107854630715125,
      "grad_norm": 9.893701553344727,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.7506,
      "step": 154
    },
    {
      "epoch": 0.3634232121922626,
      "grad_norm": 8.016587257385254,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 3.564,
      "step": 155
    },
    {
      "epoch": 0.365767878077374,
      "grad_norm": 7.139926433563232,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.7387,
      "step": 156
    },
    {
      "epoch": 0.36811254396248533,
      "grad_norm": 6.891546726226807,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.4805,
      "step": 157
    },
    {
      "epoch": 0.3704572098475967,
      "grad_norm": 6.198111057281494,
      "learning_rate": 1.157994445715706e-05,
      "loss": 3.3239,
      "step": 158
    },
    {
      "epoch": 0.37280187573270807,
      "grad_norm": 6.7223286628723145,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.4887,
      "step": 159
    },
    {
      "epoch": 0.37514654161781946,
      "grad_norm": 6.473415374755859,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.3225,
      "step": 160
    },
    {
      "epoch": 0.37749120750293086,
      "grad_norm": 7.057542324066162,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 3.8811,
      "step": 161
    },
    {
      "epoch": 0.3798358733880422,
      "grad_norm": 7.749269008636475,
      "learning_rate": 9.549150281252633e-06,
      "loss": 3.6982,
      "step": 162
    },
    {
      "epoch": 0.3821805392731536,
      "grad_norm": 6.553416728973389,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.1394,
      "step": 163
    },
    {
      "epoch": 0.38452520515826494,
      "grad_norm": 7.439867973327637,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.4977,
      "step": 164
    },
    {
      "epoch": 0.38686987104337633,
      "grad_norm": 7.266504764556885,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.24,
      "step": 165
    },
    {
      "epoch": 0.3892145369284877,
      "grad_norm": 7.101572036743164,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.3716,
      "step": 166
    },
    {
      "epoch": 0.39155920281359907,
      "grad_norm": 6.140331268310547,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.2937,
      "step": 167
    },
    {
      "epoch": 0.3939038686987104,
      "grad_norm": 6.884533882141113,
      "learning_rate": 6.837175952121306e-06,
      "loss": 3.1861,
      "step": 168
    },
    {
      "epoch": 0.3962485345838218,
      "grad_norm": 7.3040852546691895,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.4212,
      "step": 169
    },
    {
      "epoch": 0.39859320046893315,
      "grad_norm": 6.858395099639893,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.3434,
      "step": 170
    },
    {
      "epoch": 0.40093786635404455,
      "grad_norm": 7.002919673919678,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.433,
      "step": 171
    },
    {
      "epoch": 0.40328253223915594,
      "grad_norm": 5.903860569000244,
      "learning_rate": 5.263533508961827e-06,
      "loss": 3.0425,
      "step": 172
    },
    {
      "epoch": 0.4056271981242673,
      "grad_norm": 6.657563209533691,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.9576,
      "step": 173
    },
    {
      "epoch": 0.4079718640093787,
      "grad_norm": 7.222111701965332,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.1969,
      "step": 174
    },
    {
      "epoch": 0.41031652989449,
      "grad_norm": 6.370119571685791,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.9986,
      "step": 175
    },
    {
      "epoch": 0.4126611957796014,
      "grad_norm": 6.047725677490234,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.9852,
      "step": 176
    },
    {
      "epoch": 0.41500586166471276,
      "grad_norm": 5.522815227508545,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.4885,
      "step": 177
    },
    {
      "epoch": 0.41735052754982416,
      "grad_norm": 5.858480453491211,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.6544,
      "step": 178
    },
    {
      "epoch": 0.4196951934349355,
      "grad_norm": 6.4223432540893555,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.7672,
      "step": 179
    },
    {
      "epoch": 0.4220398593200469,
      "grad_norm": 6.036847114562988,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.804,
      "step": 180
    },
    {
      "epoch": 0.4243845252051583,
      "grad_norm": 6.438906669616699,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 3.0929,
      "step": 181
    },
    {
      "epoch": 0.42672919109026963,
      "grad_norm": 6.418200969696045,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.5931,
      "step": 182
    },
    {
      "epoch": 0.42907385697538103,
      "grad_norm": 5.831960678100586,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.6685,
      "step": 183
    },
    {
      "epoch": 0.43141852286049237,
      "grad_norm": 6.494126319885254,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 3.0425,
      "step": 184
    },
    {
      "epoch": 0.43376318874560377,
      "grad_norm": 6.834192752838135,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.784,
      "step": 185
    },
    {
      "epoch": 0.4361078546307151,
      "grad_norm": 6.280160427093506,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.787,
      "step": 186
    },
    {
      "epoch": 0.4384525205158265,
      "grad_norm": 7.3110151290893555,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.5908,
      "step": 187
    },
    {
      "epoch": 0.44079718640093785,
      "grad_norm": 6.006848335266113,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.5029,
      "step": 188
    },
    {
      "epoch": 0.44314185228604924,
      "grad_norm": 6.280737400054932,
      "learning_rate": 8.247462563808817e-07,
      "loss": 3.003,
      "step": 189
    },
    {
      "epoch": 0.4454865181711606,
      "grad_norm": 6.811750888824463,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.0713,
      "step": 190
    },
    {
      "epoch": 0.447831184056272,
      "grad_norm": 6.924781322479248,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.738,
      "step": 191
    },
    {
      "epoch": 0.4501758499413834,
      "grad_norm": 7.4093475341796875,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.0687,
      "step": 192
    },
    {
      "epoch": 0.4525205158264947,
      "grad_norm": 7.605098247528076,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.9803,
      "step": 193
    },
    {
      "epoch": 0.4548651817116061,
      "grad_norm": 5.863836288452148,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.0715,
      "step": 194
    },
    {
      "epoch": 0.45720984759671746,
      "grad_norm": 6.435384750366211,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.1854,
      "step": 195
    },
    {
      "epoch": 0.45955451348182885,
      "grad_norm": 7.628547191619873,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.5558,
      "step": 196
    },
    {
      "epoch": 0.4618991793669402,
      "grad_norm": 7.008022308349609,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.8638,
      "step": 197
    },
    {
      "epoch": 0.4642438452520516,
      "grad_norm": 8.426461219787598,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.6995,
      "step": 198
    },
    {
      "epoch": 0.46658851113716293,
      "grad_norm": 8.745497703552246,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.4873,
      "step": 199
    },
    {
      "epoch": 0.46893317702227433,
      "grad_norm": 9.842741966247559,
      "learning_rate": 0.0,
      "loss": 2.5042,
      "step": 200
    },
    {
      "epoch": 0.46893317702227433,
      "eval_loss": 0.7726338505744934,
      "eval_runtime": 53.7376,
      "eval_samples_per_second": 13.361,
      "eval_steps_per_second": 3.35,
      "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": 2.876309235396772e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}