{
  "best_metric": 3.264791250228882,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.20865936358894105,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010432968179447052,
      "grad_norm": 1.3536896705627441,
      "learning_rate": 1e-05,
      "loss": 3.9159,
      "step": 1
    },
    {
      "epoch": 0.0010432968179447052,
      "eval_loss": 3.9926509857177734,
      "eval_runtime": 45.3778,
      "eval_samples_per_second": 35.568,
      "eval_steps_per_second": 8.903,
      "step": 1
    },
    {
      "epoch": 0.0020865936358894104,
      "grad_norm": 1.3918986320495605,
      "learning_rate": 2e-05,
      "loss": 3.9771,
      "step": 2
    },
    {
      "epoch": 0.003129890453834116,
      "grad_norm": 1.3853706121444702,
      "learning_rate": 3e-05,
      "loss": 4.0231,
      "step": 3
    },
    {
      "epoch": 0.004173187271778821,
      "grad_norm": 1.126749873161316,
      "learning_rate": 4e-05,
      "loss": 3.9471,
      "step": 4
    },
    {
      "epoch": 0.005216484089723527,
      "grad_norm": 0.9558578133583069,
      "learning_rate": 5e-05,
      "loss": 3.9798,
      "step": 5
    },
    {
      "epoch": 0.006259780907668232,
      "grad_norm": 1.1789942979812622,
      "learning_rate": 6e-05,
      "loss": 3.8752,
      "step": 6
    },
    {
      "epoch": 0.007303077725612937,
      "grad_norm": 1.2972698211669922,
      "learning_rate": 7e-05,
      "loss": 3.8675,
      "step": 7
    },
    {
      "epoch": 0.008346374543557642,
      "grad_norm": 0.9673587679862976,
      "learning_rate": 8e-05,
      "loss": 3.8187,
      "step": 8
    },
    {
      "epoch": 0.009389671361502348,
      "grad_norm": 0.8321828842163086,
      "learning_rate": 9e-05,
      "loss": 3.7787,
      "step": 9
    },
    {
      "epoch": 0.010432968179447054,
      "grad_norm": 0.8781248331069946,
      "learning_rate": 0.0001,
      "loss": 3.7405,
      "step": 10
    },
    {
      "epoch": 0.011476264997391758,
      "grad_norm": 0.8188990950584412,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.7719,
      "step": 11
    },
    {
      "epoch": 0.012519561815336464,
      "grad_norm": 0.6031615138053894,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.6993,
      "step": 12
    },
    {
      "epoch": 0.013562858633281168,
      "grad_norm": 0.6562380790710449,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.6596,
      "step": 13
    },
    {
      "epoch": 0.014606155451225874,
      "grad_norm": 0.6179069876670837,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.6793,
      "step": 14
    },
    {
      "epoch": 0.01564945226917058,
      "grad_norm": 0.5190150141716003,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.6748,
      "step": 15
    },
    {
      "epoch": 0.016692749087115284,
      "grad_norm": 0.5107501149177551,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.6221,
      "step": 16
    },
    {
      "epoch": 0.01773604590505999,
      "grad_norm": 0.4812328517436981,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.6798,
      "step": 17
    },
    {
      "epoch": 0.018779342723004695,
      "grad_norm": 0.5446358919143677,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.5747,
      "step": 18
    },
    {
      "epoch": 0.0198226395409494,
      "grad_norm": 0.48601874709129333,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.5853,
      "step": 19
    },
    {
      "epoch": 0.020865936358894107,
      "grad_norm": 0.47902897000312805,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.6475,
      "step": 20
    },
    {
      "epoch": 0.02190923317683881,
      "grad_norm": 0.46482163667678833,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.4818,
      "step": 21
    },
    {
      "epoch": 0.022952529994783515,
      "grad_norm": 0.5013744235038757,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.5941,
      "step": 22
    },
    {
      "epoch": 0.023995826812728223,
      "grad_norm": 0.47039228677749634,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.6272,
      "step": 23
    },
    {
      "epoch": 0.025039123630672927,
      "grad_norm": 0.474264532327652,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.5004,
      "step": 24
    },
    {
      "epoch": 0.02608242044861763,
      "grad_norm": 0.457125186920166,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.4437,
      "step": 25
    },
    {
      "epoch": 0.027125717266562335,
      "grad_norm": 0.4855491518974304,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.5872,
      "step": 26
    },
    {
      "epoch": 0.028169014084507043,
      "grad_norm": 0.5192623138427734,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.5309,
      "step": 27
    },
    {
      "epoch": 0.029212310902451747,
      "grad_norm": 0.5076422691345215,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.535,
      "step": 28
    },
    {
      "epoch": 0.03025560772039645,
      "grad_norm": 0.484314501285553,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.5004,
      "step": 29
    },
    {
      "epoch": 0.03129890453834116,
      "grad_norm": 0.5465930104255676,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.5917,
      "step": 30
    },
    {
      "epoch": 0.03234220135628586,
      "grad_norm": 0.5205236077308655,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.4095,
      "step": 31
    },
    {
      "epoch": 0.03338549817423057,
      "grad_norm": 0.5593637228012085,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.5034,
      "step": 32
    },
    {
      "epoch": 0.03442879499217527,
      "grad_norm": 0.5713189244270325,
      "learning_rate": 9.642770192448536e-05,
      "loss": 3.4703,
      "step": 33
    },
    {
      "epoch": 0.03547209181011998,
      "grad_norm": 0.5576574802398682,
      "learning_rate": 9.611448774886924e-05,
      "loss": 3.4392,
      "step": 34
    },
    {
      "epoch": 0.036515388628064686,
      "grad_norm": 0.5582655668258667,
      "learning_rate": 9.578866633275288e-05,
      "loss": 3.3823,
      "step": 35
    },
    {
      "epoch": 0.03755868544600939,
      "grad_norm": 0.5881125926971436,
      "learning_rate": 9.545032675245813e-05,
      "loss": 3.4655,
      "step": 36
    },
    {
      "epoch": 0.038601982263954095,
      "grad_norm": 0.610686182975769,
      "learning_rate": 9.509956150664796e-05,
      "loss": 3.3674,
      "step": 37
    },
    {
      "epoch": 0.0396452790818988,
      "grad_norm": 0.5850858688354492,
      "learning_rate": 9.473646649103818e-05,
      "loss": 3.453,
      "step": 38
    },
    {
      "epoch": 0.0406885758998435,
      "grad_norm": 0.602027177810669,
      "learning_rate": 9.43611409721806e-05,
      "loss": 3.466,
      "step": 39
    },
    {
      "epoch": 0.041731872717788214,
      "grad_norm": 0.6175333857536316,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.4765,
      "step": 40
    },
    {
      "epoch": 0.04277516953573292,
      "grad_norm": 0.6265369057655334,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.3938,
      "step": 41
    },
    {
      "epoch": 0.04381846635367762,
      "grad_norm": 0.6787665486335754,
      "learning_rate": 9.316282404787871e-05,
      "loss": 3.3514,
      "step": 42
    },
    {
      "epoch": 0.044861763171622326,
      "grad_norm": 0.653775155544281,
      "learning_rate": 9.273963562927695e-05,
      "loss": 3.2032,
      "step": 43
    },
    {
      "epoch": 0.04590505998956703,
      "grad_norm": 0.6895964741706848,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.2933,
      "step": 44
    },
    {
      "epoch": 0.046948356807511735,
      "grad_norm": 0.7148680090904236,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.2034,
      "step": 45
    },
    {
      "epoch": 0.047991653625456446,
      "grad_norm": 0.8246645927429199,
      "learning_rate": 9.140044155740101e-05,
      "loss": 3.1824,
      "step": 46
    },
    {
      "epoch": 0.04903495044340115,
      "grad_norm": 0.8079376220703125,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.1431,
      "step": 47
    },
    {
      "epoch": 0.050078247261345854,
      "grad_norm": 0.9385347962379456,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.1974,
      "step": 48
    },
    {
      "epoch": 0.05112154407929056,
      "grad_norm": 1.0700104236602783,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.9287,
      "step": 49
    },
    {
      "epoch": 0.05216484089723526,
      "grad_norm": 1.6941026449203491,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.9044,
      "step": 50
    },
    {
      "epoch": 0.05216484089723526,
      "eval_loss": 3.4272310733795166,
      "eval_runtime": 45.7283,
      "eval_samples_per_second": 35.295,
      "eval_steps_per_second": 8.835,
      "step": 50
    },
    {
      "epoch": 0.053208137715179966,
      "grad_norm": 1.5002392530441284,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.5938,
      "step": 51
    },
    {
      "epoch": 0.05425143453312467,
      "grad_norm": 0.9698646664619446,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.5924,
      "step": 52
    },
    {
      "epoch": 0.05529473135106938,
      "grad_norm": 0.586738646030426,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.4994,
      "step": 53
    },
    {
      "epoch": 0.056338028169014086,
      "grad_norm": 0.6129284501075745,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.5114,
      "step": 54
    },
    {
      "epoch": 0.05738132498695879,
      "grad_norm": 0.6397743821144104,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.4698,
      "step": 55
    },
    {
      "epoch": 0.058424621804903494,
      "grad_norm": 0.5118110775947571,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.5357,
      "step": 56
    },
    {
      "epoch": 0.0594679186228482,
      "grad_norm": 0.5006720423698425,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.5414,
      "step": 57
    },
    {
      "epoch": 0.0605112154407929,
      "grad_norm": 0.5622320175170898,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.4434,
      "step": 58
    },
    {
      "epoch": 0.06155451225873761,
      "grad_norm": 0.5230792164802551,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.4259,
      "step": 59
    },
    {
      "epoch": 0.06259780907668232,
      "grad_norm": 0.46814659237861633,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.4231,
      "step": 60
    },
    {
      "epoch": 0.06364110589462701,
      "grad_norm": 0.4484866261482239,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.5024,
      "step": 61
    },
    {
      "epoch": 0.06468440271257173,
      "grad_norm": 0.45325762033462524,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.4605,
      "step": 62
    },
    {
      "epoch": 0.06572769953051644,
      "grad_norm": 0.4095498323440552,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.393,
      "step": 63
    },
    {
      "epoch": 0.06677099634846113,
      "grad_norm": 0.42198482155799866,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.5101,
      "step": 64
    },
    {
      "epoch": 0.06781429316640585,
      "grad_norm": 0.4125232994556427,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.407,
      "step": 65
    },
    {
      "epoch": 0.06885758998435054,
      "grad_norm": 0.4005553722381592,
      "learning_rate": 8.005405736415126e-05,
      "loss": 3.4294,
      "step": 66
    },
    {
      "epoch": 0.06990088680229525,
      "grad_norm": 0.39542457461357117,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.5029,
      "step": 67
    },
    {
      "epoch": 0.07094418362023996,
      "grad_norm": 0.4077182412147522,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.3977,
      "step": 68
    },
    {
      "epoch": 0.07198748043818466,
      "grad_norm": 0.4182034432888031,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.5822,
      "step": 69
    },
    {
      "epoch": 0.07303077725612937,
      "grad_norm": 0.41805437207221985,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.4612,
      "step": 70
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 0.42146679759025574,
      "learning_rate": 7.66515864363997e-05,
      "loss": 3.434,
      "step": 71
    },
    {
      "epoch": 0.07511737089201878,
      "grad_norm": 0.4041714072227478,
      "learning_rate": 7.594847868906076e-05,
      "loss": 3.4305,
      "step": 72
    },
    {
      "epoch": 0.07616066770996348,
      "grad_norm": 0.40653282403945923,
      "learning_rate": 7.52382768867422e-05,
      "loss": 3.3609,
      "step": 73
    },
    {
      "epoch": 0.07720396452790819,
      "grad_norm": 0.4181981086730957,
      "learning_rate": 7.452117519152542e-05,
      "loss": 3.5083,
      "step": 74
    },
    {
      "epoch": 0.0782472613458529,
      "grad_norm": 0.41767004132270813,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.3453,
      "step": 75
    },
    {
      "epoch": 0.0792905581637976,
      "grad_norm": 0.417631596326828,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.3323,
      "step": 76
    },
    {
      "epoch": 0.08033385498174231,
      "grad_norm": 0.4180872440338135,
      "learning_rate": 7.233044034264034e-05,
      "loss": 3.244,
      "step": 77
    },
    {
      "epoch": 0.081377151799687,
      "grad_norm": 0.4374276399612427,
      "learning_rate": 7.158771761692464e-05,
      "loss": 3.4003,
      "step": 78
    },
    {
      "epoch": 0.08242044861763172,
      "grad_norm": 0.4511520564556122,
      "learning_rate": 7.083909302476453e-05,
      "loss": 3.3814,
      "step": 79
    },
    {
      "epoch": 0.08346374543557643,
      "grad_norm": 0.4532424211502075,
      "learning_rate": 7.008477123264848e-05,
      "loss": 3.3354,
      "step": 80
    },
    {
      "epoch": 0.08450704225352113,
      "grad_norm": 0.447500079870224,
      "learning_rate": 6.932495846462261e-05,
      "loss": 3.3178,
      "step": 81
    },
    {
      "epoch": 0.08555033907146584,
      "grad_norm": 0.46799296140670776,
      "learning_rate": 6.855986244591104e-05,
      "loss": 3.3545,
      "step": 82
    },
    {
      "epoch": 0.08659363588941053,
      "grad_norm": 0.4701199531555176,
      "learning_rate": 6.778969234612584e-05,
      "loss": 3.231,
      "step": 83
    },
    {
      "epoch": 0.08763693270735524,
      "grad_norm": 0.4922150671482086,
      "learning_rate": 6.701465872208216e-05,
      "loss": 3.2622,
      "step": 84
    },
    {
      "epoch": 0.08868022952529994,
      "grad_norm": 0.49328702688217163,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.1954,
      "step": 85
    },
    {
      "epoch": 0.08972352634324465,
      "grad_norm": 0.4895753860473633,
      "learning_rate": 6.545084971874738e-05,
      "loss": 3.2107,
      "step": 86
    },
    {
      "epoch": 0.09076682316118936,
      "grad_norm": 0.5267639756202698,
      "learning_rate": 6.466250186922325e-05,
      "loss": 3.3637,
      "step": 87
    },
    {
      "epoch": 0.09181011997913406,
      "grad_norm": 0.5301526784896851,
      "learning_rate": 6.387014543809223e-05,
      "loss": 3.2647,
      "step": 88
    },
    {
      "epoch": 0.09285341679707877,
      "grad_norm": 0.5323478579521179,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.2916,
      "step": 89
    },
    {
      "epoch": 0.09389671361502347,
      "grad_norm": 0.5398712158203125,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.1855,
      "step": 90
    },
    {
      "epoch": 0.09494001043296818,
      "grad_norm": 0.5690402984619141,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.2861,
      "step": 91
    },
    {
      "epoch": 0.09598330725091289,
      "grad_norm": 0.5927231311798096,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.1404,
      "step": 92
    },
    {
      "epoch": 0.09702660406885759,
      "grad_norm": 0.6183264255523682,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.0545,
      "step": 93
    },
    {
      "epoch": 0.0980699008868023,
      "grad_norm": 0.6504936218261719,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.2365,
      "step": 94
    },
    {
      "epoch": 0.099113197704747,
      "grad_norm": 0.6703303456306458,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.1377,
      "step": 95
    },
    {
      "epoch": 0.10015649452269171,
      "grad_norm": 0.7078514695167542,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.0835,
      "step": 96
    },
    {
      "epoch": 0.1011997913406364,
      "grad_norm": 0.7266568541526794,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 3.1361,
      "step": 97
    },
    {
      "epoch": 0.10224308815858112,
      "grad_norm": 0.7982112169265747,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.8827,
      "step": 98
    },
    {
      "epoch": 0.10328638497652583,
      "grad_norm": 0.9347984790802002,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.9432,
      "step": 99
    },
    {
      "epoch": 0.10432968179447052,
      "grad_norm": 1.5412112474441528,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.8334,
      "step": 100
    },
    {
      "epoch": 0.10432968179447052,
      "eval_loss": 3.3252243995666504,
      "eval_runtime": 45.7641,
      "eval_samples_per_second": 35.268,
      "eval_steps_per_second": 8.828,
      "step": 100
    },
    {
      "epoch": 0.10537297861241524,
      "grad_norm": 0.787086546421051,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.4292,
      "step": 101
    },
    {
      "epoch": 0.10641627543035993,
      "grad_norm": 0.7174277901649475,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.49,
      "step": 102
    },
    {
      "epoch": 0.10745957224830464,
      "grad_norm": 0.5412613153457642,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.3275,
      "step": 103
    },
    {
      "epoch": 0.10850286906624934,
      "grad_norm": 0.4861738085746765,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.5012,
      "step": 104
    },
    {
      "epoch": 0.10954616588419405,
      "grad_norm": 0.4842031002044678,
      "learning_rate": 5e-05,
      "loss": 3.4971,
      "step": 105
    },
    {
      "epoch": 0.11058946270213876,
      "grad_norm": 0.47764962911605835,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.3493,
      "step": 106
    },
    {
      "epoch": 0.11163275952008346,
      "grad_norm": 0.5093321800231934,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.4019,
      "step": 107
    },
    {
      "epoch": 0.11267605633802817,
      "grad_norm": 0.5084569454193115,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.3828,
      "step": 108
    },
    {
      "epoch": 0.11371935315597287,
      "grad_norm": 0.4781957268714905,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.4065,
      "step": 109
    },
    {
      "epoch": 0.11476264997391758,
      "grad_norm": 0.43107083439826965,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.3369,
      "step": 110
    },
    {
      "epoch": 0.11580594679186229,
      "grad_norm": 0.44881919026374817,
      "learning_rate": 4.504772348747687e-05,
      "loss": 3.4299,
      "step": 111
    },
    {
      "epoch": 0.11684924360980699,
      "grad_norm": 0.4160110056400299,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.4753,
      "step": 112
    },
    {
      "epoch": 0.1178925404277517,
      "grad_norm": 0.4867398738861084,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 3.3588,
      "step": 113
    },
    {
      "epoch": 0.1189358372456964,
      "grad_norm": 0.47459399700164795,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.3675,
      "step": 114
    },
    {
      "epoch": 0.11997913406364111,
      "grad_norm": 0.40645161271095276,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.3632,
      "step": 115
    },
    {
      "epoch": 0.1210224308815858,
      "grad_norm": 0.39582309126853943,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.2547,
      "step": 116
    },
    {
      "epoch": 0.12206572769953052,
      "grad_norm": 0.3737109899520874,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.3265,
      "step": 117
    },
    {
      "epoch": 0.12310902451747523,
      "grad_norm": 0.39096322655677795,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.2082,
      "step": 118
    },
    {
      "epoch": 0.12415232133541992,
      "grad_norm": 0.4134329557418823,
      "learning_rate": 3.852880399766243e-05,
      "loss": 3.4101,
      "step": 119
    },
    {
      "epoch": 0.12519561815336464,
      "grad_norm": 0.41859692335128784,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.3879,
      "step": 120
    },
    {
      "epoch": 0.12623891497130935,
      "grad_norm": 0.427964448928833,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.4571,
      "step": 121
    },
    {
      "epoch": 0.12728221178925403,
      "grad_norm": 0.4351688027381897,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.3502,
      "step": 122
    },
    {
      "epoch": 0.12832550860719874,
      "grad_norm": 0.42806118726730347,
      "learning_rate": 3.533749813077677e-05,
      "loss": 3.275,
      "step": 123
    },
    {
      "epoch": 0.12936880542514345,
      "grad_norm": 0.4266616106033325,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 3.1303,
      "step": 124
    },
    {
      "epoch": 0.13041210224308816,
      "grad_norm": 0.43820124864578247,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 3.3103,
      "step": 125
    },
    {
      "epoch": 0.13145539906103287,
      "grad_norm": 0.4153601825237274,
      "learning_rate": 3.298534127791785e-05,
      "loss": 3.3252,
      "step": 126
    },
    {
      "epoch": 0.13249869587897756,
      "grad_norm": 0.453664630651474,
      "learning_rate": 3.221030765387417e-05,
      "loss": 3.2869,
      "step": 127
    },
    {
      "epoch": 0.13354199269692227,
      "grad_norm": 0.46373143792152405,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.3462,
      "step": 128
    },
    {
      "epoch": 0.13458528951486698,
      "grad_norm": 0.47818803787231445,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.3275,
      "step": 129
    },
    {
      "epoch": 0.1356285863328117,
      "grad_norm": 0.4792380928993225,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.2778,
      "step": 130
    },
    {
      "epoch": 0.1366718831507564,
      "grad_norm": 0.5010649561882019,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.2362,
      "step": 131
    },
    {
      "epoch": 0.13771517996870108,
      "grad_norm": 0.4995712637901306,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 3.3166,
      "step": 132
    },
    {
      "epoch": 0.1387584767866458,
      "grad_norm": 0.51168292760849,
      "learning_rate": 2.766955965735968e-05,
      "loss": 3.2292,
      "step": 133
    },
    {
      "epoch": 0.1398017736045905,
      "grad_norm": 0.5347958207130432,
      "learning_rate": 2.693294185106562e-05,
      "loss": 3.2483,
      "step": 134
    },
    {
      "epoch": 0.14084507042253522,
      "grad_norm": 0.5277102589607239,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 3.2252,
      "step": 135
    },
    {
      "epoch": 0.14188836724047993,
      "grad_norm": 0.5362873673439026,
      "learning_rate": 2.547882480847461e-05,
      "loss": 3.2038,
      "step": 136
    },
    {
      "epoch": 0.1429316640584246,
      "grad_norm": 0.5422380566596985,
      "learning_rate": 2.476172311325783e-05,
      "loss": 3.2725,
      "step": 137
    },
    {
      "epoch": 0.14397496087636932,
      "grad_norm": 0.552033543586731,
      "learning_rate": 2.405152131093926e-05,
      "loss": 3.1894,
      "step": 138
    },
    {
      "epoch": 0.14501825769431403,
      "grad_norm": 0.5851796865463257,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 3.2272,
      "step": 139
    },
    {
      "epoch": 0.14606155451225875,
      "grad_norm": 0.628825306892395,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 3.2583,
      "step": 140
    },
    {
      "epoch": 0.14710485133020346,
      "grad_norm": 0.5927334427833557,
      "learning_rate": 2.196424713241637e-05,
      "loss": 3.276,
      "step": 141
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 0.6273190975189209,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.1073,
      "step": 142
    },
    {
      "epoch": 0.14919144496609285,
      "grad_norm": 0.624865710735321,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.1138,
      "step": 143
    },
    {
      "epoch": 0.15023474178403756,
      "grad_norm": 0.627785861492157,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 3.0413,
      "step": 144
    },
    {
      "epoch": 0.15127803860198227,
      "grad_norm": 0.6595969200134277,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.9438,
      "step": 145
    },
    {
      "epoch": 0.15232133541992696,
      "grad_norm": 0.6877166032791138,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 3.1219,
      "step": 146
    },
    {
      "epoch": 0.15336463223787167,
      "grad_norm": 0.7304136157035828,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.1321,
      "step": 147
    },
    {
      "epoch": 0.15440792905581638,
      "grad_norm": 0.7825625538825989,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.9733,
      "step": 148
    },
    {
      "epoch": 0.1554512258737611,
      "grad_norm": 0.9954061508178711,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.9893,
      "step": 149
    },
    {
      "epoch": 0.1564945226917058,
      "grad_norm": 1.6247068643569946,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.6556,
      "step": 150
    },
    {
      "epoch": 0.1564945226917058,
      "eval_loss": 3.280613899230957,
      "eval_runtime": 45.8112,
      "eval_samples_per_second": 35.232,
      "eval_steps_per_second": 8.819,
      "step": 150
    },
    {
      "epoch": 0.15753781950965048,
      "grad_norm": 0.5763651728630066,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.3166,
      "step": 151
    },
    {
      "epoch": 0.1585811163275952,
      "grad_norm": 0.6027952432632446,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.2649,
      "step": 152
    },
    {
      "epoch": 0.1596244131455399,
      "grad_norm": 0.5729192495346069,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.4582,
      "step": 153
    },
    {
      "epoch": 0.16066770996348462,
      "grad_norm": 0.5555835366249084,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.3367,
      "step": 154
    },
    {
      "epoch": 0.16171100678142933,
      "grad_norm": 0.48806819319725037,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 3.5038,
      "step": 155
    },
    {
      "epoch": 0.162754303599374,
      "grad_norm": 0.4888882637023926,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.4001,
      "step": 156
    },
    {
      "epoch": 0.16379760041731872,
      "grad_norm": 0.44786500930786133,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.3959,
      "step": 157
    },
    {
      "epoch": 0.16484089723526343,
      "grad_norm": 0.427033394575119,
      "learning_rate": 1.157994445715706e-05,
      "loss": 3.3204,
      "step": 158
    },
    {
      "epoch": 0.16588419405320814,
      "grad_norm": 0.42746180295944214,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.2912,
      "step": 159
    },
    {
      "epoch": 0.16692749087115286,
      "grad_norm": 0.41423723101615906,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.2636,
      "step": 160
    },
    {
      "epoch": 0.16797078768909754,
      "grad_norm": 0.39984095096588135,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 3.3372,
      "step": 161
    },
    {
      "epoch": 0.16901408450704225,
      "grad_norm": 0.415336012840271,
      "learning_rate": 9.549150281252633e-06,
      "loss": 3.4322,
      "step": 162
    },
    {
      "epoch": 0.17005738132498696,
      "grad_norm": 0.4239465296268463,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.258,
      "step": 163
    },
    {
      "epoch": 0.17110067814293167,
      "grad_norm": 0.4182020127773285,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.4261,
      "step": 164
    },
    {
      "epoch": 0.17214397496087636,
      "grad_norm": 0.415536105632782,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.4219,
      "step": 165
    },
    {
      "epoch": 0.17318727177882107,
      "grad_norm": 0.431723028421402,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.4468,
      "step": 166
    },
    {
      "epoch": 0.17423056859676578,
      "grad_norm": 0.42299938201904297,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.2718,
      "step": 167
    },
    {
      "epoch": 0.1752738654147105,
      "grad_norm": 0.41531872749328613,
      "learning_rate": 6.837175952121306e-06,
      "loss": 3.2628,
      "step": 168
    },
    {
      "epoch": 0.1763171622326552,
      "grad_norm": 0.4174998998641968,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.3129,
      "step": 169
    },
    {
      "epoch": 0.17736045905059988,
      "grad_norm": 0.44035670161247253,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.1502,
      "step": 170
    },
    {
      "epoch": 0.1784037558685446,
      "grad_norm": 0.4416235685348511,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.293,
      "step": 171
    },
    {
      "epoch": 0.1794470526864893,
      "grad_norm": 0.4279758036136627,
      "learning_rate": 5.263533508961827e-06,
      "loss": 3.3121,
      "step": 172
    },
    {
      "epoch": 0.18049034950443402,
      "grad_norm": 0.4702642858028412,
      "learning_rate": 4.900438493352055e-06,
      "loss": 3.3587,
      "step": 173
    },
    {
      "epoch": 0.18153364632237873,
      "grad_norm": 0.45080041885375977,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.2199,
      "step": 174
    },
    {
      "epoch": 0.1825769431403234,
      "grad_norm": 0.44568002223968506,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 3.1716,
      "step": 175
    },
    {
      "epoch": 0.18362023995826812,
      "grad_norm": 0.45467543601989746,
      "learning_rate": 3.885512251130763e-06,
      "loss": 3.3798,
      "step": 176
    },
    {
      "epoch": 0.18466353677621283,
      "grad_norm": 0.5169409513473511,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 3.4454,
      "step": 177
    },
    {
      "epoch": 0.18570683359415754,
      "grad_norm": 0.45263707637786865,
      "learning_rate": 3.271776770026963e-06,
      "loss": 3.307,
      "step": 178
    },
    {
      "epoch": 0.18675013041210226,
      "grad_norm": 0.4634079337120056,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 3.2359,
      "step": 179
    },
    {
      "epoch": 0.18779342723004694,
      "grad_norm": 0.48052823543548584,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 3.3297,
      "step": 180
    },
    {
      "epoch": 0.18883672404799165,
      "grad_norm": 0.48120859265327454,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 3.1815,
      "step": 181
    },
    {
      "epoch": 0.18988002086593636,
      "grad_norm": 0.49086952209472656,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 3.2424,
      "step": 182
    },
    {
      "epoch": 0.19092331768388107,
      "grad_norm": 0.5069617033004761,
      "learning_rate": 1.962316193157593e-06,
      "loss": 3.4037,
      "step": 183
    },
    {
      "epoch": 0.19196661450182578,
      "grad_norm": 0.5208487510681152,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 3.2318,
      "step": 184
    },
    {
      "epoch": 0.19300991131977047,
      "grad_norm": 0.5285236239433289,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 3.2106,
      "step": 185
    },
    {
      "epoch": 0.19405320813771518,
      "grad_norm": 0.5508992075920105,
      "learning_rate": 1.333670137599713e-06,
      "loss": 3.2486,
      "step": 186
    },
    {
      "epoch": 0.1950965049556599,
      "grad_norm": 0.571118950843811,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 3.1736,
      "step": 187
    },
    {
      "epoch": 0.1961398017736046,
      "grad_norm": 0.5668531060218811,
      "learning_rate": 9.810017062595322e-07,
      "loss": 3.2182,
      "step": 188
    },
    {
      "epoch": 0.19718309859154928,
      "grad_norm": 0.5890854001045227,
      "learning_rate": 8.247462563808817e-07,
      "loss": 3.2562,
      "step": 189
    },
    {
      "epoch": 0.198226395409494,
      "grad_norm": 0.5864028930664062,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.1508,
      "step": 190
    },
    {
      "epoch": 0.1992696922274387,
      "grad_norm": 0.6390986442565918,
      "learning_rate": 5.526064699265753e-07,
      "loss": 3.2076,
      "step": 191
    },
    {
      "epoch": 0.20031298904538342,
      "grad_norm": 0.6285101771354675,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.1137,
      "step": 192
    },
    {
      "epoch": 0.20135628586332813,
      "grad_norm": 0.6637088060379028,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.1291,
      "step": 193
    },
    {
      "epoch": 0.2023995826812728,
      "grad_norm": 0.6833120584487915,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.2401,
      "step": 194
    },
    {
      "epoch": 0.20344287949921752,
      "grad_norm": 0.7075033783912659,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.9934,
      "step": 195
    },
    {
      "epoch": 0.20448617631716223,
      "grad_norm": 0.8524454832077026,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 3.0609,
      "step": 196
    },
    {
      "epoch": 0.20552947313510694,
      "grad_norm": 0.7844312191009521,
      "learning_rate": 6.150154258476315e-08,
      "loss": 3.0781,
      "step": 197
    },
    {
      "epoch": 0.20657276995305165,
      "grad_norm": 0.8506727814674377,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.9746,
      "step": 198
    },
    {
      "epoch": 0.20761606677099634,
      "grad_norm": 1.05426824092865,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.7736,
      "step": 199
    },
    {
      "epoch": 0.20865936358894105,
      "grad_norm": 1.592360258102417,
      "learning_rate": 0.0,
      "loss": 2.5687,
      "step": 200
    },
    {
      "epoch": 0.20865936358894105,
      "eval_loss": 3.264791250228882,
      "eval_runtime": 45.7239,
      "eval_samples_per_second": 35.299,
      "eval_steps_per_second": 8.836,
      "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": 5.450257781489664e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}