{
  "best_metric": 1.4084903001785278,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.1172504763300601,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005862523816503004,
      "grad_norm": 10.974843978881836,
      "learning_rate": 1e-05,
      "loss": 7.5692,
      "step": 1
    },
    {
      "epoch": 0.0005862523816503004,
      "eval_loss": 1.6656559705734253,
      "eval_runtime": 216.4511,
      "eval_samples_per_second": 13.273,
      "eval_steps_per_second": 3.322,
      "step": 1
    },
    {
      "epoch": 0.0011725047633006009,
      "grad_norm": 13.693845748901367,
      "learning_rate": 2e-05,
      "loss": 7.597,
      "step": 2
    },
    {
      "epoch": 0.0017587571449509014,
      "grad_norm": 7.404031276702881,
      "learning_rate": 3e-05,
      "loss": 6.9445,
      "step": 3
    },
    {
      "epoch": 0.0023450095266012018,
      "grad_norm": 4.048150539398193,
      "learning_rate": 4e-05,
      "loss": 5.9878,
      "step": 4
    },
    {
      "epoch": 0.0029312619082515023,
      "grad_norm": 4.052788257598877,
      "learning_rate": 5e-05,
      "loss": 5.7518,
      "step": 5
    },
    {
      "epoch": 0.003517514289901803,
      "grad_norm": 5.494608402252197,
      "learning_rate": 6e-05,
      "loss": 5.2256,
      "step": 6
    },
    {
      "epoch": 0.004103766671552103,
      "grad_norm": 3.2354836463928223,
      "learning_rate": 7e-05,
      "loss": 5.2484,
      "step": 7
    },
    {
      "epoch": 0.0046900190532024036,
      "grad_norm": 3.3828489780426025,
      "learning_rate": 8e-05,
      "loss": 5.5789,
      "step": 8
    },
    {
      "epoch": 0.005276271434852704,
      "grad_norm": 5.21439266204834,
      "learning_rate": 9e-05,
      "loss": 5.3438,
      "step": 9
    },
    {
      "epoch": 0.005862523816503005,
      "grad_norm": 3.431241273880005,
      "learning_rate": 0.0001,
      "loss": 5.2426,
      "step": 10
    },
    {
      "epoch": 0.006448776198153305,
      "grad_norm": 3.713555335998535,
      "learning_rate": 9.999316524962345e-05,
      "loss": 4.6382,
      "step": 11
    },
    {
      "epoch": 0.007035028579803606,
      "grad_norm": 2.907348394393921,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.4706,
      "step": 12
    },
    {
      "epoch": 0.007621280961453906,
      "grad_norm": 3.0908288955688477,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.2979,
      "step": 13
    },
    {
      "epoch": 0.008207533343104206,
      "grad_norm": 3.0675079822540283,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.1071,
      "step": 14
    },
    {
      "epoch": 0.008793785724754507,
      "grad_norm": 3.3359196186065674,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.1516,
      "step": 15
    },
    {
      "epoch": 0.009380038106404807,
      "grad_norm": 3.675565004348755,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.7218,
      "step": 16
    },
    {
      "epoch": 0.009966290488055108,
      "grad_norm": 3.150296211242676,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.6728,
      "step": 17
    },
    {
      "epoch": 0.010552542869705408,
      "grad_norm": 3.276887893676758,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.2585,
      "step": 18
    },
    {
      "epoch": 0.011138795251355709,
      "grad_norm": 2.844019889831543,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.5302,
      "step": 19
    },
    {
      "epoch": 0.01172504763300601,
      "grad_norm": 4.640794277191162,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.4776,
      "step": 20
    },
    {
      "epoch": 0.01231130001465631,
      "grad_norm": 7.7061285972595215,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.1846,
      "step": 21
    },
    {
      "epoch": 0.01289755239630661,
      "grad_norm": 3.5862724781036377,
      "learning_rate": 9.901899829374047e-05,
      "loss": 4.8898,
      "step": 22
    },
    {
      "epoch": 0.013483804777956911,
      "grad_norm": 4.39357852935791,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.5521,
      "step": 23
    },
    {
      "epoch": 0.014070057159607212,
      "grad_norm": 4.294259548187256,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.4554,
      "step": 24
    },
    {
      "epoch": 0.014656309541257512,
      "grad_norm": 3.6883585453033447,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.8094,
      "step": 25
    },
    {
      "epoch": 0.015242561922907813,
      "grad_norm": 3.103388547897339,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.6657,
      "step": 26
    },
    {
      "epoch": 0.01582881430455811,
      "grad_norm": 3.55566143989563,
      "learning_rate": 9.803768380684242e-05,
      "loss": 6.0851,
      "step": 27
    },
    {
      "epoch": 0.016415066686208412,
      "grad_norm": 3.5435686111450195,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.6171,
      "step": 28
    },
    {
      "epoch": 0.017001319067858713,
      "grad_norm": 3.454345703125,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.7035,
      "step": 29
    },
    {
      "epoch": 0.017587571449509013,
      "grad_norm": 3.257582187652588,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.8846,
      "step": 30
    },
    {
      "epoch": 0.018173823831159314,
      "grad_norm": 3.9570600986480713,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.4967,
      "step": 31
    },
    {
      "epoch": 0.018760076212809614,
      "grad_norm": 3.395634174346924,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.3168,
      "step": 32
    },
    {
      "epoch": 0.019346328594459915,
      "grad_norm": 3.533217668533325,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.5111,
      "step": 33
    },
    {
      "epoch": 0.019932580976110215,
      "grad_norm": 3.7604002952575684,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.5875,
      "step": 34
    },
    {
      "epoch": 0.020518833357760516,
      "grad_norm": 3.635335683822632,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.0752,
      "step": 35
    },
    {
      "epoch": 0.021105085739410816,
      "grad_norm": 4.102715015411377,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.188,
      "step": 36
    },
    {
      "epoch": 0.021691338121061117,
      "grad_norm": 4.521205902099609,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.0285,
      "step": 37
    },
    {
      "epoch": 0.022277590502711418,
      "grad_norm": 4.065936088562012,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.2301,
      "step": 38
    },
    {
      "epoch": 0.022863842884361718,
      "grad_norm": 3.8373758792877197,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.0672,
      "step": 39
    },
    {
      "epoch": 0.02345009526601202,
      "grad_norm": 3.5169742107391357,
      "learning_rate": 9.397368756032445e-05,
      "loss": 5.7006,
      "step": 40
    },
    {
      "epoch": 0.02403634764766232,
      "grad_norm": 4.195425987243652,
      "learning_rate": 9.357421218136386e-05,
      "loss": 5.0392,
      "step": 41
    },
    {
      "epoch": 0.02462260002931262,
      "grad_norm": 4.197497367858887,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.0235,
      "step": 42
    },
    {
      "epoch": 0.02520885241096292,
      "grad_norm": 3.9651541709899902,
      "learning_rate": 9.273963562927695e-05,
      "loss": 6.3518,
      "step": 43
    },
    {
      "epoch": 0.02579510479261322,
      "grad_norm": 3.6959288120269775,
      "learning_rate": 9.230476262104677e-05,
      "loss": 6.1257,
      "step": 44
    },
    {
      "epoch": 0.02638135717426352,
      "grad_norm": 4.428595066070557,
      "learning_rate": 9.185832391312644e-05,
      "loss": 5.7394,
      "step": 45
    },
    {
      "epoch": 0.026967609555913822,
      "grad_norm": 4.286281108856201,
      "learning_rate": 9.140044155740101e-05,
      "loss": 5.7254,
      "step": 46
    },
    {
      "epoch": 0.027553861937564122,
      "grad_norm": 4.481366157531738,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.0267,
      "step": 47
    },
    {
      "epoch": 0.028140114319214423,
      "grad_norm": 5.23274564743042,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.6619,
      "step": 48
    },
    {
      "epoch": 0.028726366700864724,
      "grad_norm": 4.575775623321533,
      "learning_rate": 8.995939984474624e-05,
      "loss": 5.8483,
      "step": 49
    },
    {
      "epoch": 0.029312619082515024,
      "grad_norm": 5.6270880699157715,
      "learning_rate": 8.945702546981969e-05,
      "loss": 7.0042,
      "step": 50
    },
    {
      "epoch": 0.029312619082515024,
      "eval_loss": 1.5384029150009155,
      "eval_runtime": 217.8333,
      "eval_samples_per_second": 13.189,
      "eval_steps_per_second": 3.301,
      "step": 50
    },
    {
      "epoch": 0.029898871464165325,
      "grad_norm": 18.780412673950195,
      "learning_rate": 8.894386393810563e-05,
      "loss": 7.297,
      "step": 51
    },
    {
      "epoch": 0.030485123845815625,
      "grad_norm": 4.128394603729248,
      "learning_rate": 8.842005554284296e-05,
      "loss": 6.3004,
      "step": 52
    },
    {
      "epoch": 0.031071376227465926,
      "grad_norm": 2.334108591079712,
      "learning_rate": 8.788574348801675e-05,
      "loss": 4.9779,
      "step": 53
    },
    {
      "epoch": 0.03165762860911622,
      "grad_norm": 2.399811267852783,
      "learning_rate": 8.73410738492077e-05,
      "loss": 4.9162,
      "step": 54
    },
    {
      "epoch": 0.03224388099076653,
      "grad_norm": 2.4325015544891357,
      "learning_rate": 8.678619553365659e-05,
      "loss": 5.3835,
      "step": 55
    },
    {
      "epoch": 0.032830133372416824,
      "grad_norm": 2.914151191711426,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.3231,
      "step": 56
    },
    {
      "epoch": 0.03341638575406713,
      "grad_norm": 2.2403950691223145,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.0804,
      "step": 57
    },
    {
      "epoch": 0.034002638135717425,
      "grad_norm": 2.0466275215148926,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.2719,
      "step": 58
    },
    {
      "epoch": 0.03458889051736773,
      "grad_norm": 2.2615773677825928,
      "learning_rate": 8.44676704559283e-05,
      "loss": 5.8687,
      "step": 59
    },
    {
      "epoch": 0.035175142899018026,
      "grad_norm": 2.258857250213623,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.5588,
      "step": 60
    },
    {
      "epoch": 0.03576139528066833,
      "grad_norm": 2.3287036418914795,
      "learning_rate": 8.32512286056924e-05,
      "loss": 4.9895,
      "step": 61
    },
    {
      "epoch": 0.03634764766231863,
      "grad_norm": 2.616114854812622,
      "learning_rate": 8.262928807620843e-05,
      "loss": 4.8476,
      "step": 62
    },
    {
      "epoch": 0.03693390004396893,
      "grad_norm": 2.285841703414917,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.595,
      "step": 63
    },
    {
      "epoch": 0.03752015242561923,
      "grad_norm": 3.2915735244750977,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.517,
      "step": 64
    },
    {
      "epoch": 0.03810640480726953,
      "grad_norm": 2.6036031246185303,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.0054,
      "step": 65
    },
    {
      "epoch": 0.03869265718891983,
      "grad_norm": 2.697678327560425,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.829,
      "step": 66
    },
    {
      "epoch": 0.039278909570570134,
      "grad_norm": 2.638777256011963,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.1254,
      "step": 67
    },
    {
      "epoch": 0.03986516195222043,
      "grad_norm": 2.8577005863189697,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.6927,
      "step": 68
    },
    {
      "epoch": 0.040451414333870735,
      "grad_norm": 2.635063886642456,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.9703,
      "step": 69
    },
    {
      "epoch": 0.04103766671552103,
      "grad_norm": 2.5592451095581055,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.926,
      "step": 70
    },
    {
      "epoch": 0.04162391909717133,
      "grad_norm": 2.348313808441162,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.4573,
      "step": 71
    },
    {
      "epoch": 0.04221017147882163,
      "grad_norm": 2.599757432937622,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.8047,
      "step": 72
    },
    {
      "epoch": 0.04279642386047193,
      "grad_norm": 2.5455238819122314,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.4577,
      "step": 73
    },
    {
      "epoch": 0.043382676242122234,
      "grad_norm": 2.3764522075653076,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.6628,
      "step": 74
    },
    {
      "epoch": 0.04396892862377253,
      "grad_norm": 2.619389057159424,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.6041,
      "step": 75
    },
    {
      "epoch": 0.044555181005422835,
      "grad_norm": 2.5849316120147705,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.2824,
      "step": 76
    },
    {
      "epoch": 0.04514143338707313,
      "grad_norm": 3.619281053543091,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.3853,
      "step": 77
    },
    {
      "epoch": 0.045727685768723436,
      "grad_norm": 5.075690269470215,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.5257,
      "step": 78
    },
    {
      "epoch": 0.04631393815037373,
      "grad_norm": 2.971221446990967,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.6833,
      "step": 79
    },
    {
      "epoch": 0.04690019053202404,
      "grad_norm": 2.8313474655151367,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.987,
      "step": 80
    },
    {
      "epoch": 0.047486442913674334,
      "grad_norm": 2.9153332710266113,
      "learning_rate": 6.932495846462261e-05,
      "loss": 6.2109,
      "step": 81
    },
    {
      "epoch": 0.04807269529532464,
      "grad_norm": 3.4069366455078125,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.6941,
      "step": 82
    },
    {
      "epoch": 0.048658947676974935,
      "grad_norm": 3.033674716949463,
      "learning_rate": 6.778969234612584e-05,
      "loss": 5.5635,
      "step": 83
    },
    {
      "epoch": 0.04924520005862524,
      "grad_norm": 3.89658260345459,
      "learning_rate": 6.701465872208216e-05,
      "loss": 5.4324,
      "step": 84
    },
    {
      "epoch": 0.04983145244027554,
      "grad_norm": 3.5395798683166504,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.9539,
      "step": 85
    },
    {
      "epoch": 0.05041770482192584,
      "grad_norm": 3.900202751159668,
      "learning_rate": 6.545084971874738e-05,
      "loss": 5.8918,
      "step": 86
    },
    {
      "epoch": 0.05100395720357614,
      "grad_norm": 2.9694056510925293,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.9982,
      "step": 87
    },
    {
      "epoch": 0.05159020958522644,
      "grad_norm": 3.4069461822509766,
      "learning_rate": 6.387014543809223e-05,
      "loss": 5.7944,
      "step": 88
    },
    {
      "epoch": 0.05217646196687674,
      "grad_norm": 3.693887233734131,
      "learning_rate": 6.307399704769099e-05,
      "loss": 5.9807,
      "step": 89
    },
    {
      "epoch": 0.05276271434852704,
      "grad_norm": 3.686582326889038,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.0766,
      "step": 90
    },
    {
      "epoch": 0.05334896673017734,
      "grad_norm": 3.2246813774108887,
      "learning_rate": 6.147119600233758e-05,
      "loss": 5.802,
      "step": 91
    },
    {
      "epoch": 0.053935219111827644,
      "grad_norm": 3.8852860927581787,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.7282,
      "step": 92
    },
    {
      "epoch": 0.05452147149347794,
      "grad_norm": 3.733841896057129,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.0389,
      "step": 93
    },
    {
      "epoch": 0.055107723875128245,
      "grad_norm": 3.5553946495056152,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.8287,
      "step": 94
    },
    {
      "epoch": 0.05569397625677854,
      "grad_norm": 3.28301739692688,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 6.225,
      "step": 95
    },
    {
      "epoch": 0.056280228638428846,
      "grad_norm": 4.250922679901123,
      "learning_rate": 5.74131823855921e-05,
      "loss": 5.8304,
      "step": 96
    },
    {
      "epoch": 0.05686648102007914,
      "grad_norm": 4.312722206115723,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.0656,
      "step": 97
    },
    {
      "epoch": 0.05745273340172945,
      "grad_norm": 4.033783912658691,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.0669,
      "step": 98
    },
    {
      "epoch": 0.058038985783379744,
      "grad_norm": 4.369868278503418,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.9125,
      "step": 99
    },
    {
      "epoch": 0.05862523816503005,
      "grad_norm": 4.5124192237854,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 6.2855,
      "step": 100
    },
    {
      "epoch": 0.05862523816503005,
      "eval_loss": 1.51200532913208,
      "eval_runtime": 217.4639,
      "eval_samples_per_second": 13.211,
      "eval_steps_per_second": 3.306,
      "step": 100
    },
    {
      "epoch": 0.059211490546680345,
      "grad_norm": 12.696165084838867,
      "learning_rate": 5.330452921628497e-05,
      "loss": 7.3548,
      "step": 101
    },
    {
      "epoch": 0.05979774292833065,
      "grad_norm": 6.682477951049805,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.8703,
      "step": 102
    },
    {
      "epoch": 0.060383995309980946,
      "grad_norm": 2.374072313308716,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.6728,
      "step": 103
    },
    {
      "epoch": 0.06097024769163125,
      "grad_norm": 2.8170857429504395,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 5.5303,
      "step": 104
    },
    {
      "epoch": 0.06155650007328155,
      "grad_norm": 2.234957218170166,
      "learning_rate": 5e-05,
      "loss": 4.8733,
      "step": 105
    },
    {
      "epoch": 0.06214275245493185,
      "grad_norm": 2.1801908016204834,
      "learning_rate": 4.917330276168208e-05,
      "loss": 5.8957,
      "step": 106
    },
    {
      "epoch": 0.06272900483658214,
      "grad_norm": 2.1377458572387695,
      "learning_rate": 4.834683153413459e-05,
      "loss": 5.0094,
      "step": 107
    },
    {
      "epoch": 0.06331525721823245,
      "grad_norm": 2.2051138877868652,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 5.0082,
      "step": 108
    },
    {
      "epoch": 0.06390150959988275,
      "grad_norm": 2.4450340270996094,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.2171,
      "step": 109
    },
    {
      "epoch": 0.06448776198153305,
      "grad_norm": 2.4907281398773193,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.5849,
      "step": 110
    },
    {
      "epoch": 0.06507401436318334,
      "grad_norm": 2.5389492511749268,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.6138,
      "step": 111
    },
    {
      "epoch": 0.06566026674483365,
      "grad_norm": 2.2127041816711426,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.6302,
      "step": 112
    },
    {
      "epoch": 0.06624651912648395,
      "grad_norm": 3.6804146766662598,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.328,
      "step": 113
    },
    {
      "epoch": 0.06683277150813426,
      "grad_norm": 2.4165918827056885,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.4437,
      "step": 114
    },
    {
      "epoch": 0.06741902388978455,
      "grad_norm": 2.409534454345703,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.5882,
      "step": 115
    },
    {
      "epoch": 0.06800527627143485,
      "grad_norm": 2.3013737201690674,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.2668,
      "step": 116
    },
    {
      "epoch": 0.06859152865308515,
      "grad_norm": 2.561699628829956,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.6068,
      "step": 117
    },
    {
      "epoch": 0.06917778103473546,
      "grad_norm": 2.6396546363830566,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.4774,
      "step": 118
    },
    {
      "epoch": 0.06976403341638575,
      "grad_norm": 2.8937535285949707,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.8109,
      "step": 119
    },
    {
      "epoch": 0.07035028579803605,
      "grad_norm": 2.7331321239471436,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.5422,
      "step": 120
    },
    {
      "epoch": 0.07093653817968636,
      "grad_norm": 2.8930675983428955,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 6.0664,
      "step": 121
    },
    {
      "epoch": 0.07152279056133666,
      "grad_norm": 4.411438465118408,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.856,
      "step": 122
    },
    {
      "epoch": 0.07210904294298695,
      "grad_norm": 2.734173536300659,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.4999,
      "step": 123
    },
    {
      "epoch": 0.07269529532463725,
      "grad_norm": 2.7933404445648193,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.9209,
      "step": 124
    },
    {
      "epoch": 0.07328154770628756,
      "grad_norm": 2.61995792388916,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.4888,
      "step": 125
    },
    {
      "epoch": 0.07386780008793786,
      "grad_norm": 2.6284143924713135,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.9275,
      "step": 126
    },
    {
      "epoch": 0.07445405246958815,
      "grad_norm": 3.0302343368530273,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.406,
      "step": 127
    },
    {
      "epoch": 0.07504030485123846,
      "grad_norm": 2.563217878341675,
      "learning_rate": 3.144013755408895e-05,
      "loss": 5.2984,
      "step": 128
    },
    {
      "epoch": 0.07562655723288876,
      "grad_norm": 2.6379778385162354,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.5382,
      "step": 129
    },
    {
      "epoch": 0.07621280961453906,
      "grad_norm": 2.4591808319091797,
      "learning_rate": 2.991522876735154e-05,
      "loss": 4.9354,
      "step": 130
    },
    {
      "epoch": 0.07679906199618936,
      "grad_norm": 2.7532567977905273,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.602,
      "step": 131
    },
    {
      "epoch": 0.07738531437783966,
      "grad_norm": 2.511444330215454,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.512,
      "step": 132
    },
    {
      "epoch": 0.07797156675948996,
      "grad_norm": 2.898660659790039,
      "learning_rate": 2.766955965735968e-05,
      "loss": 6.0167,
      "step": 133
    },
    {
      "epoch": 0.07855781914114027,
      "grad_norm": 2.723018169403076,
      "learning_rate": 2.693294185106562e-05,
      "loss": 6.5618,
      "step": 134
    },
    {
      "epoch": 0.07914407152279056,
      "grad_norm": 7.0018181800842285,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.0152,
      "step": 135
    },
    {
      "epoch": 0.07973032390444086,
      "grad_norm": 3.099264144897461,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.4558,
      "step": 136
    },
    {
      "epoch": 0.08031657628609117,
      "grad_norm": 3.587735891342163,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.5688,
      "step": 137
    },
    {
      "epoch": 0.08090282866774147,
      "grad_norm": 2.6472346782684326,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.5873,
      "step": 138
    },
    {
      "epoch": 0.08148908104939176,
      "grad_norm": 3.097477436065674,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.6604,
      "step": 139
    },
    {
      "epoch": 0.08207533343104206,
      "grad_norm": 3.1957669258117676,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.4788,
      "step": 140
    },
    {
      "epoch": 0.08266158581269237,
      "grad_norm": 3.3281352519989014,
      "learning_rate": 2.196424713241637e-05,
      "loss": 6.0756,
      "step": 141
    },
    {
      "epoch": 0.08324783819434266,
      "grad_norm": 5.745474815368652,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.6218,
      "step": 142
    },
    {
      "epoch": 0.08383409057599296,
      "grad_norm": 3.3978636264801025,
      "learning_rate": 2.061073738537635e-05,
      "loss": 6.1861,
      "step": 143
    },
    {
      "epoch": 0.08442034295764327,
      "grad_norm": 4.226555824279785,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 6.1142,
      "step": 144
    },
    {
      "epoch": 0.08500659533929357,
      "grad_norm": 10.205087661743164,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.8142,
      "step": 145
    },
    {
      "epoch": 0.08559284772094386,
      "grad_norm": 3.6200766563415527,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 6.3254,
      "step": 146
    },
    {
      "epoch": 0.08617910010259416,
      "grad_norm": 3.5822296142578125,
      "learning_rate": 1.800157297483417e-05,
      "loss": 6.6021,
      "step": 147
    },
    {
      "epoch": 0.08676535248424447,
      "grad_norm": 4.1829633712768555,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 6.3512,
      "step": 148
    },
    {
      "epoch": 0.08735160486589477,
      "grad_norm": 3.749767780303955,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.7628,
      "step": 149
    },
    {
      "epoch": 0.08793785724754506,
      "grad_norm": 4.500880241394043,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 7.3724,
      "step": 150
    },
    {
      "epoch": 0.08793785724754506,
      "eval_loss": 1.4424562454223633,
      "eval_runtime": 217.7477,
      "eval_samples_per_second": 13.194,
      "eval_steps_per_second": 3.302,
      "step": 150
    },
    {
      "epoch": 0.08852410962919537,
      "grad_norm": 3.762939453125,
      "learning_rate": 1.553232954407171e-05,
      "loss": 7.615,
      "step": 151
    },
    {
      "epoch": 0.08911036201084567,
      "grad_norm": 4.316311359405518,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 6.3368,
      "step": 152
    },
    {
      "epoch": 0.08969661439249597,
      "grad_norm": 4.307602882385254,
      "learning_rate": 1.435357758543015e-05,
      "loss": 6.1641,
      "step": 153
    },
    {
      "epoch": 0.09028286677414626,
      "grad_norm": 3.0188310146331787,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 5.5818,
      "step": 154
    },
    {
      "epoch": 0.09086911915579657,
      "grad_norm": 2.0442826747894287,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 5.421,
      "step": 155
    },
    {
      "epoch": 0.09145537153744687,
      "grad_norm": 2.219541311264038,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 4.9162,
      "step": 156
    },
    {
      "epoch": 0.09204162391909718,
      "grad_norm": 1.8453906774520874,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.8445,
      "step": 157
    },
    {
      "epoch": 0.09262787630074747,
      "grad_norm": 1.90604829788208,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.2768,
      "step": 158
    },
    {
      "epoch": 0.09321412868239777,
      "grad_norm": 2.1838622093200684,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 4.3835,
      "step": 159
    },
    {
      "epoch": 0.09380038106404807,
      "grad_norm": 2.372401237487793,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 4.6348,
      "step": 160
    },
    {
      "epoch": 0.09438663344569838,
      "grad_norm": 2.498532295227051,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.7201,
      "step": 161
    },
    {
      "epoch": 0.09497288582734867,
      "grad_norm": 2.3159775733947754,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.1931,
      "step": 162
    },
    {
      "epoch": 0.09555913820899897,
      "grad_norm": 2.7434277534484863,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.0402,
      "step": 163
    },
    {
      "epoch": 0.09614539059064928,
      "grad_norm": 2.5860517024993896,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.5664,
      "step": 164
    },
    {
      "epoch": 0.09673164297229958,
      "grad_norm": 2.1623849868774414,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.6256,
      "step": 165
    },
    {
      "epoch": 0.09731789535394987,
      "grad_norm": 2.36757493019104,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.4054,
      "step": 166
    },
    {
      "epoch": 0.09790414773560017,
      "grad_norm": 2.847468852996826,
      "learning_rate": 7.260364370723044e-06,
      "loss": 4.9477,
      "step": 167
    },
    {
      "epoch": 0.09849040011725048,
      "grad_norm": 2.4264841079711914,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.3519,
      "step": 168
    },
    {
      "epoch": 0.09907665249890078,
      "grad_norm": 3.365696668624878,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.2746,
      "step": 169
    },
    {
      "epoch": 0.09966290488055107,
      "grad_norm": 2.4804673194885254,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.9586,
      "step": 170
    },
    {
      "epoch": 0.10024915726220138,
      "grad_norm": 2.3575639724731445,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.751,
      "step": 171
    },
    {
      "epoch": 0.10083540964385168,
      "grad_norm": 2.990222692489624,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.3137,
      "step": 172
    },
    {
      "epoch": 0.10142166202550199,
      "grad_norm": 2.1931540966033936,
      "learning_rate": 4.900438493352055e-06,
      "loss": 5.4717,
      "step": 173
    },
    {
      "epoch": 0.10200791440715228,
      "grad_norm": 2.6367037296295166,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.3212,
      "step": 174
    },
    {
      "epoch": 0.10259416678880258,
      "grad_norm": 2.5608620643615723,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 5.4856,
      "step": 175
    },
    {
      "epoch": 0.10318041917045288,
      "grad_norm": 2.548870325088501,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.6953,
      "step": 176
    },
    {
      "epoch": 0.10376667155210317,
      "grad_norm": 2.5035483837127686,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.976,
      "step": 177
    },
    {
      "epoch": 0.10435292393375348,
      "grad_norm": 2.705728530883789,
      "learning_rate": 3.271776770026963e-06,
      "loss": 6.1103,
      "step": 178
    },
    {
      "epoch": 0.10493917631540378,
      "grad_norm": 3.1232120990753174,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 6.3249,
      "step": 179
    },
    {
      "epoch": 0.10552542869705409,
      "grad_norm": 2.642483711242676,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.5657,
      "step": 180
    },
    {
      "epoch": 0.10611168107870438,
      "grad_norm": 2.887054443359375,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 6.1515,
      "step": 181
    },
    {
      "epoch": 0.10669793346035468,
      "grad_norm": 2.6230509281158447,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.7204,
      "step": 182
    },
    {
      "epoch": 0.10728418584200498,
      "grad_norm": 3.0708298683166504,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.5642,
      "step": 183
    },
    {
      "epoch": 0.10787043822365529,
      "grad_norm": 2.7584431171417236,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.8433,
      "step": 184
    },
    {
      "epoch": 0.10845669060530558,
      "grad_norm": 2.7982194423675537,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.2949,
      "step": 185
    },
    {
      "epoch": 0.10904294298695588,
      "grad_norm": 2.8667845726013184,
      "learning_rate": 1.333670137599713e-06,
      "loss": 6.0126,
      "step": 186
    },
    {
      "epoch": 0.10962919536860619,
      "grad_norm": 3.1963698863983154,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.8353,
      "step": 187
    },
    {
      "epoch": 0.11021544775025649,
      "grad_norm": 3.4900310039520264,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.9996,
      "step": 188
    },
    {
      "epoch": 0.11080170013190678,
      "grad_norm": 3.445439100265503,
      "learning_rate": 8.247462563808817e-07,
      "loss": 6.6388,
      "step": 189
    },
    {
      "epoch": 0.11138795251355708,
      "grad_norm": 3.0243945121765137,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.6927,
      "step": 190
    },
    {
      "epoch": 0.11197420489520739,
      "grad_norm": 3.9486701488494873,
      "learning_rate": 5.526064699265753e-07,
      "loss": 6.2194,
      "step": 191
    },
    {
      "epoch": 0.11256045727685769,
      "grad_norm": 4.305104732513428,
      "learning_rate": 4.367965336512403e-07,
      "loss": 6.0657,
      "step": 192
    },
    {
      "epoch": 0.11314670965850798,
      "grad_norm": 3.078662633895874,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 6.0356,
      "step": 193
    },
    {
      "epoch": 0.11373296204015829,
      "grad_norm": 3.592252254486084,
      "learning_rate": 2.458548727494292e-07,
      "loss": 6.2811,
      "step": 194
    },
    {
      "epoch": 0.11431921442180859,
      "grad_norm": 3.661208152770996,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 6.0149,
      "step": 195
    },
    {
      "epoch": 0.1149054668034589,
      "grad_norm": 3.6830203533172607,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 6.016,
      "step": 196
    },
    {
      "epoch": 0.11549171918510918,
      "grad_norm": 3.70001482963562,
      "learning_rate": 6.150154258476315e-08,
      "loss": 7.0228,
      "step": 197
    },
    {
      "epoch": 0.11607797156675949,
      "grad_norm": 3.76037335395813,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 6.1553,
      "step": 198
    },
    {
      "epoch": 0.11666422394840979,
      "grad_norm": 5.183948993682861,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.0909,
      "step": 199
    },
    {
      "epoch": 0.1172504763300601,
      "grad_norm": 4.9264421463012695,
      "learning_rate": 0.0,
      "loss": 6.9053,
      "step": 200
    },
    {
      "epoch": 0.1172504763300601,
      "eval_loss": 1.4084903001785278,
      "eval_runtime": 217.6395,
      "eval_samples_per_second": 13.201,
      "eval_steps_per_second": 3.304,
      "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": 2.9455509164025446e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}