{
  "best_metric": 1.2369053363800049,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.3420265070542967,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0017101325352714834,
      "grad_norm": 14.640414237976074,
      "learning_rate": 1e-05,
      "loss": 5.4392,
      "step": 1
    },
    {
      "epoch": 0.0017101325352714834,
      "eval_loss": 2.389522075653076,
      "eval_runtime": 69.5278,
      "eval_samples_per_second": 14.167,
      "eval_steps_per_second": 3.553,
      "step": 1
    },
    {
      "epoch": 0.003420265070542967,
      "grad_norm": 20.288902282714844,
      "learning_rate": 2e-05,
      "loss": 6.512,
      "step": 2
    },
    {
      "epoch": 0.005130397605814451,
      "grad_norm": 18.305891036987305,
      "learning_rate": 3e-05,
      "loss": 6.8107,
      "step": 3
    },
    {
      "epoch": 0.006840530141085934,
      "grad_norm": 16.469608306884766,
      "learning_rate": 4e-05,
      "loss": 7.0771,
      "step": 4
    },
    {
      "epoch": 0.008550662676357419,
      "grad_norm": 16.029294967651367,
      "learning_rate": 5e-05,
      "loss": 6.9115,
      "step": 5
    },
    {
      "epoch": 0.010260795211628902,
      "grad_norm": 13.417438507080078,
      "learning_rate": 6e-05,
      "loss": 6.5294,
      "step": 6
    },
    {
      "epoch": 0.011970927746900385,
      "grad_norm": 13.111316680908203,
      "learning_rate": 7e-05,
      "loss": 6.4824,
      "step": 7
    },
    {
      "epoch": 0.013681060282171868,
      "grad_norm": 12.419403076171875,
      "learning_rate": 8e-05,
      "loss": 6.3141,
      "step": 8
    },
    {
      "epoch": 0.015391192817443352,
      "grad_norm": 11.091198921203613,
      "learning_rate": 9e-05,
      "loss": 6.2618,
      "step": 9
    },
    {
      "epoch": 0.017101325352714837,
      "grad_norm": 10.392128944396973,
      "learning_rate": 0.0001,
      "loss": 6.3153,
      "step": 10
    },
    {
      "epoch": 0.01881145788798632,
      "grad_norm": 9.125272750854492,
      "learning_rate": 9.999316524962345e-05,
      "loss": 5.9951,
      "step": 11
    },
    {
      "epoch": 0.020521590423257803,
      "grad_norm": 9.861510276794434,
      "learning_rate": 9.997266286704631e-05,
      "loss": 6.1288,
      "step": 12
    },
    {
      "epoch": 0.022231722958529286,
      "grad_norm": 8.92580509185791,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.7963,
      "step": 13
    },
    {
      "epoch": 0.02394185549380077,
      "grad_norm": 9.063461303710938,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.8087,
      "step": 14
    },
    {
      "epoch": 0.025651988029072252,
      "grad_norm": 9.55776596069336,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.9158,
      "step": 15
    },
    {
      "epoch": 0.027362120564343735,
      "grad_norm": 11.029727935791016,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.8921,
      "step": 16
    },
    {
      "epoch": 0.02907225309961522,
      "grad_norm": 9.607908248901367,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.6036,
      "step": 17
    },
    {
      "epoch": 0.030782385634886705,
      "grad_norm": 9.793913841247559,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.6222,
      "step": 18
    },
    {
      "epoch": 0.032492518170158184,
      "grad_norm": 8.611642837524414,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.7555,
      "step": 19
    },
    {
      "epoch": 0.034202650705429674,
      "grad_norm": 8.97571849822998,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.6779,
      "step": 20
    },
    {
      "epoch": 0.03591278324070116,
      "grad_norm": 9.353657722473145,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.8973,
      "step": 21
    },
    {
      "epoch": 0.03762291577597264,
      "grad_norm": 9.22663688659668,
      "learning_rate": 9.901899829374047e-05,
      "loss": 5.8711,
      "step": 22
    },
    {
      "epoch": 0.03933304831124412,
      "grad_norm": 8.428504943847656,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.8102,
      "step": 23
    },
    {
      "epoch": 0.041043180846515606,
      "grad_norm": 7.905302047729492,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.4847,
      "step": 24
    },
    {
      "epoch": 0.04275331338178709,
      "grad_norm": 8.299999237060547,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.8853,
      "step": 25
    },
    {
      "epoch": 0.04446344591705857,
      "grad_norm": 8.881180763244629,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.7532,
      "step": 26
    },
    {
      "epoch": 0.046173578452330055,
      "grad_norm": 8.510998725891113,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.2993,
      "step": 27
    },
    {
      "epoch": 0.04788371098760154,
      "grad_norm": 8.11540412902832,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.3116,
      "step": 28
    },
    {
      "epoch": 0.04959384352287302,
      "grad_norm": 8.081546783447266,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.7509,
      "step": 29
    },
    {
      "epoch": 0.051303976058144504,
      "grad_norm": 8.157670021057129,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.7746,
      "step": 30
    },
    {
      "epoch": 0.05301410859341599,
      "grad_norm": 8.227839469909668,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.8473,
      "step": 31
    },
    {
      "epoch": 0.05472424112868747,
      "grad_norm": 9.177515983581543,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.8004,
      "step": 32
    },
    {
      "epoch": 0.05643437366395896,
      "grad_norm": 8.275994300842285,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.0828,
      "step": 33
    },
    {
      "epoch": 0.05814450619923044,
      "grad_norm": 8.03842544555664,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.2698,
      "step": 34
    },
    {
      "epoch": 0.059854638734501926,
      "grad_norm": 9.835810661315918,
      "learning_rate": 9.578866633275288e-05,
      "loss": 5.5907,
      "step": 35
    },
    {
      "epoch": 0.06156477126977341,
      "grad_norm": 8.554304122924805,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.8263,
      "step": 36
    },
    {
      "epoch": 0.06327490380504489,
      "grad_norm": 9.51391315460205,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.0017,
      "step": 37
    },
    {
      "epoch": 0.06498503634031637,
      "grad_norm": 8.211365699768066,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.254,
      "step": 38
    },
    {
      "epoch": 0.06669516887558785,
      "grad_norm": 8.579219818115234,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.7198,
      "step": 39
    },
    {
      "epoch": 0.06840530141085935,
      "grad_norm": 8.614923477172852,
      "learning_rate": 9.397368756032445e-05,
      "loss": 5.6871,
      "step": 40
    },
    {
      "epoch": 0.07011543394613083,
      "grad_norm": 8.860878944396973,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.0608,
      "step": 41
    },
    {
      "epoch": 0.07182556648140231,
      "grad_norm": 8.788886070251465,
      "learning_rate": 9.316282404787871e-05,
      "loss": 5.9044,
      "step": 42
    },
    {
      "epoch": 0.0735356990166738,
      "grad_norm": 8.671177864074707,
      "learning_rate": 9.273963562927695e-05,
      "loss": 5.4956,
      "step": 43
    },
    {
      "epoch": 0.07524583155194528,
      "grad_norm": 8.375917434692383,
      "learning_rate": 9.230476262104677e-05,
      "loss": 5.711,
      "step": 44
    },
    {
      "epoch": 0.07695596408721676,
      "grad_norm": 8.910039901733398,
      "learning_rate": 9.185832391312644e-05,
      "loss": 5.908,
      "step": 45
    },
    {
      "epoch": 0.07866609662248825,
      "grad_norm": 9.672930717468262,
      "learning_rate": 9.140044155740101e-05,
      "loss": 5.9495,
      "step": 46
    },
    {
      "epoch": 0.08037622915775973,
      "grad_norm": 10.40195369720459,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.3817,
      "step": 47
    },
    {
      "epoch": 0.08208636169303121,
      "grad_norm": 10.205389976501465,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.1616,
      "step": 48
    },
    {
      "epoch": 0.0837964942283027,
      "grad_norm": 10.02330207824707,
      "learning_rate": 8.995939984474624e-05,
      "loss": 6.5455,
      "step": 49
    },
    {
      "epoch": 0.08550662676357418,
      "grad_norm": 11.856907844543457,
      "learning_rate": 8.945702546981969e-05,
      "loss": 6.6414,
      "step": 50
    },
    {
      "epoch": 0.08550662676357418,
      "eval_loss": 1.563002109527588,
      "eval_runtime": 70.9713,
      "eval_samples_per_second": 13.879,
      "eval_steps_per_second": 3.48,
      "step": 50
    },
    {
      "epoch": 0.08721675929884566,
      "grad_norm": 11.785542488098145,
      "learning_rate": 8.894386393810563e-05,
      "loss": 4.8819,
      "step": 51
    },
    {
      "epoch": 0.08892689183411714,
      "grad_norm": 10.87253475189209,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.6858,
      "step": 52
    },
    {
      "epoch": 0.09063702436938863,
      "grad_norm": 7.662486553192139,
      "learning_rate": 8.788574348801675e-05,
      "loss": 5.5863,
      "step": 53
    },
    {
      "epoch": 0.09234715690466011,
      "grad_norm": 5.953812599182129,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.0459,
      "step": 54
    },
    {
      "epoch": 0.0940572894399316,
      "grad_norm": 6.940700054168701,
      "learning_rate": 8.678619553365659e-05,
      "loss": 5.6281,
      "step": 55
    },
    {
      "epoch": 0.09576742197520308,
      "grad_norm": 6.840532302856445,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.4419,
      "step": 56
    },
    {
      "epoch": 0.09747755451047456,
      "grad_norm": 6.621339797973633,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.5128,
      "step": 57
    },
    {
      "epoch": 0.09918768704574604,
      "grad_norm": 6.133358478546143,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.0469,
      "step": 58
    },
    {
      "epoch": 0.10089781958101753,
      "grad_norm": 6.378569602966309,
      "learning_rate": 8.44676704559283e-05,
      "loss": 5.0157,
      "step": 59
    },
    {
      "epoch": 0.10260795211628901,
      "grad_norm": 7.036313056945801,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.2026,
      "step": 60
    },
    {
      "epoch": 0.10431808465156049,
      "grad_norm": 6.539918899536133,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.2365,
      "step": 61
    },
    {
      "epoch": 0.10602821718683197,
      "grad_norm": 6.326776504516602,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.17,
      "step": 62
    },
    {
      "epoch": 0.10773834972210346,
      "grad_norm": 6.600303649902344,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.4918,
      "step": 63
    },
    {
      "epoch": 0.10944848225737494,
      "grad_norm": 6.266218185424805,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.1133,
      "step": 64
    },
    {
      "epoch": 0.11115861479264642,
      "grad_norm": 6.132507801055908,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.1749,
      "step": 65
    },
    {
      "epoch": 0.11286874732791792,
      "grad_norm": 6.380084037780762,
      "learning_rate": 8.005405736415126e-05,
      "loss": 5.1313,
      "step": 66
    },
    {
      "epoch": 0.1145788798631894,
      "grad_norm": 6.203830242156982,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.1846,
      "step": 67
    },
    {
      "epoch": 0.11628901239846089,
      "grad_norm": 6.547234535217285,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.0656,
      "step": 68
    },
    {
      "epoch": 0.11799914493373237,
      "grad_norm": 6.533773422241211,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.0749,
      "step": 69
    },
    {
      "epoch": 0.11970927746900385,
      "grad_norm": 7.144689083099365,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.3617,
      "step": 70
    },
    {
      "epoch": 0.12141941000427534,
      "grad_norm": 7.121635437011719,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.5717,
      "step": 71
    },
    {
      "epoch": 0.12312954253954682,
      "grad_norm": 6.523452281951904,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.0268,
      "step": 72
    },
    {
      "epoch": 0.1248396750748183,
      "grad_norm": 6.674250602722168,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.5013,
      "step": 73
    },
    {
      "epoch": 0.12654980761008977,
      "grad_norm": 6.321826934814453,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.2318,
      "step": 74
    },
    {
      "epoch": 0.12825994014536127,
      "grad_norm": 6.396578788757324,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.3862,
      "step": 75
    },
    {
      "epoch": 0.12997007268063274,
      "grad_norm": 6.630756378173828,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.5383,
      "step": 76
    },
    {
      "epoch": 0.13168020521590423,
      "grad_norm": 6.786897659301758,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.6795,
      "step": 77
    },
    {
      "epoch": 0.1333903377511757,
      "grad_norm": 6.6722025871276855,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.5393,
      "step": 78
    },
    {
      "epoch": 0.1351004702864472,
      "grad_norm": 6.680119037628174,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.5529,
      "step": 79
    },
    {
      "epoch": 0.1368106028217187,
      "grad_norm": 7.221715450286865,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.4678,
      "step": 80
    },
    {
      "epoch": 0.13852073535699017,
      "grad_norm": 6.539522171020508,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.1555,
      "step": 81
    },
    {
      "epoch": 0.14023086789226166,
      "grad_norm": 6.531249523162842,
      "learning_rate": 6.855986244591104e-05,
      "loss": 4.9377,
      "step": 82
    },
    {
      "epoch": 0.14194100042753313,
      "grad_norm": 7.449616432189941,
      "learning_rate": 6.778969234612584e-05,
      "loss": 5.4777,
      "step": 83
    },
    {
      "epoch": 0.14365113296280463,
      "grad_norm": 7.076162338256836,
      "learning_rate": 6.701465872208216e-05,
      "loss": 5.1931,
      "step": 84
    },
    {
      "epoch": 0.1453612654980761,
      "grad_norm": 6.946402549743652,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.3551,
      "step": 85
    },
    {
      "epoch": 0.1470713980333476,
      "grad_norm": 7.516958713531494,
      "learning_rate": 6.545084971874738e-05,
      "loss": 5.3214,
      "step": 86
    },
    {
      "epoch": 0.14878153056861906,
      "grad_norm": 8.02112865447998,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.5993,
      "step": 87
    },
    {
      "epoch": 0.15049166310389056,
      "grad_norm": 7.544280052185059,
      "learning_rate": 6.387014543809223e-05,
      "loss": 5.3809,
      "step": 88
    },
    {
      "epoch": 0.15220179563916203,
      "grad_norm": 6.987375259399414,
      "learning_rate": 6.307399704769099e-05,
      "loss": 5.187,
      "step": 89
    },
    {
      "epoch": 0.15391192817443353,
      "grad_norm": 7.001568794250488,
      "learning_rate": 6.227427435703997e-05,
      "loss": 5.6348,
      "step": 90
    },
    {
      "epoch": 0.155622060709705,
      "grad_norm": 7.184516906738281,
      "learning_rate": 6.147119600233758e-05,
      "loss": 5.4621,
      "step": 91
    },
    {
      "epoch": 0.1573321932449765,
      "grad_norm": 6.922083377838135,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.2475,
      "step": 92
    },
    {
      "epoch": 0.15904232578024796,
      "grad_norm": 7.21707820892334,
      "learning_rate": 5.985585137257401e-05,
      "loss": 5.465,
      "step": 93
    },
    {
      "epoch": 0.16075245831551946,
      "grad_norm": 7.266740798950195,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.1979,
      "step": 94
    },
    {
      "epoch": 0.16246259085079093,
      "grad_norm": 8.004600524902344,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 5.5634,
      "step": 95
    },
    {
      "epoch": 0.16417272338606242,
      "grad_norm": 8.2750825881958,
      "learning_rate": 5.74131823855921e-05,
      "loss": 5.8962,
      "step": 96
    },
    {
      "epoch": 0.1658828559213339,
      "grad_norm": 8.06142807006836,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 5.5395,
      "step": 97
    },
    {
      "epoch": 0.1675929884566054,
      "grad_norm": 9.92174243927002,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.3507,
      "step": 98
    },
    {
      "epoch": 0.16930312099187686,
      "grad_norm": 9.798441886901855,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.3473,
      "step": 99
    },
    {
      "epoch": 0.17101325352714836,
      "grad_norm": 11.396499633789062,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 6.199,
      "step": 100
    },
    {
      "epoch": 0.17101325352714836,
      "eval_loss": 1.3837385177612305,
      "eval_runtime": 71.0039,
      "eval_samples_per_second": 13.872,
      "eval_steps_per_second": 3.479,
      "step": 100
    },
    {
      "epoch": 0.17272338606241983,
      "grad_norm": 6.1180596351623535,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.1311,
      "step": 101
    },
    {
      "epoch": 0.17443351859769132,
      "grad_norm": 6.361010551452637,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.5486,
      "step": 102
    },
    {
      "epoch": 0.1761436511329628,
      "grad_norm": 6.283280372619629,
      "learning_rate": 5.165316846586541e-05,
      "loss": 4.9742,
      "step": 103
    },
    {
      "epoch": 0.1778537836682343,
      "grad_norm": 5.876176357269287,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.9586,
      "step": 104
    },
    {
      "epoch": 0.17956391620350576,
      "grad_norm": 5.371125221252441,
      "learning_rate": 5e-05,
      "loss": 4.7882,
      "step": 105
    },
    {
      "epoch": 0.18127404873877725,
      "grad_norm": 5.151697158813477,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.9775,
      "step": 106
    },
    {
      "epoch": 0.18298418127404875,
      "grad_norm": 5.419619560241699,
      "learning_rate": 4.834683153413459e-05,
      "loss": 5.2023,
      "step": 107
    },
    {
      "epoch": 0.18469431380932022,
      "grad_norm": 5.808403491973877,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 5.1266,
      "step": 108
    },
    {
      "epoch": 0.18640444634459172,
      "grad_norm": 5.708106517791748,
      "learning_rate": 4.669547078371504e-05,
      "loss": 4.7815,
      "step": 109
    },
    {
      "epoch": 0.1881145788798632,
      "grad_norm": 5.750987529754639,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.1993,
      "step": 110
    },
    {
      "epoch": 0.18982471141513468,
      "grad_norm": 5.842879772186279,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.0834,
      "step": 111
    },
    {
      "epoch": 0.19153484395040615,
      "grad_norm": 5.688647747039795,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.1542,
      "step": 112
    },
    {
      "epoch": 0.19324497648567765,
      "grad_norm": 6.085207939147949,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.3708,
      "step": 113
    },
    {
      "epoch": 0.19495510902094912,
      "grad_norm": 5.445624351501465,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 4.8629,
      "step": 114
    },
    {
      "epoch": 0.19666524155622062,
      "grad_norm": 5.688288688659668,
      "learning_rate": 4.17702704859633e-05,
      "loss": 4.9789,
      "step": 115
    },
    {
      "epoch": 0.19837537409149208,
      "grad_norm": 6.147029399871826,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.0329,
      "step": 116
    },
    {
      "epoch": 0.20008550662676358,
      "grad_norm": 5.8157196044921875,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 4.9307,
      "step": 117
    },
    {
      "epoch": 0.20179563916203505,
      "grad_norm": 5.963608741760254,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.11,
      "step": 118
    },
    {
      "epoch": 0.20350577169730655,
      "grad_norm": 5.836517810821533,
      "learning_rate": 3.852880399766243e-05,
      "loss": 4.9338,
      "step": 119
    },
    {
      "epoch": 0.20521590423257802,
      "grad_norm": 6.028753280639648,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.018,
      "step": 120
    },
    {
      "epoch": 0.2069260367678495,
      "grad_norm": 5.885068893432617,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 4.7368,
      "step": 121
    },
    {
      "epoch": 0.20863616930312098,
      "grad_norm": 6.388590335845947,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.1395,
      "step": 122
    },
    {
      "epoch": 0.21034630183839248,
      "grad_norm": 6.1952385902404785,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.1751,
      "step": 123
    },
    {
      "epoch": 0.21205643437366395,
      "grad_norm": 6.2135443687438965,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.012,
      "step": 124
    },
    {
      "epoch": 0.21376656690893545,
      "grad_norm": 6.214436054229736,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.0691,
      "step": 125
    },
    {
      "epoch": 0.21547669944420692,
      "grad_norm": 6.635720252990723,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.3663,
      "step": 126
    },
    {
      "epoch": 0.2171868319794784,
      "grad_norm": 6.239476203918457,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.0767,
      "step": 127
    },
    {
      "epoch": 0.21889696451474988,
      "grad_norm": 6.299355983734131,
      "learning_rate": 3.144013755408895e-05,
      "loss": 5.2056,
      "step": 128
    },
    {
      "epoch": 0.22060709705002138,
      "grad_norm": 6.062747955322266,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 4.7423,
      "step": 129
    },
    {
      "epoch": 0.22231722958529285,
      "grad_norm": 6.572103023529053,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.1702,
      "step": 130
    },
    {
      "epoch": 0.22402736212056434,
      "grad_norm": 6.492870330810547,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.5961,
      "step": 131
    },
    {
      "epoch": 0.22573749465583584,
      "grad_norm": 6.8842997550964355,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.5233,
      "step": 132
    },
    {
      "epoch": 0.2274476271911073,
      "grad_norm": 6.2079973220825195,
      "learning_rate": 2.766955965735968e-05,
      "loss": 4.9688,
      "step": 133
    },
    {
      "epoch": 0.2291577597263788,
      "grad_norm": 6.081884384155273,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.1224,
      "step": 134
    },
    {
      "epoch": 0.23086789226165028,
      "grad_norm": 6.672243595123291,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.1432,
      "step": 135
    },
    {
      "epoch": 0.23257802479692177,
      "grad_norm": 6.71869421005249,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.3586,
      "step": 136
    },
    {
      "epoch": 0.23428815733219324,
      "grad_norm": 6.9483537673950195,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.4318,
      "step": 137
    },
    {
      "epoch": 0.23599828986746474,
      "grad_norm": 6.899090766906738,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.2644,
      "step": 138
    },
    {
      "epoch": 0.2377084224027362,
      "grad_norm": 7.219109535217285,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.4093,
      "step": 139
    },
    {
      "epoch": 0.2394185549380077,
      "grad_norm": 7.686084270477295,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.2123,
      "step": 140
    },
    {
      "epoch": 0.24112868747327917,
      "grad_norm": 6.914151668548584,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.3619,
      "step": 141
    },
    {
      "epoch": 0.24283882000855067,
      "grad_norm": 7.846366882324219,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.3426,
      "step": 142
    },
    {
      "epoch": 0.24454895254382214,
      "grad_norm": 7.538131237030029,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.6459,
      "step": 143
    },
    {
      "epoch": 0.24625908507909364,
      "grad_norm": 7.204062461853027,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.6492,
      "step": 144
    },
    {
      "epoch": 0.2479692176143651,
      "grad_norm": 7.883687496185303,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.4739,
      "step": 145
    },
    {
      "epoch": 0.2496793501496366,
      "grad_norm": 7.9401350021362305,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.3524,
      "step": 146
    },
    {
      "epoch": 0.25138948268490807,
      "grad_norm": 7.743297100067139,
      "learning_rate": 1.800157297483417e-05,
      "loss": 5.1239,
      "step": 147
    },
    {
      "epoch": 0.25309961522017954,
      "grad_norm": 8.06339168548584,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 5.6102,
      "step": 148
    },
    {
      "epoch": 0.25480974775545107,
      "grad_norm": 8.759541511535645,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.1376,
      "step": 149
    },
    {
      "epoch": 0.25651988029072254,
      "grad_norm": 11.883188247680664,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 6.0998,
      "step": 150
    },
    {
      "epoch": 0.25651988029072254,
      "eval_loss": 1.2809289693832397,
      "eval_runtime": 71.0145,
      "eval_samples_per_second": 13.87,
      "eval_steps_per_second": 3.478,
      "step": 150
    },
    {
      "epoch": 0.258230012825994,
      "grad_norm": 4.4953413009643555,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.7884,
      "step": 151
    },
    {
      "epoch": 0.2599401453612655,
      "grad_norm": 5.304215431213379,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.2682,
      "step": 152
    },
    {
      "epoch": 0.261650277896537,
      "grad_norm": 5.452952861785889,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.6377,
      "step": 153
    },
    {
      "epoch": 0.26336041043180847,
      "grad_norm": 6.52045202255249,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 4.8163,
      "step": 154
    },
    {
      "epoch": 0.26507054296707994,
      "grad_norm": 6.399850845336914,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.8225,
      "step": 155
    },
    {
      "epoch": 0.2667806755023514,
      "grad_norm": 6.2536749839782715,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.0223,
      "step": 156
    },
    {
      "epoch": 0.26849080803762293,
      "grad_norm": 6.324411392211914,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.4117,
      "step": 157
    },
    {
      "epoch": 0.2702009405728944,
      "grad_norm": 5.510275363922119,
      "learning_rate": 1.157994445715706e-05,
      "loss": 4.8219,
      "step": 158
    },
    {
      "epoch": 0.27191107310816587,
      "grad_norm": 5.786247730255127,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.2305,
      "step": 159
    },
    {
      "epoch": 0.2736212056434374,
      "grad_norm": 5.628164291381836,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.0321,
      "step": 160
    },
    {
      "epoch": 0.27533133817870886,
      "grad_norm": 5.594108581542969,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.8256,
      "step": 161
    },
    {
      "epoch": 0.27704147071398033,
      "grad_norm": 5.432445049285889,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.6674,
      "step": 162
    },
    {
      "epoch": 0.2787516032492518,
      "grad_norm": 5.55719518661499,
      "learning_rate": 9.068759265665384e-06,
      "loss": 4.6913,
      "step": 163
    },
    {
      "epoch": 0.2804617357845233,
      "grad_norm": 5.277993202209473,
      "learning_rate": 8.599558442598998e-06,
      "loss": 4.8593,
      "step": 164
    },
    {
      "epoch": 0.2821718683197948,
      "grad_norm": 5.357100963592529,
      "learning_rate": 8.141676086873572e-06,
      "loss": 4.8837,
      "step": 165
    },
    {
      "epoch": 0.28388200085506626,
      "grad_norm": 5.337014675140381,
      "learning_rate": 7.695237378953223e-06,
      "loss": 4.8787,
      "step": 166
    },
    {
      "epoch": 0.28559213339033773,
      "grad_norm": 5.865081310272217,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.0684,
      "step": 167
    },
    {
      "epoch": 0.28730226592560926,
      "grad_norm": 5.6288161277771,
      "learning_rate": 6.837175952121306e-06,
      "loss": 4.8526,
      "step": 168
    },
    {
      "epoch": 0.2890123984608807,
      "grad_norm": 5.587432384490967,
      "learning_rate": 6.425787818636131e-06,
      "loss": 4.6702,
      "step": 169
    },
    {
      "epoch": 0.2907225309961522,
      "grad_norm": 5.569537162780762,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.821,
      "step": 170
    },
    {
      "epoch": 0.29243266353142366,
      "grad_norm": 5.6782989501953125,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.0815,
      "step": 171
    },
    {
      "epoch": 0.2941427960666952,
      "grad_norm": 5.838556289672852,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.0719,
      "step": 172
    },
    {
      "epoch": 0.29585292860196666,
      "grad_norm": 5.764434814453125,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.8881,
      "step": 173
    },
    {
      "epoch": 0.2975630611372381,
      "grad_norm": 5.603269577026367,
      "learning_rate": 4.549673247541875e-06,
      "loss": 4.9162,
      "step": 174
    },
    {
      "epoch": 0.2992731936725096,
      "grad_norm": 5.911520481109619,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.9465,
      "step": 175
    },
    {
      "epoch": 0.3009833262077811,
      "grad_norm": 6.086195945739746,
      "learning_rate": 3.885512251130763e-06,
      "loss": 4.8536,
      "step": 176
    },
    {
      "epoch": 0.3026934587430526,
      "grad_norm": 6.540954113006592,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.3478,
      "step": 177
    },
    {
      "epoch": 0.30440359127832406,
      "grad_norm": 6.309847354888916,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.2735,
      "step": 178
    },
    {
      "epoch": 0.30611372381359553,
      "grad_norm": 6.2063775062561035,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 4.9368,
      "step": 179
    },
    {
      "epoch": 0.30782385634886705,
      "grad_norm": 6.023078441619873,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.1261,
      "step": 180
    },
    {
      "epoch": 0.3095339888841385,
      "grad_norm": 6.490723133087158,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.1628,
      "step": 181
    },
    {
      "epoch": 0.31124412141941,
      "grad_norm": 6.0313873291015625,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 4.9787,
      "step": 182
    },
    {
      "epoch": 0.31295425395468146,
      "grad_norm": 6.360531330108643,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.0005,
      "step": 183
    },
    {
      "epoch": 0.314664386489953,
      "grad_norm": 6.467580318450928,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.0739,
      "step": 184
    },
    {
      "epoch": 0.31637451902522445,
      "grad_norm": 6.431023597717285,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.0959,
      "step": 185
    },
    {
      "epoch": 0.3180846515604959,
      "grad_norm": 6.457474708557129,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.2955,
      "step": 186
    },
    {
      "epoch": 0.31979478409576745,
      "grad_norm": 6.595992565155029,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.1507,
      "step": 187
    },
    {
      "epoch": 0.3215049166310389,
      "grad_norm": 7.012680530548096,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.3988,
      "step": 188
    },
    {
      "epoch": 0.3232150491663104,
      "grad_norm": 6.616781711578369,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.2315,
      "step": 189
    },
    {
      "epoch": 0.32492518170158186,
      "grad_norm": 7.558001518249512,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.758,
      "step": 190
    },
    {
      "epoch": 0.3266353142368534,
      "grad_norm": 6.658862590789795,
      "learning_rate": 5.526064699265753e-07,
      "loss": 5.1068,
      "step": 191
    },
    {
      "epoch": 0.32834544677212485,
      "grad_norm": 7.025755882263184,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.3845,
      "step": 192
    },
    {
      "epoch": 0.3300555793073963,
      "grad_norm": 7.665297508239746,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.4417,
      "step": 193
    },
    {
      "epoch": 0.3317657118426678,
      "grad_norm": 7.272705554962158,
      "learning_rate": 2.458548727494292e-07,
      "loss": 5.2698,
      "step": 194
    },
    {
      "epoch": 0.3334758443779393,
      "grad_norm": 8.50095272064209,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.3532,
      "step": 195
    },
    {
      "epoch": 0.3351859769132108,
      "grad_norm": 7.427955627441406,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 5.291,
      "step": 196
    },
    {
      "epoch": 0.33689610944848225,
      "grad_norm": 8.0869779586792,
      "learning_rate": 6.150154258476315e-08,
      "loss": 5.5174,
      "step": 197
    },
    {
      "epoch": 0.3386062419837537,
      "grad_norm": 9.528826713562012,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 6.0544,
      "step": 198
    },
    {
      "epoch": 0.34031637451902524,
      "grad_norm": 9.277633666992188,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.0752,
      "step": 199
    },
    {
      "epoch": 0.3420265070542967,
      "grad_norm": 11.991414070129395,
      "learning_rate": 0.0,
      "loss": 6.2945,
      "step": 200
    },
    {
      "epoch": 0.3420265070542967,
      "eval_loss": 1.2369053363800049,
      "eval_runtime": 71.0035,
      "eval_samples_per_second": 13.873,
      "eval_steps_per_second": 3.479,
      "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": 2.863232968556544e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}