{
  "best_metric": 1.2364895343780518,
  "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": 14.647956848144531,
      "learning_rate": 1e-05,
      "loss": 5.4392,
      "step": 1
    },
    {
      "epoch": 0.0017101325352714834,
      "eval_loss": 2.389522075653076,
      "eval_runtime": 70.0498,
      "eval_samples_per_second": 14.061,
      "eval_steps_per_second": 3.526,
      "step": 1
    },
    {
      "epoch": 0.003420265070542967,
      "grad_norm": 20.373123168945312,
      "learning_rate": 2e-05,
      "loss": 6.512,
      "step": 2
    },
    {
      "epoch": 0.005130397605814451,
      "grad_norm": 18.366159439086914,
      "learning_rate": 3e-05,
      "loss": 6.811,
      "step": 3
    },
    {
      "epoch": 0.006840530141085934,
      "grad_norm": 16.541837692260742,
      "learning_rate": 4e-05,
      "loss": 7.0807,
      "step": 4
    },
    {
      "epoch": 0.008550662676357419,
      "grad_norm": 16.134265899658203,
      "learning_rate": 5e-05,
      "loss": 6.9222,
      "step": 5
    },
    {
      "epoch": 0.010260795211628902,
      "grad_norm": 13.496607780456543,
      "learning_rate": 6e-05,
      "loss": 6.535,
      "step": 6
    },
    {
      "epoch": 0.011970927746900385,
      "grad_norm": 12.634574890136719,
      "learning_rate": 7e-05,
      "loss": 6.4888,
      "step": 7
    },
    {
      "epoch": 0.013681060282171868,
      "grad_norm": 12.351262092590332,
      "learning_rate": 8e-05,
      "loss": 6.2987,
      "step": 8
    },
    {
      "epoch": 0.015391192817443352,
      "grad_norm": 10.987244606018066,
      "learning_rate": 9e-05,
      "loss": 6.2655,
      "step": 9
    },
    {
      "epoch": 0.017101325352714837,
      "grad_norm": 10.413485527038574,
      "learning_rate": 0.0001,
      "loss": 6.3182,
      "step": 10
    },
    {
      "epoch": 0.01881145788798632,
      "grad_norm": 9.127406120300293,
      "learning_rate": 9.999316524962345e-05,
      "loss": 5.9941,
      "step": 11
    },
    {
      "epoch": 0.020521590423257803,
      "grad_norm": 10.038461685180664,
      "learning_rate": 9.997266286704631e-05,
      "loss": 6.13,
      "step": 12
    },
    {
      "epoch": 0.022231722958529286,
      "grad_norm": 8.865433692932129,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.7875,
      "step": 13
    },
    {
      "epoch": 0.02394185549380077,
      "grad_norm": 9.137649536132812,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.8117,
      "step": 14
    },
    {
      "epoch": 0.025651988029072252,
      "grad_norm": 9.664212226867676,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.9177,
      "step": 15
    },
    {
      "epoch": 0.027362120564343735,
      "grad_norm": 11.016179084777832,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.8847,
      "step": 16
    },
    {
      "epoch": 0.02907225309961522,
      "grad_norm": 9.859611511230469,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.6148,
      "step": 17
    },
    {
      "epoch": 0.030782385634886705,
      "grad_norm": 9.732139587402344,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.6255,
      "step": 18
    },
    {
      "epoch": 0.032492518170158184,
      "grad_norm": 8.641677856445312,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.7555,
      "step": 19
    },
    {
      "epoch": 0.034202650705429674,
      "grad_norm": 8.856379508972168,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.6799,
      "step": 20
    },
    {
      "epoch": 0.03591278324070116,
      "grad_norm": 9.629952430725098,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.9026,
      "step": 21
    },
    {
      "epoch": 0.03762291577597264,
      "grad_norm": 9.597189903259277,
      "learning_rate": 9.901899829374047e-05,
      "loss": 5.8821,
      "step": 22
    },
    {
      "epoch": 0.03933304831124412,
      "grad_norm": 8.446582794189453,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.8156,
      "step": 23
    },
    {
      "epoch": 0.041043180846515606,
      "grad_norm": 7.964376926422119,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.4823,
      "step": 24
    },
    {
      "epoch": 0.04275331338178709,
      "grad_norm": 8.392891883850098,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.8879,
      "step": 25
    },
    {
      "epoch": 0.04446344591705857,
      "grad_norm": 9.026515007019043,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.7524,
      "step": 26
    },
    {
      "epoch": 0.046173578452330055,
      "grad_norm": 8.499021530151367,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.3138,
      "step": 27
    },
    {
      "epoch": 0.04788371098760154,
      "grad_norm": 9.321228981018066,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.3142,
      "step": 28
    },
    {
      "epoch": 0.04959384352287302,
      "grad_norm": 8.147331237792969,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.7566,
      "step": 29
    },
    {
      "epoch": 0.051303976058144504,
      "grad_norm": 8.19190502166748,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.7796,
      "step": 30
    },
    {
      "epoch": 0.05301410859341599,
      "grad_norm": 8.113025665283203,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.8476,
      "step": 31
    },
    {
      "epoch": 0.05472424112868747,
      "grad_norm": 9.073809623718262,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.7916,
      "step": 32
    },
    {
      "epoch": 0.05643437366395896,
      "grad_norm": 8.299051284790039,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.0941,
      "step": 33
    },
    {
      "epoch": 0.05814450619923044,
      "grad_norm": 8.128596305847168,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.2846,
      "step": 34
    },
    {
      "epoch": 0.059854638734501926,
      "grad_norm": 9.958587646484375,
      "learning_rate": 9.578866633275288e-05,
      "loss": 5.5907,
      "step": 35
    },
    {
      "epoch": 0.06156477126977341,
      "grad_norm": 8.5099515914917,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.8238,
      "step": 36
    },
    {
      "epoch": 0.06327490380504489,
      "grad_norm": 9.221707344055176,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.0025,
      "step": 37
    },
    {
      "epoch": 0.06498503634031637,
      "grad_norm": 8.256094932556152,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.2493,
      "step": 38
    },
    {
      "epoch": 0.06669516887558785,
      "grad_norm": 8.52989673614502,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.725,
      "step": 39
    },
    {
      "epoch": 0.06840530141085935,
      "grad_norm": 8.509799003601074,
      "learning_rate": 9.397368756032445e-05,
      "loss": 5.6857,
      "step": 40
    },
    {
      "epoch": 0.07011543394613083,
      "grad_norm": 8.758727073669434,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.0527,
      "step": 41
    },
    {
      "epoch": 0.07182556648140231,
      "grad_norm": 8.856103897094727,
      "learning_rate": 9.316282404787871e-05,
      "loss": 5.8952,
      "step": 42
    },
    {
      "epoch": 0.0735356990166738,
      "grad_norm": 8.542580604553223,
      "learning_rate": 9.273963562927695e-05,
      "loss": 5.5058,
      "step": 43
    },
    {
      "epoch": 0.07524583155194528,
      "grad_norm": 8.372459411621094,
      "learning_rate": 9.230476262104677e-05,
      "loss": 5.697,
      "step": 44
    },
    {
      "epoch": 0.07695596408721676,
      "grad_norm": 8.722542762756348,
      "learning_rate": 9.185832391312644e-05,
      "loss": 5.9055,
      "step": 45
    },
    {
      "epoch": 0.07866609662248825,
      "grad_norm": 9.559613227844238,
      "learning_rate": 9.140044155740101e-05,
      "loss": 5.9552,
      "step": 46
    },
    {
      "epoch": 0.08037622915775973,
      "grad_norm": 10.255206108093262,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.3674,
      "step": 47
    },
    {
      "epoch": 0.08208636169303121,
      "grad_norm": 10.23193359375,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.1721,
      "step": 48
    },
    {
      "epoch": 0.0837964942283027,
      "grad_norm": 9.893390655517578,
      "learning_rate": 8.995939984474624e-05,
      "loss": 6.5567,
      "step": 49
    },
    {
      "epoch": 0.08550662676357418,
      "grad_norm": 11.815103530883789,
      "learning_rate": 8.945702546981969e-05,
      "loss": 6.6826,
      "step": 50
    },
    {
      "epoch": 0.08550662676357418,
      "eval_loss": 1.5590479373931885,
      "eval_runtime": 71.164,
      "eval_samples_per_second": 13.841,
      "eval_steps_per_second": 3.471,
      "step": 50
    },
    {
      "epoch": 0.08721675929884566,
      "grad_norm": 11.670133590698242,
      "learning_rate": 8.894386393810563e-05,
      "loss": 4.8665,
      "step": 51
    },
    {
      "epoch": 0.08892689183411714,
      "grad_norm": 10.765670776367188,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.6777,
      "step": 52
    },
    {
      "epoch": 0.09063702436938863,
      "grad_norm": 7.763930797576904,
      "learning_rate": 8.788574348801675e-05,
      "loss": 5.6151,
      "step": 53
    },
    {
      "epoch": 0.09234715690466011,
      "grad_norm": 5.661214828491211,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.0488,
      "step": 54
    },
    {
      "epoch": 0.0940572894399316,
      "grad_norm": 6.892148017883301,
      "learning_rate": 8.678619553365659e-05,
      "loss": 5.642,
      "step": 55
    },
    {
      "epoch": 0.09576742197520308,
      "grad_norm": 6.748917102813721,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.4259,
      "step": 56
    },
    {
      "epoch": 0.09747755451047456,
      "grad_norm": 6.506460189819336,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.5046,
      "step": 57
    },
    {
      "epoch": 0.09918768704574604,
      "grad_norm": 6.217710971832275,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.0724,
      "step": 58
    },
    {
      "epoch": 0.10089781958101753,
      "grad_norm": 6.443725109100342,
      "learning_rate": 8.44676704559283e-05,
      "loss": 5.0304,
      "step": 59
    },
    {
      "epoch": 0.10260795211628901,
      "grad_norm": 6.875208377838135,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.1855,
      "step": 60
    },
    {
      "epoch": 0.10431808465156049,
      "grad_norm": 6.576230049133301,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.2583,
      "step": 61
    },
    {
      "epoch": 0.10602821718683197,
      "grad_norm": 6.3259429931640625,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.1767,
      "step": 62
    },
    {
      "epoch": 0.10773834972210346,
      "grad_norm": 6.573944091796875,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.4993,
      "step": 63
    },
    {
      "epoch": 0.10944848225737494,
      "grad_norm": 6.2207231521606445,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.1262,
      "step": 64
    },
    {
      "epoch": 0.11115861479264642,
      "grad_norm": 6.0934553146362305,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.1846,
      "step": 65
    },
    {
      "epoch": 0.11286874732791792,
      "grad_norm": 6.30827522277832,
      "learning_rate": 8.005405736415126e-05,
      "loss": 5.1388,
      "step": 66
    },
    {
      "epoch": 0.1145788798631894,
      "grad_norm": 6.256842136383057,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.1978,
      "step": 67
    },
    {
      "epoch": 0.11628901239846089,
      "grad_norm": 6.585797309875488,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.0545,
      "step": 68
    },
    {
      "epoch": 0.11799914493373237,
      "grad_norm": 6.512587070465088,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.0827,
      "step": 69
    },
    {
      "epoch": 0.11970927746900385,
      "grad_norm": 7.0146260261535645,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.3634,
      "step": 70
    },
    {
      "epoch": 0.12141941000427534,
      "grad_norm": 7.003252983093262,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.5665,
      "step": 71
    },
    {
      "epoch": 0.12312954253954682,
      "grad_norm": 6.686169147491455,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.0258,
      "step": 72
    },
    {
      "epoch": 0.1248396750748183,
      "grad_norm": 6.716172218322754,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.4769,
      "step": 73
    },
    {
      "epoch": 0.12654980761008977,
      "grad_norm": 6.407820701599121,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.2346,
      "step": 74
    },
    {
      "epoch": 0.12825994014536127,
      "grad_norm": 6.440876483917236,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.3857,
      "step": 75
    },
    {
      "epoch": 0.12997007268063274,
      "grad_norm": 6.7987236976623535,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.5307,
      "step": 76
    },
    {
      "epoch": 0.13168020521590423,
      "grad_norm": 6.891758441925049,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.6692,
      "step": 77
    },
    {
      "epoch": 0.1333903377511757,
      "grad_norm": 6.718648910522461,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.5319,
      "step": 78
    },
    {
      "epoch": 0.1351004702864472,
      "grad_norm": 6.724673271179199,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.5484,
      "step": 79
    },
    {
      "epoch": 0.1368106028217187,
      "grad_norm": 7.216301441192627,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.4996,
      "step": 80
    },
    {
      "epoch": 0.13852073535699017,
      "grad_norm": 6.592189788818359,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.1591,
      "step": 81
    },
    {
      "epoch": 0.14023086789226166,
      "grad_norm": 6.543399333953857,
      "learning_rate": 6.855986244591104e-05,
      "loss": 4.927,
      "step": 82
    },
    {
      "epoch": 0.14194100042753313,
      "grad_norm": 7.43008279800415,
      "learning_rate": 6.778969234612584e-05,
      "loss": 5.4799,
      "step": 83
    },
    {
      "epoch": 0.14365113296280463,
      "grad_norm": 7.017084121704102,
      "learning_rate": 6.701465872208216e-05,
      "loss": 5.1831,
      "step": 84
    },
    {
      "epoch": 0.1453612654980761,
      "grad_norm": 7.088075637817383,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.3733,
      "step": 85
    },
    {
      "epoch": 0.1470713980333476,
      "grad_norm": 7.448460102081299,
      "learning_rate": 6.545084971874738e-05,
      "loss": 5.3094,
      "step": 86
    },
    {
      "epoch": 0.14878153056861906,
      "grad_norm": 7.983953475952148,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.5923,
      "step": 87
    },
    {
      "epoch": 0.15049166310389056,
      "grad_norm": 7.6488261222839355,
      "learning_rate": 6.387014543809223e-05,
      "loss": 5.4021,
      "step": 88
    },
    {
      "epoch": 0.15220179563916203,
      "grad_norm": 7.249059677124023,
      "learning_rate": 6.307399704769099e-05,
      "loss": 5.1865,
      "step": 89
    },
    {
      "epoch": 0.15391192817443353,
      "grad_norm": 7.189175605773926,
      "learning_rate": 6.227427435703997e-05,
      "loss": 5.6287,
      "step": 90
    },
    {
      "epoch": 0.155622060709705,
      "grad_norm": 7.227424144744873,
      "learning_rate": 6.147119600233758e-05,
      "loss": 5.4788,
      "step": 91
    },
    {
      "epoch": 0.1573321932449765,
      "grad_norm": 7.495189189910889,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.2419,
      "step": 92
    },
    {
      "epoch": 0.15904232578024796,
      "grad_norm": 7.223033905029297,
      "learning_rate": 5.985585137257401e-05,
      "loss": 5.4925,
      "step": 93
    },
    {
      "epoch": 0.16075245831551946,
      "grad_norm": 7.196437835693359,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.1935,
      "step": 94
    },
    {
      "epoch": 0.16246259085079093,
      "grad_norm": 8.053117752075195,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 5.5453,
      "step": 95
    },
    {
      "epoch": 0.16417272338606242,
      "grad_norm": 8.177412033081055,
      "learning_rate": 5.74131823855921e-05,
      "loss": 5.8953,
      "step": 96
    },
    {
      "epoch": 0.1658828559213339,
      "grad_norm": 8.064677238464355,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 5.5499,
      "step": 97
    },
    {
      "epoch": 0.1675929884566054,
      "grad_norm": 9.8474760055542,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.354,
      "step": 98
    },
    {
      "epoch": 0.16930312099187686,
      "grad_norm": 9.995335578918457,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.3403,
      "step": 99
    },
    {
      "epoch": 0.17101325352714836,
      "grad_norm": 11.376033782958984,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 6.1669,
      "step": 100
    },
    {
      "epoch": 0.17101325352714836,
      "eval_loss": 1.382412314414978,
      "eval_runtime": 71.2056,
      "eval_samples_per_second": 13.833,
      "eval_steps_per_second": 3.469,
      "step": 100
    },
    {
      "epoch": 0.17272338606241983,
      "grad_norm": 5.957542896270752,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.1305,
      "step": 101
    },
    {
      "epoch": 0.17443351859769132,
      "grad_norm": 6.4855732917785645,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.5345,
      "step": 102
    },
    {
      "epoch": 0.1761436511329628,
      "grad_norm": 6.2549519538879395,
      "learning_rate": 5.165316846586541e-05,
      "loss": 4.9724,
      "step": 103
    },
    {
      "epoch": 0.1778537836682343,
      "grad_norm": 5.879164218902588,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.9681,
      "step": 104
    },
    {
      "epoch": 0.17956391620350576,
      "grad_norm": 5.322740077972412,
      "learning_rate": 5e-05,
      "loss": 4.7922,
      "step": 105
    },
    {
      "epoch": 0.18127404873877725,
      "grad_norm": 5.165755271911621,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.9845,
      "step": 106
    },
    {
      "epoch": 0.18298418127404875,
      "grad_norm": 5.382776260375977,
      "learning_rate": 4.834683153413459e-05,
      "loss": 5.2145,
      "step": 107
    },
    {
      "epoch": 0.18469431380932022,
      "grad_norm": 5.669363975524902,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 5.1168,
      "step": 108
    },
    {
      "epoch": 0.18640444634459172,
      "grad_norm": 5.768848896026611,
      "learning_rate": 4.669547078371504e-05,
      "loss": 4.7837,
      "step": 109
    },
    {
      "epoch": 0.1881145788798632,
      "grad_norm": 5.866610527038574,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.1897,
      "step": 110
    },
    {
      "epoch": 0.18982471141513468,
      "grad_norm": 5.909623622894287,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.0922,
      "step": 111
    },
    {
      "epoch": 0.19153484395040615,
      "grad_norm": 5.68808126449585,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.1415,
      "step": 112
    },
    {
      "epoch": 0.19324497648567765,
      "grad_norm": 5.953550338745117,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.3953,
      "step": 113
    },
    {
      "epoch": 0.19495510902094912,
      "grad_norm": 5.5413665771484375,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 4.8691,
      "step": 114
    },
    {
      "epoch": 0.19666524155622062,
      "grad_norm": 5.718400955200195,
      "learning_rate": 4.17702704859633e-05,
      "loss": 4.9716,
      "step": 115
    },
    {
      "epoch": 0.19837537409149208,
      "grad_norm": 6.088201522827148,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.0398,
      "step": 116
    },
    {
      "epoch": 0.20008550662676358,
      "grad_norm": 5.787220478057861,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 4.921,
      "step": 117
    },
    {
      "epoch": 0.20179563916203505,
      "grad_norm": 6.0000319480896,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.1257,
      "step": 118
    },
    {
      "epoch": 0.20350577169730655,
      "grad_norm": 5.856043815612793,
      "learning_rate": 3.852880399766243e-05,
      "loss": 4.9343,
      "step": 119
    },
    {
      "epoch": 0.20521590423257802,
      "grad_norm": 5.995108604431152,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.0141,
      "step": 120
    },
    {
      "epoch": 0.2069260367678495,
      "grad_norm": 5.846787929534912,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 4.7522,
      "step": 121
    },
    {
      "epoch": 0.20863616930312098,
      "grad_norm": 6.454184532165527,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.1337,
      "step": 122
    },
    {
      "epoch": 0.21034630183839248,
      "grad_norm": 6.224088668823242,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.1751,
      "step": 123
    },
    {
      "epoch": 0.21205643437366395,
      "grad_norm": 6.210409164428711,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.0064,
      "step": 124
    },
    {
      "epoch": 0.21376656690893545,
      "grad_norm": 6.251707553863525,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.0831,
      "step": 125
    },
    {
      "epoch": 0.21547669944420692,
      "grad_norm": 6.671133518218994,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.3674,
      "step": 126
    },
    {
      "epoch": 0.2171868319794784,
      "grad_norm": 6.227624893188477,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.0815,
      "step": 127
    },
    {
      "epoch": 0.21889696451474988,
      "grad_norm": 6.270323276519775,
      "learning_rate": 3.144013755408895e-05,
      "loss": 5.2016,
      "step": 128
    },
    {
      "epoch": 0.22060709705002138,
      "grad_norm": 6.1814284324646,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 4.7255,
      "step": 129
    },
    {
      "epoch": 0.22231722958529285,
      "grad_norm": 6.574129581451416,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.1615,
      "step": 130
    },
    {
      "epoch": 0.22402736212056434,
      "grad_norm": 6.457674026489258,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.606,
      "step": 131
    },
    {
      "epoch": 0.22573749465583584,
      "grad_norm": 6.920170783996582,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.5128,
      "step": 132
    },
    {
      "epoch": 0.2274476271911073,
      "grad_norm": 6.169193744659424,
      "learning_rate": 2.766955965735968e-05,
      "loss": 4.9557,
      "step": 133
    },
    {
      "epoch": 0.2291577597263788,
      "grad_norm": 6.105835914611816,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.1213,
      "step": 134
    },
    {
      "epoch": 0.23086789226165028,
      "grad_norm": 6.616054534912109,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.1441,
      "step": 135
    },
    {
      "epoch": 0.23257802479692177,
      "grad_norm": 6.647872447967529,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.3619,
      "step": 136
    },
    {
      "epoch": 0.23428815733219324,
      "grad_norm": 6.889586448669434,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.4472,
      "step": 137
    },
    {
      "epoch": 0.23599828986746474,
      "grad_norm": 6.807645320892334,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.2571,
      "step": 138
    },
    {
      "epoch": 0.2377084224027362,
      "grad_norm": 7.267321586608887,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.4178,
      "step": 139
    },
    {
      "epoch": 0.2394185549380077,
      "grad_norm": 7.587237358093262,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.2038,
      "step": 140
    },
    {
      "epoch": 0.24112868747327917,
      "grad_norm": 6.994505882263184,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.3623,
      "step": 141
    },
    {
      "epoch": 0.24283882000855067,
      "grad_norm": 7.791904926300049,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.3432,
      "step": 142
    },
    {
      "epoch": 0.24454895254382214,
      "grad_norm": 7.552254676818848,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.6422,
      "step": 143
    },
    {
      "epoch": 0.24625908507909364,
      "grad_norm": 7.126366138458252,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.6549,
      "step": 144
    },
    {
      "epoch": 0.2479692176143651,
      "grad_norm": 8.048868179321289,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.4633,
      "step": 145
    },
    {
      "epoch": 0.2496793501496366,
      "grad_norm": 7.92402982711792,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.3619,
      "step": 146
    },
    {
      "epoch": 0.25138948268490807,
      "grad_norm": 7.74806547164917,
      "learning_rate": 1.800157297483417e-05,
      "loss": 5.1205,
      "step": 147
    },
    {
      "epoch": 0.25309961522017954,
      "grad_norm": 8.040278434753418,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 5.6218,
      "step": 148
    },
    {
      "epoch": 0.25480974775545107,
      "grad_norm": 8.707082748413086,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.0942,
      "step": 149
    },
    {
      "epoch": 0.25651988029072254,
      "grad_norm": 11.835229873657227,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 6.0573,
      "step": 150
    },
    {
      "epoch": 0.25651988029072254,
      "eval_loss": 1.2828744649887085,
      "eval_runtime": 71.1991,
      "eval_samples_per_second": 13.834,
      "eval_steps_per_second": 3.469,
      "step": 150
    },
    {
      "epoch": 0.258230012825994,
      "grad_norm": 4.506284236907959,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.7777,
      "step": 151
    },
    {
      "epoch": 0.2599401453612655,
      "grad_norm": 5.299330234527588,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.2758,
      "step": 152
    },
    {
      "epoch": 0.261650277896537,
      "grad_norm": 5.438675403594971,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.649,
      "step": 153
    },
    {
      "epoch": 0.26336041043180847,
      "grad_norm": 6.68652868270874,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 4.8162,
      "step": 154
    },
    {
      "epoch": 0.26507054296707994,
      "grad_norm": 6.45853853225708,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.8465,
      "step": 155
    },
    {
      "epoch": 0.2667806755023514,
      "grad_norm": 6.240671634674072,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.0581,
      "step": 156
    },
    {
      "epoch": 0.26849080803762293,
      "grad_norm": 6.3597331047058105,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.442,
      "step": 157
    },
    {
      "epoch": 0.2702009405728944,
      "grad_norm": 5.745226860046387,
      "learning_rate": 1.157994445715706e-05,
      "loss": 4.8162,
      "step": 158
    },
    {
      "epoch": 0.27191107310816587,
      "grad_norm": 5.929082870483398,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.2591,
      "step": 159
    },
    {
      "epoch": 0.2736212056434374,
      "grad_norm": 5.630734920501709,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.033,
      "step": 160
    },
    {
      "epoch": 0.27533133817870886,
      "grad_norm": 5.56561279296875,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.8257,
      "step": 161
    },
    {
      "epoch": 0.27704147071398033,
      "grad_norm": 5.408742904663086,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.6649,
      "step": 162
    },
    {
      "epoch": 0.2787516032492518,
      "grad_norm": 5.555550575256348,
      "learning_rate": 9.068759265665384e-06,
      "loss": 4.7131,
      "step": 163
    },
    {
      "epoch": 0.2804617357845233,
      "grad_norm": 5.2834014892578125,
      "learning_rate": 8.599558442598998e-06,
      "loss": 4.8631,
      "step": 164
    },
    {
      "epoch": 0.2821718683197948,
      "grad_norm": 5.39652156829834,
      "learning_rate": 8.141676086873572e-06,
      "loss": 4.8836,
      "step": 165
    },
    {
      "epoch": 0.28388200085506626,
      "grad_norm": 5.338542938232422,
      "learning_rate": 7.695237378953223e-06,
      "loss": 4.8909,
      "step": 166
    },
    {
      "epoch": 0.28559213339033773,
      "grad_norm": 5.798850059509277,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.0611,
      "step": 167
    },
    {
      "epoch": 0.28730226592560926,
      "grad_norm": 5.571052074432373,
      "learning_rate": 6.837175952121306e-06,
      "loss": 4.8516,
      "step": 168
    },
    {
      "epoch": 0.2890123984608807,
      "grad_norm": 5.58050537109375,
      "learning_rate": 6.425787818636131e-06,
      "loss": 4.6637,
      "step": 169
    },
    {
      "epoch": 0.2907225309961522,
      "grad_norm": 5.560103893280029,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.8089,
      "step": 170
    },
    {
      "epoch": 0.29243266353142366,
      "grad_norm": 5.654231071472168,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.0751,
      "step": 171
    },
    {
      "epoch": 0.2941427960666952,
      "grad_norm": 5.748037338256836,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.0826,
      "step": 172
    },
    {
      "epoch": 0.29585292860196666,
      "grad_norm": 5.714996814727783,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.8812,
      "step": 173
    },
    {
      "epoch": 0.2975630611372381,
      "grad_norm": 5.568888187408447,
      "learning_rate": 4.549673247541875e-06,
      "loss": 4.9249,
      "step": 174
    },
    {
      "epoch": 0.2992731936725096,
      "grad_norm": 5.951992034912109,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.95,
      "step": 175
    },
    {
      "epoch": 0.3009833262077811,
      "grad_norm": 6.12166690826416,
      "learning_rate": 3.885512251130763e-06,
      "loss": 4.8529,
      "step": 176
    },
    {
      "epoch": 0.3026934587430526,
      "grad_norm": 6.451945781707764,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.3523,
      "step": 177
    },
    {
      "epoch": 0.30440359127832406,
      "grad_norm": 6.277839183807373,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.2762,
      "step": 178
    },
    {
      "epoch": 0.30611372381359553,
      "grad_norm": 6.195245265960693,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 4.9359,
      "step": 179
    },
    {
      "epoch": 0.30782385634886705,
      "grad_norm": 5.98644495010376,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.1152,
      "step": 180
    },
    {
      "epoch": 0.3095339888841385,
      "grad_norm": 6.431697845458984,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.1676,
      "step": 181
    },
    {
      "epoch": 0.31124412141941,
      "grad_norm": 5.98140811920166,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 4.9523,
      "step": 182
    },
    {
      "epoch": 0.31295425395468146,
      "grad_norm": 6.290094375610352,
      "learning_rate": 1.962316193157593e-06,
      "loss": 4.9884,
      "step": 183
    },
    {
      "epoch": 0.314664386489953,
      "grad_norm": 6.388920783996582,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.0668,
      "step": 184
    },
    {
      "epoch": 0.31637451902522445,
      "grad_norm": 6.373295307159424,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.0978,
      "step": 185
    },
    {
      "epoch": 0.3180846515604959,
      "grad_norm": 6.373231410980225,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.2877,
      "step": 186
    },
    {
      "epoch": 0.31979478409576745,
      "grad_norm": 6.621831893920898,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.1399,
      "step": 187
    },
    {
      "epoch": 0.3215049166310389,
      "grad_norm": 7.05682897567749,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.4095,
      "step": 188
    },
    {
      "epoch": 0.3232150491663104,
      "grad_norm": 6.611572265625,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.2426,
      "step": 189
    },
    {
      "epoch": 0.32492518170158186,
      "grad_norm": 7.512360572814941,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.7162,
      "step": 190
    },
    {
      "epoch": 0.3266353142368534,
      "grad_norm": 6.702224254608154,
      "learning_rate": 5.526064699265753e-07,
      "loss": 5.1063,
      "step": 191
    },
    {
      "epoch": 0.32834544677212485,
      "grad_norm": 7.078572750091553,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.3977,
      "step": 192
    },
    {
      "epoch": 0.3300555793073963,
      "grad_norm": 7.9836955070495605,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.4685,
      "step": 193
    },
    {
      "epoch": 0.3317657118426678,
      "grad_norm": 7.350215911865234,
      "learning_rate": 2.458548727494292e-07,
      "loss": 5.2704,
      "step": 194
    },
    {
      "epoch": 0.3334758443779393,
      "grad_norm": 8.438230514526367,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.3578,
      "step": 195
    },
    {
      "epoch": 0.3351859769132108,
      "grad_norm": 7.477552890777588,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 5.3017,
      "step": 196
    },
    {
      "epoch": 0.33689610944848225,
      "grad_norm": 7.963863372802734,
      "learning_rate": 6.150154258476315e-08,
      "loss": 5.5124,
      "step": 197
    },
    {
      "epoch": 0.3386062419837537,
      "grad_norm": 9.603384017944336,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 6.0447,
      "step": 198
    },
    {
      "epoch": 0.34031637451902524,
      "grad_norm": 9.086202621459961,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.057,
      "step": 199
    },
    {
      "epoch": 0.3420265070542967,
      "grad_norm": 12.168228149414062,
      "learning_rate": 0.0,
      "loss": 6.3207,
      "step": 200
    },
    {
      "epoch": 0.3420265070542967,
      "eval_loss": 1.2364895343780518,
      "eval_runtime": 71.0504,
      "eval_samples_per_second": 13.863,
      "eval_steps_per_second": 3.476,
      "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
}