prxy5607's picture
Training in progress, step 200, checkpoint
cebede3 verified
Raw
History Blame Contribute Delete
36.6 kB
{
"best_metric": 0.3544212281703949,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.32653061224489793,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0016326530612244899,
"grad_norm": 0.3510968089103699,
"learning_rate": 1e-05,
"loss": 0.409,
"step": 1
},
{
"epoch": 0.0016326530612244899,
"eval_loss": 0.7382236123085022,
"eval_runtime": 79.8162,
"eval_samples_per_second": 12.93,
"eval_steps_per_second": 3.232,
"step": 1
},
{
"epoch": 0.0032653061224489797,
"grad_norm": 0.4350750744342804,
"learning_rate": 2e-05,
"loss": 0.4735,
"step": 2
},
{
"epoch": 0.004897959183673469,
"grad_norm": 0.45716866850852966,
"learning_rate": 3e-05,
"loss": 0.5291,
"step": 3
},
{
"epoch": 0.006530612244897959,
"grad_norm": 0.4768833518028259,
"learning_rate": 4e-05,
"loss": 0.477,
"step": 4
},
{
"epoch": 0.00816326530612245,
"grad_norm": 0.45378586649894714,
"learning_rate": 5e-05,
"loss": 0.4724,
"step": 5
},
{
"epoch": 0.009795918367346938,
"grad_norm": 0.4014195203781128,
"learning_rate": 6e-05,
"loss": 0.4745,
"step": 6
},
{
"epoch": 0.011428571428571429,
"grad_norm": 0.4592766761779785,
"learning_rate": 7e-05,
"loss": 0.441,
"step": 7
},
{
"epoch": 0.013061224489795919,
"grad_norm": 0.5126383304595947,
"learning_rate": 8e-05,
"loss": 0.4637,
"step": 8
},
{
"epoch": 0.014693877551020407,
"grad_norm": 0.4321953058242798,
"learning_rate": 9e-05,
"loss": 0.383,
"step": 9
},
{
"epoch": 0.0163265306122449,
"grad_norm": 0.37420108914375305,
"learning_rate": 0.0001,
"loss": 0.3889,
"step": 10
},
{
"epoch": 0.017959183673469388,
"grad_norm": 0.37791600823402405,
"learning_rate": 9.999316524962345e-05,
"loss": 0.4265,
"step": 11
},
{
"epoch": 0.019591836734693877,
"grad_norm": 0.37340274453163147,
"learning_rate": 9.997266286704631e-05,
"loss": 0.369,
"step": 12
},
{
"epoch": 0.02122448979591837,
"grad_norm": 0.3388529419898987,
"learning_rate": 9.993849845741524e-05,
"loss": 0.3558,
"step": 13
},
{
"epoch": 0.022857142857142857,
"grad_norm": 0.3248995244503021,
"learning_rate": 9.989068136093873e-05,
"loss": 0.3386,
"step": 14
},
{
"epoch": 0.024489795918367346,
"grad_norm": 0.34298235177993774,
"learning_rate": 9.98292246503335e-05,
"loss": 0.3501,
"step": 15
},
{
"epoch": 0.026122448979591838,
"grad_norm": 0.3550427258014679,
"learning_rate": 9.975414512725057e-05,
"loss": 0.3972,
"step": 16
},
{
"epoch": 0.027755102040816326,
"grad_norm": 0.3637966811656952,
"learning_rate": 9.966546331768191e-05,
"loss": 0.4402,
"step": 17
},
{
"epoch": 0.029387755102040815,
"grad_norm": 0.3442937135696411,
"learning_rate": 9.956320346634876e-05,
"loss": 0.4042,
"step": 18
},
{
"epoch": 0.031020408163265307,
"grad_norm": 0.33977586030960083,
"learning_rate": 9.944739353007344e-05,
"loss": 0.3521,
"step": 19
},
{
"epoch": 0.0326530612244898,
"grad_norm": 0.3380832374095917,
"learning_rate": 9.931806517013612e-05,
"loss": 0.3651,
"step": 20
},
{
"epoch": 0.03428571428571429,
"grad_norm": 0.4377959370613098,
"learning_rate": 9.917525374361912e-05,
"loss": 0.3126,
"step": 21
},
{
"epoch": 0.035918367346938776,
"grad_norm": 0.3442164957523346,
"learning_rate": 9.901899829374047e-05,
"loss": 0.3224,
"step": 22
},
{
"epoch": 0.037551020408163265,
"grad_norm": 0.3431764245033264,
"learning_rate": 9.884934153917997e-05,
"loss": 0.3145,
"step": 23
},
{
"epoch": 0.03918367346938775,
"grad_norm": 0.37815147638320923,
"learning_rate": 9.86663298624003e-05,
"loss": 0.3353,
"step": 24
},
{
"epoch": 0.04081632653061224,
"grad_norm": 0.2946655750274658,
"learning_rate": 9.847001329696653e-05,
"loss": 0.2841,
"step": 25
},
{
"epoch": 0.04244897959183674,
"grad_norm": 0.3283737301826477,
"learning_rate": 9.826044551386744e-05,
"loss": 0.334,
"step": 26
},
{
"epoch": 0.044081632653061226,
"grad_norm": 0.3182757794857025,
"learning_rate": 9.803768380684242e-05,
"loss": 0.3216,
"step": 27
},
{
"epoch": 0.045714285714285714,
"grad_norm": 0.371548593044281,
"learning_rate": 9.780178907671789e-05,
"loss": 0.3709,
"step": 28
},
{
"epoch": 0.0473469387755102,
"grad_norm": 0.3637164533138275,
"learning_rate": 9.755282581475769e-05,
"loss": 0.4134,
"step": 29
},
{
"epoch": 0.04897959183673469,
"grad_norm": 0.36473169922828674,
"learning_rate": 9.729086208503174e-05,
"loss": 0.4028,
"step": 30
},
{
"epoch": 0.05061224489795919,
"grad_norm": 0.36588460206985474,
"learning_rate": 9.701596950580806e-05,
"loss": 0.3777,
"step": 31
},
{
"epoch": 0.052244897959183675,
"grad_norm": 0.37372252345085144,
"learning_rate": 9.672822322997305e-05,
"loss": 0.364,
"step": 32
},
{
"epoch": 0.053877551020408164,
"grad_norm": 0.3478218615055084,
"learning_rate": 9.642770192448536e-05,
"loss": 0.3483,
"step": 33
},
{
"epoch": 0.05551020408163265,
"grad_norm": 0.39115580916404724,
"learning_rate": 9.611448774886924e-05,
"loss": 0.3958,
"step": 34
},
{
"epoch": 0.05714285714285714,
"grad_norm": 0.40638843178749084,
"learning_rate": 9.578866633275288e-05,
"loss": 0.4227,
"step": 35
},
{
"epoch": 0.05877551020408163,
"grad_norm": 0.38316670060157776,
"learning_rate": 9.545032675245813e-05,
"loss": 0.4209,
"step": 36
},
{
"epoch": 0.060408163265306125,
"grad_norm": 0.38507622480392456,
"learning_rate": 9.509956150664796e-05,
"loss": 0.4024,
"step": 37
},
{
"epoch": 0.062040816326530614,
"grad_norm": 0.39556100964546204,
"learning_rate": 9.473646649103818e-05,
"loss": 0.4229,
"step": 38
},
{
"epoch": 0.0636734693877551,
"grad_norm": 0.4122062921524048,
"learning_rate": 9.43611409721806e-05,
"loss": 0.4432,
"step": 39
},
{
"epoch": 0.0653061224489796,
"grad_norm": 0.4069179892539978,
"learning_rate": 9.397368756032445e-05,
"loss": 0.3953,
"step": 40
},
{
"epoch": 0.06693877551020408,
"grad_norm": 0.4790738821029663,
"learning_rate": 9.357421218136386e-05,
"loss": 0.534,
"step": 41
},
{
"epoch": 0.06857142857142857,
"grad_norm": 0.4197099506855011,
"learning_rate": 9.316282404787871e-05,
"loss": 0.4043,
"step": 42
},
{
"epoch": 0.07020408163265306,
"grad_norm": 0.5003912448883057,
"learning_rate": 9.273963562927695e-05,
"loss": 0.4769,
"step": 43
},
{
"epoch": 0.07183673469387755,
"grad_norm": 0.5060177445411682,
"learning_rate": 9.230476262104677e-05,
"loss": 0.4443,
"step": 44
},
{
"epoch": 0.07346938775510205,
"grad_norm": 0.5015014410018921,
"learning_rate": 9.185832391312644e-05,
"loss": 0.4429,
"step": 45
},
{
"epoch": 0.07510204081632653,
"grad_norm": 0.5699446797370911,
"learning_rate": 9.140044155740101e-05,
"loss": 0.4227,
"step": 46
},
{
"epoch": 0.07673469387755102,
"grad_norm": 0.5878002047538757,
"learning_rate": 9.093124073433463e-05,
"loss": 0.4662,
"step": 47
},
{
"epoch": 0.0783673469387755,
"grad_norm": 0.6678152680397034,
"learning_rate": 9.045084971874738e-05,
"loss": 0.5286,
"step": 48
},
{
"epoch": 0.08,
"grad_norm": 0.6475111842155457,
"learning_rate": 8.995939984474624e-05,
"loss": 0.4761,
"step": 49
},
{
"epoch": 0.08163265306122448,
"grad_norm": 1.006797432899475,
"learning_rate": 8.945702546981969e-05,
"loss": 0.5642,
"step": 50
},
{
"epoch": 0.08163265306122448,
"eval_loss": 0.42069658637046814,
"eval_runtime": 80.3874,
"eval_samples_per_second": 12.838,
"eval_steps_per_second": 3.209,
"step": 50
},
{
"epoch": 0.08326530612244898,
"grad_norm": 0.4333561360836029,
"learning_rate": 8.894386393810563e-05,
"loss": 0.4622,
"step": 51
},
{
"epoch": 0.08489795918367347,
"grad_norm": 0.37180453538894653,
"learning_rate": 8.842005554284296e-05,
"loss": 0.403,
"step": 52
},
{
"epoch": 0.08653061224489796,
"grad_norm": 0.33497190475463867,
"learning_rate": 8.788574348801675e-05,
"loss": 0.4218,
"step": 53
},
{
"epoch": 0.08816326530612245,
"grad_norm": 0.27931031584739685,
"learning_rate": 8.73410738492077e-05,
"loss": 0.329,
"step": 54
},
{
"epoch": 0.08979591836734693,
"grad_norm": 0.2886022925376892,
"learning_rate": 8.678619553365659e-05,
"loss": 0.4065,
"step": 55
},
{
"epoch": 0.09142857142857143,
"grad_norm": 0.2585863769054413,
"learning_rate": 8.622126023955446e-05,
"loss": 0.3281,
"step": 56
},
{
"epoch": 0.09306122448979592,
"grad_norm": 0.2792675793170929,
"learning_rate": 8.564642241456986e-05,
"loss": 0.3653,
"step": 57
},
{
"epoch": 0.0946938775510204,
"grad_norm": 0.2525157928466797,
"learning_rate": 8.506183921362443e-05,
"loss": 0.3582,
"step": 58
},
{
"epoch": 0.0963265306122449,
"grad_norm": 0.2777102291584015,
"learning_rate": 8.44676704559283e-05,
"loss": 0.3467,
"step": 59
},
{
"epoch": 0.09795918367346938,
"grad_norm": 0.2691451907157898,
"learning_rate": 8.386407858128706e-05,
"loss": 0.3209,
"step": 60
},
{
"epoch": 0.09959183673469388,
"grad_norm": 0.27784255146980286,
"learning_rate": 8.32512286056924e-05,
"loss": 0.3179,
"step": 61
},
{
"epoch": 0.10122448979591837,
"grad_norm": 0.2905835211277008,
"learning_rate": 8.262928807620843e-05,
"loss": 0.3529,
"step": 62
},
{
"epoch": 0.10285714285714286,
"grad_norm": 0.3017609119415283,
"learning_rate": 8.199842702516583e-05,
"loss": 0.3405,
"step": 63
},
{
"epoch": 0.10448979591836735,
"grad_norm": 0.25602033734321594,
"learning_rate": 8.135881792367686e-05,
"loss": 0.2581,
"step": 64
},
{
"epoch": 0.10612244897959183,
"grad_norm": 0.29537057876586914,
"learning_rate": 8.07106356344834e-05,
"loss": 0.2792,
"step": 65
},
{
"epoch": 0.10775510204081633,
"grad_norm": 0.26765191555023193,
"learning_rate": 8.005405736415126e-05,
"loss": 0.2807,
"step": 66
},
{
"epoch": 0.10938775510204081,
"grad_norm": 0.3034214377403259,
"learning_rate": 7.938926261462366e-05,
"loss": 0.349,
"step": 67
},
{
"epoch": 0.1110204081632653,
"grad_norm": 0.3050192594528198,
"learning_rate": 7.871643313414718e-05,
"loss": 0.2992,
"step": 68
},
{
"epoch": 0.1126530612244898,
"grad_norm": 0.3208230137825012,
"learning_rate": 7.803575286758364e-05,
"loss": 0.3312,
"step": 69
},
{
"epoch": 0.11428571428571428,
"grad_norm": 0.35363858938217163,
"learning_rate": 7.734740790612136e-05,
"loss": 0.3699,
"step": 70
},
{
"epoch": 0.11591836734693878,
"grad_norm": 0.3319184184074402,
"learning_rate": 7.66515864363997e-05,
"loss": 0.3166,
"step": 71
},
{
"epoch": 0.11755102040816326,
"grad_norm": 0.2948947548866272,
"learning_rate": 7.594847868906076e-05,
"loss": 0.304,
"step": 72
},
{
"epoch": 0.11918367346938775,
"grad_norm": 0.31342217326164246,
"learning_rate": 7.52382768867422e-05,
"loss": 0.2951,
"step": 73
},
{
"epoch": 0.12081632653061225,
"grad_norm": 0.3001481592655182,
"learning_rate": 7.452117519152542e-05,
"loss": 0.3218,
"step": 74
},
{
"epoch": 0.12244897959183673,
"grad_norm": 0.3259095251560211,
"learning_rate": 7.379736965185368e-05,
"loss": 0.3272,
"step": 75
},
{
"epoch": 0.12408163265306123,
"grad_norm": 0.3134114742279053,
"learning_rate": 7.30670581489344e-05,
"loss": 0.316,
"step": 76
},
{
"epoch": 0.12571428571428572,
"grad_norm": 0.3459464907646179,
"learning_rate": 7.233044034264034e-05,
"loss": 0.3658,
"step": 77
},
{
"epoch": 0.1273469387755102,
"grad_norm": 0.35104259848594666,
"learning_rate": 7.158771761692464e-05,
"loss": 0.3175,
"step": 78
},
{
"epoch": 0.12897959183673469,
"grad_norm": 0.36668387055397034,
"learning_rate": 7.083909302476453e-05,
"loss": 0.3419,
"step": 79
},
{
"epoch": 0.1306122448979592,
"grad_norm": 0.3350326716899872,
"learning_rate": 7.008477123264848e-05,
"loss": 0.3238,
"step": 80
},
{
"epoch": 0.13224489795918368,
"grad_norm": 0.3355366289615631,
"learning_rate": 6.932495846462261e-05,
"loss": 0.3219,
"step": 81
},
{
"epoch": 0.13387755102040816,
"grad_norm": 0.3617432117462158,
"learning_rate": 6.855986244591104e-05,
"loss": 0.3385,
"step": 82
},
{
"epoch": 0.13551020408163264,
"grad_norm": 0.30388301610946655,
"learning_rate": 6.778969234612584e-05,
"loss": 0.2998,
"step": 83
},
{
"epoch": 0.13714285714285715,
"grad_norm": 0.38138654828071594,
"learning_rate": 6.701465872208216e-05,
"loss": 0.349,
"step": 84
},
{
"epoch": 0.13877551020408163,
"grad_norm": 0.3741893470287323,
"learning_rate": 6.623497346023418e-05,
"loss": 0.3715,
"step": 85
},
{
"epoch": 0.1404081632653061,
"grad_norm": 0.38264286518096924,
"learning_rate": 6.545084971874738e-05,
"loss": 0.4331,
"step": 86
},
{
"epoch": 0.14204081632653062,
"grad_norm": 0.3760170042514801,
"learning_rate": 6.466250186922325e-05,
"loss": 0.3296,
"step": 87
},
{
"epoch": 0.1436734693877551,
"grad_norm": 0.40451616048812866,
"learning_rate": 6.387014543809223e-05,
"loss": 0.3631,
"step": 88
},
{
"epoch": 0.14530612244897959,
"grad_norm": 0.3783271014690399,
"learning_rate": 6.307399704769099e-05,
"loss": 0.34,
"step": 89
},
{
"epoch": 0.1469387755102041,
"grad_norm": 0.384853720664978,
"learning_rate": 6.227427435703997e-05,
"loss": 0.4087,
"step": 90
},
{
"epoch": 0.14857142857142858,
"grad_norm": 0.41267094016075134,
"learning_rate": 6.147119600233758e-05,
"loss": 0.4032,
"step": 91
},
{
"epoch": 0.15020408163265306,
"grad_norm": 0.4150213599205017,
"learning_rate": 6.066498153718735e-05,
"loss": 0.4642,
"step": 92
},
{
"epoch": 0.15183673469387754,
"grad_norm": 0.42586472630500793,
"learning_rate": 5.985585137257401e-05,
"loss": 0.3882,
"step": 93
},
{
"epoch": 0.15346938775510205,
"grad_norm": 0.43605440855026245,
"learning_rate": 5.90440267166055e-05,
"loss": 0.4418,
"step": 94
},
{
"epoch": 0.15510204081632653,
"grad_norm": 0.4620981514453888,
"learning_rate": 5.8229729514036705e-05,
"loss": 0.4303,
"step": 95
},
{
"epoch": 0.156734693877551,
"grad_norm": 0.4591740667819977,
"learning_rate": 5.74131823855921e-05,
"loss": 0.4269,
"step": 96
},
{
"epoch": 0.15836734693877552,
"grad_norm": 0.7421184182167053,
"learning_rate": 5.6594608567103456e-05,
"loss": 0.4068,
"step": 97
},
{
"epoch": 0.16,
"grad_norm": 0.5088761448860168,
"learning_rate": 5.577423184847932e-05,
"loss": 0.4195,
"step": 98
},
{
"epoch": 0.16163265306122448,
"grad_norm": 0.55223149061203,
"learning_rate": 5.495227651252315e-05,
"loss": 0.3962,
"step": 99
},
{
"epoch": 0.16326530612244897,
"grad_norm": 0.7844134569168091,
"learning_rate": 5.4128967273616625e-05,
"loss": 0.5604,
"step": 100
},
{
"epoch": 0.16326530612244897,
"eval_loss": 0.3941538631916046,
"eval_runtime": 80.4222,
"eval_samples_per_second": 12.832,
"eval_steps_per_second": 3.208,
"step": 100
},
{
"epoch": 0.16489795918367348,
"grad_norm": 0.32274824380874634,
"learning_rate": 5.330452921628497e-05,
"loss": 0.389,
"step": 101
},
{
"epoch": 0.16653061224489796,
"grad_norm": 0.31748849153518677,
"learning_rate": 5.247918773366112e-05,
"loss": 0.389,
"step": 102
},
{
"epoch": 0.16816326530612244,
"grad_norm": 0.34383273124694824,
"learning_rate": 5.165316846586541e-05,
"loss": 0.387,
"step": 103
},
{
"epoch": 0.16979591836734695,
"grad_norm": 0.2862667441368103,
"learning_rate": 5.0826697238317935e-05,
"loss": 0.3286,
"step": 104
},
{
"epoch": 0.17142857142857143,
"grad_norm": 0.2569122314453125,
"learning_rate": 5e-05,
"loss": 0.3253,
"step": 105
},
{
"epoch": 0.1730612244897959,
"grad_norm": 0.2409268468618393,
"learning_rate": 4.917330276168208e-05,
"loss": 0.3357,
"step": 106
},
{
"epoch": 0.17469387755102042,
"grad_norm": 0.24116161465644836,
"learning_rate": 4.834683153413459e-05,
"loss": 0.3578,
"step": 107
},
{
"epoch": 0.1763265306122449,
"grad_norm": 0.23487241566181183,
"learning_rate": 4.7520812266338885e-05,
"loss": 0.3245,
"step": 108
},
{
"epoch": 0.17795918367346938,
"grad_norm": 0.25777649879455566,
"learning_rate": 4.669547078371504e-05,
"loss": 0.3144,
"step": 109
},
{
"epoch": 0.17959183673469387,
"grad_norm": 0.29614609479904175,
"learning_rate": 4.5871032726383386e-05,
"loss": 0.358,
"step": 110
},
{
"epoch": 0.18122448979591838,
"grad_norm": 0.2606280446052551,
"learning_rate": 4.504772348747687e-05,
"loss": 0.327,
"step": 111
},
{
"epoch": 0.18285714285714286,
"grad_norm": 0.2707988917827606,
"learning_rate": 4.4225768151520694e-05,
"loss": 0.3376,
"step": 112
},
{
"epoch": 0.18448979591836734,
"grad_norm": 0.27944138646125793,
"learning_rate": 4.3405391432896555e-05,
"loss": 0.3615,
"step": 113
},
{
"epoch": 0.18612244897959185,
"grad_norm": 0.2461269646883011,
"learning_rate": 4.2586817614407895e-05,
"loss": 0.2699,
"step": 114
},
{
"epoch": 0.18775510204081633,
"grad_norm": 0.261223167181015,
"learning_rate": 4.17702704859633e-05,
"loss": 0.2932,
"step": 115
},
{
"epoch": 0.1893877551020408,
"grad_norm": 0.2908478081226349,
"learning_rate": 4.095597328339452e-05,
"loss": 0.3649,
"step": 116
},
{
"epoch": 0.1910204081632653,
"grad_norm": 0.26536840200424194,
"learning_rate": 4.0144148627425993e-05,
"loss": 0.3482,
"step": 117
},
{
"epoch": 0.1926530612244898,
"grad_norm": 0.2866617441177368,
"learning_rate": 3.933501846281267e-05,
"loss": 0.3617,
"step": 118
},
{
"epoch": 0.19428571428571428,
"grad_norm": 0.32360154390335083,
"learning_rate": 3.852880399766243e-05,
"loss": 0.2958,
"step": 119
},
{
"epoch": 0.19591836734693877,
"grad_norm": 0.2834181785583496,
"learning_rate": 3.772572564296005e-05,
"loss": 0.333,
"step": 120
},
{
"epoch": 0.19755102040816327,
"grad_norm": 0.272098183631897,
"learning_rate": 3.6926002952309016e-05,
"loss": 0.3404,
"step": 121
},
{
"epoch": 0.19918367346938776,
"grad_norm": 0.2827979028224945,
"learning_rate": 3.612985456190778e-05,
"loss": 0.2693,
"step": 122
},
{
"epoch": 0.20081632653061224,
"grad_norm": 0.3093973398208618,
"learning_rate": 3.533749813077677e-05,
"loss": 0.3784,
"step": 123
},
{
"epoch": 0.20244897959183675,
"grad_norm": 0.30716672539711,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.3589,
"step": 124
},
{
"epoch": 0.20408163265306123,
"grad_norm": 0.27633604407310486,
"learning_rate": 3.3765026539765834e-05,
"loss": 0.2691,
"step": 125
},
{
"epoch": 0.2057142857142857,
"grad_norm": 0.28618523478507996,
"learning_rate": 3.298534127791785e-05,
"loss": 0.2855,
"step": 126
},
{
"epoch": 0.2073469387755102,
"grad_norm": 0.3070104420185089,
"learning_rate": 3.221030765387417e-05,
"loss": 0.3006,
"step": 127
},
{
"epoch": 0.2089795918367347,
"grad_norm": 0.3096109926700592,
"learning_rate": 3.144013755408895e-05,
"loss": 0.3242,
"step": 128
},
{
"epoch": 0.21061224489795918,
"grad_norm": 0.28949832916259766,
"learning_rate": 3.0675041535377405e-05,
"loss": 0.2732,
"step": 129
},
{
"epoch": 0.21224489795918366,
"grad_norm": 0.36503884196281433,
"learning_rate": 2.991522876735154e-05,
"loss": 0.3519,
"step": 130
},
{
"epoch": 0.21387755102040817,
"grad_norm": 0.32110676169395447,
"learning_rate": 2.916090697523549e-05,
"loss": 0.3283,
"step": 131
},
{
"epoch": 0.21551020408163266,
"grad_norm": 0.3160181939601898,
"learning_rate": 2.8412282383075363e-05,
"loss": 0.3514,
"step": 132
},
{
"epoch": 0.21714285714285714,
"grad_norm": 0.3308194577693939,
"learning_rate": 2.766955965735968e-05,
"loss": 0.3177,
"step": 133
},
{
"epoch": 0.21877551020408162,
"grad_norm": 0.33339327573776245,
"learning_rate": 2.693294185106562e-05,
"loss": 0.358,
"step": 134
},
{
"epoch": 0.22040816326530613,
"grad_norm": 0.30465736985206604,
"learning_rate": 2.6202630348146324e-05,
"loss": 0.2781,
"step": 135
},
{
"epoch": 0.2220408163265306,
"grad_norm": 0.3235703408718109,
"learning_rate": 2.547882480847461e-05,
"loss": 0.3622,
"step": 136
},
{
"epoch": 0.2236734693877551,
"grad_norm": 0.33747023344039917,
"learning_rate": 2.476172311325783e-05,
"loss": 0.292,
"step": 137
},
{
"epoch": 0.2253061224489796,
"grad_norm": 0.34918656945228577,
"learning_rate": 2.405152131093926e-05,
"loss": 0.3606,
"step": 138
},
{
"epoch": 0.22693877551020408,
"grad_norm": 0.3670065701007843,
"learning_rate": 2.3348413563600325e-05,
"loss": 0.3508,
"step": 139
},
{
"epoch": 0.22857142857142856,
"grad_norm": 0.3757942020893097,
"learning_rate": 2.2652592093878666e-05,
"loss": 0.3681,
"step": 140
},
{
"epoch": 0.23020408163265307,
"grad_norm": 0.3256301283836365,
"learning_rate": 2.196424713241637e-05,
"loss": 0.3008,
"step": 141
},
{
"epoch": 0.23183673469387756,
"grad_norm": 0.4201648235321045,
"learning_rate": 2.128356686585282e-05,
"loss": 0.4037,
"step": 142
},
{
"epoch": 0.23346938775510204,
"grad_norm": 0.3878253698348999,
"learning_rate": 2.061073738537635e-05,
"loss": 0.3978,
"step": 143
},
{
"epoch": 0.23510204081632652,
"grad_norm": 0.4525003731250763,
"learning_rate": 1.9945942635848748e-05,
"loss": 0.4136,
"step": 144
},
{
"epoch": 0.23673469387755103,
"grad_norm": 0.4201924800872803,
"learning_rate": 1.928936436551661e-05,
"loss": 0.3684,
"step": 145
},
{
"epoch": 0.2383673469387755,
"grad_norm": 0.4143880009651184,
"learning_rate": 1.8641182076323148e-05,
"loss": 0.3675,
"step": 146
},
{
"epoch": 0.24,
"grad_norm": 0.3984711766242981,
"learning_rate": 1.800157297483417e-05,
"loss": 0.3324,
"step": 147
},
{
"epoch": 0.2416326530612245,
"grad_norm": 0.5180820822715759,
"learning_rate": 1.7370711923791567e-05,
"loss": 0.4623,
"step": 148
},
{
"epoch": 0.24326530612244898,
"grad_norm": 0.5790478587150574,
"learning_rate": 1.6748771394307585e-05,
"loss": 0.4403,
"step": 149
},
{
"epoch": 0.24489795918367346,
"grad_norm": 0.6000601053237915,
"learning_rate": 1.6135921418712956e-05,
"loss": 0.4514,
"step": 150
},
{
"epoch": 0.24489795918367346,
"eval_loss": 0.3595074415206909,
"eval_runtime": 80.45,
"eval_samples_per_second": 12.828,
"eval_steps_per_second": 3.207,
"step": 150
},
{
"epoch": 0.24653061224489795,
"grad_norm": 0.15959057211875916,
"learning_rate": 1.553232954407171e-05,
"loss": 0.2581,
"step": 151
},
{
"epoch": 0.24816326530612245,
"grad_norm": 0.19398391246795654,
"learning_rate": 1.4938160786375572e-05,
"loss": 0.3288,
"step": 152
},
{
"epoch": 0.24979591836734694,
"grad_norm": 0.2184728980064392,
"learning_rate": 1.435357758543015e-05,
"loss": 0.3689,
"step": 153
},
{
"epoch": 0.25142857142857145,
"grad_norm": 0.2182149440050125,
"learning_rate": 1.3778739760445552e-05,
"loss": 0.3051,
"step": 154
},
{
"epoch": 0.2530612244897959,
"grad_norm": 0.23768407106399536,
"learning_rate": 1.3213804466343421e-05,
"loss": 0.3503,
"step": 155
},
{
"epoch": 0.2546938775510204,
"grad_norm": 0.2353631854057312,
"learning_rate": 1.2658926150792322e-05,
"loss": 0.3175,
"step": 156
},
{
"epoch": 0.2563265306122449,
"grad_norm": 0.2338859736919403,
"learning_rate": 1.2114256511983274e-05,
"loss": 0.3637,
"step": 157
},
{
"epoch": 0.25795918367346937,
"grad_norm": 0.24875757098197937,
"learning_rate": 1.157994445715706e-05,
"loss": 0.3764,
"step": 158
},
{
"epoch": 0.2595918367346939,
"grad_norm": 0.2550694942474365,
"learning_rate": 1.1056136061894384e-05,
"loss": 0.3226,
"step": 159
},
{
"epoch": 0.2612244897959184,
"grad_norm": 0.2498476654291153,
"learning_rate": 1.0542974530180327e-05,
"loss": 0.3054,
"step": 160
},
{
"epoch": 0.26285714285714284,
"grad_norm": 0.25995540618896484,
"learning_rate": 1.0040600155253765e-05,
"loss": 0.3153,
"step": 161
},
{
"epoch": 0.26448979591836735,
"grad_norm": 0.28035736083984375,
"learning_rate": 9.549150281252633e-06,
"loss": 0.335,
"step": 162
},
{
"epoch": 0.26612244897959186,
"grad_norm": 0.2477700114250183,
"learning_rate": 9.068759265665384e-06,
"loss": 0.3191,
"step": 163
},
{
"epoch": 0.2677551020408163,
"grad_norm": 0.2767393887042999,
"learning_rate": 8.599558442598998e-06,
"loss": 0.332,
"step": 164
},
{
"epoch": 0.2693877551020408,
"grad_norm": 0.26257529854774475,
"learning_rate": 8.141676086873572e-06,
"loss": 0.3275,
"step": 165
},
{
"epoch": 0.2710204081632653,
"grad_norm": 0.2755727767944336,
"learning_rate": 7.695237378953223e-06,
"loss": 0.3532,
"step": 166
},
{
"epoch": 0.2726530612244898,
"grad_norm": 0.28371289372444153,
"learning_rate": 7.260364370723044e-06,
"loss": 0.3107,
"step": 167
},
{
"epoch": 0.2742857142857143,
"grad_norm": 0.2793286144733429,
"learning_rate": 6.837175952121306e-06,
"loss": 0.3291,
"step": 168
},
{
"epoch": 0.27591836734693875,
"grad_norm": 0.271729975938797,
"learning_rate": 6.425787818636131e-06,
"loss": 0.301,
"step": 169
},
{
"epoch": 0.27755102040816326,
"grad_norm": 0.26402413845062256,
"learning_rate": 6.026312439675552e-06,
"loss": 0.2728,
"step": 170
},
{
"epoch": 0.27918367346938777,
"grad_norm": 0.26793578267097473,
"learning_rate": 5.6388590278194096e-06,
"loss": 0.3357,
"step": 171
},
{
"epoch": 0.2808163265306122,
"grad_norm": 0.2868030071258545,
"learning_rate": 5.263533508961827e-06,
"loss": 0.3239,
"step": 172
},
{
"epoch": 0.28244897959183674,
"grad_norm": 0.27149003744125366,
"learning_rate": 4.900438493352055e-06,
"loss": 0.2743,
"step": 173
},
{
"epoch": 0.28408163265306124,
"grad_norm": 0.2867920696735382,
"learning_rate": 4.549673247541875e-06,
"loss": 0.2889,
"step": 174
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.2825607657432556,
"learning_rate": 4.2113336672471245e-06,
"loss": 0.3089,
"step": 175
},
{
"epoch": 0.2873469387755102,
"grad_norm": 0.2784520387649536,
"learning_rate": 3.885512251130763e-06,
"loss": 0.2768,
"step": 176
},
{
"epoch": 0.2889795918367347,
"grad_norm": 0.2805027365684509,
"learning_rate": 3.5722980755146517e-06,
"loss": 0.2791,
"step": 177
},
{
"epoch": 0.29061224489795917,
"grad_norm": 0.2990702688694,
"learning_rate": 3.271776770026963e-06,
"loss": 0.2974,
"step": 178
},
{
"epoch": 0.2922448979591837,
"grad_norm": 0.28951507806777954,
"learning_rate": 2.9840304941919415e-06,
"loss": 0.2985,
"step": 179
},
{
"epoch": 0.2938775510204082,
"grad_norm": 0.30154889822006226,
"learning_rate": 2.7091379149682685e-06,
"loss": 0.3478,
"step": 180
},
{
"epoch": 0.29551020408163264,
"grad_norm": 0.31735241413116455,
"learning_rate": 2.4471741852423237e-06,
"loss": 0.3384,
"step": 181
},
{
"epoch": 0.29714285714285715,
"grad_norm": 0.31430307030677795,
"learning_rate": 2.1982109232821178e-06,
"loss": 0.3304,
"step": 182
},
{
"epoch": 0.2987755102040816,
"grad_norm": 0.3241609036922455,
"learning_rate": 1.962316193157593e-06,
"loss": 0.361,
"step": 183
},
{
"epoch": 0.3004081632653061,
"grad_norm": 0.3565089702606201,
"learning_rate": 1.7395544861325718e-06,
"loss": 0.4251,
"step": 184
},
{
"epoch": 0.3020408163265306,
"grad_norm": 0.3391473889350891,
"learning_rate": 1.5299867030334814e-06,
"loss": 0.3795,
"step": 185
},
{
"epoch": 0.3036734693877551,
"grad_norm": 0.33540958166122437,
"learning_rate": 1.333670137599713e-06,
"loss": 0.3129,
"step": 186
},
{
"epoch": 0.3053061224489796,
"grad_norm": 0.3401593863964081,
"learning_rate": 1.1506584608200367e-06,
"loss": 0.3505,
"step": 187
},
{
"epoch": 0.3069387755102041,
"grad_norm": 0.33632588386535645,
"learning_rate": 9.810017062595322e-07,
"loss": 0.3439,
"step": 188
},
{
"epoch": 0.30857142857142855,
"grad_norm": 0.3870463967323303,
"learning_rate": 8.247462563808817e-07,
"loss": 0.4092,
"step": 189
},
{
"epoch": 0.31020408163265306,
"grad_norm": 0.3882526159286499,
"learning_rate": 6.819348298638839e-07,
"loss": 0.3325,
"step": 190
},
{
"epoch": 0.31183673469387757,
"grad_norm": 0.40052559971809387,
"learning_rate": 5.526064699265753e-07,
"loss": 0.4308,
"step": 191
},
{
"epoch": 0.313469387755102,
"grad_norm": 0.4461832344532013,
"learning_rate": 4.367965336512403e-07,
"loss": 0.4173,
"step": 192
},
{
"epoch": 0.31510204081632653,
"grad_norm": 0.3978295922279358,
"learning_rate": 3.3453668231809286e-07,
"loss": 0.3777,
"step": 193
},
{
"epoch": 0.31673469387755104,
"grad_norm": 0.4485456645488739,
"learning_rate": 2.458548727494292e-07,
"loss": 0.4082,
"step": 194
},
{
"epoch": 0.3183673469387755,
"grad_norm": 0.4283614158630371,
"learning_rate": 1.7077534966650766e-07,
"loss": 0.4387,
"step": 195
},
{
"epoch": 0.32,
"grad_norm": 0.5072599649429321,
"learning_rate": 1.0931863906127327e-07,
"loss": 0.4489,
"step": 196
},
{
"epoch": 0.3216326530612245,
"grad_norm": 0.49709612131118774,
"learning_rate": 6.150154258476315e-08,
"loss": 0.4588,
"step": 197
},
{
"epoch": 0.32326530612244897,
"grad_norm": 0.5073265433311462,
"learning_rate": 2.7337132953697554e-08,
"loss": 0.394,
"step": 198
},
{
"epoch": 0.3248979591836735,
"grad_norm": 0.6317926645278931,
"learning_rate": 6.834750376549792e-09,
"loss": 0.5317,
"step": 199
},
{
"epoch": 0.32653061224489793,
"grad_norm": 0.7481065392494202,
"learning_rate": 0.0,
"loss": 0.5658,
"step": 200
},
{
"epoch": 0.32653061224489793,
"eval_loss": 0.3544212281703949,
"eval_runtime": 80.6876,
"eval_samples_per_second": 12.79,
"eval_steps_per_second": 3.198,
"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": 3.066053848800952e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}