abaddon182's picture
Training in progress, step 200, checkpoint
36dee34 verified
{
"best_metric": 1.4371888637542725,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.23460410557184752,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0011730205278592375,
"grad_norm": 8.663331031799316,
"learning_rate": 1e-05,
"loss": 13.3816,
"step": 1
},
{
"epoch": 0.0011730205278592375,
"eval_loss": 1.7326703071594238,
"eval_runtime": 53.9567,
"eval_samples_per_second": 26.614,
"eval_steps_per_second": 6.653,
"step": 1
},
{
"epoch": 0.002346041055718475,
"grad_norm": 51.352439880371094,
"learning_rate": 2e-05,
"loss": 15.251,
"step": 2
},
{
"epoch": 0.0035190615835777126,
"grad_norm": 26.00502586364746,
"learning_rate": 3e-05,
"loss": 16.2881,
"step": 3
},
{
"epoch": 0.00469208211143695,
"grad_norm": 10.644500732421875,
"learning_rate": 4e-05,
"loss": 16.008,
"step": 4
},
{
"epoch": 0.005865102639296188,
"grad_norm": 9.421217918395996,
"learning_rate": 5e-05,
"loss": 13.3204,
"step": 5
},
{
"epoch": 0.007038123167155425,
"grad_norm": 13.056597709655762,
"learning_rate": 6e-05,
"loss": 15.3015,
"step": 6
},
{
"epoch": 0.008211143695014663,
"grad_norm": 14.21023178100586,
"learning_rate": 7e-05,
"loss": 14.4311,
"step": 7
},
{
"epoch": 0.0093841642228739,
"grad_norm": 8.397007942199707,
"learning_rate": 8e-05,
"loss": 12.7272,
"step": 8
},
{
"epoch": 0.010557184750733138,
"grad_norm": 9.917564392089844,
"learning_rate": 9e-05,
"loss": 11.9593,
"step": 9
},
{
"epoch": 0.011730205278592375,
"grad_norm": 10.105253219604492,
"learning_rate": 0.0001,
"loss": 10.8888,
"step": 10
},
{
"epoch": 0.012903225806451613,
"grad_norm": 9.945771217346191,
"learning_rate": 9.999316524962345e-05,
"loss": 11.1743,
"step": 11
},
{
"epoch": 0.01407624633431085,
"grad_norm": 12.888904571533203,
"learning_rate": 9.997266286704631e-05,
"loss": 11.0644,
"step": 12
},
{
"epoch": 0.015249266862170088,
"grad_norm": 7.4127044677734375,
"learning_rate": 9.993849845741524e-05,
"loss": 9.0855,
"step": 13
},
{
"epoch": 0.016422287390029325,
"grad_norm": 7.778149127960205,
"learning_rate": 9.989068136093873e-05,
"loss": 9.4258,
"step": 14
},
{
"epoch": 0.017595307917888565,
"grad_norm": 8.04168701171875,
"learning_rate": 9.98292246503335e-05,
"loss": 9.2789,
"step": 15
},
{
"epoch": 0.0187683284457478,
"grad_norm": 6.784926414489746,
"learning_rate": 9.975414512725057e-05,
"loss": 9.1474,
"step": 16
},
{
"epoch": 0.01994134897360704,
"grad_norm": 6.441762924194336,
"learning_rate": 9.966546331768191e-05,
"loss": 8.7631,
"step": 17
},
{
"epoch": 0.021114369501466276,
"grad_norm": 9.08105754852295,
"learning_rate": 9.956320346634876e-05,
"loss": 8.476,
"step": 18
},
{
"epoch": 0.022287390029325515,
"grad_norm": 5.7696123123168945,
"learning_rate": 9.944739353007344e-05,
"loss": 8.6486,
"step": 19
},
{
"epoch": 0.02346041055718475,
"grad_norm": 5.676098823547363,
"learning_rate": 9.931806517013612e-05,
"loss": 8.7613,
"step": 20
},
{
"epoch": 0.02463343108504399,
"grad_norm": 5.42555570602417,
"learning_rate": 9.917525374361912e-05,
"loss": 8.5526,
"step": 21
},
{
"epoch": 0.025806451612903226,
"grad_norm": 5.8337321281433105,
"learning_rate": 9.901899829374047e-05,
"loss": 8.6943,
"step": 22
},
{
"epoch": 0.026979472140762465,
"grad_norm": 6.1469855308532715,
"learning_rate": 9.884934153917997e-05,
"loss": 8.8221,
"step": 23
},
{
"epoch": 0.0281524926686217,
"grad_norm": 5.11215353012085,
"learning_rate": 9.86663298624003e-05,
"loss": 8.7282,
"step": 24
},
{
"epoch": 0.02932551319648094,
"grad_norm": 5.950553894042969,
"learning_rate": 9.847001329696653e-05,
"loss": 8.0898,
"step": 25
},
{
"epoch": 0.030498533724340176,
"grad_norm": 6.3572492599487305,
"learning_rate": 9.826044551386744e-05,
"loss": 8.7058,
"step": 26
},
{
"epoch": 0.03167155425219941,
"grad_norm": 5.663265228271484,
"learning_rate": 9.803768380684242e-05,
"loss": 7.8918,
"step": 27
},
{
"epoch": 0.03284457478005865,
"grad_norm": 6.1261725425720215,
"learning_rate": 9.780178907671789e-05,
"loss": 8.3618,
"step": 28
},
{
"epoch": 0.03401759530791789,
"grad_norm": 11.174363136291504,
"learning_rate": 9.755282581475769e-05,
"loss": 7.6953,
"step": 29
},
{
"epoch": 0.03519061583577713,
"grad_norm": 6.395217418670654,
"learning_rate": 9.729086208503174e-05,
"loss": 8.6185,
"step": 30
},
{
"epoch": 0.03636363636363636,
"grad_norm": 6.65997314453125,
"learning_rate": 9.701596950580806e-05,
"loss": 8.0256,
"step": 31
},
{
"epoch": 0.0375366568914956,
"grad_norm": 6.332151889801025,
"learning_rate": 9.672822322997305e-05,
"loss": 7.8024,
"step": 32
},
{
"epoch": 0.03870967741935484,
"grad_norm": 14.636702537536621,
"learning_rate": 9.642770192448536e-05,
"loss": 8.6931,
"step": 33
},
{
"epoch": 0.03988269794721408,
"grad_norm": 6.699878215789795,
"learning_rate": 9.611448774886924e-05,
"loss": 7.633,
"step": 34
},
{
"epoch": 0.04105571847507331,
"grad_norm": 10.704326629638672,
"learning_rate": 9.578866633275288e-05,
"loss": 8.519,
"step": 35
},
{
"epoch": 0.04222873900293255,
"grad_norm": 7.794402122497559,
"learning_rate": 9.545032675245813e-05,
"loss": 8.2599,
"step": 36
},
{
"epoch": 0.04340175953079179,
"grad_norm": 7.623858451843262,
"learning_rate": 9.509956150664796e-05,
"loss": 8.5064,
"step": 37
},
{
"epoch": 0.04457478005865103,
"grad_norm": 7.614893436431885,
"learning_rate": 9.473646649103818e-05,
"loss": 7.8677,
"step": 38
},
{
"epoch": 0.04574780058651026,
"grad_norm": 19.778417587280273,
"learning_rate": 9.43611409721806e-05,
"loss": 8.3847,
"step": 39
},
{
"epoch": 0.0469208211143695,
"grad_norm": 9.958346366882324,
"learning_rate": 9.397368756032445e-05,
"loss": 7.5411,
"step": 40
},
{
"epoch": 0.04809384164222874,
"grad_norm": 10.39439582824707,
"learning_rate": 9.357421218136386e-05,
"loss": 8.6156,
"step": 41
},
{
"epoch": 0.04926686217008798,
"grad_norm": 9.843864440917969,
"learning_rate": 9.316282404787871e-05,
"loss": 8.73,
"step": 42
},
{
"epoch": 0.05043988269794721,
"grad_norm": 12.815672874450684,
"learning_rate": 9.273963562927695e-05,
"loss": 9.2569,
"step": 43
},
{
"epoch": 0.05161290322580645,
"grad_norm": 9.977739334106445,
"learning_rate": 9.230476262104677e-05,
"loss": 7.4537,
"step": 44
},
{
"epoch": 0.05278592375366569,
"grad_norm": 12.918895721435547,
"learning_rate": 9.185832391312644e-05,
"loss": 7.1692,
"step": 45
},
{
"epoch": 0.05395894428152493,
"grad_norm": 14.040423393249512,
"learning_rate": 9.140044155740101e-05,
"loss": 8.3064,
"step": 46
},
{
"epoch": 0.05513196480938416,
"grad_norm": 18.069244384765625,
"learning_rate": 9.093124073433463e-05,
"loss": 9.3387,
"step": 47
},
{
"epoch": 0.0563049853372434,
"grad_norm": 21.009868621826172,
"learning_rate": 9.045084971874738e-05,
"loss": 9.4285,
"step": 48
},
{
"epoch": 0.05747800586510264,
"grad_norm": 22.109519958496094,
"learning_rate": 8.995939984474624e-05,
"loss": 9.3635,
"step": 49
},
{
"epoch": 0.05865102639296188,
"grad_norm": 32.68623352050781,
"learning_rate": 8.945702546981969e-05,
"loss": 14.1459,
"step": 50
},
{
"epoch": 0.05865102639296188,
"eval_loss": 1.6147716045379639,
"eval_runtime": 54.9368,
"eval_samples_per_second": 26.139,
"eval_steps_per_second": 6.535,
"step": 50
},
{
"epoch": 0.05982404692082111,
"grad_norm": 11.36614990234375,
"learning_rate": 8.894386393810563e-05,
"loss": 7.8323,
"step": 51
},
{
"epoch": 0.06099706744868035,
"grad_norm": 14.187142372131348,
"learning_rate": 8.842005554284296e-05,
"loss": 8.1578,
"step": 52
},
{
"epoch": 0.06217008797653959,
"grad_norm": 8.44095516204834,
"learning_rate": 8.788574348801675e-05,
"loss": 7.6084,
"step": 53
},
{
"epoch": 0.06334310850439882,
"grad_norm": 6.0896830558776855,
"learning_rate": 8.73410738492077e-05,
"loss": 7.398,
"step": 54
},
{
"epoch": 0.06451612903225806,
"grad_norm": 6.624449253082275,
"learning_rate": 8.678619553365659e-05,
"loss": 7.1689,
"step": 55
},
{
"epoch": 0.0656891495601173,
"grad_norm": 4.72271728515625,
"learning_rate": 8.622126023955446e-05,
"loss": 7.3829,
"step": 56
},
{
"epoch": 0.06686217008797654,
"grad_norm": 11.342972755432129,
"learning_rate": 8.564642241456986e-05,
"loss": 7.1249,
"step": 57
},
{
"epoch": 0.06803519061583578,
"grad_norm": 4.155811309814453,
"learning_rate": 8.506183921362443e-05,
"loss": 7.2666,
"step": 58
},
{
"epoch": 0.06920821114369502,
"grad_norm": 4.137184143066406,
"learning_rate": 8.44676704559283e-05,
"loss": 7.1229,
"step": 59
},
{
"epoch": 0.07038123167155426,
"grad_norm": 7.632533073425293,
"learning_rate": 8.386407858128706e-05,
"loss": 7.0888,
"step": 60
},
{
"epoch": 0.07155425219941348,
"grad_norm": 4.366072654724121,
"learning_rate": 8.32512286056924e-05,
"loss": 7.3379,
"step": 61
},
{
"epoch": 0.07272727272727272,
"grad_norm": 3.8060083389282227,
"learning_rate": 8.262928807620843e-05,
"loss": 7.0819,
"step": 62
},
{
"epoch": 0.07390029325513196,
"grad_norm": 3.99633526802063,
"learning_rate": 8.199842702516583e-05,
"loss": 6.9591,
"step": 63
},
{
"epoch": 0.0750733137829912,
"grad_norm": 6.5943379402160645,
"learning_rate": 8.135881792367686e-05,
"loss": 6.9471,
"step": 64
},
{
"epoch": 0.07624633431085044,
"grad_norm": 4.896180629730225,
"learning_rate": 8.07106356344834e-05,
"loss": 7.2951,
"step": 65
},
{
"epoch": 0.07741935483870968,
"grad_norm": 4.8060479164123535,
"learning_rate": 8.005405736415126e-05,
"loss": 7.1263,
"step": 66
},
{
"epoch": 0.07859237536656892,
"grad_norm": 3.955714225769043,
"learning_rate": 7.938926261462366e-05,
"loss": 6.816,
"step": 67
},
{
"epoch": 0.07976539589442816,
"grad_norm": 4.465524673461914,
"learning_rate": 7.871643313414718e-05,
"loss": 7.3137,
"step": 68
},
{
"epoch": 0.08093841642228738,
"grad_norm": 4.67534065246582,
"learning_rate": 7.803575286758364e-05,
"loss": 6.8736,
"step": 69
},
{
"epoch": 0.08211143695014662,
"grad_norm": 4.462228298187256,
"learning_rate": 7.734740790612136e-05,
"loss": 7.3674,
"step": 70
},
{
"epoch": 0.08328445747800586,
"grad_norm": 5.6989312171936035,
"learning_rate": 7.66515864363997e-05,
"loss": 7.346,
"step": 71
},
{
"epoch": 0.0844574780058651,
"grad_norm": 4.496844291687012,
"learning_rate": 7.594847868906076e-05,
"loss": 7.2749,
"step": 72
},
{
"epoch": 0.08563049853372434,
"grad_norm": 5.543951988220215,
"learning_rate": 7.52382768867422e-05,
"loss": 7.2962,
"step": 73
},
{
"epoch": 0.08680351906158358,
"grad_norm": 4.945760726928711,
"learning_rate": 7.452117519152542e-05,
"loss": 7.3482,
"step": 74
},
{
"epoch": 0.08797653958944282,
"grad_norm": 5.323139667510986,
"learning_rate": 7.379736965185368e-05,
"loss": 7.1538,
"step": 75
},
{
"epoch": 0.08914956011730206,
"grad_norm": 5.503332138061523,
"learning_rate": 7.30670581489344e-05,
"loss": 7.7196,
"step": 76
},
{
"epoch": 0.09032258064516129,
"grad_norm": 5.515101909637451,
"learning_rate": 7.233044034264034e-05,
"loss": 6.8976,
"step": 77
},
{
"epoch": 0.09149560117302052,
"grad_norm": 6.265590190887451,
"learning_rate": 7.158771761692464e-05,
"loss": 7.1142,
"step": 78
},
{
"epoch": 0.09266862170087976,
"grad_norm": 5.986038684844971,
"learning_rate": 7.083909302476453e-05,
"loss": 7.4168,
"step": 79
},
{
"epoch": 0.093841642228739,
"grad_norm": 5.8563456535339355,
"learning_rate": 7.008477123264848e-05,
"loss": 7.7562,
"step": 80
},
{
"epoch": 0.09501466275659824,
"grad_norm": 7.149432182312012,
"learning_rate": 6.932495846462261e-05,
"loss": 6.9403,
"step": 81
},
{
"epoch": 0.09618768328445748,
"grad_norm": 5.055361747741699,
"learning_rate": 6.855986244591104e-05,
"loss": 7.4671,
"step": 82
},
{
"epoch": 0.09736070381231672,
"grad_norm": 5.610112190246582,
"learning_rate": 6.778969234612584e-05,
"loss": 7.4979,
"step": 83
},
{
"epoch": 0.09853372434017596,
"grad_norm": 5.591436862945557,
"learning_rate": 6.701465872208216e-05,
"loss": 7.4036,
"step": 84
},
{
"epoch": 0.09970674486803519,
"grad_norm": 5.245117664337158,
"learning_rate": 6.623497346023418e-05,
"loss": 7.4849,
"step": 85
},
{
"epoch": 0.10087976539589442,
"grad_norm": 6.682361602783203,
"learning_rate": 6.545084971874738e-05,
"loss": 7.4913,
"step": 86
},
{
"epoch": 0.10205278592375366,
"grad_norm": 7.1743927001953125,
"learning_rate": 6.466250186922325e-05,
"loss": 7.2481,
"step": 87
},
{
"epoch": 0.1032258064516129,
"grad_norm": 6.011048793792725,
"learning_rate": 6.387014543809223e-05,
"loss": 7.2815,
"step": 88
},
{
"epoch": 0.10439882697947214,
"grad_norm": 6.022535800933838,
"learning_rate": 6.307399704769099e-05,
"loss": 6.7935,
"step": 89
},
{
"epoch": 0.10557184750733138,
"grad_norm": 10.157115936279297,
"learning_rate": 6.227427435703997e-05,
"loss": 7.3842,
"step": 90
},
{
"epoch": 0.10674486803519062,
"grad_norm": 8.966989517211914,
"learning_rate": 6.147119600233758e-05,
"loss": 7.5907,
"step": 91
},
{
"epoch": 0.10791788856304986,
"grad_norm": 7.883905410766602,
"learning_rate": 6.066498153718735e-05,
"loss": 7.9347,
"step": 92
},
{
"epoch": 0.10909090909090909,
"grad_norm": 9.566387176513672,
"learning_rate": 5.985585137257401e-05,
"loss": 7.9712,
"step": 93
},
{
"epoch": 0.11026392961876832,
"grad_norm": 9.400655746459961,
"learning_rate": 5.90440267166055e-05,
"loss": 6.5513,
"step": 94
},
{
"epoch": 0.11143695014662756,
"grad_norm": 11.420488357543945,
"learning_rate": 5.8229729514036705e-05,
"loss": 6.9325,
"step": 95
},
{
"epoch": 0.1126099706744868,
"grad_norm": 15.522323608398438,
"learning_rate": 5.74131823855921e-05,
"loss": 8.2028,
"step": 96
},
{
"epoch": 0.11378299120234604,
"grad_norm": 14.273778915405273,
"learning_rate": 5.6594608567103456e-05,
"loss": 7.4233,
"step": 97
},
{
"epoch": 0.11495601173020528,
"grad_norm": 17.384851455688477,
"learning_rate": 5.577423184847932e-05,
"loss": 7.3731,
"step": 98
},
{
"epoch": 0.11612903225806452,
"grad_norm": 21.28325843811035,
"learning_rate": 5.495227651252315e-05,
"loss": 7.5629,
"step": 99
},
{
"epoch": 0.11730205278592376,
"grad_norm": 23.694461822509766,
"learning_rate": 5.4128967273616625e-05,
"loss": 9.0356,
"step": 100
},
{
"epoch": 0.11730205278592376,
"eval_loss": 1.5398024320602417,
"eval_runtime": 54.9395,
"eval_samples_per_second": 26.138,
"eval_steps_per_second": 6.534,
"step": 100
},
{
"epoch": 0.11847507331378299,
"grad_norm": 6.170729160308838,
"learning_rate": 5.330452921628497e-05,
"loss": 7.0788,
"step": 101
},
{
"epoch": 0.11964809384164223,
"grad_norm": 5.670572757720947,
"learning_rate": 5.247918773366112e-05,
"loss": 6.8288,
"step": 102
},
{
"epoch": 0.12082111436950146,
"grad_norm": 5.664034843444824,
"learning_rate": 5.165316846586541e-05,
"loss": 7.3678,
"step": 103
},
{
"epoch": 0.1219941348973607,
"grad_norm": 4.554018020629883,
"learning_rate": 5.0826697238317935e-05,
"loss": 6.9508,
"step": 104
},
{
"epoch": 0.12316715542521994,
"grad_norm": 3.6928629875183105,
"learning_rate": 5e-05,
"loss": 6.5448,
"step": 105
},
{
"epoch": 0.12434017595307918,
"grad_norm": 3.7579400539398193,
"learning_rate": 4.917330276168208e-05,
"loss": 6.329,
"step": 106
},
{
"epoch": 0.12551319648093842,
"grad_norm": 3.062084197998047,
"learning_rate": 4.834683153413459e-05,
"loss": 6.2742,
"step": 107
},
{
"epoch": 0.12668621700879765,
"grad_norm": 3.557678699493408,
"learning_rate": 4.7520812266338885e-05,
"loss": 7.0237,
"step": 108
},
{
"epoch": 0.1278592375366569,
"grad_norm": 3.9020395278930664,
"learning_rate": 4.669547078371504e-05,
"loss": 6.8212,
"step": 109
},
{
"epoch": 0.12903225806451613,
"grad_norm": 3.2508432865142822,
"learning_rate": 4.5871032726383386e-05,
"loss": 6.4402,
"step": 110
},
{
"epoch": 0.13020527859237538,
"grad_norm": 3.8005776405334473,
"learning_rate": 4.504772348747687e-05,
"loss": 7.13,
"step": 111
},
{
"epoch": 0.1313782991202346,
"grad_norm": 3.552947521209717,
"learning_rate": 4.4225768151520694e-05,
"loss": 6.3852,
"step": 112
},
{
"epoch": 0.13255131964809383,
"grad_norm": 3.8805272579193115,
"learning_rate": 4.3405391432896555e-05,
"loss": 6.9415,
"step": 113
},
{
"epoch": 0.13372434017595308,
"grad_norm": 4.0383806228637695,
"learning_rate": 4.2586817614407895e-05,
"loss": 6.7668,
"step": 114
},
{
"epoch": 0.1348973607038123,
"grad_norm": 3.9695963859558105,
"learning_rate": 4.17702704859633e-05,
"loss": 6.8736,
"step": 115
},
{
"epoch": 0.13607038123167156,
"grad_norm": 3.815753936767578,
"learning_rate": 4.095597328339452e-05,
"loss": 6.8127,
"step": 116
},
{
"epoch": 0.1372434017595308,
"grad_norm": 3.8608906269073486,
"learning_rate": 4.0144148627425993e-05,
"loss": 6.7532,
"step": 117
},
{
"epoch": 0.13841642228739004,
"grad_norm": 4.177197456359863,
"learning_rate": 3.933501846281267e-05,
"loss": 6.8221,
"step": 118
},
{
"epoch": 0.13958944281524927,
"grad_norm": 4.147789001464844,
"learning_rate": 3.852880399766243e-05,
"loss": 7.1846,
"step": 119
},
{
"epoch": 0.14076246334310852,
"grad_norm": 4.26793909072876,
"learning_rate": 3.772572564296005e-05,
"loss": 7.0007,
"step": 120
},
{
"epoch": 0.14193548387096774,
"grad_norm": 3.9579102993011475,
"learning_rate": 3.6926002952309016e-05,
"loss": 6.755,
"step": 121
},
{
"epoch": 0.14310850439882697,
"grad_norm": 4.167984485626221,
"learning_rate": 3.612985456190778e-05,
"loss": 7.0235,
"step": 122
},
{
"epoch": 0.14428152492668622,
"grad_norm": 4.612884044647217,
"learning_rate": 3.533749813077677e-05,
"loss": 6.6028,
"step": 123
},
{
"epoch": 0.14545454545454545,
"grad_norm": 4.473079681396484,
"learning_rate": 3.4549150281252636e-05,
"loss": 7.1627,
"step": 124
},
{
"epoch": 0.1466275659824047,
"grad_norm": 4.8667893409729,
"learning_rate": 3.3765026539765834e-05,
"loss": 7.169,
"step": 125
},
{
"epoch": 0.14780058651026393,
"grad_norm": 4.755546569824219,
"learning_rate": 3.298534127791785e-05,
"loss": 7.1955,
"step": 126
},
{
"epoch": 0.14897360703812318,
"grad_norm": 4.763340950012207,
"learning_rate": 3.221030765387417e-05,
"loss": 6.8214,
"step": 127
},
{
"epoch": 0.1501466275659824,
"grad_norm": 5.186800479888916,
"learning_rate": 3.144013755408895e-05,
"loss": 7.5825,
"step": 128
},
{
"epoch": 0.15131964809384163,
"grad_norm": 4.935817718505859,
"learning_rate": 3.0675041535377405e-05,
"loss": 6.8696,
"step": 129
},
{
"epoch": 0.15249266862170088,
"grad_norm": 4.84830904006958,
"learning_rate": 2.991522876735154e-05,
"loss": 6.6679,
"step": 130
},
{
"epoch": 0.1536656891495601,
"grad_norm": 5.055245876312256,
"learning_rate": 2.916090697523549e-05,
"loss": 6.9668,
"step": 131
},
{
"epoch": 0.15483870967741936,
"grad_norm": 5.089951038360596,
"learning_rate": 2.8412282383075363e-05,
"loss": 6.8507,
"step": 132
},
{
"epoch": 0.1560117302052786,
"grad_norm": 5.252066135406494,
"learning_rate": 2.766955965735968e-05,
"loss": 6.6759,
"step": 133
},
{
"epoch": 0.15718475073313784,
"grad_norm": 5.395578861236572,
"learning_rate": 2.693294185106562e-05,
"loss": 7.5199,
"step": 134
},
{
"epoch": 0.15835777126099707,
"grad_norm": 5.481312274932861,
"learning_rate": 2.6202630348146324e-05,
"loss": 7.0232,
"step": 135
},
{
"epoch": 0.15953079178885632,
"grad_norm": 6.59968900680542,
"learning_rate": 2.547882480847461e-05,
"loss": 6.421,
"step": 136
},
{
"epoch": 0.16070381231671554,
"grad_norm": 5.844842433929443,
"learning_rate": 2.476172311325783e-05,
"loss": 6.9942,
"step": 137
},
{
"epoch": 0.16187683284457477,
"grad_norm": 6.2577056884765625,
"learning_rate": 2.405152131093926e-05,
"loss": 6.4066,
"step": 138
},
{
"epoch": 0.16304985337243402,
"grad_norm": 7.429435729980469,
"learning_rate": 2.3348413563600325e-05,
"loss": 7.0302,
"step": 139
},
{
"epoch": 0.16422287390029325,
"grad_norm": 7.769280910491943,
"learning_rate": 2.2652592093878666e-05,
"loss": 6.7873,
"step": 140
},
{
"epoch": 0.1653958944281525,
"grad_norm": 7.310204982757568,
"learning_rate": 2.196424713241637e-05,
"loss": 7.0036,
"step": 141
},
{
"epoch": 0.16656891495601173,
"grad_norm": 9.804479598999023,
"learning_rate": 2.128356686585282e-05,
"loss": 7.0607,
"step": 142
},
{
"epoch": 0.16774193548387098,
"grad_norm": 8.672088623046875,
"learning_rate": 2.061073738537635e-05,
"loss": 7.1593,
"step": 143
},
{
"epoch": 0.1689149560117302,
"grad_norm": 12.077096939086914,
"learning_rate": 1.9945942635848748e-05,
"loss": 7.5432,
"step": 144
},
{
"epoch": 0.17008797653958943,
"grad_norm": 12.569242477416992,
"learning_rate": 1.928936436551661e-05,
"loss": 6.4717,
"step": 145
},
{
"epoch": 0.17126099706744868,
"grad_norm": 13.25029468536377,
"learning_rate": 1.8641182076323148e-05,
"loss": 7.6049,
"step": 146
},
{
"epoch": 0.1724340175953079,
"grad_norm": 15.713706016540527,
"learning_rate": 1.800157297483417e-05,
"loss": 7.0083,
"step": 147
},
{
"epoch": 0.17360703812316716,
"grad_norm": 18.675628662109375,
"learning_rate": 1.7370711923791567e-05,
"loss": 7.4158,
"step": 148
},
{
"epoch": 0.1747800586510264,
"grad_norm": 20.445755004882812,
"learning_rate": 1.6748771394307585e-05,
"loss": 8.0984,
"step": 149
},
{
"epoch": 0.17595307917888564,
"grad_norm": 31.651071548461914,
"learning_rate": 1.6135921418712956e-05,
"loss": 9.0892,
"step": 150
},
{
"epoch": 0.17595307917888564,
"eval_loss": 1.4641185998916626,
"eval_runtime": 54.969,
"eval_samples_per_second": 26.124,
"eval_steps_per_second": 6.531,
"step": 150
},
{
"epoch": 0.17712609970674487,
"grad_norm": 3.545945882797241,
"learning_rate": 1.553232954407171e-05,
"loss": 6.3979,
"step": 151
},
{
"epoch": 0.17829912023460412,
"grad_norm": 3.4330179691314697,
"learning_rate": 1.4938160786375572e-05,
"loss": 5.8609,
"step": 152
},
{
"epoch": 0.17947214076246334,
"grad_norm": 3.8319311141967773,
"learning_rate": 1.435357758543015e-05,
"loss": 6.8796,
"step": 153
},
{
"epoch": 0.18064516129032257,
"grad_norm": 4.39741849899292,
"learning_rate": 1.3778739760445552e-05,
"loss": 6.2018,
"step": 154
},
{
"epoch": 0.18181818181818182,
"grad_norm": 4.3072357177734375,
"learning_rate": 1.3213804466343421e-05,
"loss": 6.1792,
"step": 155
},
{
"epoch": 0.18299120234604105,
"grad_norm": 3.892882823944092,
"learning_rate": 1.2658926150792322e-05,
"loss": 6.3373,
"step": 156
},
{
"epoch": 0.1841642228739003,
"grad_norm": 3.4460794925689697,
"learning_rate": 1.2114256511983274e-05,
"loss": 6.3294,
"step": 157
},
{
"epoch": 0.18533724340175953,
"grad_norm": 3.7324860095977783,
"learning_rate": 1.157994445715706e-05,
"loss": 6.3531,
"step": 158
},
{
"epoch": 0.18651026392961878,
"grad_norm": 3.388925552368164,
"learning_rate": 1.1056136061894384e-05,
"loss": 6.4866,
"step": 159
},
{
"epoch": 0.187683284457478,
"grad_norm": 3.5137898921966553,
"learning_rate": 1.0542974530180327e-05,
"loss": 6.6119,
"step": 160
},
{
"epoch": 0.18885630498533723,
"grad_norm": 3.3227107524871826,
"learning_rate": 1.0040600155253765e-05,
"loss": 6.3487,
"step": 161
},
{
"epoch": 0.19002932551319648,
"grad_norm": 3.202991485595703,
"learning_rate": 9.549150281252633e-06,
"loss": 6.5303,
"step": 162
},
{
"epoch": 0.1912023460410557,
"grad_norm": 4.062410831451416,
"learning_rate": 9.068759265665384e-06,
"loss": 6.5675,
"step": 163
},
{
"epoch": 0.19237536656891496,
"grad_norm": 3.620666265487671,
"learning_rate": 8.599558442598998e-06,
"loss": 6.4053,
"step": 164
},
{
"epoch": 0.1935483870967742,
"grad_norm": 3.3483715057373047,
"learning_rate": 8.141676086873572e-06,
"loss": 6.7736,
"step": 165
},
{
"epoch": 0.19472140762463344,
"grad_norm": 3.302370309829712,
"learning_rate": 7.695237378953223e-06,
"loss": 6.3168,
"step": 166
},
{
"epoch": 0.19589442815249267,
"grad_norm": 4.081686973571777,
"learning_rate": 7.260364370723044e-06,
"loss": 7.0721,
"step": 167
},
{
"epoch": 0.19706744868035192,
"grad_norm": 3.7871413230895996,
"learning_rate": 6.837175952121306e-06,
"loss": 6.6491,
"step": 168
},
{
"epoch": 0.19824046920821115,
"grad_norm": 3.5725629329681396,
"learning_rate": 6.425787818636131e-06,
"loss": 7.232,
"step": 169
},
{
"epoch": 0.19941348973607037,
"grad_norm": 3.952382802963257,
"learning_rate": 6.026312439675552e-06,
"loss": 6.9474,
"step": 170
},
{
"epoch": 0.20058651026392962,
"grad_norm": 3.558424711227417,
"learning_rate": 5.6388590278194096e-06,
"loss": 6.8185,
"step": 171
},
{
"epoch": 0.20175953079178885,
"grad_norm": 4.020036697387695,
"learning_rate": 5.263533508961827e-06,
"loss": 6.9452,
"step": 172
},
{
"epoch": 0.2029325513196481,
"grad_norm": 4.385979175567627,
"learning_rate": 4.900438493352055e-06,
"loss": 6.7301,
"step": 173
},
{
"epoch": 0.20410557184750733,
"grad_norm": 4.704169750213623,
"learning_rate": 4.549673247541875e-06,
"loss": 6.6311,
"step": 174
},
{
"epoch": 0.20527859237536658,
"grad_norm": 4.294439315795898,
"learning_rate": 4.2113336672471245e-06,
"loss": 7.0591,
"step": 175
},
{
"epoch": 0.2064516129032258,
"grad_norm": 4.28928279876709,
"learning_rate": 3.885512251130763e-06,
"loss": 6.6714,
"step": 176
},
{
"epoch": 0.20762463343108503,
"grad_norm": 4.021836757659912,
"learning_rate": 3.5722980755146517e-06,
"loss": 6.8467,
"step": 177
},
{
"epoch": 0.20879765395894428,
"grad_norm": 4.013036727905273,
"learning_rate": 3.271776770026963e-06,
"loss": 6.4059,
"step": 178
},
{
"epoch": 0.2099706744868035,
"grad_norm": 4.504580020904541,
"learning_rate": 2.9840304941919415e-06,
"loss": 6.9156,
"step": 179
},
{
"epoch": 0.21114369501466276,
"grad_norm": 4.332256317138672,
"learning_rate": 2.7091379149682685e-06,
"loss": 6.9564,
"step": 180
},
{
"epoch": 0.212316715542522,
"grad_norm": 4.6176347732543945,
"learning_rate": 2.4471741852423237e-06,
"loss": 6.9044,
"step": 181
},
{
"epoch": 0.21348973607038124,
"grad_norm": 5.39663028717041,
"learning_rate": 2.1982109232821178e-06,
"loss": 7.2944,
"step": 182
},
{
"epoch": 0.21466275659824047,
"grad_norm": 4.653443813323975,
"learning_rate": 1.962316193157593e-06,
"loss": 7.0987,
"step": 183
},
{
"epoch": 0.21583577712609972,
"grad_norm": 6.479732513427734,
"learning_rate": 1.7395544861325718e-06,
"loss": 6.1803,
"step": 184
},
{
"epoch": 0.21700879765395895,
"grad_norm": 6.245946884155273,
"learning_rate": 1.5299867030334814e-06,
"loss": 6.7305,
"step": 185
},
{
"epoch": 0.21818181818181817,
"grad_norm": 6.659122943878174,
"learning_rate": 1.333670137599713e-06,
"loss": 7.5387,
"step": 186
},
{
"epoch": 0.21935483870967742,
"grad_norm": 5.82586669921875,
"learning_rate": 1.1506584608200367e-06,
"loss": 6.5154,
"step": 187
},
{
"epoch": 0.22052785923753665,
"grad_norm": 6.484030723571777,
"learning_rate": 9.810017062595322e-07,
"loss": 6.526,
"step": 188
},
{
"epoch": 0.2217008797653959,
"grad_norm": 5.605183124542236,
"learning_rate": 8.247462563808817e-07,
"loss": 6.6999,
"step": 189
},
{
"epoch": 0.22287390029325513,
"grad_norm": 6.196300983428955,
"learning_rate": 6.819348298638839e-07,
"loss": 6.4769,
"step": 190
},
{
"epoch": 0.22404692082111438,
"grad_norm": 7.179093837738037,
"learning_rate": 5.526064699265753e-07,
"loss": 7.2343,
"step": 191
},
{
"epoch": 0.2252199413489736,
"grad_norm": 7.460024356842041,
"learning_rate": 4.367965336512403e-07,
"loss": 7.5183,
"step": 192
},
{
"epoch": 0.22639296187683283,
"grad_norm": 9.167801856994629,
"learning_rate": 3.3453668231809286e-07,
"loss": 6.9693,
"step": 193
},
{
"epoch": 0.22756598240469209,
"grad_norm": 8.804646492004395,
"learning_rate": 2.458548727494292e-07,
"loss": 7.5381,
"step": 194
},
{
"epoch": 0.2287390029325513,
"grad_norm": 9.841207504272461,
"learning_rate": 1.7077534966650766e-07,
"loss": 7.295,
"step": 195
},
{
"epoch": 0.22991202346041056,
"grad_norm": 13.219451904296875,
"learning_rate": 1.0931863906127327e-07,
"loss": 7.4361,
"step": 196
},
{
"epoch": 0.2310850439882698,
"grad_norm": 15.988734245300293,
"learning_rate": 6.150154258476315e-08,
"loss": 7.2263,
"step": 197
},
{
"epoch": 0.23225806451612904,
"grad_norm": 15.204376220703125,
"learning_rate": 2.7337132953697554e-08,
"loss": 9.2538,
"step": 198
},
{
"epoch": 0.23343108504398827,
"grad_norm": 22.735008239746094,
"learning_rate": 6.834750376549792e-09,
"loss": 9.2901,
"step": 199
},
{
"epoch": 0.23460410557184752,
"grad_norm": 72.2196273803711,
"learning_rate": 0.0,
"loss": 10.607,
"step": 200
},
{
"epoch": 0.23460410557184752,
"eval_loss": 1.4371888637542725,
"eval_runtime": 54.9673,
"eval_samples_per_second": 26.125,
"eval_steps_per_second": 6.531,
"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": 1.169711928705024e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}