aleegis12's picture
Training in progress, step 200, checkpoint
db9b40c verified
{
"best_metric": 3.0534605979919434,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.011819981679028398,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 5.909990839514199e-05,
"grad_norm": 703.1734619140625,
"learning_rate": 1e-05,
"loss": 16.5184,
"step": 1
},
{
"epoch": 5.909990839514199e-05,
"eval_loss": 4.820686340332031,
"eval_runtime": 161.1367,
"eval_samples_per_second": 176.856,
"eval_steps_per_second": 44.217,
"step": 1
},
{
"epoch": 0.00011819981679028398,
"grad_norm": 717.785888671875,
"learning_rate": 2e-05,
"loss": 15.8334,
"step": 2
},
{
"epoch": 0.00017729972518542596,
"grad_norm": 582.750244140625,
"learning_rate": 3e-05,
"loss": 14.9435,
"step": 3
},
{
"epoch": 0.00023639963358056796,
"grad_norm": 630.3521118164062,
"learning_rate": 4e-05,
"loss": 16.1476,
"step": 4
},
{
"epoch": 0.00029549954197570993,
"grad_norm": 1142.2156982421875,
"learning_rate": 5e-05,
"loss": 15.2671,
"step": 5
},
{
"epoch": 0.0003545994503708519,
"grad_norm": 568.1258544921875,
"learning_rate": 6e-05,
"loss": 16.4606,
"step": 6
},
{
"epoch": 0.0004136993587659939,
"grad_norm": 1064.763671875,
"learning_rate": 7e-05,
"loss": 14.8942,
"step": 7
},
{
"epoch": 0.0004727992671611359,
"grad_norm": 501.0908508300781,
"learning_rate": 8e-05,
"loss": 14.5017,
"step": 8
},
{
"epoch": 0.0005318991755562779,
"grad_norm": 541.7283935546875,
"learning_rate": 9e-05,
"loss": 15.1543,
"step": 9
},
{
"epoch": 0.0005909990839514199,
"grad_norm": 645.4530029296875,
"learning_rate": 0.0001,
"loss": 14.116,
"step": 10
},
{
"epoch": 0.0006500989923465619,
"grad_norm": 678.97021484375,
"learning_rate": 9.999316524962345e-05,
"loss": 13.5733,
"step": 11
},
{
"epoch": 0.0007091989007417038,
"grad_norm": 922.26171875,
"learning_rate": 9.997266286704631e-05,
"loss": 12.9788,
"step": 12
},
{
"epoch": 0.0007682988091368458,
"grad_norm": 3544.84326171875,
"learning_rate": 9.993849845741524e-05,
"loss": 11.5049,
"step": 13
},
{
"epoch": 0.0008273987175319878,
"grad_norm": 1291.6978759765625,
"learning_rate": 9.989068136093873e-05,
"loss": 11.9413,
"step": 14
},
{
"epoch": 0.0008864986259271298,
"grad_norm": 997.0775756835938,
"learning_rate": 9.98292246503335e-05,
"loss": 12.7419,
"step": 15
},
{
"epoch": 0.0009455985343222718,
"grad_norm": 1210.50537109375,
"learning_rate": 9.975414512725057e-05,
"loss": 15.0342,
"step": 16
},
{
"epoch": 0.0010046984427174138,
"grad_norm": 1125.9130859375,
"learning_rate": 9.966546331768191e-05,
"loss": 12.9054,
"step": 17
},
{
"epoch": 0.0010637983511125558,
"grad_norm": 881.341064453125,
"learning_rate": 9.956320346634876e-05,
"loss": 12.0417,
"step": 18
},
{
"epoch": 0.0011228982595076977,
"grad_norm": 892.4519653320312,
"learning_rate": 9.944739353007344e-05,
"loss": 12.221,
"step": 19
},
{
"epoch": 0.0011819981679028397,
"grad_norm": 812.6408081054688,
"learning_rate": 9.931806517013612e-05,
"loss": 12.0205,
"step": 20
},
{
"epoch": 0.0012410980762979818,
"grad_norm": 830.1318359375,
"learning_rate": 9.917525374361912e-05,
"loss": 12.2189,
"step": 21
},
{
"epoch": 0.0013001979846931238,
"grad_norm": 1028.381103515625,
"learning_rate": 9.901899829374047e-05,
"loss": 11.5536,
"step": 22
},
{
"epoch": 0.0013592978930882656,
"grad_norm": 1027.7178955078125,
"learning_rate": 9.884934153917997e-05,
"loss": 12.3491,
"step": 23
},
{
"epoch": 0.0014183978014834077,
"grad_norm": 1146.283447265625,
"learning_rate": 9.86663298624003e-05,
"loss": 12.1409,
"step": 24
},
{
"epoch": 0.0014774977098785497,
"grad_norm": 1509.816162109375,
"learning_rate": 9.847001329696653e-05,
"loss": 12.3984,
"step": 25
},
{
"epoch": 0.0015365976182736916,
"grad_norm": 433.6212463378906,
"learning_rate": 9.826044551386744e-05,
"loss": 11.5371,
"step": 26
},
{
"epoch": 0.0015956975266688336,
"grad_norm": 841.4713745117188,
"learning_rate": 9.803768380684242e-05,
"loss": 10.9008,
"step": 27
},
{
"epoch": 0.0016547974350639757,
"grad_norm": 941.1351928710938,
"learning_rate": 9.780178907671789e-05,
"loss": 11.1156,
"step": 28
},
{
"epoch": 0.0017138973434591177,
"grad_norm": 1331.7652587890625,
"learning_rate": 9.755282581475769e-05,
"loss": 11.9987,
"step": 29
},
{
"epoch": 0.0017729972518542596,
"grad_norm": 1087.48583984375,
"learning_rate": 9.729086208503174e-05,
"loss": 12.9907,
"step": 30
},
{
"epoch": 0.0018320971602494016,
"grad_norm": 793.1517333984375,
"learning_rate": 9.701596950580806e-05,
"loss": 12.6025,
"step": 31
},
{
"epoch": 0.0018911970686445437,
"grad_norm": 751.8848266601562,
"learning_rate": 9.672822322997305e-05,
"loss": 11.7924,
"step": 32
},
{
"epoch": 0.0019502969770396855,
"grad_norm": 1595.2109375,
"learning_rate": 9.642770192448536e-05,
"loss": 11.3823,
"step": 33
},
{
"epoch": 0.0020093968854348276,
"grad_norm": 944.9501953125,
"learning_rate": 9.611448774886924e-05,
"loss": 12.6833,
"step": 34
},
{
"epoch": 0.0020684967938299696,
"grad_norm": 878.3543090820312,
"learning_rate": 9.578866633275288e-05,
"loss": 12.3733,
"step": 35
},
{
"epoch": 0.0021275967022251117,
"grad_norm": 953.1663208007812,
"learning_rate": 9.545032675245813e-05,
"loss": 13.5458,
"step": 36
},
{
"epoch": 0.0021866966106202537,
"grad_norm": 493.4095764160156,
"learning_rate": 9.509956150664796e-05,
"loss": 14.9294,
"step": 37
},
{
"epoch": 0.0022457965190153953,
"grad_norm": 777.3856811523438,
"learning_rate": 9.473646649103818e-05,
"loss": 12.7389,
"step": 38
},
{
"epoch": 0.0023048964274105374,
"grad_norm": 1160.2825927734375,
"learning_rate": 9.43611409721806e-05,
"loss": 13.0212,
"step": 39
},
{
"epoch": 0.0023639963358056794,
"grad_norm": 1605.1435546875,
"learning_rate": 9.397368756032445e-05,
"loss": 14.4828,
"step": 40
},
{
"epoch": 0.0024230962442008215,
"grad_norm": 3370.127685546875,
"learning_rate": 9.357421218136386e-05,
"loss": 11.6989,
"step": 41
},
{
"epoch": 0.0024821961525959635,
"grad_norm": 1803.8907470703125,
"learning_rate": 9.316282404787871e-05,
"loss": 12.2736,
"step": 42
},
{
"epoch": 0.0025412960609911056,
"grad_norm": 745.3320922851562,
"learning_rate": 9.273963562927695e-05,
"loss": 11.5669,
"step": 43
},
{
"epoch": 0.0026003959693862476,
"grad_norm": 1559.7301025390625,
"learning_rate": 9.230476262104677e-05,
"loss": 11.9983,
"step": 44
},
{
"epoch": 0.0026594958777813892,
"grad_norm": 754.876953125,
"learning_rate": 9.185832391312644e-05,
"loss": 12.4019,
"step": 45
},
{
"epoch": 0.0027185957861765313,
"grad_norm": 1105.93310546875,
"learning_rate": 9.140044155740101e-05,
"loss": 13.3279,
"step": 46
},
{
"epoch": 0.0027776956945716733,
"grad_norm": 1047.6898193359375,
"learning_rate": 9.093124073433463e-05,
"loss": 12.7022,
"step": 47
},
{
"epoch": 0.0028367956029668154,
"grad_norm": 1175.58984375,
"learning_rate": 9.045084971874738e-05,
"loss": 13.9106,
"step": 48
},
{
"epoch": 0.0028958955113619574,
"grad_norm": 1044.486328125,
"learning_rate": 8.995939984474624e-05,
"loss": 12.6573,
"step": 49
},
{
"epoch": 0.0029549954197570995,
"grad_norm": 824.6028442382812,
"learning_rate": 8.945702546981969e-05,
"loss": 14.7135,
"step": 50
},
{
"epoch": 0.0029549954197570995,
"eval_loss": 3.2013654708862305,
"eval_runtime": 160.9184,
"eval_samples_per_second": 177.096,
"eval_steps_per_second": 44.277,
"step": 50
},
{
"epoch": 0.0030140953281522415,
"grad_norm": 2924.23095703125,
"learning_rate": 8.894386393810563e-05,
"loss": 13.4507,
"step": 51
},
{
"epoch": 0.003073195236547383,
"grad_norm": 2628.776611328125,
"learning_rate": 8.842005554284296e-05,
"loss": 13.3128,
"step": 52
},
{
"epoch": 0.003132295144942525,
"grad_norm": 2073.83544921875,
"learning_rate": 8.788574348801675e-05,
"loss": 13.352,
"step": 53
},
{
"epoch": 0.0031913950533376673,
"grad_norm": 2465.630859375,
"learning_rate": 8.73410738492077e-05,
"loss": 12.9891,
"step": 54
},
{
"epoch": 0.0032504949617328093,
"grad_norm": 2529.382568359375,
"learning_rate": 8.678619553365659e-05,
"loss": 12.5111,
"step": 55
},
{
"epoch": 0.0033095948701279514,
"grad_norm": 1743.16162109375,
"learning_rate": 8.622126023955446e-05,
"loss": 14.8992,
"step": 56
},
{
"epoch": 0.0033686947785230934,
"grad_norm": 1476.960205078125,
"learning_rate": 8.564642241456986e-05,
"loss": 13.4571,
"step": 57
},
{
"epoch": 0.0034277946869182355,
"grad_norm": 1699.7786865234375,
"learning_rate": 8.506183921362443e-05,
"loss": 13.4452,
"step": 58
},
{
"epoch": 0.003486894595313377,
"grad_norm": 2552.3583984375,
"learning_rate": 8.44676704559283e-05,
"loss": 13.4764,
"step": 59
},
{
"epoch": 0.003545994503708519,
"grad_norm": 1311.1492919921875,
"learning_rate": 8.386407858128706e-05,
"loss": 13.1806,
"step": 60
},
{
"epoch": 0.003605094412103661,
"grad_norm": 1845.13427734375,
"learning_rate": 8.32512286056924e-05,
"loss": 13.3577,
"step": 61
},
{
"epoch": 0.0036641943204988032,
"grad_norm": 1508.0977783203125,
"learning_rate": 8.262928807620843e-05,
"loss": 12.7121,
"step": 62
},
{
"epoch": 0.0037232942288939453,
"grad_norm": 1883.4854736328125,
"learning_rate": 8.199842702516583e-05,
"loss": 13.0846,
"step": 63
},
{
"epoch": 0.0037823941372890873,
"grad_norm": 2035.174560546875,
"learning_rate": 8.135881792367686e-05,
"loss": 12.7806,
"step": 64
},
{
"epoch": 0.0038414940456842294,
"grad_norm": 1674.7613525390625,
"learning_rate": 8.07106356344834e-05,
"loss": 12.8371,
"step": 65
},
{
"epoch": 0.003900593954079371,
"grad_norm": 1369.8162841796875,
"learning_rate": 8.005405736415126e-05,
"loss": 12.3912,
"step": 66
},
{
"epoch": 0.003959693862474513,
"grad_norm": 1284.155029296875,
"learning_rate": 7.938926261462366e-05,
"loss": 12.5755,
"step": 67
},
{
"epoch": 0.004018793770869655,
"grad_norm": 1361.718994140625,
"learning_rate": 7.871643313414718e-05,
"loss": 12.3395,
"step": 68
},
{
"epoch": 0.004077893679264797,
"grad_norm": 1595.88671875,
"learning_rate": 7.803575286758364e-05,
"loss": 12.0747,
"step": 69
},
{
"epoch": 0.004136993587659939,
"grad_norm": 1556.0238037109375,
"learning_rate": 7.734740790612136e-05,
"loss": 14.2545,
"step": 70
},
{
"epoch": 0.004196093496055081,
"grad_norm": 2383.250732421875,
"learning_rate": 7.66515864363997e-05,
"loss": 13.1199,
"step": 71
},
{
"epoch": 0.004255193404450223,
"grad_norm": 952.619140625,
"learning_rate": 7.594847868906076e-05,
"loss": 13.0062,
"step": 72
},
{
"epoch": 0.004314293312845365,
"grad_norm": 2889.347900390625,
"learning_rate": 7.52382768867422e-05,
"loss": 11.9664,
"step": 73
},
{
"epoch": 0.004373393221240507,
"grad_norm": 1298.2371826171875,
"learning_rate": 7.452117519152542e-05,
"loss": 12.3684,
"step": 74
},
{
"epoch": 0.0044324931296356495,
"grad_norm": 1176.57080078125,
"learning_rate": 7.379736965185368e-05,
"loss": 13.1936,
"step": 75
},
{
"epoch": 0.004491593038030791,
"grad_norm": 2327.53662109375,
"learning_rate": 7.30670581489344e-05,
"loss": 12.5427,
"step": 76
},
{
"epoch": 0.004550692946425933,
"grad_norm": 2644.671142578125,
"learning_rate": 7.233044034264034e-05,
"loss": 13.1643,
"step": 77
},
{
"epoch": 0.004609792854821075,
"grad_norm": 1783.341064453125,
"learning_rate": 7.158771761692464e-05,
"loss": 11.3356,
"step": 78
},
{
"epoch": 0.004668892763216217,
"grad_norm": 2049.7216796875,
"learning_rate": 7.083909302476453e-05,
"loss": 11.5555,
"step": 79
},
{
"epoch": 0.004727992671611359,
"grad_norm": 1044.854248046875,
"learning_rate": 7.008477123264848e-05,
"loss": 11.2852,
"step": 80
},
{
"epoch": 0.004787092580006501,
"grad_norm": 2483.975830078125,
"learning_rate": 6.932495846462261e-05,
"loss": 12.8895,
"step": 81
},
{
"epoch": 0.004846192488401643,
"grad_norm": 1112.5596923828125,
"learning_rate": 6.855986244591104e-05,
"loss": 12.2603,
"step": 82
},
{
"epoch": 0.004905292396796785,
"grad_norm": 961.6146850585938,
"learning_rate": 6.778969234612584e-05,
"loss": 13.121,
"step": 83
},
{
"epoch": 0.004964392305191927,
"grad_norm": 1315.409423828125,
"learning_rate": 6.701465872208216e-05,
"loss": 12.461,
"step": 84
},
{
"epoch": 0.005023492213587069,
"grad_norm": 1070.69287109375,
"learning_rate": 6.623497346023418e-05,
"loss": 11.3974,
"step": 85
},
{
"epoch": 0.005082592121982211,
"grad_norm": 1133.186279296875,
"learning_rate": 6.545084971874738e-05,
"loss": 13.2055,
"step": 86
},
{
"epoch": 0.005141692030377353,
"grad_norm": 1054.05517578125,
"learning_rate": 6.466250186922325e-05,
"loss": 12.3085,
"step": 87
},
{
"epoch": 0.005200791938772495,
"grad_norm": 780.37353515625,
"learning_rate": 6.387014543809223e-05,
"loss": 11.6469,
"step": 88
},
{
"epoch": 0.005259891847167637,
"grad_norm": 1564.8665771484375,
"learning_rate": 6.307399704769099e-05,
"loss": 11.4207,
"step": 89
},
{
"epoch": 0.0053189917555627785,
"grad_norm": 1111.5013427734375,
"learning_rate": 6.227427435703997e-05,
"loss": 13.2361,
"step": 90
},
{
"epoch": 0.0053780916639579205,
"grad_norm": 1811.0882568359375,
"learning_rate": 6.147119600233758e-05,
"loss": 11.8001,
"step": 91
},
{
"epoch": 0.005437191572353063,
"grad_norm": 943.4735717773438,
"learning_rate": 6.066498153718735e-05,
"loss": 12.7045,
"step": 92
},
{
"epoch": 0.005496291480748205,
"grad_norm": 907.6875610351562,
"learning_rate": 5.985585137257401e-05,
"loss": 10.8836,
"step": 93
},
{
"epoch": 0.005555391389143347,
"grad_norm": 1158.319091796875,
"learning_rate": 5.90440267166055e-05,
"loss": 11.5309,
"step": 94
},
{
"epoch": 0.005614491297538489,
"grad_norm": 1613.986572265625,
"learning_rate": 5.8229729514036705e-05,
"loss": 10.7983,
"step": 95
},
{
"epoch": 0.005673591205933631,
"grad_norm": 2079.008544921875,
"learning_rate": 5.74131823855921e-05,
"loss": 13.062,
"step": 96
},
{
"epoch": 0.005732691114328773,
"grad_norm": 819.364990234375,
"learning_rate": 5.6594608567103456e-05,
"loss": 13.1781,
"step": 97
},
{
"epoch": 0.005791791022723915,
"grad_norm": 833.9022827148438,
"learning_rate": 5.577423184847932e-05,
"loss": 12.5504,
"step": 98
},
{
"epoch": 0.005850890931119057,
"grad_norm": 1711.939697265625,
"learning_rate": 5.495227651252315e-05,
"loss": 13.0599,
"step": 99
},
{
"epoch": 0.005909990839514199,
"grad_norm": 1087.9351806640625,
"learning_rate": 5.4128967273616625e-05,
"loss": 13.9811,
"step": 100
},
{
"epoch": 0.005909990839514199,
"eval_loss": 3.1570539474487305,
"eval_runtime": 161.0068,
"eval_samples_per_second": 176.999,
"eval_steps_per_second": 44.253,
"step": 100
},
{
"epoch": 0.005969090747909341,
"grad_norm": 2899.854248046875,
"learning_rate": 5.330452921628497e-05,
"loss": 14.1056,
"step": 101
},
{
"epoch": 0.006028190656304483,
"grad_norm": 2942.791015625,
"learning_rate": 5.247918773366112e-05,
"loss": 14.5698,
"step": 102
},
{
"epoch": 0.006087290564699625,
"grad_norm": 3781.53125,
"learning_rate": 5.165316846586541e-05,
"loss": 13.4829,
"step": 103
},
{
"epoch": 0.006146390473094766,
"grad_norm": 4378.0537109375,
"learning_rate": 5.0826697238317935e-05,
"loss": 15.5151,
"step": 104
},
{
"epoch": 0.006205490381489908,
"grad_norm": 2870.376220703125,
"learning_rate": 5e-05,
"loss": 15.1653,
"step": 105
},
{
"epoch": 0.00626459028988505,
"grad_norm": 2293.7763671875,
"learning_rate": 4.917330276168208e-05,
"loss": 11.638,
"step": 106
},
{
"epoch": 0.0063236901982801925,
"grad_norm": 3085.556884765625,
"learning_rate": 4.834683153413459e-05,
"loss": 13.4016,
"step": 107
},
{
"epoch": 0.0063827901066753345,
"grad_norm": 3198.798095703125,
"learning_rate": 4.7520812266338885e-05,
"loss": 11.9692,
"step": 108
},
{
"epoch": 0.006441890015070477,
"grad_norm": 6041.2138671875,
"learning_rate": 4.669547078371504e-05,
"loss": 12.9008,
"step": 109
},
{
"epoch": 0.006500989923465619,
"grad_norm": 3914.65185546875,
"learning_rate": 4.5871032726383386e-05,
"loss": 13.4266,
"step": 110
},
{
"epoch": 0.006560089831860761,
"grad_norm": 2149.8701171875,
"learning_rate": 4.504772348747687e-05,
"loss": 13.9688,
"step": 111
},
{
"epoch": 0.006619189740255903,
"grad_norm": 1904.322509765625,
"learning_rate": 4.4225768151520694e-05,
"loss": 12.669,
"step": 112
},
{
"epoch": 0.006678289648651045,
"grad_norm": 2146.552734375,
"learning_rate": 4.3405391432896555e-05,
"loss": 12.6239,
"step": 113
},
{
"epoch": 0.006737389557046187,
"grad_norm": 1337.752685546875,
"learning_rate": 4.2586817614407895e-05,
"loss": 11.5385,
"step": 114
},
{
"epoch": 0.006796489465441329,
"grad_norm": 1304.51513671875,
"learning_rate": 4.17702704859633e-05,
"loss": 13.6479,
"step": 115
},
{
"epoch": 0.006855589373836471,
"grad_norm": 2028.5693359375,
"learning_rate": 4.095597328339452e-05,
"loss": 12.5706,
"step": 116
},
{
"epoch": 0.006914689282231612,
"grad_norm": 1726.325927734375,
"learning_rate": 4.0144148627425993e-05,
"loss": 11.7618,
"step": 117
},
{
"epoch": 0.006973789190626754,
"grad_norm": 1568.72900390625,
"learning_rate": 3.933501846281267e-05,
"loss": 12.5052,
"step": 118
},
{
"epoch": 0.007032889099021896,
"grad_norm": 1082.642578125,
"learning_rate": 3.852880399766243e-05,
"loss": 13.6664,
"step": 119
},
{
"epoch": 0.007091989007417038,
"grad_norm": 3831.130126953125,
"learning_rate": 3.772572564296005e-05,
"loss": 12.6727,
"step": 120
},
{
"epoch": 0.00715108891581218,
"grad_norm": 1993.9852294921875,
"learning_rate": 3.6926002952309016e-05,
"loss": 12.7497,
"step": 121
},
{
"epoch": 0.007210188824207322,
"grad_norm": 6213.880859375,
"learning_rate": 3.612985456190778e-05,
"loss": 11.8832,
"step": 122
},
{
"epoch": 0.007269288732602464,
"grad_norm": 1710.1378173828125,
"learning_rate": 3.533749813077677e-05,
"loss": 11.4527,
"step": 123
},
{
"epoch": 0.0073283886409976065,
"grad_norm": 1587.0687255859375,
"learning_rate": 3.4549150281252636e-05,
"loss": 12.9201,
"step": 124
},
{
"epoch": 0.0073874885493927485,
"grad_norm": 1373.242919921875,
"learning_rate": 3.3765026539765834e-05,
"loss": 13.4101,
"step": 125
},
{
"epoch": 0.007446588457787891,
"grad_norm": 2989.611083984375,
"learning_rate": 3.298534127791785e-05,
"loss": 13.1978,
"step": 126
},
{
"epoch": 0.007505688366183033,
"grad_norm": 1946.5069580078125,
"learning_rate": 3.221030765387417e-05,
"loss": 12.5673,
"step": 127
},
{
"epoch": 0.007564788274578175,
"grad_norm": 3161.564453125,
"learning_rate": 3.144013755408895e-05,
"loss": 13.4249,
"step": 128
},
{
"epoch": 0.007623888182973317,
"grad_norm": 2255.56591796875,
"learning_rate": 3.0675041535377405e-05,
"loss": 12.1847,
"step": 129
},
{
"epoch": 0.007682988091368459,
"grad_norm": 2352.259033203125,
"learning_rate": 2.991522876735154e-05,
"loss": 13.2541,
"step": 130
},
{
"epoch": 0.0077420879997636,
"grad_norm": 1452.216796875,
"learning_rate": 2.916090697523549e-05,
"loss": 12.5034,
"step": 131
},
{
"epoch": 0.007801187908158742,
"grad_norm": 3818.702392578125,
"learning_rate": 2.8412282383075363e-05,
"loss": 12.3143,
"step": 132
},
{
"epoch": 0.007860287816553884,
"grad_norm": 1497.1851806640625,
"learning_rate": 2.766955965735968e-05,
"loss": 13.3143,
"step": 133
},
{
"epoch": 0.007919387724949026,
"grad_norm": 1135.6204833984375,
"learning_rate": 2.693294185106562e-05,
"loss": 11.5733,
"step": 134
},
{
"epoch": 0.007978487633344168,
"grad_norm": 628.8428955078125,
"learning_rate": 2.6202630348146324e-05,
"loss": 10.7102,
"step": 135
},
{
"epoch": 0.00803758754173931,
"grad_norm": 1791.0352783203125,
"learning_rate": 2.547882480847461e-05,
"loss": 12.5881,
"step": 136
},
{
"epoch": 0.008096687450134452,
"grad_norm": 1191.6285400390625,
"learning_rate": 2.476172311325783e-05,
"loss": 12.8862,
"step": 137
},
{
"epoch": 0.008155787358529594,
"grad_norm": 972.80712890625,
"learning_rate": 2.405152131093926e-05,
"loss": 11.3009,
"step": 138
},
{
"epoch": 0.008214887266924736,
"grad_norm": 1287.54345703125,
"learning_rate": 2.3348413563600325e-05,
"loss": 12.695,
"step": 139
},
{
"epoch": 0.008273987175319878,
"grad_norm": 1702.5732421875,
"learning_rate": 2.2652592093878666e-05,
"loss": 11.7642,
"step": 140
},
{
"epoch": 0.00833308708371502,
"grad_norm": 993.0706176757812,
"learning_rate": 2.196424713241637e-05,
"loss": 12.2068,
"step": 141
},
{
"epoch": 0.008392186992110163,
"grad_norm": 1540.261962890625,
"learning_rate": 2.128356686585282e-05,
"loss": 11.7752,
"step": 142
},
{
"epoch": 0.008451286900505305,
"grad_norm": 2070.322021484375,
"learning_rate": 2.061073738537635e-05,
"loss": 11.652,
"step": 143
},
{
"epoch": 0.008510386808900447,
"grad_norm": 1121.17578125,
"learning_rate": 1.9945942635848748e-05,
"loss": 11.9319,
"step": 144
},
{
"epoch": 0.008569486717295589,
"grad_norm": 1420.2528076171875,
"learning_rate": 1.928936436551661e-05,
"loss": 11.1072,
"step": 145
},
{
"epoch": 0.00862858662569073,
"grad_norm": 1496.6824951171875,
"learning_rate": 1.8641182076323148e-05,
"loss": 12.0373,
"step": 146
},
{
"epoch": 0.008687686534085873,
"grad_norm": 698.1117553710938,
"learning_rate": 1.800157297483417e-05,
"loss": 11.1577,
"step": 147
},
{
"epoch": 0.008746786442481015,
"grad_norm": 2615.68115234375,
"learning_rate": 1.7370711923791567e-05,
"loss": 12.5535,
"step": 148
},
{
"epoch": 0.008805886350876157,
"grad_norm": 1780.2978515625,
"learning_rate": 1.6748771394307585e-05,
"loss": 12.251,
"step": 149
},
{
"epoch": 0.008864986259271299,
"grad_norm": 1657.8831787109375,
"learning_rate": 1.6135921418712956e-05,
"loss": 13.837,
"step": 150
},
{
"epoch": 0.008864986259271299,
"eval_loss": 3.0583012104034424,
"eval_runtime": 161.2511,
"eval_samples_per_second": 176.731,
"eval_steps_per_second": 44.186,
"step": 150
},
{
"epoch": 0.008924086167666441,
"grad_norm": 4426.115234375,
"learning_rate": 1.553232954407171e-05,
"loss": 12.1362,
"step": 151
},
{
"epoch": 0.008983186076061581,
"grad_norm": 2777.5380859375,
"learning_rate": 1.4938160786375572e-05,
"loss": 11.8313,
"step": 152
},
{
"epoch": 0.009042285984456723,
"grad_norm": 2583.085205078125,
"learning_rate": 1.435357758543015e-05,
"loss": 12.2082,
"step": 153
},
{
"epoch": 0.009101385892851865,
"grad_norm": 3110.58935546875,
"learning_rate": 1.3778739760445552e-05,
"loss": 13.014,
"step": 154
},
{
"epoch": 0.009160485801247007,
"grad_norm": 5781.47607421875,
"learning_rate": 1.3213804466343421e-05,
"loss": 13.058,
"step": 155
},
{
"epoch": 0.00921958570964215,
"grad_norm": 2532.0634765625,
"learning_rate": 1.2658926150792322e-05,
"loss": 11.984,
"step": 156
},
{
"epoch": 0.009278685618037292,
"grad_norm": 1893.6016845703125,
"learning_rate": 1.2114256511983274e-05,
"loss": 13.2671,
"step": 157
},
{
"epoch": 0.009337785526432434,
"grad_norm": 2647.750732421875,
"learning_rate": 1.157994445715706e-05,
"loss": 12.3813,
"step": 158
},
{
"epoch": 0.009396885434827576,
"grad_norm": 3568.959716796875,
"learning_rate": 1.1056136061894384e-05,
"loss": 13.6605,
"step": 159
},
{
"epoch": 0.009455985343222718,
"grad_norm": 2706.551513671875,
"learning_rate": 1.0542974530180327e-05,
"loss": 12.8709,
"step": 160
},
{
"epoch": 0.00951508525161786,
"grad_norm": 1961.4954833984375,
"learning_rate": 1.0040600155253765e-05,
"loss": 12.6101,
"step": 161
},
{
"epoch": 0.009574185160013002,
"grad_norm": 1534.4534912109375,
"learning_rate": 9.549150281252633e-06,
"loss": 12.9987,
"step": 162
},
{
"epoch": 0.009633285068408144,
"grad_norm": 1704.029296875,
"learning_rate": 9.068759265665384e-06,
"loss": 12.3731,
"step": 163
},
{
"epoch": 0.009692384976803286,
"grad_norm": 2597.5283203125,
"learning_rate": 8.599558442598998e-06,
"loss": 12.8223,
"step": 164
},
{
"epoch": 0.009751484885198428,
"grad_norm": 1731.814453125,
"learning_rate": 8.141676086873572e-06,
"loss": 12.4875,
"step": 165
},
{
"epoch": 0.00981058479359357,
"grad_norm": 2835.732177734375,
"learning_rate": 7.695237378953223e-06,
"loss": 11.8968,
"step": 166
},
{
"epoch": 0.009869684701988712,
"grad_norm": 1887.0316162109375,
"learning_rate": 7.260364370723044e-06,
"loss": 13.3503,
"step": 167
},
{
"epoch": 0.009928784610383854,
"grad_norm": 1545.7318115234375,
"learning_rate": 6.837175952121306e-06,
"loss": 12.5817,
"step": 168
},
{
"epoch": 0.009987884518778996,
"grad_norm": 999.0375366210938,
"learning_rate": 6.425787818636131e-06,
"loss": 11.8021,
"step": 169
},
{
"epoch": 0.010046984427174138,
"grad_norm": 1047.11083984375,
"learning_rate": 6.026312439675552e-06,
"loss": 12.902,
"step": 170
},
{
"epoch": 0.01010608433556928,
"grad_norm": 1775.0853271484375,
"learning_rate": 5.6388590278194096e-06,
"loss": 12.2674,
"step": 171
},
{
"epoch": 0.010165184243964422,
"grad_norm": 1113.0733642578125,
"learning_rate": 5.263533508961827e-06,
"loss": 11.7824,
"step": 172
},
{
"epoch": 0.010224284152359564,
"grad_norm": 2153.2080078125,
"learning_rate": 4.900438493352055e-06,
"loss": 11.6983,
"step": 173
},
{
"epoch": 0.010283384060754706,
"grad_norm": 2669.885498046875,
"learning_rate": 4.549673247541875e-06,
"loss": 11.7384,
"step": 174
},
{
"epoch": 0.010342483969149848,
"grad_norm": 1753.4554443359375,
"learning_rate": 4.2113336672471245e-06,
"loss": 12.7409,
"step": 175
},
{
"epoch": 0.01040158387754499,
"grad_norm": 1929.54931640625,
"learning_rate": 3.885512251130763e-06,
"loss": 11.1606,
"step": 176
},
{
"epoch": 0.010460683785940133,
"grad_norm": 1568.9322509765625,
"learning_rate": 3.5722980755146517e-06,
"loss": 13.9085,
"step": 177
},
{
"epoch": 0.010519783694335275,
"grad_norm": 1925.484130859375,
"learning_rate": 3.271776770026963e-06,
"loss": 12.5627,
"step": 178
},
{
"epoch": 0.010578883602730415,
"grad_norm": 1441.8690185546875,
"learning_rate": 2.9840304941919415e-06,
"loss": 12.2827,
"step": 179
},
{
"epoch": 0.010637983511125557,
"grad_norm": 880.8626098632812,
"learning_rate": 2.7091379149682685e-06,
"loss": 12.22,
"step": 180
},
{
"epoch": 0.010697083419520699,
"grad_norm": 1387.427001953125,
"learning_rate": 2.4471741852423237e-06,
"loss": 11.9367,
"step": 181
},
{
"epoch": 0.010756183327915841,
"grad_norm": 1900.5921630859375,
"learning_rate": 2.1982109232821178e-06,
"loss": 11.8286,
"step": 182
},
{
"epoch": 0.010815283236310983,
"grad_norm": 2867.70703125,
"learning_rate": 1.962316193157593e-06,
"loss": 12.1023,
"step": 183
},
{
"epoch": 0.010874383144706125,
"grad_norm": 1134.2127685546875,
"learning_rate": 1.7395544861325718e-06,
"loss": 13.4676,
"step": 184
},
{
"epoch": 0.010933483053101267,
"grad_norm": 1464.923095703125,
"learning_rate": 1.5299867030334814e-06,
"loss": 12.2917,
"step": 185
},
{
"epoch": 0.01099258296149641,
"grad_norm": 1759.43896484375,
"learning_rate": 1.333670137599713e-06,
"loss": 11.2761,
"step": 186
},
{
"epoch": 0.011051682869891551,
"grad_norm": 1418.9720458984375,
"learning_rate": 1.1506584608200367e-06,
"loss": 12.2346,
"step": 187
},
{
"epoch": 0.011110782778286693,
"grad_norm": 1523.892822265625,
"learning_rate": 9.810017062595322e-07,
"loss": 13.3752,
"step": 188
},
{
"epoch": 0.011169882686681835,
"grad_norm": 1805.3973388671875,
"learning_rate": 8.247462563808817e-07,
"loss": 12.3709,
"step": 189
},
{
"epoch": 0.011228982595076977,
"grad_norm": 1991.1004638671875,
"learning_rate": 6.819348298638839e-07,
"loss": 11.3256,
"step": 190
},
{
"epoch": 0.01128808250347212,
"grad_norm": 733.4898071289062,
"learning_rate": 5.526064699265753e-07,
"loss": 13.5104,
"step": 191
},
{
"epoch": 0.011347182411867262,
"grad_norm": 865.744140625,
"learning_rate": 4.367965336512403e-07,
"loss": 11.7907,
"step": 192
},
{
"epoch": 0.011406282320262404,
"grad_norm": 1221.997314453125,
"learning_rate": 3.3453668231809286e-07,
"loss": 13.722,
"step": 193
},
{
"epoch": 0.011465382228657546,
"grad_norm": 1322.4814453125,
"learning_rate": 2.458548727494292e-07,
"loss": 11.0568,
"step": 194
},
{
"epoch": 0.011524482137052688,
"grad_norm": 1005.72607421875,
"learning_rate": 1.7077534966650766e-07,
"loss": 12.7148,
"step": 195
},
{
"epoch": 0.01158358204544783,
"grad_norm": 994.3577880859375,
"learning_rate": 1.0931863906127327e-07,
"loss": 12.0398,
"step": 196
},
{
"epoch": 0.011642681953842972,
"grad_norm": 2119.787841796875,
"learning_rate": 6.150154258476315e-08,
"loss": 12.6347,
"step": 197
},
{
"epoch": 0.011701781862238114,
"grad_norm": 734.9210205078125,
"learning_rate": 2.7337132953697554e-08,
"loss": 11.8656,
"step": 198
},
{
"epoch": 0.011760881770633256,
"grad_norm": 1884.077880859375,
"learning_rate": 6.834750376549792e-09,
"loss": 12.5588,
"step": 199
},
{
"epoch": 0.011819981679028398,
"grad_norm": 1439.942138671875,
"learning_rate": 0.0,
"loss": 14.5914,
"step": 200
},
{
"epoch": 0.011819981679028398,
"eval_loss": 3.0534605979919434,
"eval_runtime": 161.3054,
"eval_samples_per_second": 176.671,
"eval_steps_per_second": 44.171,
"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": 5234753116569600.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}