| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.10739294134205989, | |
| "eval_steps": 500, | |
| "global_step": 3052, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010908195221506739, | |
| "grad_norm": 27.24474334716797, | |
| "learning_rate": 1.0157273918741808e-06, | |
| "loss": 9.0396, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0021816390443013477, | |
| "grad_norm": 16.773351669311523, | |
| "learning_rate": 2.0314547837483616e-06, | |
| "loss": 7.9471, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0032724585664520214, | |
| "grad_norm": 15.084428787231445, | |
| "learning_rate": 3.0471821756225426e-06, | |
| "loss": 6.5868, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0043632780886026954, | |
| "grad_norm": 6.7851786613464355, | |
| "learning_rate": 4.062909567496723e-06, | |
| "loss": 5.3744, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.0054540976107533695, | |
| "grad_norm": 7.081949234008789, | |
| "learning_rate": 5.078636959370905e-06, | |
| "loss": 4.6378, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.006544917132904043, | |
| "grad_norm": 8.945405960083008, | |
| "learning_rate": 6.094364351245085e-06, | |
| "loss": 4.2292, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.007635736655054717, | |
| "grad_norm": 15.078816413879395, | |
| "learning_rate": 7.110091743119267e-06, | |
| "loss": 3.9767, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.008726556177205391, | |
| "grad_norm": 8.799291610717773, | |
| "learning_rate": 8.125819134993446e-06, | |
| "loss": 3.765, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.009817375699356064, | |
| "grad_norm": 9.741459846496582, | |
| "learning_rate": 9.141546526867629e-06, | |
| "loss": 3.6496, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.010908195221506739, | |
| "grad_norm": 9.403132438659668, | |
| "learning_rate": 1.015727391874181e-05, | |
| "loss": 3.4936, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.011999014743657412, | |
| "grad_norm": 8.479109764099121, | |
| "learning_rate": 1.117300131061599e-05, | |
| "loss": 3.3879, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.013089834265808085, | |
| "grad_norm": 4.988598823547363, | |
| "learning_rate": 1.218872870249017e-05, | |
| "loss": 3.3147, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.01418065378795876, | |
| "grad_norm": 5.518200874328613, | |
| "learning_rate": 1.3204456094364351e-05, | |
| "loss": 3.2022, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.015271473310109434, | |
| "grad_norm": 5.685606002807617, | |
| "learning_rate": 1.4220183486238533e-05, | |
| "loss": 3.1393, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.01636229283226011, | |
| "grad_norm": 5.6373491287231445, | |
| "learning_rate": 1.5235910878112714e-05, | |
| "loss": 3.05, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.017453112354410782, | |
| "grad_norm": 5.174935340881348, | |
| "learning_rate": 1.6251638269986893e-05, | |
| "loss": 2.992, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.018543931876561455, | |
| "grad_norm": 5.266006946563721, | |
| "learning_rate": 1.7267365661861077e-05, | |
| "loss": 2.9558, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.019634751398712128, | |
| "grad_norm": 4.111412048339844, | |
| "learning_rate": 1.8283093053735257e-05, | |
| "loss": 2.9089, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.0207255709208628, | |
| "grad_norm": 4.149659156799316, | |
| "learning_rate": 1.9298820445609438e-05, | |
| "loss": 2.879, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.021816390443013478, | |
| "grad_norm": 4.711772441864014, | |
| "learning_rate": 2.031454783748362e-05, | |
| "loss": 2.8057, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.02290720996516415, | |
| "grad_norm": 4.372698783874512, | |
| "learning_rate": 2.13302752293578e-05, | |
| "loss": 2.7954, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.023998029487314824, | |
| "grad_norm": 3.9052813053131104, | |
| "learning_rate": 2.234600262123198e-05, | |
| "loss": 2.7286, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.025088849009465498, | |
| "grad_norm": 3.544926404953003, | |
| "learning_rate": 2.336173001310616e-05, | |
| "loss": 2.6875, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.02617966853161617, | |
| "grad_norm": 4.380484104156494, | |
| "learning_rate": 2.437745740498034e-05, | |
| "loss": 2.6703, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.027270488053766848, | |
| "grad_norm": 4.843706130981445, | |
| "learning_rate": 2.5393184796854525e-05, | |
| "loss": 2.6387, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.02836130757591752, | |
| "grad_norm": 3.555110454559326, | |
| "learning_rate": 2.6408912188728702e-05, | |
| "loss": 2.5996, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.029452127098068194, | |
| "grad_norm": 3.3113982677459717, | |
| "learning_rate": 2.7424639580602886e-05, | |
| "loss": 2.5823, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.030542946620218867, | |
| "grad_norm": 3.4552953243255615, | |
| "learning_rate": 2.8440366972477066e-05, | |
| "loss": 2.5643, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.03163376614236954, | |
| "grad_norm": 2.9786770343780518, | |
| "learning_rate": 2.9456094364351244e-05, | |
| "loss": 2.5426, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.03272458566452022, | |
| "grad_norm": 3.227999210357666, | |
| "learning_rate": 3.0471821756225428e-05, | |
| "loss": 2.4834, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.03381540518667089, | |
| "grad_norm": 3.0225250720977783, | |
| "learning_rate": 3.148754914809961e-05, | |
| "loss": 2.4755, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.034906224708821564, | |
| "grad_norm": 2.6292648315429688, | |
| "learning_rate": 3.2503276539973785e-05, | |
| "loss": 2.4393, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.03599704423097224, | |
| "grad_norm": 2.767667055130005, | |
| "learning_rate": 3.351900393184797e-05, | |
| "loss": 2.4124, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.03708786375312291, | |
| "grad_norm": 2.6588900089263916, | |
| "learning_rate": 3.453473132372215e-05, | |
| "loss": 2.3551, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.03817868327527359, | |
| "grad_norm": 2.7497403621673584, | |
| "learning_rate": 3.555045871559633e-05, | |
| "loss": 2.3576, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.039269502797424256, | |
| "grad_norm": 2.5148332118988037, | |
| "learning_rate": 3.6566186107470514e-05, | |
| "loss": 2.3078, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.04036032231957493, | |
| "grad_norm": 3.0025594234466553, | |
| "learning_rate": 3.7581913499344695e-05, | |
| "loss": 2.303, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 0.0414511418417256, | |
| "grad_norm": 2.6871118545532227, | |
| "learning_rate": 3.8597640891218876e-05, | |
| "loss": 2.3062, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 0.04254196136387628, | |
| "grad_norm": 10.630929946899414, | |
| "learning_rate": 3.9613368283093056e-05, | |
| "loss": 2.307, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 0.043632780886026956, | |
| "grad_norm": 2.571220874786377, | |
| "learning_rate": 4.062909567496724e-05, | |
| "loss": 2.2837, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.044723600408177626, | |
| "grad_norm": 3.5510289669036865, | |
| "learning_rate": 4.164482306684142e-05, | |
| "loss": 2.2797, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 0.0458144199303283, | |
| "grad_norm": 2.794853925704956, | |
| "learning_rate": 4.26605504587156e-05, | |
| "loss": 2.2797, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 0.04690523945247897, | |
| "grad_norm": 6.132640838623047, | |
| "learning_rate": 4.367627785058978e-05, | |
| "loss": 2.2504, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 0.04799605897462965, | |
| "grad_norm": 6.17495584487915, | |
| "learning_rate": 4.469200524246396e-05, | |
| "loss": 2.2414, | |
| "step": 1364 | |
| }, | |
| { | |
| "epoch": 0.049086878496780326, | |
| "grad_norm": 2.3518564701080322, | |
| "learning_rate": 4.570773263433814e-05, | |
| "loss": 2.2198, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.050177698018930995, | |
| "grad_norm": 2.2073466777801514, | |
| "learning_rate": 4.672346002621232e-05, | |
| "loss": 2.2057, | |
| "step": 1426 | |
| }, | |
| { | |
| "epoch": 0.05126851754108167, | |
| "grad_norm": 2.018749237060547, | |
| "learning_rate": 4.77391874180865e-05, | |
| "loss": 2.1965, | |
| "step": 1457 | |
| }, | |
| { | |
| "epoch": 0.05235933706323234, | |
| "grad_norm": 2.7125244140625, | |
| "learning_rate": 4.875491480996068e-05, | |
| "loss": 2.19, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 0.05345015658538302, | |
| "grad_norm": 1.9725641012191772, | |
| "learning_rate": 4.977064220183487e-05, | |
| "loss": 2.1669, | |
| "step": 1519 | |
| }, | |
| { | |
| "epoch": 0.054540976107533695, | |
| "grad_norm": 2.0198206901550293, | |
| "learning_rate": 4.9999915451558777e-05, | |
| "loss": 2.1486, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.055631795629684365, | |
| "grad_norm": 2.068424940109253, | |
| "learning_rate": 4.999955597496219e-05, | |
| "loss": 2.1476, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 0.05672261515183504, | |
| "grad_norm": 1.8509161472320557, | |
| "learning_rate": 4.9998914381774255e-05, | |
| "loss": 2.1129, | |
| "step": 1612 | |
| }, | |
| { | |
| "epoch": 0.05781343467398571, | |
| "grad_norm": 1.805915355682373, | |
| "learning_rate": 4.999799067923527e-05, | |
| "loss": 2.1014, | |
| "step": 1643 | |
| }, | |
| { | |
| "epoch": 0.05890425419613639, | |
| "grad_norm": 1.9447969198226929, | |
| "learning_rate": 4.999678487776908e-05, | |
| "loss": 2.1067, | |
| "step": 1674 | |
| }, | |
| { | |
| "epoch": 0.059995073718287065, | |
| "grad_norm": 2.131700277328491, | |
| "learning_rate": 4.9995296990983006e-05, | |
| "loss": 2.0817, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.061085893240437734, | |
| "grad_norm": 1.9113484621047974, | |
| "learning_rate": 4.999352703566763e-05, | |
| "loss": 2.067, | |
| "step": 1736 | |
| }, | |
| { | |
| "epoch": 0.06217671276258841, | |
| "grad_norm": 1.8802495002746582, | |
| "learning_rate": 4.999147503179668e-05, | |
| "loss": 2.0617, | |
| "step": 1767 | |
| }, | |
| { | |
| "epoch": 0.06326753228473908, | |
| "grad_norm": 1.6704238653182983, | |
| "learning_rate": 4.998914100252672e-05, | |
| "loss": 2.0671, | |
| "step": 1798 | |
| }, | |
| { | |
| "epoch": 0.06435835180688976, | |
| "grad_norm": 4.303380966186523, | |
| "learning_rate": 4.998652497419696e-05, | |
| "loss": 2.0436, | |
| "step": 1829 | |
| }, | |
| { | |
| "epoch": 0.06544917132904043, | |
| "grad_norm": 1.8054529428482056, | |
| "learning_rate": 4.9983626976328927e-05, | |
| "loss": 2.0499, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.06653999085119111, | |
| "grad_norm": 1.663872241973877, | |
| "learning_rate": 4.998044704162613e-05, | |
| "loss": 2.0227, | |
| "step": 1891 | |
| }, | |
| { | |
| "epoch": 0.06763081037334177, | |
| "grad_norm": 10.130581855773926, | |
| "learning_rate": 4.9976985205973705e-05, | |
| "loss": 2.0022, | |
| "step": 1922 | |
| }, | |
| { | |
| "epoch": 0.06872162989549245, | |
| "grad_norm": 3.7727627754211426, | |
| "learning_rate": 4.997324150843799e-05, | |
| "loss": 2.0172, | |
| "step": 1953 | |
| }, | |
| { | |
| "epoch": 0.06981244941764313, | |
| "grad_norm": 1.8669508695602417, | |
| "learning_rate": 4.99692159912661e-05, | |
| "loss": 2.0214, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 0.0709032689397938, | |
| "grad_norm": 17.45224952697754, | |
| "learning_rate": 4.996490869988546e-05, | |
| "loss": 1.9959, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 0.07199408846194448, | |
| "grad_norm": 1.7007793188095093, | |
| "learning_rate": 4.996031968290326e-05, | |
| "loss": 2.0139, | |
| "step": 2046 | |
| }, | |
| { | |
| "epoch": 0.07308490798409514, | |
| "grad_norm": 1.6912076473236084, | |
| "learning_rate": 4.995544899210594e-05, | |
| "loss": 1.9963, | |
| "step": 2077 | |
| }, | |
| { | |
| "epoch": 0.07417572750624582, | |
| "grad_norm": 1.6827666759490967, | |
| "learning_rate": 4.9950296682458583e-05, | |
| "loss": 1.973, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 0.0752665470283965, | |
| "grad_norm": 1.767269253730774, | |
| "learning_rate": 4.994486281210429e-05, | |
| "loss": 1.9949, | |
| "step": 2139 | |
| }, | |
| { | |
| "epoch": 0.07635736655054717, | |
| "grad_norm": 7.507472038269043, | |
| "learning_rate": 4.9939147442363566e-05, | |
| "loss": 1.9837, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.07744818607269784, | |
| "grad_norm": 1.6840234994888306, | |
| "learning_rate": 4.9933150637733574e-05, | |
| "loss": 1.9789, | |
| "step": 2201 | |
| }, | |
| { | |
| "epoch": 0.07853900559484851, | |
| "grad_norm": 1.6442159414291382, | |
| "learning_rate": 4.992687246588743e-05, | |
| "loss": 1.9436, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 0.07962982511699919, | |
| "grad_norm": 1.546859622001648, | |
| "learning_rate": 4.992031299767347e-05, | |
| "loss": 1.9341, | |
| "step": 2263 | |
| }, | |
| { | |
| "epoch": 0.08072064463914987, | |
| "grad_norm": 1.4932371377944946, | |
| "learning_rate": 4.9913472307114386e-05, | |
| "loss": 1.9583, | |
| "step": 2294 | |
| }, | |
| { | |
| "epoch": 0.08181146416130054, | |
| "grad_norm": 1.4531840085983276, | |
| "learning_rate": 4.9906350471406446e-05, | |
| "loss": 1.9414, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.0829022836834512, | |
| "grad_norm": 1.5285487174987793, | |
| "learning_rate": 4.989894757091861e-05, | |
| "loss": 1.9502, | |
| "step": 2356 | |
| }, | |
| { | |
| "epoch": 0.08399310320560188, | |
| "grad_norm": 1.4159945249557495, | |
| "learning_rate": 4.989126368919158e-05, | |
| "loss": 1.9299, | |
| "step": 2387 | |
| }, | |
| { | |
| "epoch": 0.08508392272775256, | |
| "grad_norm": 1.4361426830291748, | |
| "learning_rate": 4.988329891293693e-05, | |
| "loss": 1.9266, | |
| "step": 2418 | |
| }, | |
| { | |
| "epoch": 0.08617474224990324, | |
| "grad_norm": 1.4310742616653442, | |
| "learning_rate": 4.987505333203608e-05, | |
| "loss": 1.9116, | |
| "step": 2449 | |
| }, | |
| { | |
| "epoch": 0.08726556177205391, | |
| "grad_norm": 1.3780311346054077, | |
| "learning_rate": 4.9866527039539276e-05, | |
| "loss": 1.9189, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.08835638129420458, | |
| "grad_norm": 1.3917787075042725, | |
| "learning_rate": 4.9857720131664594e-05, | |
| "loss": 1.9107, | |
| "step": 2511 | |
| }, | |
| { | |
| "epoch": 0.08944720081635525, | |
| "grad_norm": 1.3963998556137085, | |
| "learning_rate": 4.9848632707796773e-05, | |
| "loss": 1.899, | |
| "step": 2542 | |
| }, | |
| { | |
| "epoch": 0.09053802033850593, | |
| "grad_norm": 1.3573299646377563, | |
| "learning_rate": 4.9839264870486155e-05, | |
| "loss": 1.9084, | |
| "step": 2573 | |
| }, | |
| { | |
| "epoch": 0.0916288398606566, | |
| "grad_norm": 1.362552523612976, | |
| "learning_rate": 4.9829616725447526e-05, | |
| "loss": 1.8954, | |
| "step": 2604 | |
| }, | |
| { | |
| "epoch": 0.09271965938280728, | |
| "grad_norm": 1.3538340330123901, | |
| "learning_rate": 4.981968838155888e-05, | |
| "loss": 1.9035, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 0.09381047890495794, | |
| "grad_norm": 1.407291054725647, | |
| "learning_rate": 4.980947995086024e-05, | |
| "loss": 1.8778, | |
| "step": 2666 | |
| }, | |
| { | |
| "epoch": 0.09490129842710862, | |
| "grad_norm": 1.3554202318191528, | |
| "learning_rate": 4.979899154855234e-05, | |
| "loss": 1.8783, | |
| "step": 2697 | |
| }, | |
| { | |
| "epoch": 0.0959921179492593, | |
| "grad_norm": 1.3513182401657104, | |
| "learning_rate": 4.9788223292995386e-05, | |
| "loss": 1.8857, | |
| "step": 2728 | |
| }, | |
| { | |
| "epoch": 0.09708293747140997, | |
| "grad_norm": 1.345706582069397, | |
| "learning_rate": 4.977717530570768e-05, | |
| "loss": 1.8882, | |
| "step": 2759 | |
| }, | |
| { | |
| "epoch": 0.09817375699356065, | |
| "grad_norm": 1.3450254201889038, | |
| "learning_rate": 4.976584771136425e-05, | |
| "loss": 1.8746, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.09926457651571131, | |
| "grad_norm": 1.7820961475372314, | |
| "learning_rate": 4.975424063779547e-05, | |
| "loss": 1.8557, | |
| "step": 2821 | |
| }, | |
| { | |
| "epoch": 0.10035539603786199, | |
| "grad_norm": 1.3493785858154297, | |
| "learning_rate": 4.974235421598557e-05, | |
| "loss": 1.8683, | |
| "step": 2852 | |
| }, | |
| { | |
| "epoch": 0.10144621556001267, | |
| "grad_norm": 1.3066306114196777, | |
| "learning_rate": 4.973018858007122e-05, | |
| "loss": 1.8777, | |
| "step": 2883 | |
| }, | |
| { | |
| "epoch": 0.10253703508216334, | |
| "grad_norm": 1.2991000413894653, | |
| "learning_rate": 4.9717743867339963e-05, | |
| "loss": 1.8605, | |
| "step": 2914 | |
| }, | |
| { | |
| "epoch": 0.10362785460431402, | |
| "grad_norm": 1.3748496770858765, | |
| "learning_rate": 4.9705020218228695e-05, | |
| "loss": 1.8683, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 0.10471867412646468, | |
| "grad_norm": 1.3229252099990845, | |
| "learning_rate": 4.969201777632205e-05, | |
| "loss": 1.858, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 0.10580949364861536, | |
| "grad_norm": 1.2874318361282349, | |
| "learning_rate": 4.9678736688350846e-05, | |
| "loss": 1.8435, | |
| "step": 3007 | |
| }, | |
| { | |
| "epoch": 0.10690031317076604, | |
| "grad_norm": 1.3341971635818481, | |
| "learning_rate": 4.966517710419033e-05, | |
| "loss": 1.8663, | |
| "step": 3038 | |
| } | |
| ], | |
| "logging_steps": 31, | |
| "max_steps": 30517, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 3052, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.131685419530584e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |