| { | |
| "best_global_step": 3000, | |
| "best_metric": 0.9244844913482666, | |
| "best_model_checkpoint": "/workspace/runs_coder_noed_10/checkpoint-3000", | |
| "epoch": 0.7334963325183375, | |
| "eval_steps": 1000, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0024449877750611247, | |
| "grad_norm": 0.7603356838226318, | |
| "learning_rate": 9.988997555012225e-06, | |
| "loss": 2.0597, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004889975550122249, | |
| "grad_norm": 0.491472065448761, | |
| "learning_rate": 9.97677261613692e-06, | |
| "loss": 1.8106, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.007334963325183374, | |
| "grad_norm": 0.421262264251709, | |
| "learning_rate": 9.964547677261615e-06, | |
| "loss": 1.6648, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.009779951100244499, | |
| "grad_norm": 0.3425203263759613, | |
| "learning_rate": 9.95232273838631e-06, | |
| "loss": 1.552, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.012224938875305624, | |
| "grad_norm": 0.3278200328350067, | |
| "learning_rate": 9.940097799511004e-06, | |
| "loss": 1.4701, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.014669926650366748, | |
| "grad_norm": 0.3467581868171692, | |
| "learning_rate": 9.927872860635697e-06, | |
| "loss": 1.3899, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.017114914425427872, | |
| "grad_norm": 0.3279109001159668, | |
| "learning_rate": 9.915647921760392e-06, | |
| "loss": 1.3596, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.019559902200488997, | |
| "grad_norm": 0.3338731825351715, | |
| "learning_rate": 9.903422982885086e-06, | |
| "loss": 1.335, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.022004889975550123, | |
| "grad_norm": 0.39886194467544556, | |
| "learning_rate": 9.891198044009781e-06, | |
| "loss": 1.3055, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02444987775061125, | |
| "grad_norm": 0.3964369297027588, | |
| "learning_rate": 9.878973105134476e-06, | |
| "loss": 1.2911, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02689486552567237, | |
| "grad_norm": 0.42521995306015015, | |
| "learning_rate": 9.866748166259169e-06, | |
| "loss": 1.2764, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.029339853300733496, | |
| "grad_norm": 0.40386906266212463, | |
| "learning_rate": 9.854523227383865e-06, | |
| "loss": 1.2573, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03178484107579462, | |
| "grad_norm": 0.42052966356277466, | |
| "learning_rate": 9.842298288508558e-06, | |
| "loss": 1.2147, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.034229828850855744, | |
| "grad_norm": 0.4453684091567993, | |
| "learning_rate": 9.830073349633253e-06, | |
| "loss": 1.2383, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03667481662591687, | |
| "grad_norm": 0.4648647606372833, | |
| "learning_rate": 9.817848410757947e-06, | |
| "loss": 1.2257, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.039119804400977995, | |
| "grad_norm": 0.4787471890449524, | |
| "learning_rate": 9.805623471882642e-06, | |
| "loss": 1.1803, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.04156479217603912, | |
| "grad_norm": 0.5450437068939209, | |
| "learning_rate": 9.793398533007335e-06, | |
| "loss": 1.1454, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.044009779951100246, | |
| "grad_norm": 0.4808744192123413, | |
| "learning_rate": 9.78117359413203e-06, | |
| "loss": 1.1936, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.04645476772616137, | |
| "grad_norm": 0.48509007692337036, | |
| "learning_rate": 9.768948655256724e-06, | |
| "loss": 1.1674, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.0488997555012225, | |
| "grad_norm": 0.5013049840927124, | |
| "learning_rate": 9.756723716381419e-06, | |
| "loss": 1.1691, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05134474327628362, | |
| "grad_norm": 0.5472636818885803, | |
| "learning_rate": 9.744498777506112e-06, | |
| "loss": 1.1596, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.05378973105134474, | |
| "grad_norm": 0.5386976599693298, | |
| "learning_rate": 9.732273838630808e-06, | |
| "loss": 1.1449, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.05623471882640587, | |
| "grad_norm": 0.5038230419158936, | |
| "learning_rate": 9.720048899755501e-06, | |
| "loss": 1.1029, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.05867970660146699, | |
| "grad_norm": 0.5687288045883179, | |
| "learning_rate": 9.707823960880196e-06, | |
| "loss": 1.1463, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.061124694376528114, | |
| "grad_norm": 0.53361576795578, | |
| "learning_rate": 9.69559902200489e-06, | |
| "loss": 1.1289, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06356968215158924, | |
| "grad_norm": 0.5407126545906067, | |
| "learning_rate": 9.683374083129585e-06, | |
| "loss": 1.0949, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.06601466992665037, | |
| "grad_norm": 0.6220566630363464, | |
| "learning_rate": 9.67114914425428e-06, | |
| "loss": 1.1131, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.06845965770171149, | |
| "grad_norm": 0.5715101361274719, | |
| "learning_rate": 9.658924205378973e-06, | |
| "loss": 1.108, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.07090464547677261, | |
| "grad_norm": 0.5447026491165161, | |
| "learning_rate": 9.646699266503668e-06, | |
| "loss": 1.0953, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.07334963325183375, | |
| "grad_norm": 0.607760488986969, | |
| "learning_rate": 9.634474327628362e-06, | |
| "loss": 1.0689, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07579462102689487, | |
| "grad_norm": 0.5735874176025391, | |
| "learning_rate": 9.622249388753057e-06, | |
| "loss": 1.0704, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.07823960880195599, | |
| "grad_norm": 0.6558325886726379, | |
| "learning_rate": 9.610024449877752e-06, | |
| "loss": 1.0834, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.08068459657701711, | |
| "grad_norm": 0.644957959651947, | |
| "learning_rate": 9.597799511002447e-06, | |
| "loss": 1.1016, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.08312958435207823, | |
| "grad_norm": 0.4896561801433563, | |
| "learning_rate": 9.58557457212714e-06, | |
| "loss": 1.0945, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.08557457212713937, | |
| "grad_norm": 0.5884416699409485, | |
| "learning_rate": 9.573349633251834e-06, | |
| "loss": 1.1114, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.08801955990220049, | |
| "grad_norm": 0.6569796204566956, | |
| "learning_rate": 9.561124694376529e-06, | |
| "loss": 1.1083, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.09046454767726161, | |
| "grad_norm": 0.6272417306900024, | |
| "learning_rate": 9.548899755501224e-06, | |
| "loss": 1.0743, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.09290953545232274, | |
| "grad_norm": 0.5897823572158813, | |
| "learning_rate": 9.536674816625916e-06, | |
| "loss": 1.0434, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.09535452322738386, | |
| "grad_norm": 0.5946202278137207, | |
| "learning_rate": 9.524449877750613e-06, | |
| "loss": 1.0474, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.097799511002445, | |
| "grad_norm": 0.5851741433143616, | |
| "learning_rate": 9.512224938875306e-06, | |
| "loss": 1.0697, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.10024449877750612, | |
| "grad_norm": 0.7547958493232727, | |
| "learning_rate": 9.5e-06, | |
| "loss": 1.0767, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.10268948655256724, | |
| "grad_norm": 0.651896595954895, | |
| "learning_rate": 9.487775061124695e-06, | |
| "loss": 1.0853, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.10513447432762836, | |
| "grad_norm": 0.6152411103248596, | |
| "learning_rate": 9.47555012224939e-06, | |
| "loss": 1.0582, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.10757946210268948, | |
| "grad_norm": 0.707961916923523, | |
| "learning_rate": 9.463325183374083e-06, | |
| "loss": 1.0658, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.1100244498777506, | |
| "grad_norm": 0.6049492359161377, | |
| "learning_rate": 9.45110024449878e-06, | |
| "loss": 1.0687, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.11246943765281174, | |
| "grad_norm": 0.6512130498886108, | |
| "learning_rate": 9.438875305623472e-06, | |
| "loss": 1.0574, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.11491442542787286, | |
| "grad_norm": 0.6922876834869385, | |
| "learning_rate": 9.426650366748167e-06, | |
| "loss": 1.0311, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.11735941320293398, | |
| "grad_norm": 0.6023642420768738, | |
| "learning_rate": 9.414425427872862e-06, | |
| "loss": 1.0098, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.1198044009779951, | |
| "grad_norm": 0.6482692360877991, | |
| "learning_rate": 9.402200488997556e-06, | |
| "loss": 1.0537, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.12224938875305623, | |
| "grad_norm": 0.6799217462539673, | |
| "learning_rate": 9.38997555012225e-06, | |
| "loss": 1.0474, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12469437652811736, | |
| "grad_norm": 0.7018452286720276, | |
| "learning_rate": 9.377750611246944e-06, | |
| "loss": 1.0273, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.1271393643031785, | |
| "grad_norm": 0.7044438123703003, | |
| "learning_rate": 9.365525672371639e-06, | |
| "loss": 1.0305, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.1295843520782396, | |
| "grad_norm": 0.7802556157112122, | |
| "learning_rate": 9.353300733496333e-06, | |
| "loss": 1.0758, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.13202933985330073, | |
| "grad_norm": 0.6845948100090027, | |
| "learning_rate": 9.341075794621028e-06, | |
| "loss": 1.0158, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.13447432762836187, | |
| "grad_norm": 0.7554173469543457, | |
| "learning_rate": 9.328850855745723e-06, | |
| "loss": 1.0307, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.13691931540342298, | |
| "grad_norm": 0.7574878334999084, | |
| "learning_rate": 9.316625916870417e-06, | |
| "loss": 0.9949, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.1393643031784841, | |
| "grad_norm": 0.6939298510551453, | |
| "learning_rate": 9.30440097799511e-06, | |
| "loss": 1.041, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.14180929095354522, | |
| "grad_norm": 0.6318020224571228, | |
| "learning_rate": 9.292176039119805e-06, | |
| "loss": 1.032, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.14425427872860636, | |
| "grad_norm": 0.7026094794273376, | |
| "learning_rate": 9.2799511002445e-06, | |
| "loss": 1.0642, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.1466992665036675, | |
| "grad_norm": 0.6520543694496155, | |
| "learning_rate": 9.267726161369194e-06, | |
| "loss": 1.0535, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1491442542787286, | |
| "grad_norm": 0.6793085336685181, | |
| "learning_rate": 9.255501222493887e-06, | |
| "loss": 1.0447, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.15158924205378974, | |
| "grad_norm": 0.6935579776763916, | |
| "learning_rate": 9.243276283618584e-06, | |
| "loss": 1.0151, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.15403422982885084, | |
| "grad_norm": 0.767677903175354, | |
| "learning_rate": 9.231051344743277e-06, | |
| "loss": 0.9973, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.15647921760391198, | |
| "grad_norm": 0.7972632646560669, | |
| "learning_rate": 9.218826405867971e-06, | |
| "loss": 0.9839, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.15892420537897312, | |
| "grad_norm": 0.716218113899231, | |
| "learning_rate": 9.206601466992666e-06, | |
| "loss": 0.9674, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.16136919315403422, | |
| "grad_norm": 0.6637817621231079, | |
| "learning_rate": 9.19437652811736e-06, | |
| "loss": 1.0171, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.16381418092909536, | |
| "grad_norm": 0.6555060148239136, | |
| "learning_rate": 9.182151589242054e-06, | |
| "loss": 1.0037, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.16625916870415647, | |
| "grad_norm": 0.7460082769393921, | |
| "learning_rate": 9.16992665036675e-06, | |
| "loss": 1.0329, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.1687041564792176, | |
| "grad_norm": 0.6652795672416687, | |
| "learning_rate": 9.157701711491443e-06, | |
| "loss": 1.041, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.17114914425427874, | |
| "grad_norm": 0.6055278182029724, | |
| "learning_rate": 9.145476772616138e-06, | |
| "loss": 1.0152, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.17359413202933985, | |
| "grad_norm": 0.7681952118873596, | |
| "learning_rate": 9.133251833740832e-06, | |
| "loss": 1.0292, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.17603911980440098, | |
| "grad_norm": 0.8185127973556519, | |
| "learning_rate": 9.121026894865527e-06, | |
| "loss": 1.002, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.1784841075794621, | |
| "grad_norm": 0.6619365811347961, | |
| "learning_rate": 9.10880195599022e-06, | |
| "loss": 1.0032, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.18092909535452323, | |
| "grad_norm": 0.6229087710380554, | |
| "learning_rate": 9.096577017114915e-06, | |
| "loss": 1.0107, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.18337408312958436, | |
| "grad_norm": 0.8737857937812805, | |
| "learning_rate": 9.08435207823961e-06, | |
| "loss": 1.0276, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.18581907090464547, | |
| "grad_norm": 0.7398986220359802, | |
| "learning_rate": 9.072127139364304e-06, | |
| "loss": 1.0185, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.1882640586797066, | |
| "grad_norm": 0.662419855594635, | |
| "learning_rate": 9.059902200488999e-06, | |
| "loss": 1.0028, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.19070904645476772, | |
| "grad_norm": 0.7412256002426147, | |
| "learning_rate": 9.047677261613693e-06, | |
| "loss": 1.0102, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.19315403422982885, | |
| "grad_norm": 0.7194265127182007, | |
| "learning_rate": 9.035452322738388e-06, | |
| "loss": 1.0024, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.19559902200489, | |
| "grad_norm": 0.6712033152580261, | |
| "learning_rate": 9.023227383863081e-06, | |
| "loss": 0.981, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.1980440097799511, | |
| "grad_norm": 0.7297273278236389, | |
| "learning_rate": 9.011002444987776e-06, | |
| "loss": 1.0087, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.20048899755501223, | |
| "grad_norm": 0.7646290063858032, | |
| "learning_rate": 8.99877750611247e-06, | |
| "loss": 1.008, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.20293398533007334, | |
| "grad_norm": 0.6914790272712708, | |
| "learning_rate": 8.986552567237165e-06, | |
| "loss": 0.9967, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.20537897310513448, | |
| "grad_norm": 0.7169461250305176, | |
| "learning_rate": 8.974327628361858e-06, | |
| "loss": 0.9945, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.2078239608801956, | |
| "grad_norm": 0.686245322227478, | |
| "learning_rate": 8.962102689486554e-06, | |
| "loss": 1.003, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.21026894865525672, | |
| "grad_norm": 0.7075187563896179, | |
| "learning_rate": 8.949877750611247e-06, | |
| "loss": 0.9875, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.21271393643031786, | |
| "grad_norm": 0.8134737610816956, | |
| "learning_rate": 8.937652811735942e-06, | |
| "loss": 1.0068, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.21515892420537897, | |
| "grad_norm": 0.6891161799430847, | |
| "learning_rate": 8.925427872860637e-06, | |
| "loss": 1.0258, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.2176039119804401, | |
| "grad_norm": 0.8024953007698059, | |
| "learning_rate": 8.913202933985331e-06, | |
| "loss": 1.0199, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.2200488997555012, | |
| "grad_norm": 0.6608093976974487, | |
| "learning_rate": 8.900977995110024e-06, | |
| "loss": 1.0042, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.22249388753056235, | |
| "grad_norm": 0.7665097117424011, | |
| "learning_rate": 8.888753056234719e-06, | |
| "loss": 1.0004, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.22493887530562348, | |
| "grad_norm": 0.7506141662597656, | |
| "learning_rate": 8.876528117359414e-06, | |
| "loss": 0.9913, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.2273838630806846, | |
| "grad_norm": 0.8530935049057007, | |
| "learning_rate": 8.864303178484108e-06, | |
| "loss": 1.0061, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.22982885085574573, | |
| "grad_norm": 0.7526981830596924, | |
| "learning_rate": 8.852078239608803e-06, | |
| "loss": 1.0343, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.23227383863080683, | |
| "grad_norm": 0.6752293109893799, | |
| "learning_rate": 8.839853300733498e-06, | |
| "loss": 1.0351, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.23471882640586797, | |
| "grad_norm": 0.7808672189712524, | |
| "learning_rate": 8.82762836185819e-06, | |
| "loss": 1.0216, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.2371638141809291, | |
| "grad_norm": 0.7220605611801147, | |
| "learning_rate": 8.815403422982885e-06, | |
| "loss": 1.0066, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.2396088019559902, | |
| "grad_norm": 0.777722954750061, | |
| "learning_rate": 8.80317848410758e-06, | |
| "loss": 0.9936, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.24205378973105135, | |
| "grad_norm": 0.7381393313407898, | |
| "learning_rate": 8.790953545232275e-06, | |
| "loss": 0.9982, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.24449877750611246, | |
| "grad_norm": 0.6590691804885864, | |
| "learning_rate": 8.77872860635697e-06, | |
| "loss": 0.9788, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.24449877750611246, | |
| "eval_loss": 0.9873583912849426, | |
| "eval_runtime": 795.9609, | |
| "eval_samples_per_second": 18.27, | |
| "eval_steps_per_second": 0.572, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2469437652811736, | |
| "grad_norm": 0.851658284664154, | |
| "learning_rate": 8.766503667481662e-06, | |
| "loss": 0.9715, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.24938875305623473, | |
| "grad_norm": 0.6827694177627563, | |
| "learning_rate": 8.754278728606359e-06, | |
| "loss": 0.9521, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.25183374083129584, | |
| "grad_norm": 0.7433952689170837, | |
| "learning_rate": 8.742053789731052e-06, | |
| "loss": 1.0064, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.254278728606357, | |
| "grad_norm": 0.7873063087463379, | |
| "learning_rate": 8.729828850855746e-06, | |
| "loss": 0.997, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.2567237163814181, | |
| "grad_norm": 0.7808490991592407, | |
| "learning_rate": 8.717603911980441e-06, | |
| "loss": 1.021, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.2591687041564792, | |
| "grad_norm": 0.7002695798873901, | |
| "learning_rate": 8.705378973105136e-06, | |
| "loss": 0.9787, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.2616136919315403, | |
| "grad_norm": 0.8023979663848877, | |
| "learning_rate": 8.693154034229829e-06, | |
| "loss": 0.9913, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.26405867970660146, | |
| "grad_norm": 0.6559330224990845, | |
| "learning_rate": 8.680929095354525e-06, | |
| "loss": 0.9755, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.2665036674816626, | |
| "grad_norm": 0.703667402267456, | |
| "learning_rate": 8.668704156479218e-06, | |
| "loss": 1.0047, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.26894865525672373, | |
| "grad_norm": 0.7463963627815247, | |
| "learning_rate": 8.656479217603913e-06, | |
| "loss": 0.9607, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2713936430317848, | |
| "grad_norm": 0.8031372427940369, | |
| "learning_rate": 8.644254278728606e-06, | |
| "loss": 0.9835, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.27383863080684595, | |
| "grad_norm": 0.7800993919372559, | |
| "learning_rate": 8.632029339853302e-06, | |
| "loss": 0.9651, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.2762836185819071, | |
| "grad_norm": 0.8147879242897034, | |
| "learning_rate": 8.619804400977995e-06, | |
| "loss": 1.029, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.2787286063569682, | |
| "grad_norm": 0.6160231828689575, | |
| "learning_rate": 8.60757946210269e-06, | |
| "loss": 0.9771, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.28117359413202936, | |
| "grad_norm": 0.8735865354537964, | |
| "learning_rate": 8.595354523227385e-06, | |
| "loss": 0.9719, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.28361858190709044, | |
| "grad_norm": 0.6970623135566711, | |
| "learning_rate": 8.58312958435208e-06, | |
| "loss": 0.9948, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.2860635696821516, | |
| "grad_norm": 0.7240091562271118, | |
| "learning_rate": 8.570904645476774e-06, | |
| "loss": 0.9798, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.2885085574572127, | |
| "grad_norm": 0.8112177848815918, | |
| "learning_rate": 8.558679706601469e-06, | |
| "loss": 0.9857, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.29095354523227385, | |
| "grad_norm": 0.7607941031455994, | |
| "learning_rate": 8.546454767726162e-06, | |
| "loss": 0.9467, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.293398533007335, | |
| "grad_norm": 0.7950695753097534, | |
| "learning_rate": 8.534229828850856e-06, | |
| "loss": 0.9691, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.29584352078239606, | |
| "grad_norm": 0.7780068516731262, | |
| "learning_rate": 8.522004889975551e-06, | |
| "loss": 0.9658, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.2982885085574572, | |
| "grad_norm": 0.7490630149841309, | |
| "learning_rate": 8.509779951100246e-06, | |
| "loss": 0.9528, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.30073349633251834, | |
| "grad_norm": 0.7754825353622437, | |
| "learning_rate": 8.49755501222494e-06, | |
| "loss": 0.995, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.30317848410757947, | |
| "grad_norm": 0.7797338366508484, | |
| "learning_rate": 8.485330073349633e-06, | |
| "loss": 0.9976, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.3056234718826406, | |
| "grad_norm": 0.7192637920379639, | |
| "learning_rate": 8.473105134474328e-06, | |
| "loss": 0.9764, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.3080684596577017, | |
| "grad_norm": 0.7009165287017822, | |
| "learning_rate": 8.460880195599023e-06, | |
| "loss": 0.9794, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.3105134474327628, | |
| "grad_norm": 0.7941039204597473, | |
| "learning_rate": 8.448655256723717e-06, | |
| "loss": 0.9701, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.31295843520782396, | |
| "grad_norm": 0.7658547163009644, | |
| "learning_rate": 8.436430317848412e-06, | |
| "loss": 0.9658, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.3154034229828851, | |
| "grad_norm": 0.8190957903862, | |
| "learning_rate": 8.424205378973107e-06, | |
| "loss": 0.9566, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.31784841075794623, | |
| "grad_norm": 0.7688964605331421, | |
| "learning_rate": 8.4119804400978e-06, | |
| "loss": 0.9743, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.3202933985330073, | |
| "grad_norm": 0.7160921096801758, | |
| "learning_rate": 8.399755501222494e-06, | |
| "loss": 0.9363, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.32273838630806845, | |
| "grad_norm": 0.6877838373184204, | |
| "learning_rate": 8.387530562347189e-06, | |
| "loss": 0.9704, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.3251833740831296, | |
| "grad_norm": 0.7970196008682251, | |
| "learning_rate": 8.375305623471884e-06, | |
| "loss": 0.9824, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.3276283618581907, | |
| "grad_norm": 0.7897329926490784, | |
| "learning_rate": 8.363080684596577e-06, | |
| "loss": 0.971, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.33007334963325186, | |
| "grad_norm": 0.7263765931129456, | |
| "learning_rate": 8.350855745721273e-06, | |
| "loss": 0.9754, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.33251833740831294, | |
| "grad_norm": 0.6933903098106384, | |
| "learning_rate": 8.338630806845966e-06, | |
| "loss": 0.9451, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.33496332518337407, | |
| "grad_norm": 0.6975806951522827, | |
| "learning_rate": 8.32640586797066e-06, | |
| "loss": 0.9538, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.3374083129584352, | |
| "grad_norm": 0.8183455467224121, | |
| "learning_rate": 8.314180929095355e-06, | |
| "loss": 0.955, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.33985330073349634, | |
| "grad_norm": 0.8085072636604309, | |
| "learning_rate": 8.30195599022005e-06, | |
| "loss": 0.9337, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.3422982885085575, | |
| "grad_norm": 0.7899576425552368, | |
| "learning_rate": 8.289731051344743e-06, | |
| "loss": 0.984, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.34474327628361856, | |
| "grad_norm": 0.8049628734588623, | |
| "learning_rate": 8.277506112469438e-06, | |
| "loss": 0.967, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.3471882640586797, | |
| "grad_norm": 0.7915717959403992, | |
| "learning_rate": 8.265281173594132e-06, | |
| "loss": 0.9379, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.34963325183374083, | |
| "grad_norm": 0.7397669553756714, | |
| "learning_rate": 8.253056234718827e-06, | |
| "loss": 0.9771, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.35207823960880197, | |
| "grad_norm": 0.7894385457038879, | |
| "learning_rate": 8.240831295843522e-06, | |
| "loss": 0.9647, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.3545232273838631, | |
| "grad_norm": 0.7507162094116211, | |
| "learning_rate": 8.228606356968216e-06, | |
| "loss": 0.9783, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.3569682151589242, | |
| "grad_norm": 0.8646982908248901, | |
| "learning_rate": 8.216381418092911e-06, | |
| "loss": 0.9817, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.3594132029339853, | |
| "grad_norm": 0.7320569753646851, | |
| "learning_rate": 8.204156479217604e-06, | |
| "loss": 0.9502, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.36185819070904646, | |
| "grad_norm": 0.7949222922325134, | |
| "learning_rate": 8.191931540342299e-06, | |
| "loss": 0.9694, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.3643031784841076, | |
| "grad_norm": 0.8175340294837952, | |
| "learning_rate": 8.179706601466993e-06, | |
| "loss": 0.972, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.36674816625916873, | |
| "grad_norm": 0.8479053974151611, | |
| "learning_rate": 8.167481662591688e-06, | |
| "loss": 1.0059, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3691931540342298, | |
| "grad_norm": 0.7485557794570923, | |
| "learning_rate": 8.155256723716381e-06, | |
| "loss": 0.9303, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.37163814180929094, | |
| "grad_norm": 0.7518207430839539, | |
| "learning_rate": 8.143031784841077e-06, | |
| "loss": 0.9572, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.3740831295843521, | |
| "grad_norm": 0.8368051052093506, | |
| "learning_rate": 8.13080684596577e-06, | |
| "loss": 0.9865, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.3765281173594132, | |
| "grad_norm": 0.8675082325935364, | |
| "learning_rate": 8.118581907090465e-06, | |
| "loss": 0.9355, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.37897310513447435, | |
| "grad_norm": 0.8239061832427979, | |
| "learning_rate": 8.10635696821516e-06, | |
| "loss": 0.9475, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.38141809290953543, | |
| "grad_norm": 0.700191080570221, | |
| "learning_rate": 8.094132029339854e-06, | |
| "loss": 0.9343, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.38386308068459657, | |
| "grad_norm": 0.789686381816864, | |
| "learning_rate": 8.081907090464547e-06, | |
| "loss": 0.9633, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.3863080684596577, | |
| "grad_norm": 0.8411287665367126, | |
| "learning_rate": 8.069682151589244e-06, | |
| "loss": 0.991, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.38875305623471884, | |
| "grad_norm": 0.7479956150054932, | |
| "learning_rate": 8.057457212713937e-06, | |
| "loss": 0.9371, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.39119804400978, | |
| "grad_norm": 0.726841151714325, | |
| "learning_rate": 8.045232273838631e-06, | |
| "loss": 0.9595, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.39364303178484106, | |
| "grad_norm": 0.7860400676727295, | |
| "learning_rate": 8.033007334963326e-06, | |
| "loss": 0.9673, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.3960880195599022, | |
| "grad_norm": 0.8316232562065125, | |
| "learning_rate": 8.02078239608802e-06, | |
| "loss": 0.9592, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.39853300733496333, | |
| "grad_norm": 0.743289053440094, | |
| "learning_rate": 8.008557457212714e-06, | |
| "loss": 0.9424, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.40097799511002447, | |
| "grad_norm": 0.7066758275032043, | |
| "learning_rate": 7.996332518337408e-06, | |
| "loss": 0.9515, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.4034229828850856, | |
| "grad_norm": 0.8527940511703491, | |
| "learning_rate": 7.984107579462103e-06, | |
| "loss": 0.9579, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.4058679706601467, | |
| "grad_norm": 0.747097909450531, | |
| "learning_rate": 7.971882640586798e-06, | |
| "loss": 0.96, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.4083129584352078, | |
| "grad_norm": 0.7440111637115479, | |
| "learning_rate": 7.959657701711492e-06, | |
| "loss": 0.9428, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.41075794621026895, | |
| "grad_norm": 0.7893000245094299, | |
| "learning_rate": 7.947432762836187e-06, | |
| "loss": 0.9775, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.4132029339853301, | |
| "grad_norm": 0.7849116325378418, | |
| "learning_rate": 7.935207823960882e-06, | |
| "loss": 0.9636, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.4156479217603912, | |
| "grad_norm": 0.8839460611343384, | |
| "learning_rate": 7.922982885085575e-06, | |
| "loss": 0.9427, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.4180929095354523, | |
| "grad_norm": 0.7875691056251526, | |
| "learning_rate": 7.91075794621027e-06, | |
| "loss": 0.9188, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.42053789731051344, | |
| "grad_norm": 0.7899657487869263, | |
| "learning_rate": 7.898533007334964e-06, | |
| "loss": 0.9337, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.4229828850855746, | |
| "grad_norm": 0.8064518570899963, | |
| "learning_rate": 7.886308068459659e-06, | |
| "loss": 0.9473, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.4254278728606357, | |
| "grad_norm": 0.9735845327377319, | |
| "learning_rate": 7.874083129584352e-06, | |
| "loss": 0.943, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.4278728606356968, | |
| "grad_norm": 0.8656879663467407, | |
| "learning_rate": 7.861858190709048e-06, | |
| "loss": 0.9276, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.43031784841075793, | |
| "grad_norm": 0.7446141242980957, | |
| "learning_rate": 7.849633251833741e-06, | |
| "loss": 0.9609, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.43276283618581907, | |
| "grad_norm": 0.7805430889129639, | |
| "learning_rate": 7.837408312958436e-06, | |
| "loss": 0.9437, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.4352078239608802, | |
| "grad_norm": 0.8328519463539124, | |
| "learning_rate": 7.82518337408313e-06, | |
| "loss": 0.9921, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.43765281173594134, | |
| "grad_norm": 0.7910134196281433, | |
| "learning_rate": 7.812958435207825e-06, | |
| "loss": 0.9392, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.4400977995110024, | |
| "grad_norm": 0.8082761168479919, | |
| "learning_rate": 7.800733496332518e-06, | |
| "loss": 0.9655, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.44254278728606355, | |
| "grad_norm": 0.7104289531707764, | |
| "learning_rate": 7.788508557457214e-06, | |
| "loss": 0.9604, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.4449877750611247, | |
| "grad_norm": 0.7942298054695129, | |
| "learning_rate": 7.776283618581907e-06, | |
| "loss": 0.9839, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.4474327628361858, | |
| "grad_norm": 0.7665939927101135, | |
| "learning_rate": 7.764058679706602e-06, | |
| "loss": 0.9907, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.44987775061124696, | |
| "grad_norm": 0.8066325187683105, | |
| "learning_rate": 7.751833740831297e-06, | |
| "loss": 0.9503, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.45232273838630804, | |
| "grad_norm": 0.7494056224822998, | |
| "learning_rate": 7.739608801955991e-06, | |
| "loss": 0.9594, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.4547677261613692, | |
| "grad_norm": 0.7743037939071655, | |
| "learning_rate": 7.727383863080684e-06, | |
| "loss": 0.9708, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.4572127139364303, | |
| "grad_norm": 0.6371968388557434, | |
| "learning_rate": 7.715158924205379e-06, | |
| "loss": 0.9247, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.45965770171149145, | |
| "grad_norm": 0.8373169898986816, | |
| "learning_rate": 7.702933985330074e-06, | |
| "loss": 0.9631, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.4621026894865526, | |
| "grad_norm": 0.898855984210968, | |
| "learning_rate": 7.690709046454768e-06, | |
| "loss": 0.9598, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.46454767726161367, | |
| "grad_norm": 0.7827709317207336, | |
| "learning_rate": 7.678484107579463e-06, | |
| "loss": 0.9407, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.4669926650366748, | |
| "grad_norm": 0.8428008556365967, | |
| "learning_rate": 7.666259168704158e-06, | |
| "loss": 0.9753, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.46943765281173594, | |
| "grad_norm": 0.7905760407447815, | |
| "learning_rate": 7.654034229828853e-06, | |
| "loss": 0.9487, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.4718826405867971, | |
| "grad_norm": 0.7715455293655396, | |
| "learning_rate": 7.641809290953546e-06, | |
| "loss": 0.9304, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.4743276283618582, | |
| "grad_norm": 0.7903933525085449, | |
| "learning_rate": 7.62958435207824e-06, | |
| "loss": 0.9277, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.4767726161369193, | |
| "grad_norm": 0.7191179990768433, | |
| "learning_rate": 7.617359413202935e-06, | |
| "loss": 0.915, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.4792176039119804, | |
| "grad_norm": 0.8574967980384827, | |
| "learning_rate": 7.605134474327629e-06, | |
| "loss": 0.918, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.48166259168704156, | |
| "grad_norm": 0.8661892414093018, | |
| "learning_rate": 7.592909535452323e-06, | |
| "loss": 0.9658, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.4841075794621027, | |
| "grad_norm": 0.8399495482444763, | |
| "learning_rate": 7.580684596577018e-06, | |
| "loss": 0.9611, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.48655256723716384, | |
| "grad_norm": 0.8675727844238281, | |
| "learning_rate": 7.568459657701712e-06, | |
| "loss": 0.9367, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.4889975550122249, | |
| "grad_norm": 0.8307384252548218, | |
| "learning_rate": 7.5562347188264065e-06, | |
| "loss": 0.9628, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4889975550122249, | |
| "eval_loss": 0.9446731805801392, | |
| "eval_runtime": 795.9747, | |
| "eval_samples_per_second": 18.269, | |
| "eval_steps_per_second": 0.572, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.49144254278728605, | |
| "grad_norm": 0.7585815191268921, | |
| "learning_rate": 7.544009779951101e-06, | |
| "loss": 0.9499, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.4938875305623472, | |
| "grad_norm": 0.870883584022522, | |
| "learning_rate": 7.531784841075795e-06, | |
| "loss": 0.9758, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.4963325183374083, | |
| "grad_norm": 0.7365143895149231, | |
| "learning_rate": 7.51955990220049e-06, | |
| "loss": 0.9429, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.49877750611246946, | |
| "grad_norm": 0.9007924199104309, | |
| "learning_rate": 7.5073349633251836e-06, | |
| "loss": 0.9208, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.5012224938875306, | |
| "grad_norm": 0.8040947914123535, | |
| "learning_rate": 7.495110024449879e-06, | |
| "loss": 0.931, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.5036674816625917, | |
| "grad_norm": 0.7994057536125183, | |
| "learning_rate": 7.482885085574573e-06, | |
| "loss": 0.9406, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.5061124694376528, | |
| "grad_norm": 0.7404657006263733, | |
| "learning_rate": 7.470660146699267e-06, | |
| "loss": 0.91, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.508557457212714, | |
| "grad_norm": 0.8444405794143677, | |
| "learning_rate": 7.458435207823962e-06, | |
| "loss": 0.9748, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.511002444987775, | |
| "grad_norm": 0.830786406993866, | |
| "learning_rate": 7.446210268948656e-06, | |
| "loss": 0.9546, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.5134474327628362, | |
| "grad_norm": 0.8706929683685303, | |
| "learning_rate": 7.43398533007335e-06, | |
| "loss": 0.9359, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.5158924205378973, | |
| "grad_norm": 0.937125563621521, | |
| "learning_rate": 7.4217603911980454e-06, | |
| "loss": 0.9367, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.5183374083129584, | |
| "grad_norm": 0.871612548828125, | |
| "learning_rate": 7.409535452322739e-06, | |
| "loss": 0.9363, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.5207823960880196, | |
| "grad_norm": 0.7397099733352661, | |
| "learning_rate": 7.397310513447433e-06, | |
| "loss": 0.9312, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.5232273838630807, | |
| "grad_norm": 0.9171463847160339, | |
| "learning_rate": 7.385085574572127e-06, | |
| "loss": 0.9657, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.5256723716381418, | |
| "grad_norm": 0.7274787425994873, | |
| "learning_rate": 7.3728606356968224e-06, | |
| "loss": 0.9359, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.5281173594132029, | |
| "grad_norm": 0.8870149850845337, | |
| "learning_rate": 7.360635696821516e-06, | |
| "loss": 0.9469, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.530562347188264, | |
| "grad_norm": 0.770986020565033, | |
| "learning_rate": 7.34841075794621e-06, | |
| "loss": 0.9529, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.5330073349633252, | |
| "grad_norm": 0.7791648507118225, | |
| "learning_rate": 7.336185819070906e-06, | |
| "loss": 0.9509, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.5354523227383863, | |
| "grad_norm": 0.8232088088989258, | |
| "learning_rate": 7.3239608801955995e-06, | |
| "loss": 0.9639, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.5378973105134475, | |
| "grad_norm": 0.8500565886497498, | |
| "learning_rate": 7.311735941320294e-06, | |
| "loss": 0.97, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.5403422982885085, | |
| "grad_norm": 0.8228991627693176, | |
| "learning_rate": 7.299511002444989e-06, | |
| "loss": 0.9514, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.5427872860635696, | |
| "grad_norm": 0.7978509068489075, | |
| "learning_rate": 7.287286063569683e-06, | |
| "loss": 0.9599, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.5452322738386308, | |
| "grad_norm": 0.7359360456466675, | |
| "learning_rate": 7.275061124694377e-06, | |
| "loss": 0.9087, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.5476772616136919, | |
| "grad_norm": 0.8995153903961182, | |
| "learning_rate": 7.262836185819071e-06, | |
| "loss": 0.9082, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.5501222493887531, | |
| "grad_norm": 0.7778546214103699, | |
| "learning_rate": 7.250611246943766e-06, | |
| "loss": 0.9599, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.5525672371638142, | |
| "grad_norm": 0.7799263596534729, | |
| "learning_rate": 7.2383863080684605e-06, | |
| "loss": 0.895, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.5550122249388753, | |
| "grad_norm": 0.831047534942627, | |
| "learning_rate": 7.226161369193154e-06, | |
| "loss": 0.9426, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.5574572127139364, | |
| "grad_norm": 0.863277792930603, | |
| "learning_rate": 7.213936430317849e-06, | |
| "loss": 0.947, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.5599022004889975, | |
| "grad_norm": 0.7998844981193542, | |
| "learning_rate": 7.201711491442544e-06, | |
| "loss": 0.9366, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.5623471882640587, | |
| "grad_norm": 0.818229079246521, | |
| "learning_rate": 7.1894865525672375e-06, | |
| "loss": 0.9524, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.5647921760391198, | |
| "grad_norm": 0.8541170954704285, | |
| "learning_rate": 7.177261613691933e-06, | |
| "loss": 0.9297, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.5672371638141809, | |
| "grad_norm": 0.7943381071090698, | |
| "learning_rate": 7.165036674816627e-06, | |
| "loss": 0.9186, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.5696821515892421, | |
| "grad_norm": 0.886789858341217, | |
| "learning_rate": 7.152811735941321e-06, | |
| "loss": 0.9551, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.5721271393643031, | |
| "grad_norm": 0.7614038586616516, | |
| "learning_rate": 7.1405867970660145e-06, | |
| "loss": 0.9427, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.5745721271393643, | |
| "grad_norm": 0.8962105512619019, | |
| "learning_rate": 7.12836185819071e-06, | |
| "loss": 0.9191, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.5770171149144254, | |
| "grad_norm": 0.9211586713790894, | |
| "learning_rate": 7.116136919315404e-06, | |
| "loss": 0.933, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.5794621026894865, | |
| "grad_norm": 0.9243327975273132, | |
| "learning_rate": 7.103911980440098e-06, | |
| "loss": 0.9152, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.5819070904645477, | |
| "grad_norm": 0.851266086101532, | |
| "learning_rate": 7.091687041564793e-06, | |
| "loss": 0.9697, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.5843520782396088, | |
| "grad_norm": 0.8055517673492432, | |
| "learning_rate": 7.079462102689487e-06, | |
| "loss": 0.921, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.58679706601467, | |
| "grad_norm": 0.9354420900344849, | |
| "learning_rate": 7.067237163814181e-06, | |
| "loss": 0.9368, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.589242053789731, | |
| "grad_norm": 0.8339366912841797, | |
| "learning_rate": 7.055012224938876e-06, | |
| "loss": 0.9387, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.5916870415647921, | |
| "grad_norm": 0.8128229975700378, | |
| "learning_rate": 7.04278728606357e-06, | |
| "loss": 0.9462, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.5941320293398533, | |
| "grad_norm": 0.7740962505340576, | |
| "learning_rate": 7.030562347188264e-06, | |
| "loss": 0.931, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.5965770171149144, | |
| "grad_norm": 0.8874317407608032, | |
| "learning_rate": 7.018337408312959e-06, | |
| "loss": 0.9809, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.5990220048899756, | |
| "grad_norm": 0.8482634425163269, | |
| "learning_rate": 7.006112469437653e-06, | |
| "loss": 0.9104, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.6014669926650367, | |
| "grad_norm": 0.7957248687744141, | |
| "learning_rate": 6.993887530562348e-06, | |
| "loss": 0.9195, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.6039119804400978, | |
| "grad_norm": 0.8182454109191895, | |
| "learning_rate": 6.981662591687042e-06, | |
| "loss": 0.9565, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.6063569682151589, | |
| "grad_norm": 0.7790459990501404, | |
| "learning_rate": 6.969437652811737e-06, | |
| "loss": 0.9114, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.60880195599022, | |
| "grad_norm": 0.7870607376098633, | |
| "learning_rate": 6.957212713936431e-06, | |
| "loss": 0.9372, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.6112469437652812, | |
| "grad_norm": 0.8847014307975769, | |
| "learning_rate": 6.944987775061125e-06, | |
| "loss": 0.9344, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6136919315403423, | |
| "grad_norm": 0.8556163907051086, | |
| "learning_rate": 6.93276283618582e-06, | |
| "loss": 0.915, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.6161369193154034, | |
| "grad_norm": 0.846760630607605, | |
| "learning_rate": 6.9205378973105144e-06, | |
| "loss": 0.9301, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.6185819070904646, | |
| "grad_norm": 0.7853952646255493, | |
| "learning_rate": 6.908312958435208e-06, | |
| "loss": 0.9395, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.6210268948655256, | |
| "grad_norm": 0.8032101392745972, | |
| "learning_rate": 6.896088019559902e-06, | |
| "loss": 0.9561, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.6234718826405868, | |
| "grad_norm": 0.7820572853088379, | |
| "learning_rate": 6.883863080684598e-06, | |
| "loss": 0.9167, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.6259168704156479, | |
| "grad_norm": 0.8646982908248901, | |
| "learning_rate": 6.8716381418092915e-06, | |
| "loss": 0.9297, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.628361858190709, | |
| "grad_norm": 0.8482604026794434, | |
| "learning_rate": 6.859413202933985e-06, | |
| "loss": 0.9254, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.6308068459657702, | |
| "grad_norm": 0.775800883769989, | |
| "learning_rate": 6.847188264058681e-06, | |
| "loss": 0.9346, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.6332518337408313, | |
| "grad_norm": 0.7460722327232361, | |
| "learning_rate": 6.834963325183375e-06, | |
| "loss": 0.9545, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.6356968215158925, | |
| "grad_norm": 0.8128139972686768, | |
| "learning_rate": 6.8227383863080685e-06, | |
| "loss": 0.9262, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.6381418092909535, | |
| "grad_norm": 0.8368150591850281, | |
| "learning_rate": 6.810513447432764e-06, | |
| "loss": 0.9123, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.6405867970660146, | |
| "grad_norm": 0.7825431227684021, | |
| "learning_rate": 6.798288508557458e-06, | |
| "loss": 0.9336, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.6430317848410758, | |
| "grad_norm": 0.7432078123092651, | |
| "learning_rate": 6.786063569682152e-06, | |
| "loss": 0.9396, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.6454767726161369, | |
| "grad_norm": 0.6994110345840454, | |
| "learning_rate": 6.773838630806846e-06, | |
| "loss": 0.9527, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.6479217603911981, | |
| "grad_norm": 0.8581116795539856, | |
| "learning_rate": 6.761613691931541e-06, | |
| "loss": 0.8979, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.6503667481662592, | |
| "grad_norm": 0.7824479937553406, | |
| "learning_rate": 6.749388753056235e-06, | |
| "loss": 0.93, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.6528117359413202, | |
| "grad_norm": 0.7407302260398865, | |
| "learning_rate": 6.7371638141809295e-06, | |
| "loss": 0.9175, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.6552567237163814, | |
| "grad_norm": 0.8635402917861938, | |
| "learning_rate": 6.724938875305624e-06, | |
| "loss": 0.951, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.6577017114914425, | |
| "grad_norm": 0.7739470601081848, | |
| "learning_rate": 6.712713936430319e-06, | |
| "loss": 0.9199, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.6601466992665037, | |
| "grad_norm": 0.879205048084259, | |
| "learning_rate": 6.700488997555013e-06, | |
| "loss": 0.9296, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.6625916870415648, | |
| "grad_norm": 0.7694469094276428, | |
| "learning_rate": 6.688264058679707e-06, | |
| "loss": 0.9312, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.6650366748166259, | |
| "grad_norm": 0.8447741270065308, | |
| "learning_rate": 6.676039119804402e-06, | |
| "loss": 0.9486, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.6674816625916871, | |
| "grad_norm": 0.7987990379333496, | |
| "learning_rate": 6.663814180929096e-06, | |
| "loss": 0.9554, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.6699266503667481, | |
| "grad_norm": 0.8286950588226318, | |
| "learning_rate": 6.65158924205379e-06, | |
| "loss": 0.9264, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.6723716381418093, | |
| "grad_norm": 0.8135730028152466, | |
| "learning_rate": 6.639364303178485e-06, | |
| "loss": 0.9183, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.6748166259168704, | |
| "grad_norm": 0.9580305218696594, | |
| "learning_rate": 6.627139364303179e-06, | |
| "loss": 0.9055, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.6772616136919315, | |
| "grad_norm": 0.7580097317695618, | |
| "learning_rate": 6.614914425427873e-06, | |
| "loss": 0.923, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.6797066014669927, | |
| "grad_norm": 0.7250223755836487, | |
| "learning_rate": 6.602689486552568e-06, | |
| "loss": 0.9472, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.6821515892420538, | |
| "grad_norm": 0.720930278301239, | |
| "learning_rate": 6.590464547677262e-06, | |
| "loss": 0.9155, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.684596577017115, | |
| "grad_norm": 0.7639275789260864, | |
| "learning_rate": 6.578239608801956e-06, | |
| "loss": 0.917, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.687041564792176, | |
| "grad_norm": 0.8890448808670044, | |
| "learning_rate": 6.5660146699266516e-06, | |
| "loss": 0.933, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.6894865525672371, | |
| "grad_norm": 0.816336452960968, | |
| "learning_rate": 6.553789731051345e-06, | |
| "loss": 0.9606, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.6919315403422983, | |
| "grad_norm": 0.8258111476898193, | |
| "learning_rate": 6.541564792176039e-06, | |
| "loss": 0.9331, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.6943765281173594, | |
| "grad_norm": 0.8009893894195557, | |
| "learning_rate": 6.529339853300734e-06, | |
| "loss": 0.918, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.6968215158924206, | |
| "grad_norm": 0.861232578754425, | |
| "learning_rate": 6.517114914425429e-06, | |
| "loss": 0.9157, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.6992665036674817, | |
| "grad_norm": 0.8324180841445923, | |
| "learning_rate": 6.504889975550122e-06, | |
| "loss": 0.9322, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.7017114914425427, | |
| "grad_norm": 0.8205760717391968, | |
| "learning_rate": 6.492665036674817e-06, | |
| "loss": 0.9401, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.7041564792176039, | |
| "grad_norm": 0.8924916386604309, | |
| "learning_rate": 6.480440097799512e-06, | |
| "loss": 0.9165, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.706601466992665, | |
| "grad_norm": 0.7587029337882996, | |
| "learning_rate": 6.468215158924206e-06, | |
| "loss": 0.9127, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.7090464547677262, | |
| "grad_norm": 0.7448475360870361, | |
| "learning_rate": 6.4559902200489e-06, | |
| "loss": 0.9366, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.7114914425427873, | |
| "grad_norm": 0.8168658018112183, | |
| "learning_rate": 6.443765281173595e-06, | |
| "loss": 0.9377, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.7139364303178484, | |
| "grad_norm": 0.7668200731277466, | |
| "learning_rate": 6.431540342298289e-06, | |
| "loss": 0.9342, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.7163814180929096, | |
| "grad_norm": 0.8051294088363647, | |
| "learning_rate": 6.4193154034229834e-06, | |
| "loss": 0.9236, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.7188264058679706, | |
| "grad_norm": 0.7371141910552979, | |
| "learning_rate": 6.407090464547677e-06, | |
| "loss": 0.9123, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.7212713936430318, | |
| "grad_norm": 0.7507117390632629, | |
| "learning_rate": 6.394865525672373e-06, | |
| "loss": 0.8978, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.7237163814180929, | |
| "grad_norm": 0.8171530365943909, | |
| "learning_rate": 6.382640586797067e-06, | |
| "loss": 0.9516, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.726161369193154, | |
| "grad_norm": 0.8637788891792297, | |
| "learning_rate": 6.3704156479217605e-06, | |
| "loss": 0.8987, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.7286063569682152, | |
| "grad_norm": 0.7505759596824646, | |
| "learning_rate": 6.358190709046456e-06, | |
| "loss": 0.9134, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.7310513447432763, | |
| "grad_norm": 0.9182484149932861, | |
| "learning_rate": 6.34596577017115e-06, | |
| "loss": 0.9351, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.7334963325183375, | |
| "grad_norm": 1.006665825843811, | |
| "learning_rate": 6.333740831295844e-06, | |
| "loss": 0.9171, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7334963325183375, | |
| "eval_loss": 0.9244844913482666, | |
| "eval_runtime": 795.2171, | |
| "eval_samples_per_second": 18.287, | |
| "eval_steps_per_second": 0.572, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 8180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.914805497032868e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |