| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.2009107956067506, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.000669702652022502, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.693440428380188e-07, | |
| "loss": 2.3893, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.001339405304045004, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.3386880856760376e-06, | |
| "loss": 2.4328, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.002009107956067506, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.0080321285140564e-06, | |
| "loss": 2.349, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.002678810608090008, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.6773761713520752e-06, | |
| "loss": 2.2695, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.00334851326011251, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.346720214190094e-06, | |
| "loss": 2.3902, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.004018215912135012, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.016064257028113e-06, | |
| "loss": 2.307, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.004687918564157514, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.685408299866132e-06, | |
| "loss": 2.2561, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.005357621216180016, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.3547523427041504e-06, | |
| "loss": 2.3053, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.006027323868202518, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.024096385542169e-06, | |
| "loss": 2.2695, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.00669702652022502, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.693440428380188e-06, | |
| "loss": 2.2224, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0073667291722475225, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.362784471218207e-06, | |
| "loss": 2.2922, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.008036431824270024, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.032128514056226e-06, | |
| "loss": 2.4112, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.008706134476292525, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.701472556894244e-06, | |
| "loss": 2.3502, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.009375837128315028, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.370816599732263e-06, | |
| "loss": 2.3654, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.01004553978033753, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0040160642570281e-05, | |
| "loss": 2.3673, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.010715242432360031, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0709504685408301e-05, | |
| "loss": 2.3255, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.011384945084382534, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.1378848728246319e-05, | |
| "loss": 2.2099, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.012054647736405036, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.2048192771084338e-05, | |
| "loss": 2.3237, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.012724350388427539, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.2717536813922356e-05, | |
| "loss": 2.3802, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.01339405304045004, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.3386880856760376e-05, | |
| "loss": 2.2796, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.014063755692472542, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.4056224899598394e-05, | |
| "loss": 2.2388, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.014733458344495045, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.4725568942436414e-05, | |
| "loss": 2.259, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.015403160996517546, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.5394912985274433e-05, | |
| "loss": 2.3091, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.016072863648540048, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.606425702811245e-05, | |
| "loss": 2.2959, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.01674256630056255, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.673360107095047e-05, | |
| "loss": 2.2937, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.01741226895258505, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.7402945113788487e-05, | |
| "loss": 2.3535, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.018081971604607554, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.8072289156626505e-05, | |
| "loss": 2.25, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.018751674256630057, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.8741633199464527e-05, | |
| "loss": 2.2753, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.019421376908652557, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.9410977242302544e-05, | |
| "loss": 2.1762, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.02009107956067506, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.0080321285140562e-05, | |
| "loss": 2.2938, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.020760782212697563, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.074966532797858e-05, | |
| "loss": 2.2703, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.021430484864720063, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.1419009370816602e-05, | |
| "loss": 2.2566, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.022100187516742566, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.208835341365462e-05, | |
| "loss": 2.2291, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.02276989016876507, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.2757697456492638e-05, | |
| "loss": 2.2852, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.023439592820787572, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.3427041499330656e-05, | |
| "loss": 2.3257, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.02410929547281007, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.4096385542168677e-05, | |
| "loss": 2.3063, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.024778998124832575, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.4765729585006695e-05, | |
| "loss": 2.2745, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.025448700776855078, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.5435073627844713e-05, | |
| "loss": 2.3215, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.026118403428877578, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.6104417670682734e-05, | |
| "loss": 2.3273, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.02678810608090008, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.6773761713520752e-05, | |
| "loss": 2.331, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.027457808732922584, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.7443105756358774e-05, | |
| "loss": 2.2648, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.028127511384945084, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.8112449799196788e-05, | |
| "loss": 2.2876, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.028797214036967587, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.878179384203481e-05, | |
| "loss": 2.3393, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.02946691668899009, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.9451137884872827e-05, | |
| "loss": 2.2856, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.03013661934101259, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.012048192771085e-05, | |
| "loss": 2.2461, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.030806321993035093, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.078982597054887e-05, | |
| "loss": 2.2753, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.031476024645057596, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.1459170013386885e-05, | |
| "loss": 2.3217, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.032145727297080096, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.21285140562249e-05, | |
| "loss": 2.1547, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.032815429949102595, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.279785809906292e-05, | |
| "loss": 2.3005, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.0334851326011251, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.346720214190094e-05, | |
| "loss": 2.2335, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0341548352531476, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.413654618473896e-05, | |
| "loss": 2.2393, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.0348245379051701, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.4805890227576974e-05, | |
| "loss": 2.2483, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.03549424055719261, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.5475234270415e-05, | |
| "loss": 2.2597, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.03616394320921511, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.614457831325301e-05, | |
| "loss": 2.2253, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.03683364586123761, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.6813922356091035e-05, | |
| "loss": 2.2051, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.037503348513260114, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.748326639892905e-05, | |
| "loss": 2.3168, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.038173051165282613, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.815261044176707e-05, | |
| "loss": 2.3519, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.03884275381730511, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.882195448460509e-05, | |
| "loss": 2.3006, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.03951245646932762, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.949129852744311e-05, | |
| "loss": 2.2757, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.04018215912135012, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.0160642570281125e-05, | |
| "loss": 2.2025, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.04085186177337262, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.082998661311915e-05, | |
| "loss": 2.302, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.041521564425395126, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.149933065595716e-05, | |
| "loss": 2.3281, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.042191267077417625, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.2168674698795186e-05, | |
| "loss": 2.2667, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.042860969729440125, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.2838018741633203e-05, | |
| "loss": 2.1984, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.04353067238146263, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.350736278447122e-05, | |
| "loss": 2.2134, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.04420037503348513, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.417670682730924e-05, | |
| "loss": 2.192, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.04487007768550764, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.484605087014726e-05, | |
| "loss": 2.3038, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.04553978033753014, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.5515394912985275e-05, | |
| "loss": 2.3534, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.04620948298955264, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.61847389558233e-05, | |
| "loss": 2.2772, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.046879185641575144, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.685408299866131e-05, | |
| "loss": 2.3267, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.047548888293597644, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.7523427041499336e-05, | |
| "loss": 2.2713, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.04821859094562014, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.8192771084337354e-05, | |
| "loss": 2.3307, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.04888829359764265, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.886211512717537e-05, | |
| "loss": 2.3222, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.04955799624966515, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.953145917001339e-05, | |
| "loss": 2.2475, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.05022769890168765, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.020080321285141e-05, | |
| "loss": 2.2681, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.050897401553710156, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.0870147255689426e-05, | |
| "loss": 2.2273, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.051567104205732656, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.1539491298527444e-05, | |
| "loss": 2.2599, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.052236806857755155, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.220883534136547e-05, | |
| "loss": 2.2973, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.05290650950977766, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.2878179384203486e-05, | |
| "loss": 2.2479, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.05357621216180016, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.3547523427041504e-05, | |
| "loss": 2.2399, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.05424591481382266, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.4216867469879516e-05, | |
| "loss": 2.3338, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.05491561746584517, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.488621151271755e-05, | |
| "loss": 2.3081, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.05558532011786767, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 2.3125, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.05625502276989017, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.6224899598393576e-05, | |
| "loss": 2.1618, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.056924725421912674, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.6894243641231594e-05, | |
| "loss": 2.2753, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.057594428073935174, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.756358768406962e-05, | |
| "loss": 2.2857, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.05826413072595767, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.823293172690764e-05, | |
| "loss": 2.2962, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.05893383337798018, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.8902275769745655e-05, | |
| "loss": 2.271, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.05960353603000268, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.9571619812583666e-05, | |
| "loss": 2.2821, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.06027323868202518, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.02409638554217e-05, | |
| "loss": 2.2425, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.060942941334047686, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.091030789825971e-05, | |
| "loss": 2.3644, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.061612643986070185, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.157965194109773e-05, | |
| "loss": 2.3292, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.062282346638092685, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.224899598393574e-05, | |
| "loss": 2.1308, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.06295204929011519, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.291834002677377e-05, | |
| "loss": 2.2449, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.06362175194213769, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.358768406961179e-05, | |
| "loss": 2.2582, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.06429145459416019, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.42570281124498e-05, | |
| "loss": 2.1761, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.06496115724618269, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.492637215528782e-05, | |
| "loss": 2.1469, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.06563085989820519, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.559571619812584e-05, | |
| "loss": 2.1693, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.0663005625502277, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.626506024096386e-05, | |
| "loss": 2.2056, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.0669702652022502, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.693440428380188e-05, | |
| "loss": 2.2208, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.0676399678542727, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.76037483266399e-05, | |
| "loss": 2.2388, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.0683096705062952, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.827309236947793e-05, | |
| "loss": 2.3245, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.0689793731583177, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.894243641231593e-05, | |
| "loss": 2.2144, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.0696490758103402, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.961178045515395e-05, | |
| "loss": 2.2641, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.07031877846236272, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.028112449799197e-05, | |
| "loss": 2.1884, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.07098848111438522, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.095046854083e-05, | |
| "loss": 2.2975, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.07165818376640772, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.161981258366802e-05, | |
| "loss": 2.2589, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.07232788641843022, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.228915662650602e-05, | |
| "loss": 2.3497, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.07299758907045271, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.295850066934404e-05, | |
| "loss": 2.3039, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.07366729172247521, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.362784471218207e-05, | |
| "loss": 2.3212, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.07433699437449773, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.429718875502009e-05, | |
| "loss": 2.3193, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.07500669702652023, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.49665327978581e-05, | |
| "loss": 2.3162, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.07567639967854273, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.563587684069612e-05, | |
| "loss": 2.3661, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.07634610233056523, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.630522088353414e-05, | |
| "loss": 2.3379, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.07701580498258773, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.697456492637216e-05, | |
| "loss": 2.3418, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.07768550763461023, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.764390896921018e-05, | |
| "loss": 2.3239, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.07835521028663274, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.83132530120482e-05, | |
| "loss": 2.2827, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.07902491293865524, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.898259705488621e-05, | |
| "loss": 2.3565, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.07969461559067774, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.965194109772423e-05, | |
| "loss": 2.298, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.08036431824270024, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.032128514056225e-05, | |
| "loss": 2.2788, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.08103402089472274, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.099062918340027e-05, | |
| "loss": 2.254, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.08170372354674524, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.16599732262383e-05, | |
| "loss": 2.2501, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.08237342619876775, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.232931726907632e-05, | |
| "loss": 2.2683, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.08304312885079025, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.299866131191432e-05, | |
| "loss": 2.2645, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.08371283150281275, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.366800535475234e-05, | |
| "loss": 2.2241, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.08438253415483525, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.433734939759037e-05, | |
| "loss": 2.2863, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.08505223680685775, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.500669344042839e-05, | |
| "loss": 2.2603, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.08572193945888025, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.567603748326641e-05, | |
| "loss": 2.2955, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.08639164211090276, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.634538152610442e-05, | |
| "loss": 2.2459, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.08706134476292526, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.701472556894244e-05, | |
| "loss": 2.2492, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.08773104741494776, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.768406961178046e-05, | |
| "loss": 2.2569, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.08840075006697026, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.835341365461848e-05, | |
| "loss": 2.3215, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.08907045271899276, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.90227576974565e-05, | |
| "loss": 2.2873, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.08974015537101528, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.969210174029451e-05, | |
| "loss": 2.3188, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.09040985802303778, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.036144578313253e-05, | |
| "loss": 2.2357, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.09107956067506028, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.103078982597055e-05, | |
| "loss": 2.3087, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.09174926332708278, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.170013386880857e-05, | |
| "loss": 2.2656, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.09241896597910527, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.23694779116466e-05, | |
| "loss": 2.2529, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.09308866863112777, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.303882195448462e-05, | |
| "loss": 2.2483, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.09375837128315029, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.370816599732262e-05, | |
| "loss": 2.2257, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.09442807393517279, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.437751004016064e-05, | |
| "loss": 2.2937, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.09509777658719529, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.504685408299867e-05, | |
| "loss": 2.2907, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.09576747923921779, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.571619812583669e-05, | |
| "loss": 2.3068, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.09643718189124029, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.638554216867471e-05, | |
| "loss": 2.2546, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.09710688454326279, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.705488621151271e-05, | |
| "loss": 2.3099, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.0977765871952853, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.772423025435074e-05, | |
| "loss": 2.3662, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.0984462898473078, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.839357429718876e-05, | |
| "loss": 2.2033, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.0991159924993303, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.906291834002678e-05, | |
| "loss": 2.2192, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.0997856951513528, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.97322623828648e-05, | |
| "loss": 2.2329, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.1004553978033753, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.999995081044314e-05, | |
| "loss": 2.2754, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1011251004553978, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.999965020794615e-05, | |
| "loss": 2.3376, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.10179480310742031, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.99990763321247e-05, | |
| "loss": 2.2637, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.10246450575944281, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.999822918611533e-05, | |
| "loss": 2.2462, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.10313420841146531, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.999710877454811e-05, | |
| "loss": 2.2325, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.10380391106348781, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.999571510354664e-05, | |
| "loss": 2.3306, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.10447361371551031, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.999404818072808e-05, | |
| "loss": 2.2797, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.10514331636753281, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.999210801520296e-05, | |
| "loss": 2.4045, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.10581301901955532, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.998989461757526e-05, | |
| "loss": 2.3031, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.10648272167157782, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.998740799994235e-05, | |
| "loss": 2.2414, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.10715242432360032, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.998464817589484e-05, | |
| "loss": 2.288, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.10782212697562282, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.998161516051656e-05, | |
| "loss": 2.2926, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.10849182962764532, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.997830897038446e-05, | |
| "loss": 2.2471, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.10916153227966782, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.997472962356854e-05, | |
| "loss": 2.2067, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.10983123493169034, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.997087713963174e-05, | |
| "loss": 2.2404, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.11050093758371284, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.996675153962984e-05, | |
| "loss": 2.2364, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.11117064023573534, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.996235284611131e-05, | |
| "loss": 2.2742, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.11184034288775783, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.995768108311722e-05, | |
| "loss": 2.3221, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.11251004553978033, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.995273627618116e-05, | |
| "loss": 2.2253, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.11317974819180283, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.994751845232894e-05, | |
| "loss": 2.2679, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.11384945084382535, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.994202764007865e-05, | |
| "loss": 2.3354, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.11451915349584785, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.993626386944031e-05, | |
| "loss": 2.3038, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.11518885614787035, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.993022717191584e-05, | |
| "loss": 2.2999, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.11585855879989285, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.992391758049889e-05, | |
| "loss": 2.2606, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.11652826145191535, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.99173351296745e-05, | |
| "loss": 2.3125, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.11719796410393785, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.99104798554191e-05, | |
| "loss": 2.2503, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.11786766675596036, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.990335179520023e-05, | |
| "loss": 2.3219, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.11853736940798286, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.989595098797635e-05, | |
| "loss": 2.2829, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.11920707206000536, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.988827747419659e-05, | |
| "loss": 2.2715, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.11987677471202786, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.988033129580059e-05, | |
| "loss": 2.2071, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.12054647736405036, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.98721124962182e-05, | |
| "loss": 2.3135, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.12121618001607286, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.986362112036935e-05, | |
| "loss": 2.2781, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.12188588266809537, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.985485721466366e-05, | |
| "loss": 2.2473, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.12255558532011787, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.984582082700029e-05, | |
| "loss": 2.3274, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.12322528797214037, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.98365120067677e-05, | |
| "loss": 2.2683, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.12389499062416287, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.982693080484327e-05, | |
| "loss": 2.2058, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.12456469327618537, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.981707727359308e-05, | |
| "loss": 2.2458, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.12523439592820787, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.98069514668717e-05, | |
| "loss": 2.3385, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.12590409858023038, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.979655344002172e-05, | |
| "loss": 2.2759, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.12657380123225287, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.978588324987365e-05, | |
| "loss": 2.2529, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.12724350388427538, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.977494095474546e-05, | |
| "loss": 2.2715, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.1279132065362979, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.97637266144423e-05, | |
| "loss": 2.3065, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.12858290918832038, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.975224029025619e-05, | |
| "loss": 2.3046, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.1292526118403429, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.974048204496572e-05, | |
| "loss": 2.412, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.12992231449236538, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.97284519428356e-05, | |
| "loss": 2.2593, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.1305920171443879, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.971615004961645e-05, | |
| "loss": 2.2236, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.13126171979641038, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.970357643254429e-05, | |
| "loss": 2.3241, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.1319314224484329, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.96907311603403e-05, | |
| "loss": 2.2731, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.1326011251004554, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.967761430321037e-05, | |
| "loss": 2.2762, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.1332708277524779, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.966422593284474e-05, | |
| "loss": 2.2341, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.1339405304045004, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.965056612241764e-05, | |
| "loss": 2.3731, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1346102330565229, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.963663494658681e-05, | |
| "loss": 2.1989, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.1352799357085454, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.962243248149314e-05, | |
| "loss": 2.4027, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.13594963836056792, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.960795880476029e-05, | |
| "loss": 2.3882, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.1366193410125904, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.95932139954942e-05, | |
| "loss": 2.2943, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.13728904366461292, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.95781981342827e-05, | |
| "loss": 2.2777, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.1379587463166354, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.956291130319506e-05, | |
| "loss": 2.2989, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.13862844896865792, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.95473535857815e-05, | |
| "loss": 2.3316, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.1392981516206804, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.953152506707283e-05, | |
| "loss": 2.1754, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.13996785427270292, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.951542583357986e-05, | |
| "loss": 2.3177, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.14063755692472543, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.949905597329306e-05, | |
| "loss": 2.1837, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.14130725957674792, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.948241557568196e-05, | |
| "loss": 2.2382, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.14197696222877043, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.946550473169474e-05, | |
| "loss": 2.28, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.14264666488079292, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.944832353375769e-05, | |
| "loss": 2.2878, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.14331636753281543, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.943087207577473e-05, | |
| "loss": 2.3153, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.14398607018483794, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.941315045312689e-05, | |
| "loss": 2.2299, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.14465577283686043, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.939515876267178e-05, | |
| "loss": 2.3591, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.14532547548888294, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.937689710274308e-05, | |
| "loss": 2.2673, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.14599517814090543, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.935836557314995e-05, | |
| "loss": 2.2735, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.14666488079292794, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.933956427517657e-05, | |
| "loss": 2.288, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.14733458344495043, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.932049331158153e-05, | |
| "loss": 2.3028, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.14800428609697294, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.930115278659727e-05, | |
| "loss": 2.2032, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.14867398874899546, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.928154280592952e-05, | |
| "loss": 2.31, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.14934369140101794, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.92616634767567e-05, | |
| "loss": 2.2859, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.15001339405304046, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.924151490772942e-05, | |
| "loss": 2.2332, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.15068309670506294, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.922109720896973e-05, | |
| "loss": 2.3205, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.15135279935708545, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.92004104920707e-05, | |
| "loss": 2.2014, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.15202250200910797, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.917945487009566e-05, | |
| "loss": 2.3219, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.15269220466113045, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.915823045757765e-05, | |
| "loss": 2.2987, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.15336190731315297, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.913673737051882e-05, | |
| "loss": 2.2566, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.15403160996517545, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.911497572638973e-05, | |
| "loss": 2.2533, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.15470131261719797, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.909294564412874e-05, | |
| "loss": 2.3904, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.15537101526922045, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.907064724414139e-05, | |
| "loss": 2.1739, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.15604071792124297, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.904808064829967e-05, | |
| "loss": 2.2503, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.15671042057326548, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.902524597994143e-05, | |
| "loss": 2.3042, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.15738012322528797, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.900214336386964e-05, | |
| "loss": 2.206, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.15804982587731048, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.897877292635179e-05, | |
| "loss": 2.3278, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.15871952852933296, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.895513479511907e-05, | |
| "loss": 2.2708, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.15938923118135548, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.893122909936583e-05, | |
| "loss": 2.1588, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.160058933833378, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.890705596974875e-05, | |
| "loss": 2.2171, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.16072863648540048, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.88826155383862e-05, | |
| "loss": 2.274, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.161398339137423, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.885790793885746e-05, | |
| "loss": 2.1989, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.16206804178944548, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.883293330620205e-05, | |
| "loss": 2.2859, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.162737744441468, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.880769177691892e-05, | |
| "loss": 2.2734, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.16340744709349048, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.878218348896577e-05, | |
| "loss": 2.2998, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.164077149745513, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.875640858175827e-05, | |
| "loss": 2.3672, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.1647468523975355, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.87303671961693e-05, | |
| "loss": 2.2789, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.165416555049558, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.870405947452819e-05, | |
| "loss": 2.3171, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.1660862577015805, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.86774855606199e-05, | |
| "loss": 2.205, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.166755960353603, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.86506455996843e-05, | |
| "loss": 2.2462, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.1674256630056255, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.862353973841526e-05, | |
| "loss": 2.2796, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.16809536565764802, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.859616812496008e-05, | |
| "loss": 2.289, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.1687650683096705, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.856853090891843e-05, | |
| "loss": 2.2444, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.16943477096169302, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.854062824134159e-05, | |
| "loss": 2.2591, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.1701044736137155, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.851246027473173e-05, | |
| "loss": 2.3646, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.17077417626573801, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.848402716304106e-05, | |
| "loss": 2.2569, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.1714438789177605, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.845532906167083e-05, | |
| "loss": 2.2114, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.17211358156978301, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.842636612747069e-05, | |
| "loss": 2.2354, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.17278328422180553, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.839713851873766e-05, | |
| "loss": 2.2848, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.173452986873828, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.836764639521539e-05, | |
| "loss": 2.2848, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.17412268952585053, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.833788991809323e-05, | |
| "loss": 2.3165, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.174792392177873, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.830786925000533e-05, | |
| "loss": 2.2351, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.17546209482989553, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.827758455502978e-05, | |
| "loss": 2.2333, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.17613179748191804, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.824703599868776e-05, | |
| "loss": 2.3464, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.17680150013394053, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.821622374794253e-05, | |
| "loss": 2.305, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.17747120278596304, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.81851479711986e-05, | |
| "loss": 2.1684, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.17814090543798553, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.815380883830076e-05, | |
| "loss": 2.2124, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.17881060809000804, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.812220652053318e-05, | |
| "loss": 2.3608, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.17948031074203055, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.80903411906185e-05, | |
| "loss": 2.2411, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.18015001339405304, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.805821302271685e-05, | |
| "loss": 2.2363, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.18081971604607555, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.802582219242484e-05, | |
| "loss": 2.3204, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.18148941869809804, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.799316887677471e-05, | |
| "loss": 2.2496, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.18215912135012055, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.796025325423334e-05, | |
| "loss": 2.2741, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.18282882400214304, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.792707550470122e-05, | |
| "loss": 2.1737, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.18349852665416555, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.789363580951145e-05, | |
| "loss": 2.2353, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.18416822930618806, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.785993435142891e-05, | |
| "loss": 2.2664, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.18483793195821055, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.782597131464901e-05, | |
| "loss": 2.3049, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.18550763461023306, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.779174688479693e-05, | |
| "loss": 2.2721, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.18617733726225555, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.775726124892646e-05, | |
| "loss": 2.5128, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.18684703991427806, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.772251459551897e-05, | |
| "loss": 2.2106, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.18751674256630058, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.768750711448249e-05, | |
| "loss": 2.2744, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.18818644521832306, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.765223899715054e-05, | |
| "loss": 2.1929, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.18885614787034558, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.761671043628124e-05, | |
| "loss": 2.3493, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.18952585052236806, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.758092162605604e-05, | |
| "loss": 2.3092, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.19019555317439057, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.754487276207889e-05, | |
| "loss": 2.2649, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.19086525582641306, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.750856404137502e-05, | |
| "loss": 2.1621, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.19153495847843557, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.747199566238991e-05, | |
| "loss": 2.2921, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.1922046611304581, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.743516782498822e-05, | |
| "loss": 2.1954, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.19287436378248057, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.739808073045264e-05, | |
| "loss": 2.3438, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.1935440664345031, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.73607345814829e-05, | |
| "loss": 2.2986, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.19421376908652557, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.732312958219453e-05, | |
| "loss": 2.3054, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.1948834717385481, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.728526593811783e-05, | |
| "loss": 2.212, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.1955531743905706, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.724714385619673e-05, | |
| "loss": 2.1972, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.19622287704259309, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.720876354478765e-05, | |
| "loss": 2.2505, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.1968925796946156, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.717012521365836e-05, | |
| "loss": 2.2319, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.19756228234663809, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.713122907398686e-05, | |
| "loss": 2.3005, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.1982319849986606, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.709207533836016e-05, | |
| "loss": 2.0689, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.19890168765068308, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.70526642207732e-05, | |
| "loss": 2.2497, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.1995713903027056, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.701299593662763e-05, | |
| "loss": 2.315, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.2002410929547281, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.697307070273062e-05, | |
| "loss": 2.2616, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.2009107956067506, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.693288873729376e-05, | |
| "loss": 2.2162, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 14932, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.106646172434432e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |