| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9998749194192412, | |
| "eval_steps": 500, | |
| "global_step": 6495, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0015394533016462528, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 1.7262, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0030789066032925055, | |
| "grad_norm": 8.375, | |
| "learning_rate": 6.153846153846155e-07, | |
| "loss": 1.7972, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.004618359904938758, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 9.230769230769232e-07, | |
| "loss": 1.768, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.006157813206585011, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 1.7936, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.007697266508231265, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.7506, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.009236719809877517, | |
| "grad_norm": 5.25, | |
| "learning_rate": 1.8461538461538465e-06, | |
| "loss": 1.7089, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01077617311152377, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 2.153846153846154e-06, | |
| "loss": 1.6747, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.012315626413170022, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 2.461538461538462e-06, | |
| "loss": 1.6836, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.013855079714816277, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 1.6626, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01539453301646253, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.6116, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01693398631810878, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 3.384615384615385e-06, | |
| "loss": 1.5566, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.018473439619755033, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 3.692307692307693e-06, | |
| "loss": 1.53, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.020012892921401286, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.5322, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02155234622304754, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 4.307692307692308e-06, | |
| "loss": 1.5002, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.02309179952469379, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.4738, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.024631252826340044, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 4.923076923076924e-06, | |
| "loss": 1.4935, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0261707061279863, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 5.230769230769232e-06, | |
| "loss": 1.446, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.027710159429632553, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 5.538461538461539e-06, | |
| "loss": 1.4611, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.029249612731278806, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.846153846153847e-06, | |
| "loss": 1.4191, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03078906603292506, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.4419, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03232851933457131, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 6.461538461538463e-06, | |
| "loss": 1.4503, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.03386797263621756, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 6.76923076923077e-06, | |
| "loss": 1.4216, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.035407425937863814, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 7.076923076923078e-06, | |
| "loss": 1.4101, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.036946879239510066, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 7.384615384615386e-06, | |
| "loss": 1.4217, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.03848633254115632, | |
| "grad_norm": 2.375, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.4087, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04002578584280257, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.3453, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.041565239144448825, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 8.307692307692309e-06, | |
| "loss": 1.358, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.04310469244609508, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.615384615384617e-06, | |
| "loss": 1.377, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.04464414574774133, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.923076923076925e-06, | |
| "loss": 1.38, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.04618359904938758, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.3745, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.047723052351033836, | |
| "grad_norm": 2.0, | |
| "learning_rate": 9.53846153846154e-06, | |
| "loss": 1.3737, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.04926250565268009, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 9.846153846153848e-06, | |
| "loss": 1.3393, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.05080195895432634, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.0153846153846154e-05, | |
| "loss": 1.3267, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.0523414122559726, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.0461538461538463e-05, | |
| "loss": 1.3294, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.053880865557618854, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 1.3504, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.05542031885926511, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.1076923076923079e-05, | |
| "loss": 1.3404, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.05695977216091136, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.1384615384615385e-05, | |
| "loss": 1.3509, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.05849922546255761, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.1692307692307694e-05, | |
| "loss": 1.3597, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.060038678764203865, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.2981, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.06157813206585012, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 1.274, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06311758536749637, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.2615384615384616e-05, | |
| "loss": 1.275, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.06465703866914262, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.2923076923076925e-05, | |
| "loss": 1.298, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.06619649197078888, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.3230769230769231e-05, | |
| "loss": 1.3105, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.06773594527243512, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.353846153846154e-05, | |
| "loss": 1.3525, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.06927539857408138, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 1.3378, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.07081485187572763, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.4153846153846156e-05, | |
| "loss": 1.2971, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.07235430517737389, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.4461538461538462e-05, | |
| "loss": 1.3607, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.07389375847902013, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 1.4769230769230772e-05, | |
| "loss": 1.3031, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.07543321178066639, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.5076923076923078e-05, | |
| "loss": 1.2899, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.07697266508231264, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 1.2925, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0785121183839589, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.5692307692307693e-05, | |
| "loss": 1.2697, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.08005157168560514, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.2969, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.0815910249872514, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.630769230769231e-05, | |
| "loss": 1.2807, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.08313047828889765, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.6615384615384618e-05, | |
| "loss": 1.2498, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.08466993159054391, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 1.285, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.08620938489219016, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.7230769230769234e-05, | |
| "loss": 1.268, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.08774883819383641, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.753846153846154e-05, | |
| "loss": 1.3036, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.08928829149548266, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.784615384615385e-05, | |
| "loss": 1.2921, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.09082774479712892, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.8153846153846155e-05, | |
| "loss": 1.2433, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.09236719809877517, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 1.2554, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.09390665140042143, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.876923076923077e-05, | |
| "loss": 1.248, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.09544610470206767, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.907692307692308e-05, | |
| "loss": 1.2987, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.09698555800371393, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.9384615384615386e-05, | |
| "loss": 1.2726, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.09852501130536018, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.9692307692307696e-05, | |
| "loss": 1.2511, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.10006446460700644, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2196, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.10160391790865268, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9999855555944584e-05, | |
| "loss": 1.264, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.10314337121029894, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.999942222795114e-05, | |
| "loss": 1.2615, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.1046828245119452, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.9998700028538e-05, | |
| "loss": 1.291, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.10622227781359145, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.9997688978568653e-05, | |
| "loss": 1.2532, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.10776173111523771, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.9996389107251126e-05, | |
| "loss": 1.264, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.10930118441688395, | |
| "grad_norm": 2.375, | |
| "learning_rate": 1.9994800452137158e-05, | |
| "loss": 1.2498, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.11084063771853021, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9992923059121107e-05, | |
| "loss": 1.2506, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.11238009102017646, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.999075698243862e-05, | |
| "loss": 1.2375, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.11391954432182272, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.998830228466508e-05, | |
| "loss": 1.2478, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.11545899762346896, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.998555903671379e-05, | |
| "loss": 1.2518, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.11699845092511522, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9982527317833922e-05, | |
| "loss": 1.2227, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.11853790422676147, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.9979207215608224e-05, | |
| "loss": 1.2787, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.12007735752840773, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9975598825950507e-05, | |
| "loss": 1.2991, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.12161681083005398, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.9971702253102856e-05, | |
| "loss": 1.2403, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.12315626413170024, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.996751760963263e-05, | |
| "loss": 1.2529, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.12469571743334648, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9963045016429202e-05, | |
| "loss": 1.2516, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.12623517073499274, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.995828460270047e-05, | |
| "loss": 1.2235, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.127774624036639, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.995323650596913e-05, | |
| "loss": 1.2124, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.12931407733828523, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.9947900872068696e-05, | |
| "loss": 1.2891, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.1308535306399315, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9942277855139284e-05, | |
| "loss": 1.2427, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.13239298394157775, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.993636761762317e-05, | |
| "loss": 1.2534, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.133932437243224, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9930170330260093e-05, | |
| "loss": 1.2416, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.13547189054487024, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.9923686172082308e-05, | |
| "loss": 1.2763, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.1370113438465165, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.991691533040944e-05, | |
| "loss": 1.2747, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.13855079714816276, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9909858000843058e-05, | |
| "loss": 1.2531, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.14009025044980902, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9902514387261017e-05, | |
| "loss": 1.3041, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.14162970375145525, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.989488470181159e-05, | |
| "loss": 1.2637, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.14316915705310151, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.9886969164907306e-05, | |
| "loss": 1.2539, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.14470861035474777, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.987876800521863e-05, | |
| "loss": 1.2786, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.14624806365639403, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9870281459667304e-05, | |
| "loss": 1.2926, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.14778751695804027, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.9861509773419544e-05, | |
| "loss": 1.2558, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.14932697025968653, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9852453199878937e-05, | |
| "loss": 1.2615, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.15086642356133279, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.9843112000679127e-05, | |
| "loss": 1.2784, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.15240587686297905, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.9833486445676245e-05, | |
| "loss": 1.2658, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.15394533016462528, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.982357681294114e-05, | |
| "loss": 1.2886, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.15548478346627154, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9813383388751313e-05, | |
| "loss": 1.2468, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.1570242367679178, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9802906467582674e-05, | |
| "loss": 1.2381, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.15856369006956406, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9792146352101017e-05, | |
| "loss": 1.2694, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.1601031433712103, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.9781103353153292e-05, | |
| "loss": 1.2631, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.16164259667285655, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.97697777897586e-05, | |
| "loss": 1.2477, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.1631820499745028, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.975816998909901e-05, | |
| "loss": 1.2353, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.16472150327614907, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.974628028651007e-05, | |
| "loss": 1.2681, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.1662609565777953, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9734109025471168e-05, | |
| "loss": 1.3015, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.16780040987944156, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.9721656557595556e-05, | |
| "loss": 1.25, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.16933986318108782, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.9708923242620228e-05, | |
| "loss": 1.2688, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.17087931648273408, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.9695909448395515e-05, | |
| "loss": 1.2725, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.1724187697843803, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.9682615550874458e-05, | |
| "loss": 1.2741, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.17395822308602657, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.966904193410195e-05, | |
| "loss": 1.272, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.17549767638767283, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9655188990203648e-05, | |
| "loss": 1.2651, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.1770371296893191, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.9641057119374623e-05, | |
| "loss": 1.2183, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.17857658299096532, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.962664672986783e-05, | |
| "loss": 1.2264, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.18011603629261158, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9611958237982285e-05, | |
| "loss": 1.2697, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.18165548959425784, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.9596992068051054e-05, | |
| "loss": 1.2831, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.1831949428959041, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.9581748652428995e-05, | |
| "loss": 1.2868, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.18473439619755033, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.9566228431480256e-05, | |
| "loss": 1.2707, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.1862738494991966, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.9550431853565577e-05, | |
| "loss": 1.2572, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.18781330280084285, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9534359375029308e-05, | |
| "loss": 1.2409, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.1893527561024891, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9518011460186246e-05, | |
| "loss": 1.2489, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.19089220940413534, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.9501388581308215e-05, | |
| "loss": 1.2392, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.1924316627057816, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.9484491218610423e-05, | |
| "loss": 1.2299, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.19397111600742786, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.946731986023759e-05, | |
| "loss": 1.2723, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.19551056930907412, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9449875002249835e-05, | |
| "loss": 1.264, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.19705002261072035, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.943215714860838e-05, | |
| "loss": 1.2613, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.19858947591236661, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.941416681116094e-05, | |
| "loss": 1.2534, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.20012892921401287, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.939590450962698e-05, | |
| "loss": 1.2561, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.20166838251565913, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9377370771582677e-05, | |
| "loss": 1.2618, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.20320783581730537, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.935856613244569e-05, | |
| "loss": 1.2488, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.20474728911895163, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.9339491135459683e-05, | |
| "loss": 1.2258, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.20628674242059789, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.932014633167864e-05, | |
| "loss": 1.2749, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.20782619572224414, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.930053227995095e-05, | |
| "loss": 1.1903, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.2093656490238904, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.928064954690324e-05, | |
| "loss": 1.3004, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.21090510232553664, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.9260498706924036e-05, | |
| "loss": 1.2255, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.2124445556271829, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.924008034214714e-05, | |
| "loss": 1.2626, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.21398400892882916, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.921939504243484e-05, | |
| "loss": 1.2563, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.21552346223047542, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.919844340536085e-05, | |
| "loss": 1.2406, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.21706291553212165, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.9177226036193058e-05, | |
| "loss": 1.2178, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.2186023688337679, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.9155743547876026e-05, | |
| "loss": 1.2385, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.22014182213541417, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.9133996561013305e-05, | |
| "loss": 1.2643, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.22168127543706043, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9111985703849493e-05, | |
| "loss": 1.2659, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.22322072873870666, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9089711612252082e-05, | |
| "loss": 1.2504, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.22476018204035292, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.9067174929693095e-05, | |
| "loss": 1.2148, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.22629963534199918, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.9044376307230496e-05, | |
| "loss": 1.3097, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.22783908864364544, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.9021316403489387e-05, | |
| "loss": 1.2139, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.22937854194529167, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.8997995884642968e-05, | |
| "loss": 1.2255, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.23091799524693793, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.8974415424393302e-05, | |
| "loss": 1.2577, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2324574485485842, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.895057570395185e-05, | |
| "loss": 1.2222, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.23399690185023045, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.8926477412019793e-05, | |
| "loss": 1.2298, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.23553635515187668, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.8902121244768135e-05, | |
| "loss": 1.2368, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.23707580845352294, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.887750790581759e-05, | |
| "loss": 1.2801, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.2386152617551692, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.8852638106218254e-05, | |
| "loss": 1.222, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.24015471505681546, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.8827512564429076e-05, | |
| "loss": 1.2335, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.2416941683584617, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.8802132006297082e-05, | |
| "loss": 1.2756, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.24323362166010795, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.8776497165036415e-05, | |
| "loss": 1.2417, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.2447730749617542, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.8750608781207172e-05, | |
| "loss": 1.2156, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.24631252826340047, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.8724467602693978e-05, | |
| "loss": 1.2686, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.2478519815650467, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8698074384684403e-05, | |
| "loss": 1.2667, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.24939143486669296, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.867142988964713e-05, | |
| "loss": 1.2481, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.2509308881683392, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.8644534887309947e-05, | |
| "loss": 1.1874, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.2524703414699855, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.8617390154637494e-05, | |
| "loss": 1.2225, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.2540097947716317, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.858999647580883e-05, | |
| "loss": 1.2612, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.255549248073278, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.856235464219476e-05, | |
| "loss": 1.2158, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.25708870137492423, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.8534465452335e-05, | |
| "loss": 1.2551, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.25862815467657047, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.850632971191508e-05, | |
| "loss": 1.2419, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.26016760797821675, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.8477948233743098e-05, | |
| "loss": 1.2247, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.261707061279863, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.8449321837726207e-05, | |
| "loss": 1.227, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.2632465145815092, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.842045135084696e-05, | |
| "loss": 1.2355, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.2647859678831555, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.8391337607139396e-05, | |
| "loss": 1.2215, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.26632542118480174, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.8361981447664955e-05, | |
| "loss": 1.2323, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.267864874486448, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8332383720488188e-05, | |
| "loss": 1.2376, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.26940432778809426, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.8302545280652234e-05, | |
| "loss": 1.2654, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.2709437810897405, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.8272466990154155e-05, | |
| "loss": 1.242, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.2724832343913868, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8242149717919993e-05, | |
| "loss": 1.208, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.274022687693033, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.8211594339779712e-05, | |
| "loss": 1.2327, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.27556214099467924, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.8180801738441843e-05, | |
| "loss": 1.275, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.2771015942963255, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.814977280346804e-05, | |
| "loss": 1.2041, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.27864104759797176, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.8118508431247338e-05, | |
| "loss": 1.2476, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.28018050089961805, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.808700952497028e-05, | |
| "loss": 1.2273, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.2817199542012643, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.8055276994602825e-05, | |
| "loss": 1.2121, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.2832594075029105, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.8023311756860036e-05, | |
| "loss": 1.2607, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.2847988608045568, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.799111473517964e-05, | |
| "loss": 1.2393, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.28633831410620303, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7958686859695307e-05, | |
| "loss": 1.2149, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.28787776740784926, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.7926029067209808e-05, | |
| "loss": 1.237, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.28941722070949555, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.789314230116794e-05, | |
| "loss": 1.2456, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.2909566740111418, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.786002751162927e-05, | |
| "loss": 1.2625, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.29249612731278807, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.7826685655240702e-05, | |
| "loss": 1.2202, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2940355806144343, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.7793117695208828e-05, | |
| "loss": 1.2714, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.29557503391608053, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.7759324601272098e-05, | |
| "loss": 1.2527, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.2971144872177268, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.7725307349672826e-05, | |
| "loss": 1.247, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.29865394051937305, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.7691066923128962e-05, | |
| "loss": 1.2389, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.3001933938210193, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.765660431080572e-05, | |
| "loss": 1.2449, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.30173284712266557, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.7621920508287e-05, | |
| "loss": 1.2037, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.3032723004243118, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.7587016517546614e-05, | |
| "loss": 1.2367, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.3048117537259581, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.7551893346919363e-05, | |
| "loss": 1.2234, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.3063512070276043, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.751655201107189e-05, | |
| "loss": 1.2423, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.30789066032925055, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.748099353097336e-05, | |
| "loss": 1.2393, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.30943011363089684, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.7445218933865994e-05, | |
| "loss": 1.2416, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.3109695669325431, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.7409229253235365e-05, | |
| "loss": 1.2058, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.31250902023418936, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.7373025528780568e-05, | |
| "loss": 1.1996, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.3140484735358356, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.733660880638415e-05, | |
| "loss": 1.229, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.3155879268374818, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.729998013808192e-05, | |
| "loss": 1.2154, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.3171273801391281, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.726314058203257e-05, | |
| "loss": 1.2293, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.31866683344077434, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.7226091202487068e-05, | |
| "loss": 1.242, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.3202062867424206, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.718883306975794e-05, | |
| "loss": 1.2323, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.32174574004406686, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.7151367260188348e-05, | |
| "loss": 1.1962, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.3232851933457131, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 1.7113694856120983e-05, | |
| "loss": 1.2355, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.3248246466473594, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.7075816945866805e-05, | |
| "loss": 1.2173, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.3263640999490056, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.7037734623673616e-05, | |
| "loss": 1.2464, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.32790355325065185, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.6999448989694424e-05, | |
| "loss": 1.2087, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.32944300655229813, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.6960961149955676e-05, | |
| "loss": 1.2423, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.33098245985394437, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 1.6922272216325302e-05, | |
| "loss": 1.2444, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.3325219131555906, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.68833833064806e-05, | |
| "loss": 1.2362, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.3340613664572369, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.6844295543875936e-05, | |
| "loss": 1.2213, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.3356008197588831, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.6805010057710306e-05, | |
| "loss": 1.2586, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.3371402730605294, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.676552798289469e-05, | |
| "loss": 1.2177, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.33867972636217564, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.6725850460019293e-05, | |
| "loss": 1.2562, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.34021917966382187, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.6685978635320578e-05, | |
| "loss": 1.2212, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.34175863296546816, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.6645913660648153e-05, | |
| "loss": 1.2413, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.3432980862671144, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.6605656693431504e-05, | |
| "loss": 1.2449, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.3448375395687606, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.6565208896646554e-05, | |
| "loss": 1.2832, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.3463769928704069, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.652457143878206e-05, | |
| "loss": 1.2153, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.34791644617205314, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.648374549380587e-05, | |
| "loss": 1.2656, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.3494558994736994, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.6442732241130987e-05, | |
| "loss": 1.2398, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.35099535277534566, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.640153286558153e-05, | |
| "loss": 1.2344, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.3525348060769919, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.6360148557358468e-05, | |
| "loss": 1.2415, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.3540742593786382, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.6318580512005266e-05, | |
| "loss": 1.2392, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3556137126802844, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.6276829930373335e-05, | |
| "loss": 1.2433, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.35715316598193064, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.6234898018587336e-05, | |
| "loss": 1.2768, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.35869261928357693, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.6192785988010354e-05, | |
| "loss": 1.2245, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.36023207258522316, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.6150495055208876e-05, | |
| "loss": 1.2241, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.36177152588686945, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.610802644191768e-05, | |
| "loss": 1.2334, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.3633109791885157, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.60653813750045e-05, | |
| "loss": 1.2278, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.3648504324901619, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.6022561086434628e-05, | |
| "loss": 1.2763, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.3663898857918082, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.5979566813235288e-05, | |
| "loss": 1.2296, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.36792933909345443, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.5936399797459916e-05, | |
| "loss": 1.201, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.36946879239510066, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.5893061286152275e-05, | |
| "loss": 1.2525, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.37100824569674695, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.5849552531310432e-05, | |
| "loss": 1.217, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.3725476989983932, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.5805874789850586e-05, | |
| "loss": 1.2368, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.37408715230003947, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.576202932357076e-05, | |
| "loss": 1.2079, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.3756266056016857, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.571801739911434e-05, | |
| "loss": 1.2388, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.37716605890333194, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.5673840287933504e-05, | |
| "loss": 1.2795, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.3787055122049782, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.562949926625247e-05, | |
| "loss": 1.2192, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.38024496550662445, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.5584995615030634e-05, | |
| "loss": 1.239, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.3817844188082707, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.554033061992557e-05, | |
| "loss": 1.2019, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.383323872109917, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.5495505571255884e-05, | |
| "loss": 1.2104, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.3848633254115632, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.5450521763963943e-05, | |
| "loss": 1.2426, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3864027787132095, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.5405380497578445e-05, | |
| "loss": 1.2542, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.3879422320148557, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.536008307617692e-05, | |
| "loss": 1.2149, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.38948168531650196, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.5314630808348013e-05, | |
| "loss": 1.2396, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.39102113861814825, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.52690250071537e-05, | |
| "loss": 1.2132, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.3925605919197945, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.522326699009136e-05, | |
| "loss": 1.2143, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.3941000452214407, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.5177358079055697e-05, | |
| "loss": 1.2405, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.395639498523087, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.513129960030058e-05, | |
| "loss": 1.2648, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.39717895182473323, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.5085092884400689e-05, | |
| "loss": 1.2312, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.3987184051263795, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.5038739266213118e-05, | |
| "loss": 1.2176, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.40025785842802575, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.4992240084838786e-05, | |
| "loss": 1.2382, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.401797311729672, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.4945596683583754e-05, | |
| "loss": 1.2135, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.40333676503131827, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.4898810409920432e-05, | |
| "loss": 1.2252, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.4048762183329645, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.485188261544864e-05, | |
| "loss": 1.2436, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.40641567163461073, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.480481465585657e-05, | |
| "loss": 1.2567, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.407955124936257, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.4757607890881615e-05, | |
| "loss": 1.2577, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.40949457823790325, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.4710263684271087e-05, | |
| "loss": 1.2255, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.41103403153954954, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.4662783403742826e-05, | |
| "loss": 1.2305, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.41257348484119577, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.4615168420945689e-05, | |
| "loss": 1.2841, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.414112938142842, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.4567420111419919e-05, | |
| "loss": 1.216, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.4156523914444883, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.4519539854557412e-05, | |
| "loss": 1.1817, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.4171918447461345, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.4471529033561856e-05, | |
| "loss": 1.2346, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.4187312980477808, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.442338903540879e-05, | |
| "loss": 1.2352, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.42027075134942704, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.4375121250805532e-05, | |
| "loss": 1.2227, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.4218102046510733, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.432672707415099e-05, | |
| "loss": 1.1777, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.42334965795271956, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.4278207903495388e-05, | |
| "loss": 1.2389, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.4248891112543658, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.4229565140499885e-05, | |
| "loss": 1.1987, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.426428564556012, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.418080019039607e-05, | |
| "loss": 1.2211, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.4279680178576583, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.4131914461945377e-05, | |
| "loss": 1.2202, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.42950747115930454, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.408290936739838e-05, | |
| "loss": 1.2389, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.43104692446095083, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.4033786322453989e-05, | |
| "loss": 1.2537, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.43258637776259706, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.3984546746218576e-05, | |
| "loss": 1.2682, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.4341258310642433, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.3935192061164957e-05, | |
| "loss": 1.2386, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.4356652843658896, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.3885723693091304e-05, | |
| "loss": 1.2681, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.4372047376675358, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.3836143071079954e-05, | |
| "loss": 1.2176, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.43874419096918205, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.3786451627456134e-05, | |
| "loss": 1.2285, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.44028364427082833, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.3736650797746568e-05, | |
| "loss": 1.2128, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.44182309757247457, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.3686742020638023e-05, | |
| "loss": 1.2559, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.44336255087412085, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.3636726737935724e-05, | |
| "loss": 1.2586, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.4449020041757671, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.3586606394521729e-05, | |
| "loss": 1.2085, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.4464414574774133, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.353638243831317e-05, | |
| "loss": 1.244, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.4479809107790596, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.348605632022043e-05, | |
| "loss": 1.2222, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.44952036408070584, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.3435629494105226e-05, | |
| "loss": 1.218, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.45105981738235207, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.3385103416738614e-05, | |
| "loss": 1.2392, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.45259927068399836, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.3334479547758894e-05, | |
| "loss": 1.2327, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.4541387239856446, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.3283759349629457e-05, | |
| "loss": 1.2302, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.4556781772872909, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.323294428759652e-05, | |
| "loss": 1.217, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.4572176305889371, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.3182035829646818e-05, | |
| "loss": 1.2292, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.45875708389058334, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.3131035446465165e-05, | |
| "loss": 1.2211, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.4602965371922296, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.3079944611392005e-05, | |
| "loss": 1.2049, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.46183599049387586, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.3028764800380818e-05, | |
| "loss": 1.2315, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4633754437955221, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.2977497491955493e-05, | |
| "loss": 1.177, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.4649148970971684, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.292614416716762e-05, | |
| "loss": 1.2158, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.4664543503988146, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2874706309553697e-05, | |
| "loss": 1.2513, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.4679938037004609, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2823185405092274e-05, | |
| "loss": 1.2859, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.46953325700210713, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.277158294216103e-05, | |
| "loss": 1.2705, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.47107271030375336, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.2719900411493764e-05, | |
| "loss": 1.2111, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.47261216360539965, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.2668139306137343e-05, | |
| "loss": 1.2126, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.4741516169070459, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.261630112140856e-05, | |
| "loss": 1.2502, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.4756910702086921, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.2564387354850949e-05, | |
| "loss": 1.2279, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.4772305235103384, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.251239950619149e-05, | |
| "loss": 1.2369, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.47876997681198463, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.2460339077297335e-05, | |
| "loss": 1.2727, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.4803094301136309, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.2408207572132367e-05, | |
| "loss": 1.2004, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.48184888341527715, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.2356006496713798e-05, | |
| "loss": 1.2237, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.4833883367169234, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.2303737359068632e-05, | |
| "loss": 1.1884, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.48492779001856967, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2251401669190115e-05, | |
| "loss": 1.2038, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.4864672433202159, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.2199000938994101e-05, | |
| "loss": 1.2383, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.48800669662186213, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.2146536682275388e-05, | |
| "loss": 1.2389, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.4895461499235084, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.2094010414663972e-05, | |
| "loss": 1.2236, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.49108560322515465, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2041423653581284e-05, | |
| "loss": 1.2085, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.49262505652680094, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.1988777918196324e-05, | |
| "loss": 1.241, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.4941645098284472, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.1936074729381795e-05, | |
| "loss": 1.2456, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.4957039631300934, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.1883315609670163e-05, | |
| "loss": 1.2387, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.4972434164317397, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.1830502083209674e-05, | |
| "loss": 1.244, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.4987828697333859, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.1777635675720313e-05, | |
| "loss": 1.2172, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.5003223230350322, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.1724717914449744e-05, | |
| "loss": 1.2698, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.5018617763366784, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.1671750328129168e-05, | |
| "loss": 1.2265, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.5034012296383247, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.161873444692918e-05, | |
| "loss": 1.194, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.504940682939971, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.1565671802415564e-05, | |
| "loss": 1.2437, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.5064801362416171, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1512563927505025e-05, | |
| "loss": 1.1946, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.5080195895432634, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.1459412356420936e-05, | |
| "loss": 1.2534, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.5095590428449097, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.1406218624648986e-05, | |
| "loss": 1.2192, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.511098496146556, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.1352984268892844e-05, | |
| "loss": 1.1832, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.5126379494482022, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.129971082702976e-05, | |
| "loss": 1.2234, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.5141774027498485, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.124639983806613e-05, | |
| "loss": 1.2296, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.5157168560514948, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.1193052842093043e-05, | |
| "loss": 1.2196, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.5172563093531409, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.1139671380241786e-05, | |
| "loss": 1.2308, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.5187957626547872, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.1086256994639328e-05, | |
| "loss": 1.217, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.5203352159564335, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.1032811228363766e-05, | |
| "loss": 1.2209, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.5218746692580797, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.0979335625399739e-05, | |
| "loss": 1.2224, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.523414122559726, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.0925831730593843e-05, | |
| "loss": 1.1895, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.5249535758613723, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.087230108960999e-05, | |
| "loss": 1.246, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.5264930291630184, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.0818745248884746e-05, | |
| "loss": 1.2262, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.5280324824646647, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.076516575558268e-05, | |
| "loss": 1.2029, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.529571935766311, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.0711564157551654e-05, | |
| "loss": 1.1688, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.5311113890679573, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.0657942003278107e-05, | |
| "loss": 1.211, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.5326508423696035, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.0604300841842324e-05, | |
| "loss": 1.2085, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.5341902956712498, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.0550642222873681e-05, | |
| "loss": 1.2051, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.535729748972896, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.0496967696505888e-05, | |
| "loss": 1.1968, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.5372692022745422, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.0443278813332197e-05, | |
| "loss": 1.2471, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.5388086555761885, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.038957712436061e-05, | |
| "loss": 1.2299, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5403481088778348, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.0335864180969072e-05, | |
| "loss": 1.2289, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.541887562179481, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.0282141534860662e-05, | |
| "loss": 1.2811, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.5434270154811273, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.022841073801875e-05, | |
| "loss": 1.2566, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.5449664687827735, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.0174673342662182e-05, | |
| "loss": 1.2111, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.5465059220844197, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.0120930901200415e-05, | |
| "loss": 1.1946, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.548045375386066, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.0067184966188687e-05, | |
| "loss": 1.209, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.5495848286877123, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.0013437090283164e-05, | |
| "loss": 1.2233, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.5511242819893585, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 9.959688826196078e-06, | |
| "loss": 1.2574, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.5526637352910048, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 9.90594172665088e-06, | |
| "loss": 1.2208, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.554203188592651, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.852197344337366e-06, | |
| "loss": 1.2289, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.5557426418942973, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.798457231866853e-06, | |
| "loss": 1.2027, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.5572820951959435, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 9.744722941727292e-06, | |
| "loss": 1.2382, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.5588215484975898, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 9.690996026238443e-06, | |
| "loss": 1.2074, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.5603610017992361, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 9.637278037507013e-06, | |
| "loss": 1.2176, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.5619004551008823, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 9.583570527381838e-06, | |
| "loss": 1.2196, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.5634399084025286, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 9.529875047409027e-06, | |
| "loss": 1.2159, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.5649793617041748, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 9.476193148787155e-06, | |
| "loss": 1.2233, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.566518815005821, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 9.422526382322454e-06, | |
| "loss": 1.2253, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.5680582683074673, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 9.368876298383999e-06, | |
| "loss": 1.1681, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.5695977216091136, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 9.315244446858932e-06, | |
| "loss": 1.2346, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5711371749107598, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 9.261632377107674e-06, | |
| "loss": 1.2196, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.5726766282124061, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 9.20804163791919e-06, | |
| "loss": 1.2159, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.5742160815140523, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 9.154473777466209e-06, | |
| "loss": 1.2226, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.5757555348156985, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 9.100930343260544e-06, | |
| "loss": 1.2466, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.5772949881173448, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 9.047412882108341e-06, | |
| "loss": 1.1791, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.5788344414189911, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 8.993922940065426e-06, | |
| "loss": 1.2507, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.5803738947206374, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 8.940462062392635e-06, | |
| "loss": 1.2668, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.5819133480222836, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 8.887031793511154e-06, | |
| "loss": 1.2039, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.5834528013239298, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 8.833633676957939e-06, | |
| "loss": 1.2296, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.5849922546255761, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 8.78026925534108e-06, | |
| "loss": 1.205, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5865317079272223, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 8.726940070295282e-06, | |
| "loss": 1.1951, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.5880711612288686, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 8.673647662437288e-06, | |
| "loss": 1.1984, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.5896106145305149, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 8.620393571321408e-06, | |
| "loss": 1.2233, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.5911500678321611, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 8.567179335395012e-06, | |
| "loss": 1.2112, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.5926895211338074, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 8.514006491954118e-06, | |
| "loss": 1.2664, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.5942289744354536, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 8.460876577098951e-06, | |
| "loss": 1.1851, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.5957684277370998, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 8.407791125689577e-06, | |
| "loss": 1.1998, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.5973078810387461, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 8.354751671301584e-06, | |
| "loss": 1.1966, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.5988473343403924, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 8.30175974618174e-06, | |
| "loss": 1.2512, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.6003867876420386, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 8.248816881203771e-06, | |
| "loss": 1.2015, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.6019262409436849, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 8.195924605824088e-06, | |
| "loss": 1.218, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.6034656942453311, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 8.143084448037654e-06, | |
| "loss": 1.1875, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.6050051475469774, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 8.090297934333795e-06, | |
| "loss": 1.2273, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.6065446008486236, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 8.037566589652141e-06, | |
| "loss": 1.1902, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.6080840541502699, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 7.984891937338546e-06, | |
| "loss": 1.2323, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.6096235074519162, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 7.932275499101081e-06, | |
| "loss": 1.2546, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.6111629607535624, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.8797187949661e-06, | |
| "loss": 1.2519, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.6127024140552086, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 7.827223343234298e-06, | |
| "loss": 1.184, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.6142418673568549, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 7.774790660436857e-06, | |
| "loss": 1.2536, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.6157813206585011, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 7.72242226129165e-06, | |
| "loss": 1.2455, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.6173207739601474, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.670119658659469e-06, | |
| "loss": 1.2347, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.6188602272617937, | |
| "grad_norm": 1.625, | |
| "learning_rate": 7.617884363500313e-06, | |
| "loss": 1.1972, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.6203996805634399, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 7.565717884829764e-06, | |
| "loss": 1.2546, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.6219391338650861, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 7.513621729675362e-06, | |
| "loss": 1.2472, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.6234785871667324, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 7.4615974030330985e-06, | |
| "loss": 1.2416, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.6250180404683787, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 7.409646407823914e-06, | |
| "loss": 1.2288, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.6265574937700249, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 7.357770244850291e-06, | |
| "loss": 1.2328, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.6280969470716712, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.30597041275291e-06, | |
| "loss": 1.2831, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.6296364003733175, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 7.254248407967323e-06, | |
| "loss": 1.2569, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.6311758536749636, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 7.202605724680769e-06, | |
| "loss": 1.2089, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.6327153069766099, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 7.151043854788959e-06, | |
| "loss": 1.2329, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.6342547602782562, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.099564287853016e-06, | |
| "loss": 1.2164, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.6357942135799024, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 7.048168511056421e-06, | |
| "loss": 1.2881, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.6373336668815487, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 6.996858009162064e-06, | |
| "loss": 1.2515, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.638873120183195, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 6.945634264469338e-06, | |
| "loss": 1.2581, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.6404125734848412, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 6.894498756771326e-06, | |
| "loss": 1.2559, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.6419520267864874, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 6.843452963312055e-06, | |
| "loss": 1.2397, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.6434914800881337, | |
| "grad_norm": 1.625, | |
| "learning_rate": 6.7924983587438e-06, | |
| "loss": 1.2234, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.6450309333897799, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 6.741636415084514e-06, | |
| "loss": 1.2609, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.6465703866914262, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 6.690868601675273e-06, | |
| "loss": 1.2129, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.6481098399930725, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 6.640196385137854e-06, | |
| "loss": 1.2219, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.6496492932947188, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 6.58962122933234e-06, | |
| "loss": 1.2781, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.6511887465963649, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 6.5391445953148615e-06, | |
| "loss": 1.2453, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.6527281998980112, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 6.4887679412953546e-06, | |
| "loss": 1.2456, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.6542676531996575, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 6.438492722595467e-06, | |
| "loss": 1.2498, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.6558071065013037, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 6.38832039160649e-06, | |
| "loss": 1.2614, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.65734655980295, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 6.338252397747417e-06, | |
| "loss": 1.2455, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.6588860131045963, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 6.2882901874230604e-06, | |
| "loss": 1.2504, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.6604254664062424, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 6.238435203982278e-06, | |
| "loss": 1.1987, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.6619649197078887, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 6.1886888876762725e-06, | |
| "loss": 1.2427, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.663504373009535, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 6.1390526756169675e-06, | |
| "loss": 1.2541, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.6650438263111812, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 6.089528001735527e-06, | |
| "loss": 1.1983, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.6665832796128275, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 6.04011629674089e-06, | |
| "loss": 1.2211, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.6681227329144738, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 5.990818988078475e-06, | |
| "loss": 1.2224, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.66966218621612, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 5.941637499888915e-06, | |
| "loss": 1.2194, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.6712016395177662, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 5.892573252966926e-06, | |
| "loss": 1.196, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.6727410928194125, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 5.843627664720279e-06, | |
| "loss": 1.2119, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.6742805461210588, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 5.794802149128818e-06, | |
| "loss": 1.2326, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.675819999422705, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 5.746098116703647e-06, | |
| "loss": 1.2473, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.6773594527243513, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 5.697516974446344e-06, | |
| "loss": 1.244, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.6788989060259976, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 5.6490601258083675e-06, | |
| "loss": 1.2076, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.6804383593276437, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 5.600728970650446e-06, | |
| "loss": 1.2208, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.68197781262929, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 5.552524905202208e-06, | |
| "loss": 1.2155, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.6835172659309363, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 5.504449322021774e-06, | |
| "loss": 1.2246, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.6850567192325825, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 5.456503609955606e-06, | |
| "loss": 1.1766, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.6865961725342288, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 5.408689154098306e-06, | |
| "loss": 1.2345, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.6881356258358751, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 5.361007335752662e-06, | |
| "loss": 1.2412, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.6896750791375212, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 5.313459532389713e-06, | |
| "loss": 1.2205, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.6912145324391675, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 5.266047117608973e-06, | |
| "loss": 1.171, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.6927539857408138, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 5.218771461098733e-06, | |
| "loss": 1.1932, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.69429343904246, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 5.1716339285965035e-06, | |
| "loss": 1.2628, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.6958328923441063, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 5.124635881849558e-06, | |
| "loss": 1.2283, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.6973723456457526, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.077778678575581e-06, | |
| "loss": 1.2474, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.6989117989473989, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 5.0310636724234796e-06, | |
| "loss": 1.2207, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.700451252249045, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 4.984492212934231e-06, | |
| "loss": 1.2369, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.7019907055506913, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 4.938065645501934e-06, | |
| "loss": 1.231, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.7035301588523376, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 4.891785311334924e-06, | |
| "loss": 1.2238, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.7050696121539838, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 4.84565254741703e-06, | |
| "loss": 1.2368, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.7066090654556301, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 4.7996686864689565e-06, | |
| "loss": 1.2171, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.7081485187572764, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 4.753835056909774e-06, | |
| "loss": 1.2543, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.7096879720589225, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 4.708152982818548e-06, | |
| "loss": 1.2192, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.7112274253605688, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 4.662623783896088e-06, | |
| "loss": 1.2161, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.7127668786622151, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 4.617248775426818e-06, | |
| "loss": 1.2368, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.7143063319638613, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.572029268240787e-06, | |
| "loss": 1.2, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.7158457852655076, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 4.526966568675802e-06, | |
| "loss": 1.2892, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.7173852385671539, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 4.482061978539665e-06, | |
| "loss": 1.1981, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.7189246918688001, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 4.437316795072605e-06, | |
| "loss": 1.217, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.7204641451704463, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 4.392732310909771e-06, | |
| "loss": 1.2843, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.7220035984720926, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 4.3483098140439066e-06, | |
| "loss": 1.2294, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.7235430517737389, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 4.304050587788129e-06, | |
| "loss": 1.2224, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.7250825050753851, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 4.2599559107388645e-06, | |
| "loss": 1.2226, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.7266219583770314, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 4.216027056738909e-06, | |
| "loss": 1.2431, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.7281614116786777, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 4.172265294840628e-06, | |
| "loss": 1.2082, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.7297008649803238, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 4.128671889269292e-06, | |
| "loss": 1.1767, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.7312403182819701, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 4.0852480993865625e-06, | |
| "loss": 1.1983, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.7327797715836164, | |
| "grad_norm": 1.625, | |
| "learning_rate": 4.0419951796540915e-06, | |
| "loss": 1.1887, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.7343192248852626, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 3.9989143795973206e-06, | |
| "loss": 1.2006, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.7358586781869089, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 3.956006943769331e-06, | |
| "loss": 1.2182, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.7373981314885552, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 3.913274111714942e-06, | |
| "loss": 1.1946, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.7389375847902013, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.870717117934857e-06, | |
| "loss": 1.2858, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.7404770380918476, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 3.8283371918500325e-06, | |
| "loss": 1.2157, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.7420164913934939, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.786135557766146e-06, | |
| "loss": 1.2178, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.7435559446951402, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 3.7441134348382345e-06, | |
| "loss": 1.2035, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.7450953979967864, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.702272037035468e-06, | |
| "loss": 1.2462, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.7466348512984327, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 3.6606125731060815e-06, | |
| "loss": 1.1922, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.7481743046000789, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 3.6191362465424617e-06, | |
| "loss": 1.2296, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.7497137579017251, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 3.5778442555463633e-06, | |
| "loss": 1.2526, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.7512532112033714, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 3.5367377929943246e-06, | |
| "loss": 1.2143, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.7527926645050177, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.4958180464031645e-06, | |
| "loss": 1.2277, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.7543321178066639, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 3.455086197895725e-06, | |
| "loss": 1.2065, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.7558715711083102, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.41454342416667e-06, | |
| "loss": 1.1921, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.7574110244099564, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 3.3741908964485415e-06, | |
| "loss": 1.2295, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.7589504777116026, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 3.3340297804778822e-06, | |
| "loss": 1.248, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.7604899310132489, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.2940612364615863e-06, | |
| "loss": 1.2379, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.7620293843148952, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 3.254286419043372e-06, | |
| "loss": 1.246, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.7635688376165414, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.214706477270424e-06, | |
| "loss": 1.2417, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.7651082909181877, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.1753225545602083e-06, | |
| "loss": 1.2571, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.766647744219834, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 3.1361357886674172e-06, | |
| "loss": 1.2244, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.7681871975214802, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 3.09714731165114e-06, | |
| "loss": 1.219, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.7697266508231264, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 3.0583582498421107e-06, | |
| "loss": 1.197, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.7712661041247727, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.019769723810222e-06, | |
| "loss": 1.1917, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.772805557426419, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 2.9813828483320983e-06, | |
| "loss": 1.2693, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.7743450107280652, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 2.9431987323589473e-06, | |
| "loss": 1.2338, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.7758844640297115, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 2.9052184789844696e-06, | |
| "loss": 1.2399, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.7774239173313577, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 2.86744318541304e-06, | |
| "loss": 1.2505, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.7789633706330039, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.829873942927971e-06, | |
| "loss": 1.2275, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.7805028239346502, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 2.7925118368600126e-06, | |
| "loss": 1.1781, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.7820422772362965, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 2.7553579465559908e-06, | |
| "loss": 1.1947, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.7835817305379427, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 2.7184133453476224e-06, | |
| "loss": 1.1985, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.785121183839589, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 2.6816791005205124e-06, | |
| "loss": 1.2575, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.7866606371412352, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.645156273283319e-06, | |
| "loss": 1.2176, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.7882000904428814, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 2.6088459187371027e-06, | |
| "loss": 1.2582, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.7897395437445277, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 2.572749085844829e-06, | |
| "loss": 1.2022, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.791278997046174, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 2.5368668174010926e-06, | |
| "loss": 1.2003, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.7928184503478203, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.501200150001961e-06, | |
| "loss": 1.2139, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.7943579036494665, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.465750114015053e-06, | |
| "loss": 1.2381, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.7958973569511127, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 2.430517733549761e-06, | |
| "loss": 1.2559, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.797436810252759, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.3955040264276684e-06, | |
| "loss": 1.2447, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.7989762635544052, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 2.3607100041531474e-06, | |
| "loss": 1.2168, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.8005157168560515, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 2.3261366718841307e-06, | |
| "loss": 1.1945, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.8020551701576978, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 2.291785028403086e-06, | |
| "loss": 1.2354, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.803594623459344, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.257656066088151e-06, | |
| "loss": 1.2296, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.8051340767609902, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 2.2237507708844707e-06, | |
| "loss": 1.1465, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.8066735300626365, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 2.1900701222757124e-06, | |
| "loss": 1.2334, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.8082129833642827, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.1566150932557737e-06, | |
| "loss": 1.1959, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.809752436665929, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.123386650300664e-06, | |
| "loss": 1.1912, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.8112918899675753, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.090385753340596e-06, | |
| "loss": 1.19, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.8128313432692215, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 2.0576133557322463e-06, | |
| "loss": 1.2264, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.8143707965708677, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 2.0250704042312198e-06, | |
| "loss": 1.2282, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.815910249872514, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.992757838964694e-06, | |
| "loss": 1.2152, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.8174497031741603, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.9606765934042637e-06, | |
| "loss": 1.2547, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.8189891564758065, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.928827594338969e-06, | |
| "loss": 1.258, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.8205286097774528, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.8972117618485286e-06, | |
| "loss": 1.2381, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.8220680630790991, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.8658300092767546e-06, | |
| "loss": 1.2604, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.8236075163807453, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.834683243205171e-06, | |
| "loss": 1.2129, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.8251469696823915, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.80377236342681e-06, | |
| "loss": 1.2199, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.8266864229840378, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.7730982629202497e-06, | |
| "loss": 1.2336, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.828225876285684, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.7426618278237761e-06, | |
| "loss": 1.2604, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.8297653295873303, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.7124639374098162e-06, | |
| "loss": 1.2213, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.8313047828889766, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.6825054640595196e-06, | |
| "loss": 1.2854, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.8328442361906228, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.652787273237565e-06, | |
| "loss": 1.196, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.834383689492269, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.6233102234671528e-06, | |
| "loss": 1.2436, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.8359231427939153, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.5940751663052046e-06, | |
| "loss": 1.2107, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.8374625960955616, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.565082946317763e-06, | |
| "loss": 1.2167, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.8390020493972078, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.5363344010555958e-06, | |
| "loss": 1.1855, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.8405415026988541, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.5078303610299972e-06, | |
| "loss": 1.2, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.8420809560005004, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.4795716496887879e-06, | |
| "loss": 1.2382, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.8436204093021465, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.4515590833925508e-06, | |
| "loss": 1.2268, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.8451598626037928, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.4237934713910118e-06, | |
| "loss": 1.1563, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.8466993159054391, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.3962756157997004e-06, | |
| "loss": 1.236, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.8482387692070853, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.3690063115767393e-06, | |
| "loss": 1.2667, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.8497782225087316, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.3419863464999138e-06, | |
| "loss": 1.2505, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.8513176758103779, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.3152165011438867e-06, | |
| "loss": 1.192, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.852857129112024, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.2886975488576637e-06, | |
| "loss": 1.2168, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.8543965824136703, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.2624302557422475e-06, | |
| "loss": 1.2354, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.8559360357153166, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.2364153806285074e-06, | |
| "loss": 1.2418, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.8574754890169628, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.2106536750552578e-06, | |
| "loss": 1.2171, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.8590149423186091, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.1851458832475371e-06, | |
| "loss": 1.25, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.8605543956202554, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.1598927420951333e-06, | |
| "loss": 1.2345, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.8620938489219017, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1348949811312616e-06, | |
| "loss": 1.2224, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.8636333022235478, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.1101533225115247e-06, | |
| "loss": 1.1752, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.8651727555251941, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.085668480993015e-06, | |
| "loss": 1.2384, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.8667122088268404, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.0614411639137034e-06, | |
| "loss": 1.2213, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.8682516621284866, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.0374720711719676e-06, | |
| "loss": 1.2478, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.8697911154301329, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.0137618952064066e-06, | |
| "loss": 1.2059, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.8713305687317792, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 9.903113209758098e-07, | |
| "loss": 1.1992, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.8728700220334253, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 9.67121025939387e-07, | |
| "loss": 1.2318, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.8744094753350716, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 9.441916800371909e-07, | |
| "loss": 1.2161, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.8759489286367179, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.215239456707636e-07, | |
| "loss": 1.2029, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.8774883819383641, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 8.991184776840034e-07, | |
| "loss": 1.1941, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.8790278352400104, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 8.769759233442366e-07, | |
| "loss": 1.221, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.8805672885416567, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 8.550969223235383e-07, | |
| "loss": 1.2539, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.8821067418433028, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 8.334821066802301e-07, | |
| "loss": 1.1944, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.8836461951449491, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 8.121321008406436e-07, | |
| "loss": 1.2218, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.8851856484465954, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 7.910475215810598e-07, | |
| "loss": 1.2036, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.8867251017482417, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 7.70228978009907e-07, | |
| "loss": 1.2703, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.8882645550498879, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 7.496770715501566e-07, | |
| "loss": 1.2755, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.8898040083515342, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 7.29392395921954e-07, | |
| "loss": 1.2312, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.8913434616531805, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 7.093755371254596e-07, | |
| "loss": 1.26, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.8928829149548266, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 6.896270734239263e-07, | |
| "loss": 1.2239, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.8944223682564729, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 6.701475753269915e-07, | |
| "loss": 1.2245, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.8959618215581192, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 6.509376055741944e-07, | |
| "loss": 1.1909, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.8975012748597654, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 6.319977191187232e-07, | |
| "loss": 1.1744, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.8990407281614117, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 6.133284631113789e-07, | |
| "loss": 1.2049, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.900580181463058, | |
| "grad_norm": 1.5, | |
| "learning_rate": 5.949303768847736e-07, | |
| "loss": 1.1935, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.9021196347647041, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 5.768039919377421e-07, | |
| "loss": 1.2084, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.9036590880663504, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 5.589498319199949e-07, | |
| "loss": 1.1963, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.9051985413679967, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 5.413684126169882e-07, | |
| "loss": 1.2019, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.9067379946696429, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 5.240602419350238e-07, | |
| "loss": 1.2063, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.9082774479712892, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 5.070258198865741e-07, | |
| "loss": 1.2011, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.9098169012729355, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 4.902656385758397e-07, | |
| "loss": 1.2129, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.9113563545745818, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.7378018218453113e-07, | |
| "loss": 1.2446, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.9128958078762279, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 4.575699269578848e-07, | |
| "loss": 1.2285, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.9144352611778742, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 4.4163534119090245e-07, | |
| "loss": 1.2069, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.9159747144795205, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.259768852148216e-07, | |
| "loss": 1.2399, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.9175141677811667, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 4.105950113838153e-07, | |
| "loss": 1.2366, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.919053621082813, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 3.9549016406193686e-07, | |
| "loss": 1.2248, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 0.9205930743844593, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 3.8066277961026134e-07, | |
| "loss": 1.2546, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.9221325276861054, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 3.661132863743011e-07, | |
| "loss": 1.2198, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 0.9236719809877517, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.518421046716147e-07, | |
| "loss": 1.186, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.925211434289398, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.378496467796766e-07, | |
| "loss": 1.2156, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 0.9267508875910442, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 3.2413631692395753e-07, | |
| "loss": 1.1959, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.9282903408926905, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 3.1070251126625427e-07, | |
| "loss": 1.2426, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 0.9298297941943368, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 2.9754861789324074e-07, | |
| "loss": 1.21, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.931369247495983, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 2.84675016805257e-07, | |
| "loss": 1.2328, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.9329087007976292, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 2.7208207990533365e-07, | |
| "loss": 1.2377, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.9344481540992755, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.597701709884426e-07, | |
| "loss": 1.2589, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 0.9359876074009218, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.4773964573099997e-07, | |
| "loss": 1.2219, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.937527060702568, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 2.359908516805731e-07, | |
| "loss": 1.2027, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 0.9390665140042143, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 2.2452412824585546e-07, | |
| "loss": 1.1985, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.9406059673058605, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.1333980668685416e-07, | |
| "loss": 1.2069, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 0.9421454206075067, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 2.0243821010532015e-07, | |
| "loss": 1.2327, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.943684873909153, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.9181965343541887e-07, | |
| "loss": 1.2065, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 0.9452243272107993, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.8148444343462857e-07, | |
| "loss": 1.2099, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.9467637805124455, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.7143287867487756e-07, | |
| "loss": 1.2252, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.9483032338140918, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.6166524953392104e-07, | |
| "loss": 1.2313, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.949842687115738, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.521818381869544e-07, | |
| "loss": 1.2127, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 0.9513821404173842, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.4298291859845215e-07, | |
| "loss": 1.195, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.9529215937190305, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.3406875651426843e-07, | |
| "loss": 1.2191, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 0.9544610470206768, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.2543960945394452e-07, | |
| "loss": 1.2605, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.9560005003223231, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.170957267032824e-07, | |
| "loss": 1.2346, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 0.9575399536239693, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.0903734930713283e-07, | |
| "loss": 1.2341, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.9590794069256156, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.0126471006243866e-07, | |
| "loss": 1.1935, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 0.9606188602272618, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 9.377803351150683e-08, | |
| "loss": 1.2718, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.962158313528908, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 8.657753593552142e-08, | |
| "loss": 1.2041, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.9636977668305543, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 7.966342534829752e-08, | |
| "loss": 1.222, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.9652372201322006, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.303590149026818e-08, | |
| "loss": 1.2267, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 0.9667766734338468, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 6.669515582271913e-08, | |
| "loss": 1.2352, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.968316126735493, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 6.064137152225425e-08, | |
| "loss": 1.2402, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 0.9698555800371393, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 5.4874723475504267e-08, | |
| "loss": 1.2357, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.9713950333387855, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 4.939537827407415e-08, | |
| "loss": 1.1789, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 0.9729344866404318, | |
| "grad_norm": 1.625, | |
| "learning_rate": 4.420349420973358e-08, | |
| "loss": 1.2433, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.9744739399420781, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.929922126983954e-08, | |
| "loss": 1.1677, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 0.9760133932437243, | |
| "grad_norm": 1.625, | |
| "learning_rate": 3.4682701133008645e-08, | |
| "loss": 1.2134, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.9775528465453706, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.035406716501599e-08, | |
| "loss": 1.2336, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.9790922998470168, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.6313444414952648e-08, | |
| "loss": 1.2341, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.9806317531486631, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 2.2560949611605264e-08, | |
| "loss": 1.1928, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 0.9821712064503093, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.90966911600865e-08, | |
| "loss": 1.2076, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.9837106597519556, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.592076913870644e-08, | |
| "loss": 1.2274, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 0.9852501130536019, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.3033275296073789e-08, | |
| "loss": 1.2503, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.9867895663552481, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.0434293048453559e-08, | |
| "loss": 1.2615, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 0.9883290196568943, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 8.123897477354537e-09, | |
| "loss": 1.2132, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.9898684729585406, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 6.102155327355475e-09, | |
| "loss": 1.2083, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 0.9914079262601868, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 4.369125004185515e-09, | |
| "loss": 1.2451, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.9929473795618331, | |
| "grad_norm": 1.75, | |
| "learning_rate": 2.924856573028878e-09, | |
| "loss": 1.2202, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.9944868328634794, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.7693917570837938e-09, | |
| "loss": 1.2657, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.9960262861651256, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 9.027639363545782e-10, | |
| "loss": 1.2212, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 0.9975657394667719, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 3.249981466901808e-10, | |
| "loss": 1.2315, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.9991051927684181, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.6111079055878294e-11, | |
| "loss": 1.22, | |
| "step": 6490 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 6495, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.756027118157693e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |