| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9338205440519691, | |
| "eval_steps": 100, | |
| "global_step": 6900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0013533631073216945, | |
| "grad_norm": 33.467918395996094, | |
| "learning_rate": 4.0595399188092017e-07, | |
| "loss": 0.7404, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.002706726214643389, | |
| "grad_norm": 22.66303253173828, | |
| "learning_rate": 8.570139828597205e-07, | |
| "loss": 0.72, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.004060089321965083, | |
| "grad_norm": 28.549816131591797, | |
| "learning_rate": 1.3080739738385204e-06, | |
| "loss": 1.0342, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.005413452429286778, | |
| "grad_norm": 26.86774253845215, | |
| "learning_rate": 1.7591339648173207e-06, | |
| "loss": 0.851, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.006766815536608472, | |
| "grad_norm": 24.830371856689453, | |
| "learning_rate": 2.210193955796121e-06, | |
| "loss": 0.8493, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.008120178643930167, | |
| "grad_norm": 31.990707397460938, | |
| "learning_rate": 2.6612539467749215e-06, | |
| "loss": 0.8196, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.009473541751251861, | |
| "grad_norm": 19.266714096069336, | |
| "learning_rate": 3.112313937753721e-06, | |
| "loss": 0.8137, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.010826904858573556, | |
| "grad_norm": 27.596582412719727, | |
| "learning_rate": 3.563373928732522e-06, | |
| "loss": 0.9199, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.012180267965895249, | |
| "grad_norm": 20.1337890625, | |
| "learning_rate": 4.014433919711322e-06, | |
| "loss": 0.6817, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.013533631073216944, | |
| "grad_norm": 30.579744338989258, | |
| "learning_rate": 4.465493910690122e-06, | |
| "loss": 0.7792, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.013533631073216944, | |
| "eval_accuracy": 0.5047184170471841, | |
| "eval_loss": 0.7254102230072021, | |
| "eval_runtime": 586.4826, | |
| "eval_samples_per_second": 5.601, | |
| "eval_steps_per_second": 5.601, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.014886994180538638, | |
| "grad_norm": 32.78731918334961, | |
| "learning_rate": 4.916553901668922e-06, | |
| "loss": 0.8336, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.016240357287860333, | |
| "grad_norm": 29.351137161254883, | |
| "learning_rate": 5.367613892647722e-06, | |
| "loss": 0.8204, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.017593720395182026, | |
| "grad_norm": 24.733667373657227, | |
| "learning_rate": 5.8186738836265224e-06, | |
| "loss": 0.9134, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.018947083502503723, | |
| "grad_norm": 27.078041076660156, | |
| "learning_rate": 6.269733874605323e-06, | |
| "loss": 0.7954, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.020300446609825416, | |
| "grad_norm": 32.13090515136719, | |
| "learning_rate": 6.7207938655841226e-06, | |
| "loss": 0.9003, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.021653809717147112, | |
| "grad_norm": 24.68378448486328, | |
| "learning_rate": 7.171853856562922e-06, | |
| "loss": 0.7541, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.023007172824468805, | |
| "grad_norm": 24.132369995117188, | |
| "learning_rate": 7.6229138475417236e-06, | |
| "loss": 0.8898, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.024360535931790498, | |
| "grad_norm": 30.671886444091797, | |
| "learning_rate": 8.073973838520524e-06, | |
| "loss": 0.8849, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.025713899039112195, | |
| "grad_norm": 30.794673919677734, | |
| "learning_rate": 8.525033829499324e-06, | |
| "loss": 0.7421, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.027067262146433888, | |
| "grad_norm": 35.09535598754883, | |
| "learning_rate": 8.976093820478123e-06, | |
| "loss": 0.7815, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.027067262146433888, | |
| "eval_accuracy": 0.6313546423135464, | |
| "eval_loss": 0.638473391532898, | |
| "eval_runtime": 586.0225, | |
| "eval_samples_per_second": 5.606, | |
| "eval_steps_per_second": 5.606, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.028420625253755584, | |
| "grad_norm": 13.783787727355957, | |
| "learning_rate": 9.427153811456923e-06, | |
| "loss": 0.7288, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.029773988361077277, | |
| "grad_norm": 29.833702087402344, | |
| "learning_rate": 9.878213802435724e-06, | |
| "loss": 0.5495, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.03112735146839897, | |
| "grad_norm": 14.958450317382812, | |
| "learning_rate": 1.0329273793414524e-05, | |
| "loss": 0.69, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.032480714575720666, | |
| "grad_norm": 28.827964782714844, | |
| "learning_rate": 1.0780333784393325e-05, | |
| "loss": 0.6468, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.03383407768304236, | |
| "grad_norm": 29.778186798095703, | |
| "learning_rate": 1.1231393775372125e-05, | |
| "loss": 0.7499, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.03518744079036405, | |
| "grad_norm": 10.802916526794434, | |
| "learning_rate": 1.1682453766350926e-05, | |
| "loss": 0.6447, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.03654080389768575, | |
| "grad_norm": 24.211503982543945, | |
| "learning_rate": 1.2133513757329726e-05, | |
| "loss": 0.6241, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.037894167005007445, | |
| "grad_norm": 26.76028060913086, | |
| "learning_rate": 1.2584573748308526e-05, | |
| "loss": 0.4461, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.039247530112329135, | |
| "grad_norm": 20.0728759765625, | |
| "learning_rate": 1.3035633739287325e-05, | |
| "loss": 0.6034, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.04060089321965083, | |
| "grad_norm": 15.319375038146973, | |
| "learning_rate": 1.3486693730266125e-05, | |
| "loss": 0.5155, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04060089321965083, | |
| "eval_accuracy": 0.7418569254185693, | |
| "eval_loss": 0.5153380632400513, | |
| "eval_runtime": 586.0685, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04195425632697253, | |
| "grad_norm": 22.047800064086914, | |
| "learning_rate": 1.3937753721244926e-05, | |
| "loss": 0.7796, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.043307619434294224, | |
| "grad_norm": 23.10489845275879, | |
| "learning_rate": 1.4388813712223726e-05, | |
| "loss": 0.5833, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.044660982541615914, | |
| "grad_norm": 34.646427154541016, | |
| "learning_rate": 1.4839873703202526e-05, | |
| "loss": 0.4678, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.04601434564893761, | |
| "grad_norm": 36.290771484375, | |
| "learning_rate": 1.529093369418133e-05, | |
| "loss": 0.5535, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.04736770875625931, | |
| "grad_norm": 4.73361873626709, | |
| "learning_rate": 1.5741993685160127e-05, | |
| "loss": 0.3168, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.048721071863580996, | |
| "grad_norm": 14.30739974975586, | |
| "learning_rate": 1.6193053676138928e-05, | |
| "loss": 0.3948, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.05007443497090269, | |
| "grad_norm": 15.850822448730469, | |
| "learning_rate": 1.664411366711773e-05, | |
| "loss": 0.6431, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.05142779807822439, | |
| "grad_norm": 40.54030990600586, | |
| "learning_rate": 1.7095173658096527e-05, | |
| "loss": 0.6361, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.05278116118554608, | |
| "grad_norm": 3.195178270339966, | |
| "learning_rate": 1.754623364907533e-05, | |
| "loss": 0.4635, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.054134524292867775, | |
| "grad_norm": 20.665943145751953, | |
| "learning_rate": 1.7997293640054127e-05, | |
| "loss": 0.5693, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.054134524292867775, | |
| "eval_accuracy": 0.7948249619482496, | |
| "eval_loss": 0.4957842528820038, | |
| "eval_runtime": 585.7637, | |
| "eval_samples_per_second": 5.608, | |
| "eval_steps_per_second": 5.608, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.05548788740018947, | |
| "grad_norm": 24.18285369873047, | |
| "learning_rate": 1.8448353631032928e-05, | |
| "loss": 0.6247, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.05684125050751117, | |
| "grad_norm": 37.275447845458984, | |
| "learning_rate": 1.8899413622011726e-05, | |
| "loss": 0.5164, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.05819461361483286, | |
| "grad_norm": 22.858854293823242, | |
| "learning_rate": 1.9350473612990527e-05, | |
| "loss": 0.4309, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.059547976722154554, | |
| "grad_norm": 11.075826644897461, | |
| "learning_rate": 1.980153360396933e-05, | |
| "loss": 0.3716, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.06090133982947625, | |
| "grad_norm": 20.966968536376953, | |
| "learning_rate": 2.025259359494813e-05, | |
| "loss": 0.5862, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.06225470293679794, | |
| "grad_norm": 16.335777282714844, | |
| "learning_rate": 2.070365358592693e-05, | |
| "loss": 0.293, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.06360806604411964, | |
| "grad_norm": 42.5803108215332, | |
| "learning_rate": 2.115471357690573e-05, | |
| "loss": 0.336, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.06496142915144133, | |
| "grad_norm": 1.7485947608947754, | |
| "learning_rate": 2.160577356788453e-05, | |
| "loss": 0.2536, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.06631479225876302, | |
| "grad_norm": 28.608144760131836, | |
| "learning_rate": 2.205683355886333e-05, | |
| "loss": 0.4017, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.06766815536608473, | |
| "grad_norm": 35.06800079345703, | |
| "learning_rate": 2.250789354984213e-05, | |
| "loss": 0.3398, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06766815536608473, | |
| "eval_accuracy": 0.8213089802130898, | |
| "eval_loss": 0.6229318380355835, | |
| "eval_runtime": 586.2164, | |
| "eval_samples_per_second": 5.604, | |
| "eval_steps_per_second": 5.604, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06902151847340642, | |
| "grad_norm": 41.761592864990234, | |
| "learning_rate": 2.2958953540820928e-05, | |
| "loss": 0.6924, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.0703748815807281, | |
| "grad_norm": 26.0762882232666, | |
| "learning_rate": 2.341001353179973e-05, | |
| "loss": 0.4984, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.07172824468804981, | |
| "grad_norm": 51.21356201171875, | |
| "learning_rate": 2.386107352277853e-05, | |
| "loss": 1.0133, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.0730816077953715, | |
| "grad_norm": 9.679603576660156, | |
| "learning_rate": 2.4312133513757332e-05, | |
| "loss": 0.7537, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.07443497090269319, | |
| "grad_norm": 8.944125175476074, | |
| "learning_rate": 2.4763193504736133e-05, | |
| "loss": 0.1747, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.07578833401001489, | |
| "grad_norm": 0.24319039285182953, | |
| "learning_rate": 2.5214253495714928e-05, | |
| "loss": 0.2228, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.07714169711733658, | |
| "grad_norm": 34.457130432128906, | |
| "learning_rate": 2.5665313486693732e-05, | |
| "loss": 0.6227, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.07849506022465827, | |
| "grad_norm": 25.974857330322266, | |
| "learning_rate": 2.6116373477672534e-05, | |
| "loss": 0.2999, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.07984842333197997, | |
| "grad_norm": 29.637929916381836, | |
| "learning_rate": 2.6567433468651332e-05, | |
| "loss": 0.4552, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.08120178643930166, | |
| "grad_norm": 22.073013305664062, | |
| "learning_rate": 2.7018493459630133e-05, | |
| "loss": 0.4963, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.08120178643930166, | |
| "eval_accuracy": 0.8307458143074582, | |
| "eval_loss": 0.4110361933708191, | |
| "eval_runtime": 586.3068, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.08255514954662337, | |
| "grad_norm": 0.12677842378616333, | |
| "learning_rate": 2.746955345060893e-05, | |
| "loss": 0.3773, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.08390851265394506, | |
| "grad_norm": 42.65321350097656, | |
| "learning_rate": 2.7920613441587736e-05, | |
| "loss": 0.6737, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.08526187576126674, | |
| "grad_norm": 26.458065032958984, | |
| "learning_rate": 2.837167343256653e-05, | |
| "loss": 0.6558, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.08661523886858845, | |
| "grad_norm": 5.073329925537109, | |
| "learning_rate": 2.8822733423545335e-05, | |
| "loss": 0.3947, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.08796860197591014, | |
| "grad_norm": 2.559771776199341, | |
| "learning_rate": 2.927379341452413e-05, | |
| "loss": 0.2376, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.08932196508323183, | |
| "grad_norm": 46.948219299316406, | |
| "learning_rate": 2.9724853405502934e-05, | |
| "loss": 0.6246, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.09067532819055353, | |
| "grad_norm": 20.228105545043945, | |
| "learning_rate": 3.0175913396481732e-05, | |
| "loss": 0.4244, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.09202869129787522, | |
| "grad_norm": 5.761535167694092, | |
| "learning_rate": 3.0626973387460534e-05, | |
| "loss": 0.356, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.09338205440519691, | |
| "grad_norm": 0.10911667346954346, | |
| "learning_rate": 3.107803337843934e-05, | |
| "loss": 0.4661, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.09473541751251861, | |
| "grad_norm": 12.342616081237793, | |
| "learning_rate": 3.1529093369418136e-05, | |
| "loss": 0.4579, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.09473541751251861, | |
| "eval_accuracy": 0.8584474885844748, | |
| "eval_loss": 0.4246827960014343, | |
| "eval_runtime": 586.4424, | |
| "eval_samples_per_second": 5.602, | |
| "eval_steps_per_second": 5.602, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.0960887806198403, | |
| "grad_norm": 42.24137496948242, | |
| "learning_rate": 3.1980153360396934e-05, | |
| "loss": 0.52, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.09744214372716199, | |
| "grad_norm": 0.19052202999591827, | |
| "learning_rate": 3.243121335137573e-05, | |
| "loss": 0.3293, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.0987955068344837, | |
| "grad_norm": 29.600139617919922, | |
| "learning_rate": 3.288227334235454e-05, | |
| "loss": 0.5003, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.10014886994180539, | |
| "grad_norm": 30.022396087646484, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.5022, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.10150223304912707, | |
| "grad_norm": 1.689180850982666, | |
| "learning_rate": 3.378439332431213e-05, | |
| "loss": 0.4278, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.10285559615644878, | |
| "grad_norm": 0.31317076086997986, | |
| "learning_rate": 3.423545331529093e-05, | |
| "loss": 0.4199, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.10420895926377047, | |
| "grad_norm": 20.078535079956055, | |
| "learning_rate": 3.4686513306269736e-05, | |
| "loss": 0.3041, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.10556232237109216, | |
| "grad_norm": 20.944692611694336, | |
| "learning_rate": 3.513757329724854e-05, | |
| "loss": 0.8085, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.10691568547841386, | |
| "grad_norm": 9.586214065551758, | |
| "learning_rate": 3.558863328822734e-05, | |
| "loss": 0.5523, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.10826904858573555, | |
| "grad_norm": 31.01270866394043, | |
| "learning_rate": 3.6039693279206136e-05, | |
| "loss": 0.472, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.10826904858573555, | |
| "eval_accuracy": 0.8681887366818873, | |
| "eval_loss": 0.32172152400016785, | |
| "eval_runtime": 586.2592, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.10962241169305725, | |
| "grad_norm": 0.7769675850868225, | |
| "learning_rate": 3.6490753270184934e-05, | |
| "loss": 0.2738, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.11097577480037894, | |
| "grad_norm": 37.9739875793457, | |
| "learning_rate": 3.694181326116374e-05, | |
| "loss": 0.4221, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.11232913790770063, | |
| "grad_norm": 4.82348108291626, | |
| "learning_rate": 3.739287325214254e-05, | |
| "loss": 0.2673, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.11368250101502234, | |
| "grad_norm": 0.9005016088485718, | |
| "learning_rate": 3.7843933243121335e-05, | |
| "loss": 0.2568, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.11503586412234403, | |
| "grad_norm": 23.65435028076172, | |
| "learning_rate": 3.829499323410013e-05, | |
| "loss": 0.4254, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.11638922722966571, | |
| "grad_norm": 0.8585798144340515, | |
| "learning_rate": 3.874605322507894e-05, | |
| "loss": 0.8521, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.11774259033698742, | |
| "grad_norm": 33.647430419921875, | |
| "learning_rate": 3.9197113216057735e-05, | |
| "loss": 1.389, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.11909595344430911, | |
| "grad_norm": 14.677156448364258, | |
| "learning_rate": 3.964817320703654e-05, | |
| "loss": 0.1391, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.1204493165516308, | |
| "grad_norm": 1.6325709819793701, | |
| "learning_rate": 4.009923319801534e-05, | |
| "loss": 0.7501, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.1218026796589525, | |
| "grad_norm": 0.28640714287757874, | |
| "learning_rate": 4.0550293188994136e-05, | |
| "loss": 0.4304, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.1218026796589525, | |
| "eval_accuracy": 0.8639269406392694, | |
| "eval_loss": 0.3807564675807953, | |
| "eval_runtime": 586.1022, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.12315604276627419, | |
| "grad_norm": 25.761606216430664, | |
| "learning_rate": 4.100135317997294e-05, | |
| "loss": 0.2124, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.12450940587359588, | |
| "grad_norm": 4.299468994140625, | |
| "learning_rate": 4.145241317095174e-05, | |
| "loss": 0.443, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.12586276898091758, | |
| "grad_norm": 2.907810688018799, | |
| "learning_rate": 4.190347316193054e-05, | |
| "loss": 0.3628, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.1272161320882393, | |
| "grad_norm": 0.42966228723526, | |
| "learning_rate": 4.2354533152909335e-05, | |
| "loss": 0.5665, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.12856949519556096, | |
| "grad_norm": 0.043091725558042526, | |
| "learning_rate": 4.280559314388814e-05, | |
| "loss": 0.6238, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.12992285830288267, | |
| "grad_norm": 4.317753791809082, | |
| "learning_rate": 4.325665313486694e-05, | |
| "loss": 0.4853, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.13127622141020437, | |
| "grad_norm": 6.607635974884033, | |
| "learning_rate": 4.370771312584574e-05, | |
| "loss": 0.5516, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.13262958451752604, | |
| "grad_norm": 5.1537861824035645, | |
| "learning_rate": 4.415877311682454e-05, | |
| "loss": 0.2681, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.13398294762484775, | |
| "grad_norm": 18.95270538330078, | |
| "learning_rate": 4.460983310780334e-05, | |
| "loss": 0.1933, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.13533631073216945, | |
| "grad_norm": 37.21832275390625, | |
| "learning_rate": 4.506089309878214e-05, | |
| "loss": 0.5274, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13533631073216945, | |
| "eval_accuracy": 0.8748858447488584, | |
| "eval_loss": 0.46359485387802124, | |
| "eval_runtime": 586.3431, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13668967383949113, | |
| "grad_norm": 31.545852661132812, | |
| "learning_rate": 4.551195308976094e-05, | |
| "loss": 0.4664, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.13804303694681283, | |
| "grad_norm": 10.666961669921875, | |
| "learning_rate": 4.596301308073974e-05, | |
| "loss": 0.3493, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.13939640005413453, | |
| "grad_norm": 18.62867546081543, | |
| "learning_rate": 4.641407307171854e-05, | |
| "loss": 0.3411, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.1407497631614562, | |
| "grad_norm": 0.038888320326805115, | |
| "learning_rate": 4.686513306269734e-05, | |
| "loss": 0.2676, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.1421031262687779, | |
| "grad_norm": 0.8312036991119385, | |
| "learning_rate": 4.731619305367614e-05, | |
| "loss": 0.577, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.14345648937609962, | |
| "grad_norm": 0.6841209530830383, | |
| "learning_rate": 4.7767253044654944e-05, | |
| "loss": 0.2822, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.1448098524834213, | |
| "grad_norm": 2.010380983352661, | |
| "learning_rate": 4.821831303563374e-05, | |
| "loss": 0.214, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.146163215590743, | |
| "grad_norm": 0.2284691035747528, | |
| "learning_rate": 4.866937302661254e-05, | |
| "loss": 0.6455, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.1475165786980647, | |
| "grad_norm": 15.199652671813965, | |
| "learning_rate": 4.9120433017591345e-05, | |
| "loss": 0.1528, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.14886994180538637, | |
| "grad_norm": 0.560634195804596, | |
| "learning_rate": 4.957149300857014e-05, | |
| "loss": 0.5175, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.14886994180538637, | |
| "eval_accuracy": 0.8745814307458143, | |
| "eval_loss": 0.37603330612182617, | |
| "eval_runtime": 586.3343, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.15022330491270808, | |
| "grad_norm": 0.30973055958747864, | |
| "learning_rate": 5.002255299954894e-05, | |
| "loss": 0.1818, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.15157666802002978, | |
| "grad_norm": 0.012624499388039112, | |
| "learning_rate": 5.047361299052774e-05, | |
| "loss": 0.6186, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.15293003112735146, | |
| "grad_norm": 32.54011917114258, | |
| "learning_rate": 5.0924672981506537e-05, | |
| "loss": 0.5022, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.15428339423467316, | |
| "grad_norm": 3.9204163551330566, | |
| "learning_rate": 5.137573297248535e-05, | |
| "loss": 0.3571, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.15563675734199486, | |
| "grad_norm": 0.25882434844970703, | |
| "learning_rate": 5.1826792963464146e-05, | |
| "loss": 0.4617, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.15699012044931654, | |
| "grad_norm": 0.029281217604875565, | |
| "learning_rate": 5.2277852954442944e-05, | |
| "loss": 0.2459, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.15834348355663824, | |
| "grad_norm": 15.895069122314453, | |
| "learning_rate": 5.272891294542175e-05, | |
| "loss": 0.3078, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.15969684666395995, | |
| "grad_norm": 15.33349323272705, | |
| "learning_rate": 5.317997293640055e-05, | |
| "loss": 0.5131, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.16105020977128162, | |
| "grad_norm": 6.226064205169678, | |
| "learning_rate": 5.3631032927379345e-05, | |
| "loss": 0.1718, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.16240357287860333, | |
| "grad_norm": 20.2957820892334, | |
| "learning_rate": 5.408209291835814e-05, | |
| "loss": 0.671, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.16240357287860333, | |
| "eval_accuracy": 0.871841704718417, | |
| "eval_loss": 0.36159640550613403, | |
| "eval_runtime": 586.2928, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.16375693598592503, | |
| "grad_norm": 0.6848016977310181, | |
| "learning_rate": 5.453315290933695e-05, | |
| "loss": 0.7948, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.16511029909324673, | |
| "grad_norm": 2.2201507091522217, | |
| "learning_rate": 5.4984212900315745e-05, | |
| "loss": 0.3166, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.1664636622005684, | |
| "grad_norm": 17.319942474365234, | |
| "learning_rate": 5.543527289129454e-05, | |
| "loss": 0.6259, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.1678170253078901, | |
| "grad_norm": 11.914515495300293, | |
| "learning_rate": 5.588633288227334e-05, | |
| "loss": 0.4776, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.16917038841521181, | |
| "grad_norm": 19.129995346069336, | |
| "learning_rate": 5.6337392873252146e-05, | |
| "loss": 0.3623, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.1705237515225335, | |
| "grad_norm": 6.791560649871826, | |
| "learning_rate": 5.6788452864230944e-05, | |
| "loss": 0.1206, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.1718771146298552, | |
| "grad_norm": 24.70437240600586, | |
| "learning_rate": 5.723951285520974e-05, | |
| "loss": 0.6828, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.1732304777371769, | |
| "grad_norm": 1.6283477544784546, | |
| "learning_rate": 5.769057284618855e-05, | |
| "loss": 0.2961, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.17458384084449857, | |
| "grad_norm": 0.2976061701774597, | |
| "learning_rate": 5.814163283716735e-05, | |
| "loss": 0.2001, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.17593720395182028, | |
| "grad_norm": 5.002806663513184, | |
| "learning_rate": 5.859269282814614e-05, | |
| "loss": 0.7133, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.17593720395182028, | |
| "eval_accuracy": 0.8916286149162862, | |
| "eval_loss": 0.37970247864723206, | |
| "eval_runtime": 586.3256, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.17729056705914198, | |
| "grad_norm": 11.172118186950684, | |
| "learning_rate": 5.904375281912494e-05, | |
| "loss": 0.2119, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.17864393016646365, | |
| "grad_norm": 16.503934860229492, | |
| "learning_rate": 5.949481281010375e-05, | |
| "loss": 0.5483, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.17999729327378536, | |
| "grad_norm": 0.00956371147185564, | |
| "learning_rate": 5.994587280108255e-05, | |
| "loss": 0.1664, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.18135065638110706, | |
| "grad_norm": 1.2320352792739868, | |
| "learning_rate": 6.039693279206135e-05, | |
| "loss": 0.5364, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.18270401948842874, | |
| "grad_norm": 2.635582208633423, | |
| "learning_rate": 6.084799278304014e-05, | |
| "loss": 0.2823, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.18405738259575044, | |
| "grad_norm": 0.07258131355047226, | |
| "learning_rate": 6.129905277401894e-05, | |
| "loss": 0.2589, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.18541074570307214, | |
| "grad_norm": 0.6865749359130859, | |
| "learning_rate": 6.175011276499775e-05, | |
| "loss": 0.5216, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.18676410881039382, | |
| "grad_norm": 0.460299015045166, | |
| "learning_rate": 6.220117275597654e-05, | |
| "loss": 0.2386, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.18811747191771552, | |
| "grad_norm": 0.32845044136047363, | |
| "learning_rate": 6.265223274695536e-05, | |
| "loss": 0.6403, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.18947083502503723, | |
| "grad_norm": 19.33603286743164, | |
| "learning_rate": 6.310329273793415e-05, | |
| "loss": 0.3134, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.18947083502503723, | |
| "eval_accuracy": 0.8964992389649924, | |
| "eval_loss": 0.2971131503582001, | |
| "eval_runtime": 586.3223, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.1908241981323589, | |
| "grad_norm": 18.192237854003906, | |
| "learning_rate": 6.355435272891294e-05, | |
| "loss": 0.308, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.1921775612396806, | |
| "grad_norm": 0.7239986658096313, | |
| "learning_rate": 6.400541271989174e-05, | |
| "loss": 0.1362, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.1935309243470023, | |
| "grad_norm": 0.13601917028427124, | |
| "learning_rate": 6.445647271087055e-05, | |
| "loss": 0.3244, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.19488428745432398, | |
| "grad_norm": 0.047646623104810715, | |
| "learning_rate": 6.490753270184935e-05, | |
| "loss": 0.4921, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.1962376505616457, | |
| "grad_norm": 0.7094109058380127, | |
| "learning_rate": 6.535859269282815e-05, | |
| "loss": 0.1284, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.1975910136689674, | |
| "grad_norm": 0.036302514374256134, | |
| "learning_rate": 6.580965268380695e-05, | |
| "loss": 0.1962, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.19894437677628907, | |
| "grad_norm": 0.2627946138381958, | |
| "learning_rate": 6.626071267478576e-05, | |
| "loss": 0.3998, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.20029773988361077, | |
| "grad_norm": 11.520143508911133, | |
| "learning_rate": 6.671177266576455e-05, | |
| "loss": 0.311, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.20165110299093247, | |
| "grad_norm": 0.5052374005317688, | |
| "learning_rate": 6.716283265674335e-05, | |
| "loss": 0.4324, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.20300446609825415, | |
| "grad_norm": 0.007840686477720737, | |
| "learning_rate": 6.761389264772216e-05, | |
| "loss": 0.0874, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.20300446609825415, | |
| "eval_accuracy": 0.8898021308980213, | |
| "eval_loss": 0.5496937036514282, | |
| "eval_runtime": 586.3431, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.20435782920557585, | |
| "grad_norm": 3.039323329925537, | |
| "learning_rate": 6.806495263870095e-05, | |
| "loss": 0.4797, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.20571119231289756, | |
| "grad_norm": 0.00013633279013447464, | |
| "learning_rate": 6.851601262967975e-05, | |
| "loss": 0.4127, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.20706455542021923, | |
| "grad_norm": 0.127132847905159, | |
| "learning_rate": 6.896707262065854e-05, | |
| "loss": 0.5609, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.20841791852754094, | |
| "grad_norm": 22.69544792175293, | |
| "learning_rate": 6.941813261163735e-05, | |
| "loss": 0.3238, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.20977128163486264, | |
| "grad_norm": 0.02062298357486725, | |
| "learning_rate": 6.986919260261615e-05, | |
| "loss": 0.4243, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.21112464474218431, | |
| "grad_norm": 23.498069763183594, | |
| "learning_rate": 7.032025259359494e-05, | |
| "loss": 0.9495, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.21247800784950602, | |
| "grad_norm": 0.8411264419555664, | |
| "learning_rate": 7.077131258457376e-05, | |
| "loss": 0.1026, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.21383137095682772, | |
| "grad_norm": 0.030804749578237534, | |
| "learning_rate": 7.122237257555255e-05, | |
| "loss": 0.4295, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.21518473406414942, | |
| "grad_norm": 2.670105218887329, | |
| "learning_rate": 7.167343256653136e-05, | |
| "loss": 0.7808, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.2165380971714711, | |
| "grad_norm": 1.184561014175415, | |
| "learning_rate": 7.212449255751015e-05, | |
| "loss": 0.4684, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.2165380971714711, | |
| "eval_accuracy": 0.8971080669710807, | |
| "eval_loss": 0.4002440571784973, | |
| "eval_runtime": 586.6131, | |
| "eval_samples_per_second": 5.6, | |
| "eval_steps_per_second": 5.6, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.2178914602787928, | |
| "grad_norm": 0.03285335749387741, | |
| "learning_rate": 7.257555254848895e-05, | |
| "loss": 0.0504, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.2192448233861145, | |
| "grad_norm": 0.07612759619951248, | |
| "learning_rate": 7.302661253946776e-05, | |
| "loss": 0.4363, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.22059818649343618, | |
| "grad_norm": 17.422269821166992, | |
| "learning_rate": 7.347767253044655e-05, | |
| "loss": 0.0434, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.22195154960075789, | |
| "grad_norm": 0.026716621592640877, | |
| "learning_rate": 7.392873252142535e-05, | |
| "loss": 0.4682, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.2233049127080796, | |
| "grad_norm": 0.9974814057350159, | |
| "learning_rate": 7.437979251240416e-05, | |
| "loss": 0.3381, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.22465827581540126, | |
| "grad_norm": 0.2876833975315094, | |
| "learning_rate": 7.483085250338295e-05, | |
| "loss": 0.5237, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.22601163892272297, | |
| "grad_norm": 0.009840277023613453, | |
| "learning_rate": 7.528191249436175e-05, | |
| "loss": 0.1314, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.22736500203004467, | |
| "grad_norm": 0.004841567948460579, | |
| "learning_rate": 7.573297248534056e-05, | |
| "loss": 0.6736, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.22871836513736635, | |
| "grad_norm": 5.4699602127075195, | |
| "learning_rate": 7.618403247631935e-05, | |
| "loss": 0.4901, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.23007172824468805, | |
| "grad_norm": 15.60845947265625, | |
| "learning_rate": 7.663509246729816e-05, | |
| "loss": 0.1759, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.23007172824468805, | |
| "eval_accuracy": 0.8943683409436834, | |
| "eval_loss": 0.33465254306793213, | |
| "eval_runtime": 586.0743, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.23142509135200975, | |
| "grad_norm": 25.261932373046875, | |
| "learning_rate": 7.708615245827695e-05, | |
| "loss": 0.347, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.23277845445933143, | |
| "grad_norm": 0.9137635231018066, | |
| "learning_rate": 7.753721244925576e-05, | |
| "loss": 0.5517, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.23413181756665313, | |
| "grad_norm": 1.7071709632873535, | |
| "learning_rate": 7.798827244023456e-05, | |
| "loss": 0.0289, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.23548518067397484, | |
| "grad_norm": 1.1686097383499146, | |
| "learning_rate": 7.843933243121335e-05, | |
| "loss": 0.5705, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.2368385437812965, | |
| "grad_norm": 6.95435094833374, | |
| "learning_rate": 7.889039242219217e-05, | |
| "loss": 0.4207, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.23819190688861822, | |
| "grad_norm": 1.6849822998046875, | |
| "learning_rate": 7.934145241317096e-05, | |
| "loss": 0.1384, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.23954526999593992, | |
| "grad_norm": 19.90876007080078, | |
| "learning_rate": 7.979251240414976e-05, | |
| "loss": 0.4826, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.2408986331032616, | |
| "grad_norm": 19.53013801574707, | |
| "learning_rate": 8.024357239512855e-05, | |
| "loss": 0.4383, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.2422519962105833, | |
| "grad_norm": 0.00940781645476818, | |
| "learning_rate": 8.069463238610736e-05, | |
| "loss": 0.1539, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.243605359317905, | |
| "grad_norm": 0.9285580515861511, | |
| "learning_rate": 8.114569237708616e-05, | |
| "loss": 0.0173, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.243605359317905, | |
| "eval_accuracy": 0.8964992389649924, | |
| "eval_loss": 0.7079048156738281, | |
| "eval_runtime": 586.1865, | |
| "eval_samples_per_second": 5.604, | |
| "eval_steps_per_second": 5.604, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.24495872242522668, | |
| "grad_norm": 1.0155157070812493e-07, | |
| "learning_rate": 8.159675236806495e-05, | |
| "loss": 0.6124, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.24631208553254838, | |
| "grad_norm": 26.104318618774414, | |
| "learning_rate": 8.204781235904376e-05, | |
| "loss": 1.1316, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.24766544863987008, | |
| "grad_norm": 36.21076202392578, | |
| "learning_rate": 8.249887235002256e-05, | |
| "loss": 1.1308, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.24901881174719176, | |
| "grad_norm": 6.247893333435059, | |
| "learning_rate": 8.294993234100135e-05, | |
| "loss": 0.3823, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.25037217485451346, | |
| "grad_norm": 2.5336642265319824, | |
| "learning_rate": 8.340099233198016e-05, | |
| "loss": 0.3386, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.25172553796183517, | |
| "grad_norm": 0.1235450878739357, | |
| "learning_rate": 8.385205232295896e-05, | |
| "loss": 0.282, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.25307890106915687, | |
| "grad_norm": 0.0033467705361545086, | |
| "learning_rate": 8.430311231393775e-05, | |
| "loss": 0.0699, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.2544322641764786, | |
| "grad_norm": 46.03356170654297, | |
| "learning_rate": 8.475417230491656e-05, | |
| "loss": 0.6011, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.2557856272838002, | |
| "grad_norm": 32.812896728515625, | |
| "learning_rate": 8.520523229589535e-05, | |
| "loss": 0.4655, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.2571389903911219, | |
| "grad_norm": 0.27037665247917175, | |
| "learning_rate": 8.565629228687417e-05, | |
| "loss": 0.2309, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2571389903911219, | |
| "eval_accuracy": 0.9007610350076104, | |
| "eval_loss": 0.35448792576789856, | |
| "eval_runtime": 585.6872, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2584923534984436, | |
| "grad_norm": 0.004123059567064047, | |
| "learning_rate": 8.610735227785296e-05, | |
| "loss": 0.4046, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.25984571660576533, | |
| "grad_norm": 11.693790435791016, | |
| "learning_rate": 8.655841226883175e-05, | |
| "loss": 0.1442, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.26119907971308703, | |
| "grad_norm": 0.026265673339366913, | |
| "learning_rate": 8.700947225981056e-05, | |
| "loss": 0.1297, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.26255244282040874, | |
| "grad_norm": 31.22381019592285, | |
| "learning_rate": 8.746053225078936e-05, | |
| "loss": 0.1566, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.2639058059277304, | |
| "grad_norm": 0.0626770630478859, | |
| "learning_rate": 8.791159224176817e-05, | |
| "loss": 1.2843, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.2652591690350521, | |
| "grad_norm": 6.931087017059326, | |
| "learning_rate": 8.836265223274696e-05, | |
| "loss": 0.8391, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.2666125321423738, | |
| "grad_norm": 1.1776670217514038, | |
| "learning_rate": 8.881371222372576e-05, | |
| "loss": 0.3079, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.2679658952496955, | |
| "grad_norm": 0.06983193755149841, | |
| "learning_rate": 8.926477221470457e-05, | |
| "loss": 0.1029, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.2693192583570172, | |
| "grad_norm": 3.716341972351074, | |
| "learning_rate": 8.971583220568336e-05, | |
| "loss": 0.3396, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.2706726214643389, | |
| "grad_norm": 35.3888053894043, | |
| "learning_rate": 9.016689219666216e-05, | |
| "loss": 0.3427, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2706726214643389, | |
| "eval_accuracy": 0.8840182648401826, | |
| "eval_loss": 0.6963387727737427, | |
| "eval_runtime": 586.0714, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.27202598457166055, | |
| "grad_norm": 0.007088725455105305, | |
| "learning_rate": 9.061795218764097e-05, | |
| "loss": 0.2362, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.27337934767898225, | |
| "grad_norm": 3.1651086807250977, | |
| "learning_rate": 9.106901217861976e-05, | |
| "loss": 0.2904, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.27473271078630396, | |
| "grad_norm": 0.019235916435718536, | |
| "learning_rate": 9.152007216959856e-05, | |
| "loss": 0.5162, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.27608607389362566, | |
| "grad_norm": 10.593847274780273, | |
| "learning_rate": 9.197113216057735e-05, | |
| "loss": 0.5283, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.27743943700094736, | |
| "grad_norm": 2.6608517169952393, | |
| "learning_rate": 9.242219215155616e-05, | |
| "loss": 0.0297, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.27879280010826907, | |
| "grad_norm": 11.656271934509277, | |
| "learning_rate": 9.287325214253496e-05, | |
| "loss": 0.6071, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.2801461632155907, | |
| "grad_norm": 3.8772504329681396, | |
| "learning_rate": 9.332431213351375e-05, | |
| "loss": 0.3125, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.2814995263229124, | |
| "grad_norm": 2.7373952865600586, | |
| "learning_rate": 9.377537212449257e-05, | |
| "loss": 0.8945, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.2828528894302341, | |
| "grad_norm": 0.14976456761360168, | |
| "learning_rate": 9.422643211547136e-05, | |
| "loss": 0.2199, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.2842062525375558, | |
| "grad_norm": 1.8281391859054565, | |
| "learning_rate": 9.467749210645016e-05, | |
| "loss": 0.5043, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2842062525375558, | |
| "eval_accuracy": 0.893455098934551, | |
| "eval_loss": 0.4096009433269501, | |
| "eval_runtime": 586.9389, | |
| "eval_samples_per_second": 5.597, | |
| "eval_steps_per_second": 5.597, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.28555961564487753, | |
| "grad_norm": 32.19676208496094, | |
| "learning_rate": 9.512855209742896e-05, | |
| "loss": 0.4847, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.28691297875219923, | |
| "grad_norm": 1.159845232963562, | |
| "learning_rate": 9.557961208840776e-05, | |
| "loss": 0.2801, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.28826634185952094, | |
| "grad_norm": 0.039890531450510025, | |
| "learning_rate": 9.603067207938657e-05, | |
| "loss": 0.3207, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.2896197049668426, | |
| "grad_norm": 0.00016595161287114024, | |
| "learning_rate": 9.648173207036536e-05, | |
| "loss": 0.2653, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.2909730680741643, | |
| "grad_norm": 0.2522432506084442, | |
| "learning_rate": 9.693279206134417e-05, | |
| "loss": 0.4832, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.292326431181486, | |
| "grad_norm": 10.649747848510742, | |
| "learning_rate": 9.738385205232297e-05, | |
| "loss": 0.1198, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.2936797942888077, | |
| "grad_norm": 22.584964752197266, | |
| "learning_rate": 9.783491204330176e-05, | |
| "loss": 0.5245, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.2950331573961294, | |
| "grad_norm": 1.6885179281234741, | |
| "learning_rate": 9.828597203428057e-05, | |
| "loss": 0.3199, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.2963865205034511, | |
| "grad_norm": 4.436751365661621, | |
| "learning_rate": 9.873703202525937e-05, | |
| "loss": 0.5085, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.29773988361077275, | |
| "grad_norm": 0.6824151873588562, | |
| "learning_rate": 9.918809201623816e-05, | |
| "loss": 0.6263, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.29773988361077275, | |
| "eval_accuracy": 0.869406392694064, | |
| "eval_loss": 0.4238719642162323, | |
| "eval_runtime": 586.4265, | |
| "eval_samples_per_second": 5.602, | |
| "eval_steps_per_second": 5.602, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.29909324671809445, | |
| "grad_norm": 0.0010402319021522999, | |
| "learning_rate": 9.963915200721697e-05, | |
| "loss": 0.391, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.30044660982541616, | |
| "grad_norm": 1.8170876502990723, | |
| "learning_rate": 9.999999752021549e-05, | |
| "loss": 0.262, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.30179997293273786, | |
| "grad_norm": 1.5593578815460205, | |
| "learning_rate": 9.999991072778369e-05, | |
| "loss": 0.217, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.30315333604005956, | |
| "grad_norm": 0.04098629578948021, | |
| "learning_rate": 9.999969994637269e-05, | |
| "loss": 0.3085, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.30450669914738127, | |
| "grad_norm": 3.2124032974243164, | |
| "learning_rate": 9.999936517650514e-05, | |
| "loss": 0.1197, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.3058600622547029, | |
| "grad_norm": 16.91176414489746, | |
| "learning_rate": 9.999890641901125e-05, | |
| "loss": 0.2527, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.3072134253620246, | |
| "grad_norm": 39.94670486450195, | |
| "learning_rate": 9.999832367502859e-05, | |
| "loss": 0.5771, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.3085667884693463, | |
| "grad_norm": 5.255732536315918, | |
| "learning_rate": 9.999761694600227e-05, | |
| "loss": 0.7849, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.309920151576668, | |
| "grad_norm": 5.741457462310791, | |
| "learning_rate": 9.999678623368483e-05, | |
| "loss": 0.6328, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.3112735146839897, | |
| "grad_norm": 2.6297736167907715, | |
| "learning_rate": 9.999583154013623e-05, | |
| "loss": 0.2523, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3112735146839897, | |
| "eval_accuracy": 0.8843226788432268, | |
| "eval_loss": 0.28884410858154297, | |
| "eval_runtime": 586.1248, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.31262687779131143, | |
| "grad_norm": 0.5248340368270874, | |
| "learning_rate": 9.999475286772394e-05, | |
| "loss": 0.2686, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.3139802408986331, | |
| "grad_norm": 0.18005584180355072, | |
| "learning_rate": 9.99935502191228e-05, | |
| "loss": 0.3142, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.3153336040059548, | |
| "grad_norm": 0.26872676610946655, | |
| "learning_rate": 9.999222359731514e-05, | |
| "loss": 0.1423, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.3166869671132765, | |
| "grad_norm": 39.6584587097168, | |
| "learning_rate": 9.99907730055907e-05, | |
| "loss": 0.5064, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.3180403302205982, | |
| "grad_norm": 27.253623962402344, | |
| "learning_rate": 9.998919844754661e-05, | |
| "loss": 0.6555, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.3193936933279199, | |
| "grad_norm": 14.004796981811523, | |
| "learning_rate": 9.998749992708744e-05, | |
| "loss": 0.3537, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.3207470564352416, | |
| "grad_norm": 10.884716033935547, | |
| "learning_rate": 9.998567744842517e-05, | |
| "loss": 0.1047, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.32210041954256324, | |
| "grad_norm": 13.677248001098633, | |
| "learning_rate": 9.998373101607915e-05, | |
| "loss": 0.4014, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.32345378264988495, | |
| "grad_norm": 0.055363189429044724, | |
| "learning_rate": 9.998166063487611e-05, | |
| "loss": 0.3669, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.32480714575720665, | |
| "grad_norm": 2.3912179470062256, | |
| "learning_rate": 9.997946630995013e-05, | |
| "loss": 0.2504, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.32480714575720665, | |
| "eval_accuracy": 0.8952815829528158, | |
| "eval_loss": 0.3796720802783966, | |
| "eval_runtime": 585.6459, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.32616050886452835, | |
| "grad_norm": 25.042545318603516, | |
| "learning_rate": 9.997714804674268e-05, | |
| "loss": 0.5136, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.32751387197185006, | |
| "grad_norm": 22.99011993408203, | |
| "learning_rate": 9.997470585100255e-05, | |
| "loss": 0.225, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.32886723507917176, | |
| "grad_norm": 18.535400390625, | |
| "learning_rate": 9.997213972878586e-05, | |
| "loss": 0.1548, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.33022059818649346, | |
| "grad_norm": 30.402786254882812, | |
| "learning_rate": 9.996944968645603e-05, | |
| "loss": 0.6825, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.3315739612938151, | |
| "grad_norm": 9.916972160339355, | |
| "learning_rate": 9.99666357306838e-05, | |
| "loss": 0.6176, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.3329273244011368, | |
| "grad_norm": 29.573097229003906, | |
| "learning_rate": 9.996369786844714e-05, | |
| "loss": 0.438, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.3342806875084585, | |
| "grad_norm": 2.503023386001587, | |
| "learning_rate": 9.996063610703137e-05, | |
| "loss": 0.464, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.3356340506157802, | |
| "grad_norm": 4.922478675842285, | |
| "learning_rate": 9.995745045402893e-05, | |
| "loss": 0.3029, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.3369874137231019, | |
| "grad_norm": 17.966007232666016, | |
| "learning_rate": 9.99541409173396e-05, | |
| "loss": 0.3314, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.33834077683042363, | |
| "grad_norm": 0.032199665904045105, | |
| "learning_rate": 9.99507075051703e-05, | |
| "loss": 0.3536, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.33834077683042363, | |
| "eval_accuracy": 0.9038051750380518, | |
| "eval_loss": 0.259170264005661, | |
| "eval_runtime": 585.8326, | |
| "eval_samples_per_second": 5.607, | |
| "eval_steps_per_second": 5.607, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3396941399377453, | |
| "grad_norm": 17.7895565032959, | |
| "learning_rate": 9.994715022603514e-05, | |
| "loss": 0.2788, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.341047503045067, | |
| "grad_norm": 0.004291262943297625, | |
| "learning_rate": 9.994346908875543e-05, | |
| "loss": 0.2134, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.3424008661523887, | |
| "grad_norm": 12.855873107910156, | |
| "learning_rate": 9.993966410245957e-05, | |
| "loss": 1.1679, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.3437542292597104, | |
| "grad_norm": 4.638855457305908, | |
| "learning_rate": 9.99357352765831e-05, | |
| "loss": 0.3796, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.3451075923670321, | |
| "grad_norm": 0.11056680232286453, | |
| "learning_rate": 9.99316826208687e-05, | |
| "loss": 0.1339, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.3464609554743538, | |
| "grad_norm": 0.006429149303585291, | |
| "learning_rate": 9.992750614536605e-05, | |
| "loss": 0.1356, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.34781431858167544, | |
| "grad_norm": 0.028468480333685875, | |
| "learning_rate": 9.992320586043192e-05, | |
| "loss": 0.1875, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.34916768168899714, | |
| "grad_norm": 0.16883939504623413, | |
| "learning_rate": 9.991878177673006e-05, | |
| "loss": 0.3505, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.35052104479631885, | |
| "grad_norm": 0.0001893436856335029, | |
| "learning_rate": 9.991423390523126e-05, | |
| "loss": 0.2705, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.35187440790364055, | |
| "grad_norm": 0.9219558835029602, | |
| "learning_rate": 9.990956225721328e-05, | |
| "loss": 0.3312, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.35187440790364055, | |
| "eval_accuracy": 0.898021308980213, | |
| "eval_loss": 0.3198848366737366, | |
| "eval_runtime": 586.1648, | |
| "eval_samples_per_second": 5.604, | |
| "eval_steps_per_second": 5.604, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.35322777101096225, | |
| "grad_norm": 0.16585978865623474, | |
| "learning_rate": 9.990476684426075e-05, | |
| "loss": 0.2798, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.35458113411828396, | |
| "grad_norm": 27.86931800842285, | |
| "learning_rate": 9.989984767826532e-05, | |
| "loss": 0.3643, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.3559344972256056, | |
| "grad_norm": 10.681389808654785, | |
| "learning_rate": 9.98948047714254e-05, | |
| "loss": 0.7329, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.3572878603329273, | |
| "grad_norm": 0.3092847168445587, | |
| "learning_rate": 9.988963813624635e-05, | |
| "loss": 0.2119, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.358641223440249, | |
| "grad_norm": 19.003339767456055, | |
| "learning_rate": 9.98843477855403e-05, | |
| "loss": 0.287, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.3599945865475707, | |
| "grad_norm": 13.935131072998047, | |
| "learning_rate": 9.987893373242616e-05, | |
| "loss": 0.2836, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.3613479496548924, | |
| "grad_norm": 0.20323041081428528, | |
| "learning_rate": 9.987339599032964e-05, | |
| "loss": 0.0202, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.3627013127622141, | |
| "grad_norm": 32.1824951171875, | |
| "learning_rate": 9.986773457298311e-05, | |
| "loss": 0.5027, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.36405467586953577, | |
| "grad_norm": 0.03664253652095795, | |
| "learning_rate": 9.986194949442568e-05, | |
| "loss": 0.4831, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.3654080389768575, | |
| "grad_norm": 0.41898396611213684, | |
| "learning_rate": 9.985604076900312e-05, | |
| "loss": 0.2259, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3654080389768575, | |
| "eval_accuracy": 0.9056316590563166, | |
| "eval_loss": 0.3357527554035187, | |
| "eval_runtime": 585.747, | |
| "eval_samples_per_second": 5.608, | |
| "eval_steps_per_second": 5.608, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3667614020841792, | |
| "grad_norm": 0.08296563476324081, | |
| "learning_rate": 9.985000841136775e-05, | |
| "loss": 0.213, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.3681147651915009, | |
| "grad_norm": 3.9525413513183594, | |
| "learning_rate": 9.984385243647855e-05, | |
| "loss": 0.1392, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.3694681282988226, | |
| "grad_norm": 0.15052391588687897, | |
| "learning_rate": 9.983757285960098e-05, | |
| "loss": 0.7221, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.3708214914061443, | |
| "grad_norm": 7.1570258140563965, | |
| "learning_rate": 9.983116969630706e-05, | |
| "loss": 0.3955, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.37217485451346594, | |
| "grad_norm": 0.4611501395702362, | |
| "learning_rate": 9.982464296247522e-05, | |
| "loss": 0.0577, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.37352821762078764, | |
| "grad_norm": 0.04361600801348686, | |
| "learning_rate": 9.981799267429037e-05, | |
| "loss": 0.3531, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.37488158072810934, | |
| "grad_norm": 5.342017650604248, | |
| "learning_rate": 9.98112188482438e-05, | |
| "loss": 0.174, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.37623494383543105, | |
| "grad_norm": 1.9316107034683228, | |
| "learning_rate": 9.980432150113311e-05, | |
| "loss": 0.5168, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.37758830694275275, | |
| "grad_norm": 0.29214081168174744, | |
| "learning_rate": 9.979730065006225e-05, | |
| "loss": 0.1109, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.37894167005007445, | |
| "grad_norm": 24.477149963378906, | |
| "learning_rate": 9.97901563124414e-05, | |
| "loss": 0.2328, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.37894167005007445, | |
| "eval_accuracy": 0.8964992389649924, | |
| "eval_loss": 0.5654892325401306, | |
| "eval_runtime": 586.152, | |
| "eval_samples_per_second": 5.604, | |
| "eval_steps_per_second": 5.604, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.38029503315739616, | |
| "grad_norm": 0.38575196266174316, | |
| "learning_rate": 9.978288850598698e-05, | |
| "loss": 0.6395, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.3816483962647178, | |
| "grad_norm": 2.8028533458709717, | |
| "learning_rate": 9.97754972487216e-05, | |
| "loss": 0.1854, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.3830017593720395, | |
| "grad_norm": 0.018388153985142708, | |
| "learning_rate": 9.976798255897394e-05, | |
| "loss": 0.3204, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.3843551224793612, | |
| "grad_norm": 0.017660358920693398, | |
| "learning_rate": 9.976034445537885e-05, | |
| "loss": 0.3979, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.3857084855866829, | |
| "grad_norm": 0.19985133409500122, | |
| "learning_rate": 9.975258295687715e-05, | |
| "loss": 0.1326, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.3870618486940046, | |
| "grad_norm": 25.679462432861328, | |
| "learning_rate": 9.97446980827157e-05, | |
| "loss": 0.6361, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.3884152118013263, | |
| "grad_norm": 25.165712356567383, | |
| "learning_rate": 9.973668985244724e-05, | |
| "loss": 0.5958, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.38976857490864797, | |
| "grad_norm": 14.897231101989746, | |
| "learning_rate": 9.972855828593051e-05, | |
| "loss": 0.462, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.3911219380159697, | |
| "grad_norm": 7.232959747314453, | |
| "learning_rate": 9.972030340333001e-05, | |
| "loss": 0.3582, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.3924753011232914, | |
| "grad_norm": 19.949562072753906, | |
| "learning_rate": 9.971192522511608e-05, | |
| "loss": 0.2967, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.3924753011232914, | |
| "eval_accuracy": 0.9068493150684932, | |
| "eval_loss": 0.2848670482635498, | |
| "eval_runtime": 585.5038, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.3938286642306131, | |
| "grad_norm": 6.787762641906738, | |
| "learning_rate": 9.970342377206477e-05, | |
| "loss": 0.3068, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.3951820273379348, | |
| "grad_norm": 0.035039156675338745, | |
| "learning_rate": 9.969479906525785e-05, | |
| "loss": 0.1438, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.3965353904452565, | |
| "grad_norm": 0.027156801894307137, | |
| "learning_rate": 9.968605112608273e-05, | |
| "loss": 0.4841, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.39788875355257813, | |
| "grad_norm": 16.14340591430664, | |
| "learning_rate": 9.967717997623245e-05, | |
| "loss": 0.1882, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.39924211665989984, | |
| "grad_norm": 3.19144606590271, | |
| "learning_rate": 9.966818563770548e-05, | |
| "loss": 0.3626, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.40059547976722154, | |
| "grad_norm": 1.6469515562057495, | |
| "learning_rate": 9.96590681328059e-05, | |
| "loss": 0.1981, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.40194884287454324, | |
| "grad_norm": 5.4307092796079814e-06, | |
| "learning_rate": 9.96498274841431e-05, | |
| "loss": 0.3645, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.40330220598186495, | |
| "grad_norm": 0.009263657964766026, | |
| "learning_rate": 9.964046371463193e-05, | |
| "loss": 0.1692, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.40465556908918665, | |
| "grad_norm": 12.660831451416016, | |
| "learning_rate": 9.963097684749251e-05, | |
| "loss": 0.2142, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.4060089321965083, | |
| "grad_norm": 3.7643866539001465, | |
| "learning_rate": 9.962136690625019e-05, | |
| "loss": 0.2978, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4060089321965083, | |
| "eval_accuracy": 0.8995433789954338, | |
| "eval_loss": 0.41810643672943115, | |
| "eval_runtime": 585.6633, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.40736229530383, | |
| "grad_norm": 25.727134704589844, | |
| "learning_rate": 9.96116339147356e-05, | |
| "loss": 0.2822, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.4087156584111517, | |
| "grad_norm": 0.004231706261634827, | |
| "learning_rate": 9.96017778970844e-05, | |
| "loss": 0.1751, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.4100690215184734, | |
| "grad_norm": 0.3602442741394043, | |
| "learning_rate": 9.959179887773744e-05, | |
| "loss": 0.1602, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.4114223846257951, | |
| "grad_norm": 46.329593658447266, | |
| "learning_rate": 9.95816968814405e-05, | |
| "loss": 1.1669, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.4127757477331168, | |
| "grad_norm": 19.935781478881836, | |
| "learning_rate": 9.957147193324434e-05, | |
| "loss": 0.7264, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.41412911084043846, | |
| "grad_norm": 3.980362892150879, | |
| "learning_rate": 9.956112405850466e-05, | |
| "loss": 0.2222, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.41548247394776017, | |
| "grad_norm": 24.50153160095215, | |
| "learning_rate": 9.955065328288193e-05, | |
| "loss": 0.3385, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.41683583705508187, | |
| "grad_norm": 28.30177879333496, | |
| "learning_rate": 9.954005963234141e-05, | |
| "loss": 0.5514, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.4181892001624036, | |
| "grad_norm": 1.61983323097229, | |
| "learning_rate": 9.952934313315306e-05, | |
| "loss": 0.0859, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.4195425632697253, | |
| "grad_norm": 6.002326011657715, | |
| "learning_rate": 9.95185038118915e-05, | |
| "loss": 0.4642, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.4195425632697253, | |
| "eval_accuracy": 0.9007610350076104, | |
| "eval_loss": 0.6233646869659424, | |
| "eval_runtime": 585.9277, | |
| "eval_samples_per_second": 5.606, | |
| "eval_steps_per_second": 5.606, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.420895926377047, | |
| "grad_norm": 1.9925620555877686, | |
| "learning_rate": 9.95075416954359e-05, | |
| "loss": 0.4925, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.42224928948436863, | |
| "grad_norm": 18.721527099609375, | |
| "learning_rate": 9.949645681096995e-05, | |
| "loss": 0.3548, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.42360265259169033, | |
| "grad_norm": 0.3973216712474823, | |
| "learning_rate": 9.948524918598175e-05, | |
| "loss": 0.085, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.42495601569901204, | |
| "grad_norm": 17.404512405395508, | |
| "learning_rate": 9.947391884826381e-05, | |
| "loss": 0.5475, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.42630937880633374, | |
| "grad_norm": 0.08265689015388489, | |
| "learning_rate": 9.94624658259129e-05, | |
| "loss": 0.6296, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.42766274191365544, | |
| "grad_norm": 1.8543760776519775, | |
| "learning_rate": 9.945089014733005e-05, | |
| "loss": 0.6776, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.42901610502097715, | |
| "grad_norm": 4.1883063316345215, | |
| "learning_rate": 9.943919184122043e-05, | |
| "loss": 0.2573, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.43036946812829885, | |
| "grad_norm": 0.24804548919200897, | |
| "learning_rate": 9.942737093659335e-05, | |
| "loss": 0.2905, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.4317228312356205, | |
| "grad_norm": 0.36752820014953613, | |
| "learning_rate": 9.941542746276207e-05, | |
| "loss": 0.5516, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.4330761943429422, | |
| "grad_norm": 0.18298649787902832, | |
| "learning_rate": 9.940336144934383e-05, | |
| "loss": 0.0876, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.4330761943429422, | |
| "eval_accuracy": 0.9068493150684932, | |
| "eval_loss": 0.3489275276660919, | |
| "eval_runtime": 586.2418, | |
| "eval_samples_per_second": 5.603, | |
| "eval_steps_per_second": 5.603, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.4344295574502639, | |
| "grad_norm": 6.957592972867133e-07, | |
| "learning_rate": 9.939117292625972e-05, | |
| "loss": 0.4436, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.4357829205575856, | |
| "grad_norm": 26.52113914489746, | |
| "learning_rate": 9.937886192373469e-05, | |
| "loss": 0.6185, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.4371362836649073, | |
| "grad_norm": 8.334674835205078, | |
| "learning_rate": 9.936642847229734e-05, | |
| "loss": 0.1342, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.438489646772229, | |
| "grad_norm": 0.1016705185174942, | |
| "learning_rate": 9.935387260277993e-05, | |
| "loss": 0.6313, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.43984300987955066, | |
| "grad_norm": 7.029394149780273, | |
| "learning_rate": 9.934119434631832e-05, | |
| "loss": 0.5876, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.44119637298687236, | |
| "grad_norm": 0.10425078123807907, | |
| "learning_rate": 9.932839373435185e-05, | |
| "loss": 0.1288, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.44254973609419407, | |
| "grad_norm": 10.366453170776367, | |
| "learning_rate": 9.931547079862329e-05, | |
| "loss": 0.4045, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.44390309920151577, | |
| "grad_norm": 0.6578564643859863, | |
| "learning_rate": 9.930242557117869e-05, | |
| "loss": 0.5181, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.4452564623088375, | |
| "grad_norm": 6.488555908203125, | |
| "learning_rate": 9.928925808436743e-05, | |
| "loss": 0.1018, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.4466098254161592, | |
| "grad_norm": 1.798299789428711, | |
| "learning_rate": 9.927596837084198e-05, | |
| "loss": 0.1241, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.4466098254161592, | |
| "eval_accuracy": 0.9156773211567732, | |
| "eval_loss": 0.298028826713562, | |
| "eval_runtime": 585.8439, | |
| "eval_samples_per_second": 5.607, | |
| "eval_steps_per_second": 5.607, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.4479631885234808, | |
| "grad_norm": 8.091084480285645, | |
| "learning_rate": 9.926255646355804e-05, | |
| "loss": 0.7205, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.44931655163080253, | |
| "grad_norm": 1.7898259162902832, | |
| "learning_rate": 9.924902239577418e-05, | |
| "loss": 0.3658, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.45066991473812423, | |
| "grad_norm": 7.857369422912598, | |
| "learning_rate": 9.923536620105201e-05, | |
| "loss": 0.1974, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.45202327784544594, | |
| "grad_norm": 1.4473187923431396, | |
| "learning_rate": 9.922158791325588e-05, | |
| "loss": 0.5274, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.45337664095276764, | |
| "grad_norm": 20.294761657714844, | |
| "learning_rate": 9.920768756655304e-05, | |
| "loss": 0.4609, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.45473000406008934, | |
| "grad_norm": 0.015980269759893417, | |
| "learning_rate": 9.919366519541332e-05, | |
| "loss": 0.1363, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.456083367167411, | |
| "grad_norm": 7.722431659698486, | |
| "learning_rate": 9.917952083460916e-05, | |
| "loss": 0.3419, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.4574367302747327, | |
| "grad_norm": 0.9992192387580872, | |
| "learning_rate": 9.91652545192155e-05, | |
| "loss": 0.6367, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.4587900933820544, | |
| "grad_norm": 30.987764358520508, | |
| "learning_rate": 9.915086628460977e-05, | |
| "loss": 0.813, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.4601434564893761, | |
| "grad_norm": 16.1995906829834, | |
| "learning_rate": 9.913635616647166e-05, | |
| "loss": 0.6461, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.4601434564893761, | |
| "eval_accuracy": 0.908675799086758, | |
| "eval_loss": 0.25029391050338745, | |
| "eval_runtime": 607.2706, | |
| "eval_samples_per_second": 5.409, | |
| "eval_steps_per_second": 5.409, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.4614968195966978, | |
| "grad_norm": 3.102372407913208, | |
| "learning_rate": 9.912172420078312e-05, | |
| "loss": 0.2572, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.4628501827040195, | |
| "grad_norm": 11.124046325683594, | |
| "learning_rate": 9.910697042382829e-05, | |
| "loss": 0.2574, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.46420354581134116, | |
| "grad_norm": 8.01544189453125, | |
| "learning_rate": 9.909209487219333e-05, | |
| "loss": 0.5683, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.46555690891866286, | |
| "grad_norm": 9.192936897277832, | |
| "learning_rate": 9.90770975827664e-05, | |
| "loss": 0.2297, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.46691027202598456, | |
| "grad_norm": 7.2200117111206055, | |
| "learning_rate": 9.906197859273753e-05, | |
| "loss": 0.169, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.46826363513330627, | |
| "grad_norm": 19.230791091918945, | |
| "learning_rate": 9.904673793959857e-05, | |
| "loss": 0.5687, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.46961699824062797, | |
| "grad_norm": 5.5020012855529785, | |
| "learning_rate": 9.903137566114304e-05, | |
| "loss": 0.5738, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.4709703613479497, | |
| "grad_norm": 8.84924030303955, | |
| "learning_rate": 9.901589179546606e-05, | |
| "loss": 0.2117, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.4723237244552714, | |
| "grad_norm": 17.913070678710938, | |
| "learning_rate": 9.900028638096428e-05, | |
| "loss": 0.3447, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.473677087562593, | |
| "grad_norm": 12.876395225524902, | |
| "learning_rate": 9.898455945633576e-05, | |
| "loss": 0.1405, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.473677087562593, | |
| "eval_accuracy": 0.9077625570776255, | |
| "eval_loss": 0.2597663700580597, | |
| "eval_runtime": 586.0918, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.47503045066991473, | |
| "grad_norm": 13.94190502166748, | |
| "learning_rate": 9.896871106057989e-05, | |
| "loss": 0.1196, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.47638381377723643, | |
| "grad_norm": 0.00028016060241498053, | |
| "learning_rate": 9.895274123299723e-05, | |
| "loss": 0.3356, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.47773717688455813, | |
| "grad_norm": 21.0711669921875, | |
| "learning_rate": 9.893665001318954e-05, | |
| "loss": 0.3105, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.47909053999187984, | |
| "grad_norm": 2.196962594985962, | |
| "learning_rate": 9.892043744105957e-05, | |
| "loss": 0.3484, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.48044390309920154, | |
| "grad_norm": 9.535185813903809, | |
| "learning_rate": 9.890410355681097e-05, | |
| "loss": 0.5164, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.4817972662065232, | |
| "grad_norm": 10.87157917022705, | |
| "learning_rate": 9.888764840094825e-05, | |
| "loss": 0.2638, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.4831506293138449, | |
| "grad_norm": 3.433655023574829, | |
| "learning_rate": 9.887107201427666e-05, | |
| "loss": 0.0772, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.4845039924211666, | |
| "grad_norm": 18.490392684936523, | |
| "learning_rate": 9.885437443790204e-05, | |
| "loss": 0.224, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.4858573555284883, | |
| "grad_norm": 36.07349395751953, | |
| "learning_rate": 9.88375557132308e-05, | |
| "loss": 0.1838, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.48721071863581, | |
| "grad_norm": 0.00971564557403326, | |
| "learning_rate": 9.882061588196971e-05, | |
| "loss": 0.0814, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.48721071863581, | |
| "eval_accuracy": 0.9053272450532724, | |
| "eval_loss": 0.39746227860450745, | |
| "eval_runtime": 585.9258, | |
| "eval_samples_per_second": 5.607, | |
| "eval_steps_per_second": 5.607, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.4885640817431317, | |
| "grad_norm": 8.707074165344238, | |
| "learning_rate": 9.880355498612593e-05, | |
| "loss": 0.1846, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.48991744485045335, | |
| "grad_norm": 0.2467634230852127, | |
| "learning_rate": 9.878637306800676e-05, | |
| "loss": 0.5841, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.49127080795777506, | |
| "grad_norm": 11.655877113342285, | |
| "learning_rate": 9.876907017021967e-05, | |
| "loss": 0.6614, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.49262417106509676, | |
| "grad_norm": 2.5910391807556152, | |
| "learning_rate": 9.87516463356721e-05, | |
| "loss": 0.3747, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.49397753417241846, | |
| "grad_norm": 0.08143145591020584, | |
| "learning_rate": 9.873410160757139e-05, | |
| "loss": 0.2551, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.49533089727974017, | |
| "grad_norm": 0.006202420219779015, | |
| "learning_rate": 9.871643602942469e-05, | |
| "loss": 0.1721, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.49668426038706187, | |
| "grad_norm": 0.4420149028301239, | |
| "learning_rate": 9.869864964503881e-05, | |
| "loss": 0.2613, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.4980376234943835, | |
| "grad_norm": 4.616814613342285, | |
| "learning_rate": 9.868074249852016e-05, | |
| "loss": 0.3163, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.4993909866017052, | |
| "grad_norm": 1.4008029699325562, | |
| "learning_rate": 9.866271463427457e-05, | |
| "loss": 0.392, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.5007443497090269, | |
| "grad_norm": 10.468450546264648, | |
| "learning_rate": 9.864456609700726e-05, | |
| "loss": 0.2675, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5007443497090269, | |
| "eval_accuracy": 0.9092846270928463, | |
| "eval_loss": 0.31681403517723083, | |
| "eval_runtime": 585.9869, | |
| "eval_samples_per_second": 5.606, | |
| "eval_steps_per_second": 5.606, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5020977128163486, | |
| "grad_norm": 3.0164525508880615, | |
| "learning_rate": 9.862629693172267e-05, | |
| "loss": 0.3723, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.5034510759236703, | |
| "grad_norm": 0.2863682210445404, | |
| "learning_rate": 9.860790718372441e-05, | |
| "loss": 0.3251, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.504804439030992, | |
| "grad_norm": 0.02388608641922474, | |
| "learning_rate": 9.858939689861506e-05, | |
| "loss": 0.2996, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.5061578021383137, | |
| "grad_norm": 0.42931804060935974, | |
| "learning_rate": 9.857076612229614e-05, | |
| "loss": 0.4254, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.5075111652456354, | |
| "grad_norm": 0.24269169569015503, | |
| "learning_rate": 9.855201490096795e-05, | |
| "loss": 0.2688, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.5088645283529571, | |
| "grad_norm": 3.338435649871826, | |
| "learning_rate": 9.853314328112947e-05, | |
| "loss": 0.0494, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.5102178914602787, | |
| "grad_norm": 0.04143134132027626, | |
| "learning_rate": 9.851415130957824e-05, | |
| "loss": 0.2798, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.5115712545676004, | |
| "grad_norm": 0.18607792258262634, | |
| "learning_rate": 9.849503903341024e-05, | |
| "loss": 0.2892, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.5129246176749221, | |
| "grad_norm": 11.773528099060059, | |
| "learning_rate": 9.84758065000198e-05, | |
| "loss": 0.4357, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.5142779807822438, | |
| "grad_norm": 0.7222903370857239, | |
| "learning_rate": 9.845645375709945e-05, | |
| "loss": 0.3765, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5142779807822438, | |
| "eval_accuracy": 0.919634703196347, | |
| "eval_loss": 0.2469821274280548, | |
| "eval_runtime": 585.5969, | |
| "eval_samples_per_second": 5.61, | |
| "eval_steps_per_second": 5.61, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5156313438895656, | |
| "grad_norm": 0.587492823600769, | |
| "learning_rate": 9.84369808526398e-05, | |
| "loss": 0.2611, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.5169847069968873, | |
| "grad_norm": 1.6213396787643433, | |
| "learning_rate": 9.841738783492944e-05, | |
| "loss": 0.4305, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.518338070104209, | |
| "grad_norm": 25.090078353881836, | |
| "learning_rate": 9.839767475255484e-05, | |
| "loss": 0.3079, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.5196914332115307, | |
| "grad_norm": 1.8611633777618408, | |
| "learning_rate": 9.837784165440018e-05, | |
| "loss": 0.312, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.5210447963188524, | |
| "grad_norm": 16.361417770385742, | |
| "learning_rate": 9.835788858964726e-05, | |
| "loss": 0.1514, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.5223981594261741, | |
| "grad_norm": 0.027297548949718475, | |
| "learning_rate": 9.833781560777537e-05, | |
| "loss": 0.8527, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.5237515225334958, | |
| "grad_norm": 9.58652400970459, | |
| "learning_rate": 9.831762275856118e-05, | |
| "loss": 0.3958, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.5251048856408175, | |
| "grad_norm": 0.11664781719446182, | |
| "learning_rate": 9.829731009207859e-05, | |
| "loss": 0.105, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.5264582487481392, | |
| "grad_norm": 17.347599029541016, | |
| "learning_rate": 9.82768776586986e-05, | |
| "loss": 0.5082, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.5278116118554608, | |
| "grad_norm": 10.610013961791992, | |
| "learning_rate": 9.825632550908926e-05, | |
| "loss": 0.3189, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.5278116118554608, | |
| "eval_accuracy": 0.9147640791476408, | |
| "eval_loss": 0.21581518650054932, | |
| "eval_runtime": 585.6064, | |
| "eval_samples_per_second": 5.61, | |
| "eval_steps_per_second": 5.61, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.5291649749627825, | |
| "grad_norm": 5.391573905944824, | |
| "learning_rate": 9.823565369421545e-05, | |
| "loss": 0.2908, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.5305183380701042, | |
| "grad_norm": 18.12822723388672, | |
| "learning_rate": 9.821486226533882e-05, | |
| "loss": 0.3401, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.5318717011774259, | |
| "grad_norm": 0.004009630065411329, | |
| "learning_rate": 9.819395127401762e-05, | |
| "loss": 0.1484, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.5332250642847476, | |
| "grad_norm": 0.21845737099647522, | |
| "learning_rate": 9.817292077210659e-05, | |
| "loss": 0.2242, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.5345784273920693, | |
| "grad_norm": 0.21392692625522614, | |
| "learning_rate": 9.81517708117568e-05, | |
| "loss": 0.3041, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.535931790499391, | |
| "grad_norm": 0.029498528689146042, | |
| "learning_rate": 9.813050144541562e-05, | |
| "loss": 0.0472, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.5372851536067127, | |
| "grad_norm": 0.0002741254575084895, | |
| "learning_rate": 9.81091127258265e-05, | |
| "loss": 0.0588, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.5386385167140344, | |
| "grad_norm": 0.002227282151579857, | |
| "learning_rate": 9.808760470602879e-05, | |
| "loss": 0.183, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.5399918798213561, | |
| "grad_norm": 11.413102149963379, | |
| "learning_rate": 9.806597743935778e-05, | |
| "loss": 0.0947, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.5413452429286778, | |
| "grad_norm": 2.3081635447397275e-07, | |
| "learning_rate": 9.804423097944439e-05, | |
| "loss": 0.1807, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5413452429286778, | |
| "eval_accuracy": 0.9141552511415525, | |
| "eval_loss": 0.40402328968048096, | |
| "eval_runtime": 586.0899, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5426986060359995, | |
| "grad_norm": 24.99509620666504, | |
| "learning_rate": 9.802236538021518e-05, | |
| "loss": 0.2574, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.5440519691433211, | |
| "grad_norm": 0.0018494591349735856, | |
| "learning_rate": 9.800038069589208e-05, | |
| "loss": 0.2827, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.5454053322506428, | |
| "grad_norm": 0.6751343011856079, | |
| "learning_rate": 9.797827698099238e-05, | |
| "loss": 0.2763, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.5467586953579645, | |
| "grad_norm": 3.1831467151641846, | |
| "learning_rate": 9.795605429032851e-05, | |
| "loss": 0.1686, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.5481120584652862, | |
| "grad_norm": 3.9633137930650264e-05, | |
| "learning_rate": 9.793371267900793e-05, | |
| "loss": 0.1481, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.5494654215726079, | |
| "grad_norm": 0.24243293702602386, | |
| "learning_rate": 9.791125220243303e-05, | |
| "loss": 0.1069, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.5508187846799296, | |
| "grad_norm": 28.994619369506836, | |
| "learning_rate": 9.788867291630091e-05, | |
| "loss": 0.7202, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.5521721477872513, | |
| "grad_norm": 23.049583435058594, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.1929, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.553525510894573, | |
| "grad_norm": 0.003061707131564617, | |
| "learning_rate": 9.784315813962662e-05, | |
| "loss": 0.0455, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.5548788740018947, | |
| "grad_norm": 0.8015981912612915, | |
| "learning_rate": 9.782022276195124e-05, | |
| "loss": 0.1293, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.5548788740018947, | |
| "eval_accuracy": 0.9080669710806697, | |
| "eval_loss": 0.5554362535476685, | |
| "eval_runtime": 585.9472, | |
| "eval_samples_per_second": 5.606, | |
| "eval_steps_per_second": 5.606, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.5562322371092164, | |
| "grad_norm": 10.175394058227539, | |
| "learning_rate": 9.7797168800452e-05, | |
| "loss": 0.7186, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.5575856002165381, | |
| "grad_norm": 2.411634341115132e-07, | |
| "learning_rate": 9.777399631229777e-05, | |
| "loss": 0.4884, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.5589389633238598, | |
| "grad_norm": 0.3643926978111267, | |
| "learning_rate": 9.775070535495132e-05, | |
| "loss": 0.453, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.5602923264311814, | |
| "grad_norm": 3.4061479254887672e-06, | |
| "learning_rate": 9.772729598616916e-05, | |
| "loss": 0.2994, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.5616456895385031, | |
| "grad_norm": 0.7406808137893677, | |
| "learning_rate": 9.77037682640015e-05, | |
| "loss": 0.428, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.5629990526458248, | |
| "grad_norm": 0.026358220726251602, | |
| "learning_rate": 9.768012224679198e-05, | |
| "loss": 0.1555, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.5643524157531465, | |
| "grad_norm": 0.08676059544086456, | |
| "learning_rate": 9.765635799317765e-05, | |
| "loss": 0.2118, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.5657057788604682, | |
| "grad_norm": 0.06001422554254532, | |
| "learning_rate": 9.76324755620887e-05, | |
| "loss": 0.1356, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.56705914196779, | |
| "grad_norm": 4.249134063720703, | |
| "learning_rate": 9.76084750127484e-05, | |
| "loss": 0.3016, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.5684125050751117, | |
| "grad_norm": 8.967272758483887, | |
| "learning_rate": 9.758435640467295e-05, | |
| "loss": 0.2838, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5684125050751117, | |
| "eval_accuracy": 0.9105022831050228, | |
| "eval_loss": 0.32533395290374756, | |
| "eval_runtime": 585.8922, | |
| "eval_samples_per_second": 5.607, | |
| "eval_steps_per_second": 5.607, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5697658681824334, | |
| "grad_norm": 7.07581090927124, | |
| "learning_rate": 9.756011979767128e-05, | |
| "loss": 0.4945, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.5711192312897551, | |
| "grad_norm": 2.3176674842834473, | |
| "learning_rate": 9.753576525184492e-05, | |
| "loss": 0.3317, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.5724725943970768, | |
| "grad_norm": 0.14022812247276306, | |
| "learning_rate": 9.751129282758791e-05, | |
| "loss": 0.2113, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.5738259575043985, | |
| "grad_norm": 5.834623516420834e-05, | |
| "learning_rate": 9.748670258558656e-05, | |
| "loss": 0.644, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.5751793206117202, | |
| "grad_norm": 7.6325764656066895, | |
| "learning_rate": 9.746199458681938e-05, | |
| "loss": 0.0884, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.5765326837190419, | |
| "grad_norm": 1.895483374595642, | |
| "learning_rate": 9.743716889255684e-05, | |
| "loss": 0.258, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.5778860468263635, | |
| "grad_norm": 1.6728992462158203, | |
| "learning_rate": 9.741222556436132e-05, | |
| "loss": 0.0853, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.5792394099336852, | |
| "grad_norm": 0.009905705228447914, | |
| "learning_rate": 9.738716466408688e-05, | |
| "loss": 0.4743, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.5805927730410069, | |
| "grad_norm": 0.7430686354637146, | |
| "learning_rate": 9.736198625387916e-05, | |
| "loss": 0.7162, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.5819461361483286, | |
| "grad_norm": 7.7744269371032715, | |
| "learning_rate": 9.733669039617513e-05, | |
| "loss": 0.2627, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.5819461361483286, | |
| "eval_accuracy": 0.915068493150685, | |
| "eval_loss": 0.23285283148288727, | |
| "eval_runtime": 585.6291, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.5832994992556503, | |
| "grad_norm": 18.253538131713867, | |
| "learning_rate": 9.731127715370311e-05, | |
| "loss": 0.5606, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.584652862362972, | |
| "grad_norm": 12.1471529006958, | |
| "learning_rate": 9.728574658948242e-05, | |
| "loss": 0.5073, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.5860062254702937, | |
| "grad_norm": 13.163832664489746, | |
| "learning_rate": 9.726009876682333e-05, | |
| "loss": 0.2834, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.5873595885776154, | |
| "grad_norm": 7.22966194152832, | |
| "learning_rate": 9.723433374932696e-05, | |
| "loss": 0.1887, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.5887129516849371, | |
| "grad_norm": 3.5801568031311035, | |
| "learning_rate": 9.720845160088492e-05, | |
| "loss": 0.1997, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.5900663147922588, | |
| "grad_norm": 14.825773239135742, | |
| "learning_rate": 9.718245238567939e-05, | |
| "loss": 0.3341, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.5914196778995805, | |
| "grad_norm": 0.009115755558013916, | |
| "learning_rate": 9.71563361681828e-05, | |
| "loss": 0.0791, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.5927730410069022, | |
| "grad_norm": 28.929086685180664, | |
| "learning_rate": 9.713010301315772e-05, | |
| "loss": 0.4627, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.5941264041142238, | |
| "grad_norm": 0.04894016310572624, | |
| "learning_rate": 9.710375298565672e-05, | |
| "loss": 0.2501, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.5954797672215455, | |
| "grad_norm": 0.1197153702378273, | |
| "learning_rate": 9.707728615102217e-05, | |
| "loss": 0.1651, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.5954797672215455, | |
| "eval_accuracy": 0.9098934550989346, | |
| "eval_loss": 0.3739698827266693, | |
| "eval_runtime": 585.3122, | |
| "eval_samples_per_second": 5.612, | |
| "eval_steps_per_second": 5.612, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.5968331303288672, | |
| "grad_norm": 18.560256958007812, | |
| "learning_rate": 9.705070257488609e-05, | |
| "loss": 0.2983, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.5981864934361889, | |
| "grad_norm": 1.4426162242889404, | |
| "learning_rate": 9.702400232317003e-05, | |
| "loss": 0.3937, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.5995398565435106, | |
| "grad_norm": 35.894256591796875, | |
| "learning_rate": 9.699718546208484e-05, | |
| "loss": 0.4369, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.6008932196508323, | |
| "grad_norm": 24.106157302856445, | |
| "learning_rate": 9.697025205813054e-05, | |
| "loss": 0.3225, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.602246582758154, | |
| "grad_norm": 20.590375900268555, | |
| "learning_rate": 9.694320217809616e-05, | |
| "loss": 0.342, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.6035999458654757, | |
| "grad_norm": 27.700132369995117, | |
| "learning_rate": 9.691603588905955e-05, | |
| "loss": 0.2763, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.6049533089727974, | |
| "grad_norm": 1.0037168264389038, | |
| "learning_rate": 9.688875325838725e-05, | |
| "loss": 0.1983, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.6063066720801191, | |
| "grad_norm": 2.838923478520883e-07, | |
| "learning_rate": 9.686135435373428e-05, | |
| "loss": 0.3122, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.6076600351874408, | |
| "grad_norm": 1.0174156427383423, | |
| "learning_rate": 9.683383924304401e-05, | |
| "loss": 0.0064, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.6090133982947625, | |
| "grad_norm": 0.011479837819933891, | |
| "learning_rate": 9.6806207994548e-05, | |
| "loss": 0.4422, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6090133982947625, | |
| "eval_accuracy": 0.9126331811263318, | |
| "eval_loss": 0.4408724009990692, | |
| "eval_runtime": 585.495, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6103667614020842, | |
| "grad_norm": 25.792768478393555, | |
| "learning_rate": 9.677846067676572e-05, | |
| "loss": 0.4849, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.6117201245094058, | |
| "grad_norm": 0.0019514035666361451, | |
| "learning_rate": 9.675059735850457e-05, | |
| "loss": 0.4136, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.6130734876167275, | |
| "grad_norm": 14.797130584716797, | |
| "learning_rate": 9.672261810885955e-05, | |
| "loss": 0.7369, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.6144268507240492, | |
| "grad_norm": 9.552214622497559, | |
| "learning_rate": 9.669452299721316e-05, | |
| "loss": 0.2675, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.6157802138313709, | |
| "grad_norm": 15.778959274291992, | |
| "learning_rate": 9.66663120932352e-05, | |
| "loss": 0.1837, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.6171335769386926, | |
| "grad_norm": 0.35701024532318115, | |
| "learning_rate": 9.663798546688262e-05, | |
| "loss": 0.0761, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.6184869400460143, | |
| "grad_norm": 31.343671798706055, | |
| "learning_rate": 9.660954318839933e-05, | |
| "loss": 0.2741, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.619840303153336, | |
| "grad_norm": 4.513859676080756e-05, | |
| "learning_rate": 9.658098532831605e-05, | |
| "loss": 0.2051, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.6211936662606578, | |
| "grad_norm": 0.13937516510486603, | |
| "learning_rate": 9.65523119574501e-05, | |
| "loss": 1.4443, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.6225470293679795, | |
| "grad_norm": 0.54905104637146, | |
| "learning_rate": 9.652352314690524e-05, | |
| "loss": 0.125, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.6225470293679795, | |
| "eval_accuracy": 0.9025875190258752, | |
| "eval_loss": 0.39096349477767944, | |
| "eval_runtime": 585.735, | |
| "eval_samples_per_second": 5.608, | |
| "eval_steps_per_second": 5.608, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.6239003924753012, | |
| "grad_norm": 0.001251073321327567, | |
| "learning_rate": 9.649461896807153e-05, | |
| "loss": 0.4539, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.6252537555826229, | |
| "grad_norm": 0.060062967240810394, | |
| "learning_rate": 9.646559949262504e-05, | |
| "loss": 0.1364, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.6266071186899446, | |
| "grad_norm": 20.359573364257812, | |
| "learning_rate": 9.643646479252784e-05, | |
| "loss": 0.4804, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.6279604817972662, | |
| "grad_norm": 22.817575454711914, | |
| "learning_rate": 9.640721494002769e-05, | |
| "loss": 0.4074, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.6293138449045879, | |
| "grad_norm": 1.5756419897079468, | |
| "learning_rate": 9.637785000765789e-05, | |
| "loss": 0.1832, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.6306672080119096, | |
| "grad_norm": 0.32080402970314026, | |
| "learning_rate": 9.634837006823714e-05, | |
| "loss": 0.266, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.6320205711192313, | |
| "grad_norm": 0.09690892696380615, | |
| "learning_rate": 9.631877519486934e-05, | |
| "loss": 0.0025, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.633373934226553, | |
| "grad_norm": 26.12261199951172, | |
| "learning_rate": 9.628906546094333e-05, | |
| "loss": 0.1316, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.6347272973338747, | |
| "grad_norm": 1.7195595502853394, | |
| "learning_rate": 9.62592409401329e-05, | |
| "loss": 0.531, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.6360806604411964, | |
| "grad_norm": 20.064390182495117, | |
| "learning_rate": 9.622930170639637e-05, | |
| "loss": 0.4875, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.6360806604411964, | |
| "eval_accuracy": 0.9141552511415525, | |
| "eval_loss": 0.31653645634651184, | |
| "eval_runtime": 585.8969, | |
| "eval_samples_per_second": 5.607, | |
| "eval_steps_per_second": 5.607, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.6374340235485181, | |
| "grad_norm": 18.887651443481445, | |
| "learning_rate": 9.619924783397661e-05, | |
| "loss": 0.1611, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.6387873866558398, | |
| "grad_norm": 0.010769748128950596, | |
| "learning_rate": 9.61690793974007e-05, | |
| "loss": 0.3398, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.6401407497631615, | |
| "grad_norm": 7.151393413543701, | |
| "learning_rate": 9.613879647147988e-05, | |
| "loss": 0.0745, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.6414941128704832, | |
| "grad_norm": 0.03005339950323105, | |
| "learning_rate": 9.610839913130923e-05, | |
| "loss": 0.2734, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.6428474759778049, | |
| "grad_norm": 3.4520528316497803, | |
| "learning_rate": 9.60778874522676e-05, | |
| "loss": 0.7906, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.6442008390851265, | |
| "grad_norm": 7.305643558502197, | |
| "learning_rate": 9.604726151001737e-05, | |
| "loss": 0.4846, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.6455542021924482, | |
| "grad_norm": 19.83353042602539, | |
| "learning_rate": 9.601652138050428e-05, | |
| "loss": 0.1922, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.6469075652997699, | |
| "grad_norm": 17.064523696899414, | |
| "learning_rate": 9.598566713995718e-05, | |
| "loss": 0.4546, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.6482609284070916, | |
| "grad_norm": 14.504972457885742, | |
| "learning_rate": 9.595469886488793e-05, | |
| "loss": 0.3455, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.6496142915144133, | |
| "grad_norm": 5.4947285652160645, | |
| "learning_rate": 9.592361663209117e-05, | |
| "loss": 0.1002, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.6496142915144133, | |
| "eval_accuracy": 0.9178082191780822, | |
| "eval_loss": 0.29287853837013245, | |
| "eval_runtime": 585.872, | |
| "eval_samples_per_second": 5.607, | |
| "eval_steps_per_second": 5.607, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.650967654621735, | |
| "grad_norm": 0.10837817192077637, | |
| "learning_rate": 9.589242051864413e-05, | |
| "loss": 0.12, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.6523210177290567, | |
| "grad_norm": 16.458295822143555, | |
| "learning_rate": 9.586111060190641e-05, | |
| "loss": 0.115, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.6536743808363784, | |
| "grad_norm": 3.44444160873536e-05, | |
| "learning_rate": 9.582968695951986e-05, | |
| "loss": 0.4776, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.6550277439437001, | |
| "grad_norm": 27.302371978759766, | |
| "learning_rate": 9.579814966940833e-05, | |
| "loss": 0.9683, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.6563811070510218, | |
| "grad_norm": 0.0014636669075116515, | |
| "learning_rate": 9.576649880977748e-05, | |
| "loss": 0.7764, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.6577344701583435, | |
| "grad_norm": 2.055596113204956, | |
| "learning_rate": 9.573473445911461e-05, | |
| "loss": 0.177, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.6590878332656652, | |
| "grad_norm": 17.790176391601562, | |
| "learning_rate": 9.570285669618842e-05, | |
| "loss": 0.2155, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.6604411963729869, | |
| "grad_norm": 1.0239602327346802, | |
| "learning_rate": 9.567086560004892e-05, | |
| "loss": 0.1939, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.6617945594803085, | |
| "grad_norm": 18.533052444458008, | |
| "learning_rate": 9.563876125002711e-05, | |
| "loss": 0.6398, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.6631479225876302, | |
| "grad_norm": 0.025914179161190987, | |
| "learning_rate": 9.560654372573481e-05, | |
| "loss": 0.1241, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.6631479225876302, | |
| "eval_accuracy": 0.9101978691019786, | |
| "eval_loss": 0.2989567220211029, | |
| "eval_runtime": 585.4692, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.6645012856949519, | |
| "grad_norm": 25.82878875732422, | |
| "learning_rate": 9.557421310706456e-05, | |
| "loss": 0.4629, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.6658546488022736, | |
| "grad_norm": 2.8251757621765137, | |
| "learning_rate": 9.554176947418931e-05, | |
| "loss": 0.5236, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.6672080119095953, | |
| "grad_norm": 1.0843572616577148, | |
| "learning_rate": 9.550921290756222e-05, | |
| "loss": 0.0906, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.668561375016917, | |
| "grad_norm": 2.299241542816162, | |
| "learning_rate": 9.54765434879166e-05, | |
| "loss": 0.1686, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.6699147381242387, | |
| "grad_norm": 0.24002690613269806, | |
| "learning_rate": 9.54437612962655e-05, | |
| "loss": 0.3661, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.6712681012315604, | |
| "grad_norm": 13.394685745239258, | |
| "learning_rate": 9.54108664139017e-05, | |
| "loss": 0.2536, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.6726214643388821, | |
| "grad_norm": 20.657196044921875, | |
| "learning_rate": 9.537785892239743e-05, | |
| "loss": 0.3507, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.6739748274462039, | |
| "grad_norm": 4.67914342880249, | |
| "learning_rate": 9.534473890360412e-05, | |
| "loss": 0.5186, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.6753281905535256, | |
| "grad_norm": 7.856649398803711, | |
| "learning_rate": 9.531150643965223e-05, | |
| "loss": 0.7312, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.6766815536608473, | |
| "grad_norm": 6.752259731292725, | |
| "learning_rate": 9.527816161295114e-05, | |
| "loss": 0.1414, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6766815536608473, | |
| "eval_accuracy": 0.9117199391171994, | |
| "eval_loss": 0.2604728937149048, | |
| "eval_runtime": 585.4721, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6780349167681688, | |
| "grad_norm": 1.1008288860321045, | |
| "learning_rate": 9.524470450618879e-05, | |
| "loss": 0.2198, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.6793882798754906, | |
| "grad_norm": 7.540661811828613, | |
| "learning_rate": 9.521113520233158e-05, | |
| "loss": 0.6508, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.6807416429828123, | |
| "grad_norm": 0.06511321663856506, | |
| "learning_rate": 9.517745378462416e-05, | |
| "loss": 0.3953, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.682095006090134, | |
| "grad_norm": 0.5184662342071533, | |
| "learning_rate": 9.514366033658914e-05, | |
| "loss": 0.3206, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.6834483691974557, | |
| "grad_norm": 4.385899543762207, | |
| "learning_rate": 9.510975494202698e-05, | |
| "loss": 0.3306, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.6848017323047774, | |
| "grad_norm": 0.368313729763031, | |
| "learning_rate": 9.507573768501574e-05, | |
| "loss": 0.0516, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.6861550954120991, | |
| "grad_norm": 27.66187286376953, | |
| "learning_rate": 9.504160864991087e-05, | |
| "loss": 0.4828, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.6875084585194208, | |
| "grad_norm": 20.792448043823242, | |
| "learning_rate": 9.500736792134501e-05, | |
| "loss": 0.751, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.6888618216267425, | |
| "grad_norm": 6.490540504455566, | |
| "learning_rate": 9.497301558422776e-05, | |
| "loss": 0.3981, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.6902151847340642, | |
| "grad_norm": 6.3285136222839355, | |
| "learning_rate": 9.493855172374551e-05, | |
| "loss": 0.2298, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.6902151847340642, | |
| "eval_accuracy": 0.9159817351598174, | |
| "eval_loss": 0.22350817918777466, | |
| "eval_runtime": 585.6563, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.6915685478413859, | |
| "grad_norm": 0.9619494676589966, | |
| "learning_rate": 9.490397642536117e-05, | |
| "loss": 0.1571, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.6929219109487076, | |
| "grad_norm": 0.28258052468299866, | |
| "learning_rate": 9.486928977481402e-05, | |
| "loss": 0.1245, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.6942752740560292, | |
| "grad_norm": 0.1793600469827652, | |
| "learning_rate": 9.483449185811948e-05, | |
| "loss": 0.5415, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.6956286371633509, | |
| "grad_norm": 23.72464942932129, | |
| "learning_rate": 9.479958276156884e-05, | |
| "loss": 0.7074, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.6969820002706726, | |
| "grad_norm": 3.1746864318847656, | |
| "learning_rate": 9.476456257172915e-05, | |
| "loss": 0.5878, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.6983353633779943, | |
| "grad_norm": 0.05973691865801811, | |
| "learning_rate": 9.47294313754429e-05, | |
| "loss": 0.0698, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.699688726485316, | |
| "grad_norm": 17.902040481567383, | |
| "learning_rate": 9.469418925982785e-05, | |
| "loss": 0.2651, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.7010420895926377, | |
| "grad_norm": 0.9632867574691772, | |
| "learning_rate": 9.465883631227686e-05, | |
| "loss": 0.3769, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.7023954526999594, | |
| "grad_norm": 1.0752183198928833, | |
| "learning_rate": 9.46233726204576e-05, | |
| "loss": 0.3865, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.7037488158072811, | |
| "grad_norm": 0.0032021338120102882, | |
| "learning_rate": 9.458779827231237e-05, | |
| "loss": 0.0677, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.7037488158072811, | |
| "eval_accuracy": 0.9147640791476408, | |
| "eval_loss": 0.33627045154571533, | |
| "eval_runtime": 585.7425, | |
| "eval_samples_per_second": 5.608, | |
| "eval_steps_per_second": 5.608, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.7051021789146028, | |
| "grad_norm": 0.574644148349762, | |
| "learning_rate": 9.455211335605784e-05, | |
| "loss": 0.0694, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.7064555420219245, | |
| "grad_norm": 13.346553802490234, | |
| "learning_rate": 9.451631796018494e-05, | |
| "loss": 0.1663, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.7078089051292462, | |
| "grad_norm": 22.342451095581055, | |
| "learning_rate": 9.448041217345851e-05, | |
| "loss": 0.8053, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.7091622682365679, | |
| "grad_norm": 13.331931114196777, | |
| "learning_rate": 9.444439608491711e-05, | |
| "loss": 0.5684, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.7105156313438896, | |
| "grad_norm": 2.19920015335083, | |
| "learning_rate": 9.440826978387289e-05, | |
| "loss": 0.4047, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.7118689944512112, | |
| "grad_norm": 0.0734083279967308, | |
| "learning_rate": 9.437203335991127e-05, | |
| "loss": 0.4998, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.7132223575585329, | |
| "grad_norm": 20.231857299804688, | |
| "learning_rate": 9.433568690289075e-05, | |
| "loss": 0.1117, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.7145757206658546, | |
| "grad_norm": 1.9291208445793018e-05, | |
| "learning_rate": 9.42992305029427e-05, | |
| "loss": 0.2927, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.7159290837731763, | |
| "grad_norm": 0.6859157681465149, | |
| "learning_rate": 9.426266425047113e-05, | |
| "loss": 0.1162, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.717282446880498, | |
| "grad_norm": 2.1293721199035645, | |
| "learning_rate": 9.42259882361524e-05, | |
| "loss": 0.2384, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.717282446880498, | |
| "eval_accuracy": 0.9205479452054794, | |
| "eval_loss": 0.3333543539047241, | |
| "eval_runtime": 585.504, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.7186358099878197, | |
| "grad_norm": 9.881598472595215, | |
| "learning_rate": 9.418920255093515e-05, | |
| "loss": 0.1674, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.7199891730951414, | |
| "grad_norm": 4.3182172775268555, | |
| "learning_rate": 9.415230728603992e-05, | |
| "loss": 0.477, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.7213425362024631, | |
| "grad_norm": 19.00124168395996, | |
| "learning_rate": 9.4115302532959e-05, | |
| "loss": 0.317, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.7226958993097848, | |
| "grad_norm": 4.142504621995613e-05, | |
| "learning_rate": 9.407818838345619e-05, | |
| "loss": 0.5577, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.7240492624171065, | |
| "grad_norm": 0.0461486354470253, | |
| "learning_rate": 9.404096492956656e-05, | |
| "loss": 0.3604, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.7254026255244282, | |
| "grad_norm": 0.5307131409645081, | |
| "learning_rate": 9.40036322635962e-05, | |
| "loss": 0.1165, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.72675598863175, | |
| "grad_norm": 0.0007177101797424257, | |
| "learning_rate": 9.396619047812213e-05, | |
| "loss": 0.7133, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.7281093517390715, | |
| "grad_norm": 0.09178412705659866, | |
| "learning_rate": 9.392863966599183e-05, | |
| "loss": 0.1665, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.7294627148463932, | |
| "grad_norm": 5.707544551114552e-05, | |
| "learning_rate": 9.389097992032325e-05, | |
| "loss": 0.3494, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.730816077953715, | |
| "grad_norm": 31.593660354614258, | |
| "learning_rate": 9.385321133450438e-05, | |
| "loss": 0.6661, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.730816077953715, | |
| "eval_accuracy": 0.9156773211567732, | |
| "eval_loss": 0.2831815481185913, | |
| "eval_runtime": 585.0863, | |
| "eval_samples_per_second": 5.615, | |
| "eval_steps_per_second": 5.615, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.7321694410610367, | |
| "grad_norm": 5.065242767333984, | |
| "learning_rate": 9.381533400219318e-05, | |
| "loss": 0.1203, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.7335228041683584, | |
| "grad_norm": 23.13774299621582, | |
| "learning_rate": 9.377734801731726e-05, | |
| "loss": 0.9495, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.7348761672756801, | |
| "grad_norm": 1.0966769456863403, | |
| "learning_rate": 9.373925347407364e-05, | |
| "loss": 0.2632, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.7362295303830018, | |
| "grad_norm": 0.010265067219734192, | |
| "learning_rate": 9.370105046692856e-05, | |
| "loss": 0.4109, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.7375828934903235, | |
| "grad_norm": 14.803862571716309, | |
| "learning_rate": 9.366273909061725e-05, | |
| "loss": 0.1273, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.7389362565976452, | |
| "grad_norm": 21.68570899963379, | |
| "learning_rate": 9.362431944014363e-05, | |
| "loss": 0.1676, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.7402896197049669, | |
| "grad_norm": 0.006488638464361429, | |
| "learning_rate": 9.358579161078013e-05, | |
| "loss": 0.512, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.7416429828122886, | |
| "grad_norm": 0.21234115958213806, | |
| "learning_rate": 9.354715569806744e-05, | |
| "loss": 0.5069, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.7429963459196103, | |
| "grad_norm": 13.795744895935059, | |
| "learning_rate": 9.350841179781431e-05, | |
| "loss": 0.48, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.7443497090269319, | |
| "grad_norm": 13.517033576965332, | |
| "learning_rate": 9.346956000609721e-05, | |
| "loss": 0.3189, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7443497090269319, | |
| "eval_accuracy": 0.9181126331811263, | |
| "eval_loss": 0.21662920713424683, | |
| "eval_runtime": 585.767, | |
| "eval_samples_per_second": 5.608, | |
| "eval_steps_per_second": 5.608, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7457030721342536, | |
| "grad_norm": 0.05359930172562599, | |
| "learning_rate": 9.343060041926021e-05, | |
| "loss": 0.3663, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.7470564352415753, | |
| "grad_norm": 9.406262397766113, | |
| "learning_rate": 9.339153313391469e-05, | |
| "loss": 0.3323, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.748409798348897, | |
| "grad_norm": 1.5440926551818848, | |
| "learning_rate": 9.335235824693904e-05, | |
| "loss": 0.223, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.7497631614562187, | |
| "grad_norm": 2.935401439666748, | |
| "learning_rate": 9.331307585547853e-05, | |
| "loss": 0.2858, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.7511165245635404, | |
| "grad_norm": 0.1924748718738556, | |
| "learning_rate": 9.327368605694502e-05, | |
| "loss": 0.5013, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.7524698876708621, | |
| "grad_norm": 16.0751895904541, | |
| "learning_rate": 9.323418894901669e-05, | |
| "loss": 0.2946, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.7538232507781838, | |
| "grad_norm": 5.284033298492432, | |
| "learning_rate": 9.319458462963785e-05, | |
| "loss": 0.4561, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.7551766138855055, | |
| "grad_norm": 0.00020838312047999352, | |
| "learning_rate": 9.315487319701865e-05, | |
| "loss": 0.1584, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.7565299769928272, | |
| "grad_norm": 6.6507532210380305e-06, | |
| "learning_rate": 9.311505474963484e-05, | |
| "loss": 0.1945, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.7578833401001489, | |
| "grad_norm": 0.00047044234815984964, | |
| "learning_rate": 9.307512938622762e-05, | |
| "loss": 0.5115, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.7578833401001489, | |
| "eval_accuracy": 0.9193302891933028, | |
| "eval_loss": 0.2699982523918152, | |
| "eval_runtime": 585.7552, | |
| "eval_samples_per_second": 5.608, | |
| "eval_steps_per_second": 5.608, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.7592367032074706, | |
| "grad_norm": 0.015827298164367676, | |
| "learning_rate": 9.303509720580324e-05, | |
| "loss": 0.131, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.7605900663147923, | |
| "grad_norm": 16.13619613647461, | |
| "learning_rate": 9.299495830763286e-05, | |
| "loss": 0.3044, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.7619434294221139, | |
| "grad_norm": 0.672073245048523, | |
| "learning_rate": 9.295471279125227e-05, | |
| "loss": 0.1588, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.7632967925294356, | |
| "grad_norm": 0.0014772227732464671, | |
| "learning_rate": 9.291436075646168e-05, | |
| "loss": 0.1987, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.7646501556367573, | |
| "grad_norm": 5.766489508118866e-08, | |
| "learning_rate": 9.28739023033254e-05, | |
| "loss": 0.2238, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.766003518744079, | |
| "grad_norm": 0.010482143610715866, | |
| "learning_rate": 9.283333753217167e-05, | |
| "loss": 0.3199, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.7673568818514007, | |
| "grad_norm": 1.591705322265625, | |
| "learning_rate": 9.279266654359235e-05, | |
| "loss": 0.2747, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.7687102449587224, | |
| "grad_norm": 0.0668347105383873, | |
| "learning_rate": 9.275188943844273e-05, | |
| "loss": 0.5941, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.7700636080660441, | |
| "grad_norm": 9.043622016906738, | |
| "learning_rate": 9.27110063178412e-05, | |
| "loss": 0.5263, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.7714169711733658, | |
| "grad_norm": 6.835690498352051, | |
| "learning_rate": 9.267001728316907e-05, | |
| "loss": 0.1861, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.7714169711733658, | |
| "eval_accuracy": 0.9108066971080669, | |
| "eval_loss": 0.26752638816833496, | |
| "eval_runtime": 585.647, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.7727703342806875, | |
| "grad_norm": 20.598922729492188, | |
| "learning_rate": 9.262892243607031e-05, | |
| "loss": 0.5422, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.7741236973880092, | |
| "grad_norm": 0.8921458721160889, | |
| "learning_rate": 9.258772187845123e-05, | |
| "loss": 0.103, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.7754770604953309, | |
| "grad_norm": 0.09386087208986282, | |
| "learning_rate": 9.254641571248035e-05, | |
| "loss": 0.3001, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.7768304236026526, | |
| "grad_norm": 0.05627532675862312, | |
| "learning_rate": 9.250500404058804e-05, | |
| "loss": 0.2537, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.7781837867099742, | |
| "grad_norm": 23.66926383972168, | |
| "learning_rate": 9.246348696546627e-05, | |
| "loss": 0.5345, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.7795371498172959, | |
| "grad_norm": 0.019794423133134842, | |
| "learning_rate": 9.242186459006845e-05, | |
| "loss": 0.227, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.7808905129246176, | |
| "grad_norm": 19.66176986694336, | |
| "learning_rate": 9.238013701760906e-05, | |
| "loss": 0.3494, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.7822438760319393, | |
| "grad_norm": 2.0934901237487793, | |
| "learning_rate": 9.233830435156348e-05, | |
| "loss": 0.0302, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.783597239139261, | |
| "grad_norm": 19.443466186523438, | |
| "learning_rate": 9.229636669566769e-05, | |
| "loss": 0.7859, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.7849506022465828, | |
| "grad_norm": 6.290703374878248e-11, | |
| "learning_rate": 9.2254324153918e-05, | |
| "loss": 0.3958, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.7849506022465828, | |
| "eval_accuracy": 0.9120243531202435, | |
| "eval_loss": 0.33059442043304443, | |
| "eval_runtime": 585.9682, | |
| "eval_samples_per_second": 5.606, | |
| "eval_steps_per_second": 5.606, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.7863039653539045, | |
| "grad_norm": 2.0353949069976807, | |
| "learning_rate": 9.221217683057083e-05, | |
| "loss": 0.2058, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.7876573284612262, | |
| "grad_norm": 30.70398712158203, | |
| "learning_rate": 9.216992483014247e-05, | |
| "loss": 0.4889, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.7890106915685479, | |
| "grad_norm": 5.967011451721191, | |
| "learning_rate": 9.212756825740873e-05, | |
| "loss": 0.1419, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.7903640546758696, | |
| "grad_norm": 0.024264423176646233, | |
| "learning_rate": 9.208510721740479e-05, | |
| "loss": 0.0395, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.7917174177831913, | |
| "grad_norm": 0.05600500851869583, | |
| "learning_rate": 9.204254181542483e-05, | |
| "loss": 0.482, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.793070780890513, | |
| "grad_norm": 2.5246126651763916, | |
| "learning_rate": 9.199987215702184e-05, | |
| "loss": 0.2441, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.7944241439978346, | |
| "grad_norm": 29.7937068939209, | |
| "learning_rate": 9.195709834800739e-05, | |
| "loss": 0.2949, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.7957775071051563, | |
| "grad_norm": 0.23492039740085602, | |
| "learning_rate": 9.191422049445128e-05, | |
| "loss": 0.2055, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.797130870212478, | |
| "grad_norm": 0.0021444044541567564, | |
| "learning_rate": 9.187123870268133e-05, | |
| "loss": 0.1847, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.7984842333197997, | |
| "grad_norm": 1.0205862963630352e-06, | |
| "learning_rate": 9.182815307928307e-05, | |
| "loss": 0.3423, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.7984842333197997, | |
| "eval_accuracy": 0.9126331811263318, | |
| "eval_loss": 0.34658023715019226, | |
| "eval_runtime": 585.7249, | |
| "eval_samples_per_second": 5.608, | |
| "eval_steps_per_second": 5.608, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.7998375964271214, | |
| "grad_norm": 0.04354476556181908, | |
| "learning_rate": 9.178496373109958e-05, | |
| "loss": 0.4167, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.8011909595344431, | |
| "grad_norm": 14.758959770202637, | |
| "learning_rate": 9.174167076523108e-05, | |
| "loss": 0.1091, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.8025443226417648, | |
| "grad_norm": 2.380498170852661, | |
| "learning_rate": 9.169827428903479e-05, | |
| "loss": 0.8495, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.8038976857490865, | |
| "grad_norm": 10.680726051330566, | |
| "learning_rate": 9.16547744101246e-05, | |
| "loss": 0.1548, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.8052510488564082, | |
| "grad_norm": 2.218698263168335, | |
| "learning_rate": 9.161117123637081e-05, | |
| "loss": 0.4776, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.8066044119637299, | |
| "grad_norm": 1.6176288681890583e-06, | |
| "learning_rate": 9.15674648758999e-05, | |
| "loss": 0.2087, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.8079577750710516, | |
| "grad_norm": 0.03002401813864708, | |
| "learning_rate": 9.152365543709416e-05, | |
| "loss": 0.4413, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 0.8093111381783733, | |
| "grad_norm": 29.92314910888672, | |
| "learning_rate": 9.147974302859157e-05, | |
| "loss": 0.58, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.810664501285695, | |
| "grad_norm": 0.7911025881767273, | |
| "learning_rate": 9.143572775928539e-05, | |
| "loss": 0.4153, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 0.8120178643930166, | |
| "grad_norm": 0.0276151355355978, | |
| "learning_rate": 9.139160973832399e-05, | |
| "loss": 0.2627, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8120178643930166, | |
| "eval_accuracy": 0.9199391171993911, | |
| "eval_loss": 0.2898274064064026, | |
| "eval_runtime": 585.4903, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8133712275003383, | |
| "grad_norm": 1.8124691247940063, | |
| "learning_rate": 9.134738907511055e-05, | |
| "loss": 0.1485, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 0.81472459060766, | |
| "grad_norm": 24.3751220703125, | |
| "learning_rate": 9.130306587930275e-05, | |
| "loss": 0.1473, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.8160779537149817, | |
| "grad_norm": 26.1238956451416, | |
| "learning_rate": 9.125864026081255e-05, | |
| "loss": 0.279, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 0.8174313168223034, | |
| "grad_norm": 0.0028976276516914368, | |
| "learning_rate": 9.121411232980588e-05, | |
| "loss": 0.1216, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.8187846799296251, | |
| "grad_norm": 0.4264237582683563, | |
| "learning_rate": 9.116948219670239e-05, | |
| "loss": 0.2344, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.8201380430369468, | |
| "grad_norm": 0.1696048080921173, | |
| "learning_rate": 9.112474997217517e-05, | |
| "loss": 1.0503, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.8214914061442685, | |
| "grad_norm": 0.2018238753080368, | |
| "learning_rate": 9.107991576715049e-05, | |
| "loss": 0.4023, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 0.8228447692515902, | |
| "grad_norm": 4.706511974334717, | |
| "learning_rate": 9.103497969280748e-05, | |
| "loss": 0.4097, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.8241981323589119, | |
| "grad_norm": 5.938671588897705, | |
| "learning_rate": 9.09899418605779e-05, | |
| "loss": 0.1898, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 0.8255514954662336, | |
| "grad_norm": 17.77438735961914, | |
| "learning_rate": 9.094480238214586e-05, | |
| "loss": 0.2257, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.8255514954662336, | |
| "eval_accuracy": 0.9187214611872146, | |
| "eval_loss": 0.3077405095100403, | |
| "eval_runtime": 586.1068, | |
| "eval_samples_per_second": 5.605, | |
| "eval_steps_per_second": 5.605, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.8269048585735553, | |
| "grad_norm": 1.0022088289260864, | |
| "learning_rate": 9.089956136944751e-05, | |
| "loss": 0.0093, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 0.8282582216808769, | |
| "grad_norm": 0.00016064182273112237, | |
| "learning_rate": 9.085421893467076e-05, | |
| "loss": 0.0219, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.8296115847881986, | |
| "grad_norm": 0.17109623551368713, | |
| "learning_rate": 9.08087751902551e-05, | |
| "loss": 0.3479, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 0.8309649478955203, | |
| "grad_norm": 20.761587142944336, | |
| "learning_rate": 9.076323024889119e-05, | |
| "loss": 0.6529, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.832318311002842, | |
| "grad_norm": 0.00015795385115779936, | |
| "learning_rate": 9.071758422352063e-05, | |
| "loss": 0.0424, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.8336716741101637, | |
| "grad_norm": 14.546133041381836, | |
| "learning_rate": 9.06718372273357e-05, | |
| "loss": 0.298, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.8350250372174854, | |
| "grad_norm": 0.030780814588069916, | |
| "learning_rate": 9.06259893737791e-05, | |
| "loss": 0.1547, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 0.8363784003248071, | |
| "grad_norm": 10.233747482299805, | |
| "learning_rate": 9.058004077654359e-05, | |
| "loss": 0.5607, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.8377317634321289, | |
| "grad_norm": 0.13436900079250336, | |
| "learning_rate": 9.053399154957176e-05, | |
| "loss": 0.5464, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 0.8390851265394506, | |
| "grad_norm": 0.34480223059654236, | |
| "learning_rate": 9.048784180705573e-05, | |
| "loss": 0.1525, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.8390851265394506, | |
| "eval_accuracy": 0.9190258751902588, | |
| "eval_loss": 0.24334192276000977, | |
| "eval_runtime": 585.6983, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.8404384896467723, | |
| "grad_norm": 0.0004358985461294651, | |
| "learning_rate": 9.044159166343695e-05, | |
| "loss": 0.2265, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 0.841791852754094, | |
| "grad_norm": 5.295452117919922, | |
| "learning_rate": 9.039524123340572e-05, | |
| "loss": 0.3064, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.8431452158614157, | |
| "grad_norm": 3.436157203395851e-05, | |
| "learning_rate": 9.034879063190116e-05, | |
| "loss": 0.212, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 0.8444985789687373, | |
| "grad_norm": 14.672189712524414, | |
| "learning_rate": 9.030223997411067e-05, | |
| "loss": 0.444, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.845851942076059, | |
| "grad_norm": 15.262385368347168, | |
| "learning_rate": 9.025558937546988e-05, | |
| "loss": 0.4573, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.8472053051833807, | |
| "grad_norm": 13.051666259765625, | |
| "learning_rate": 9.020883895166214e-05, | |
| "loss": 0.4719, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.8485586682907024, | |
| "grad_norm": 4.082529544830322, | |
| "learning_rate": 9.016198881861845e-05, | |
| "loss": 0.0876, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 0.8499120313980241, | |
| "grad_norm": 0.01626618765294552, | |
| "learning_rate": 9.0115039092517e-05, | |
| "loss": 0.2108, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.8512653945053458, | |
| "grad_norm": 2.203218698501587, | |
| "learning_rate": 9.006798988978298e-05, | |
| "loss": 0.0862, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 0.8526187576126675, | |
| "grad_norm": 7.0966997146606445, | |
| "learning_rate": 9.002084132708823e-05, | |
| "loss": 0.5941, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.8526187576126675, | |
| "eval_accuracy": 0.9123287671232877, | |
| "eval_loss": 0.4010595679283142, | |
| "eval_runtime": 585.622, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.8539721207199892, | |
| "grad_norm": 0.00025126567925326526, | |
| "learning_rate": 8.997359352135105e-05, | |
| "loss": 0.3468, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 0.8553254838273109, | |
| "grad_norm": 0.005354021675884724, | |
| "learning_rate": 8.992624658973574e-05, | |
| "loss": 0.2055, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.8566788469346326, | |
| "grad_norm": 20.462688446044922, | |
| "learning_rate": 8.987880064965249e-05, | |
| "loss": 0.316, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 0.8580322100419543, | |
| "grad_norm": 6.252931118011475, | |
| "learning_rate": 8.983125581875698e-05, | |
| "loss": 0.348, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.859385573149276, | |
| "grad_norm": 12.182239532470703, | |
| "learning_rate": 8.978361221495013e-05, | |
| "loss": 0.1299, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.8607389362565977, | |
| "grad_norm": 16.952531814575195, | |
| "learning_rate": 8.973586995637778e-05, | |
| "loss": 0.1622, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.8620922993639193, | |
| "grad_norm": 1.5264373359968886e-05, | |
| "learning_rate": 8.968802916143039e-05, | |
| "loss": 0.2455, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 0.863445662471241, | |
| "grad_norm": 0.003773150034248829, | |
| "learning_rate": 8.964008994874285e-05, | |
| "loss": 0.3139, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.8647990255785627, | |
| "grad_norm": 5.219977378845215, | |
| "learning_rate": 8.959205243719402e-05, | |
| "loss": 0.5018, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 0.8661523886858844, | |
| "grad_norm": 0.11269356310367584, | |
| "learning_rate": 8.954391674590656e-05, | |
| "loss": 0.0723, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.8661523886858844, | |
| "eval_accuracy": 0.9208523592085236, | |
| "eval_loss": 0.26359477639198303, | |
| "eval_runtime": 585.3471, | |
| "eval_samples_per_second": 5.612, | |
| "eval_steps_per_second": 5.612, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.8675057517932061, | |
| "grad_norm": 0.05086831748485565, | |
| "learning_rate": 8.949568299424659e-05, | |
| "loss": 0.1954, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 0.8688591149005278, | |
| "grad_norm": 2.221594149887096e-05, | |
| "learning_rate": 8.94473513018234e-05, | |
| "loss": 0.352, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.8702124780078495, | |
| "grad_norm": 0.014133933931589127, | |
| "learning_rate": 8.939892178848914e-05, | |
| "loss": 0.054, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 0.8715658411151712, | |
| "grad_norm": 3.5625367164611816, | |
| "learning_rate": 8.935039457433857e-05, | |
| "loss": 0.217, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.8729192042224929, | |
| "grad_norm": 0.15607240796089172, | |
| "learning_rate": 8.930176977970864e-05, | |
| "loss": 0.2095, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.8742725673298146, | |
| "grad_norm": 47.42460250854492, | |
| "learning_rate": 8.92530475251784e-05, | |
| "loss": 0.7065, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.8756259304371363, | |
| "grad_norm": 21.992969512939453, | |
| "learning_rate": 8.920422793156847e-05, | |
| "loss": 0.5967, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 0.876979293544458, | |
| "grad_norm": 12.470071792602539, | |
| "learning_rate": 8.915531111994094e-05, | |
| "loss": 0.2819, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.8783326566517796, | |
| "grad_norm": 12.152610778808594, | |
| "learning_rate": 8.910629721159892e-05, | |
| "loss": 0.2951, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 0.8796860197591013, | |
| "grad_norm": 18.442096710205078, | |
| "learning_rate": 8.90571863280863e-05, | |
| "loss": 0.2415, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.8796860197591013, | |
| "eval_accuracy": 0.9178082191780822, | |
| "eval_loss": 0.24678540229797363, | |
| "eval_runtime": 585.4998, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.881039382866423, | |
| "grad_norm": 14.306652069091797, | |
| "learning_rate": 8.900797859118748e-05, | |
| "loss": 0.4111, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 0.8823927459737447, | |
| "grad_norm": 15.398430824279785, | |
| "learning_rate": 8.895867412292702e-05, | |
| "loss": 0.1817, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 0.8837461090810664, | |
| "grad_norm": 0.8829298615455627, | |
| "learning_rate": 8.890927304556935e-05, | |
| "loss": 0.0695, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 0.8850994721883881, | |
| "grad_norm": 19.324243545532227, | |
| "learning_rate": 8.885977548161848e-05, | |
| "loss": 0.2991, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 0.8864528352957098, | |
| "grad_norm": 4.953662872314453, | |
| "learning_rate": 8.881018155381766e-05, | |
| "loss": 0.6572, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.8878061984030315, | |
| "grad_norm": 5.32732629776001, | |
| "learning_rate": 8.876049138514912e-05, | |
| "loss": 0.3841, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.8891595615103532, | |
| "grad_norm": 9.986154556274414, | |
| "learning_rate": 8.871070509883377e-05, | |
| "loss": 0.1994, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 0.890512924617675, | |
| "grad_norm": 16.808300018310547, | |
| "learning_rate": 8.866082281833081e-05, | |
| "loss": 0.2473, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 0.8918662877249967, | |
| "grad_norm": 12.585075378417969, | |
| "learning_rate": 8.861084466733757e-05, | |
| "loss": 0.7061, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 0.8932196508323184, | |
| "grad_norm": 1.7754485607147217, | |
| "learning_rate": 8.856077076978902e-05, | |
| "loss": 0.214, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.8932196508323184, | |
| "eval_accuracy": 0.9114155251141552, | |
| "eval_loss": 0.24676522612571716, | |
| "eval_runtime": 585.6426, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.8945730139396401, | |
| "grad_norm": 21.269773483276367, | |
| "learning_rate": 8.851060124985765e-05, | |
| "loss": 0.3384, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 0.8959263770469617, | |
| "grad_norm": 12.510165214538574, | |
| "learning_rate": 8.8460336231953e-05, | |
| "loss": 0.4515, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 0.8972797401542834, | |
| "grad_norm": 0.671404242515564, | |
| "learning_rate": 8.840997584072149e-05, | |
| "loss": 0.2762, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 0.8986331032616051, | |
| "grad_norm": 0.46358880400657654, | |
| "learning_rate": 8.835952020104598e-05, | |
| "loss": 0.0521, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 0.8999864663689268, | |
| "grad_norm": 18.224842071533203, | |
| "learning_rate": 8.830896943804557e-05, | |
| "loss": 0.4169, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.9013398294762485, | |
| "grad_norm": 0.3978288471698761, | |
| "learning_rate": 8.825832367707525e-05, | |
| "loss": 0.2784, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 0.9026931925835702, | |
| "grad_norm": 20.021671295166016, | |
| "learning_rate": 8.820758304372557e-05, | |
| "loss": 0.2109, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 0.9040465556908919, | |
| "grad_norm": 1.5909063816070557, | |
| "learning_rate": 8.815674766382232e-05, | |
| "loss": 0.2103, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 0.9053999187982136, | |
| "grad_norm": 1.8711028099060059, | |
| "learning_rate": 8.810581766342626e-05, | |
| "loss": 0.4238, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 0.9067532819055353, | |
| "grad_norm": 31.497356414794922, | |
| "learning_rate": 8.80547931688328e-05, | |
| "loss": 0.2431, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.9067532819055353, | |
| "eval_accuracy": 0.919634703196347, | |
| "eval_loss": 0.2998688519001007, | |
| "eval_runtime": 585.4623, | |
| "eval_samples_per_second": 5.611, | |
| "eval_steps_per_second": 5.611, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.908106645012857, | |
| "grad_norm": 12.628376960754395, | |
| "learning_rate": 8.800367430657168e-05, | |
| "loss": 0.1661, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 0.9094600081201787, | |
| "grad_norm": 0.07297774404287338, | |
| "learning_rate": 8.795246120340663e-05, | |
| "loss": 0.0459, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.9108133712275004, | |
| "grad_norm": 28.95281982421875, | |
| "learning_rate": 8.790115398633507e-05, | |
| "loss": 0.567, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 0.912166734334822, | |
| "grad_norm": 12.253477096557617, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.226, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 0.9135200974421437, | |
| "grad_norm": 19.2222900390625, | |
| "learning_rate": 8.779825771962879e-05, | |
| "loss": 0.1918, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.9148734605494654, | |
| "grad_norm": 0.03731995075941086, | |
| "learning_rate": 8.774666892515458e-05, | |
| "loss": 0.1789, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 0.9162268236567871, | |
| "grad_norm": 1.9367414712905884, | |
| "learning_rate": 8.769498652709427e-05, | |
| "loss": 0.1, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 0.9175801867641088, | |
| "grad_norm": 4.96419534101733e-06, | |
| "learning_rate": 8.764321065360905e-05, | |
| "loss": 0.9138, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 0.9189335498714305, | |
| "grad_norm": 9.806981086730957, | |
| "learning_rate": 8.759134143309189e-05, | |
| "loss": 0.2375, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 0.9202869129787522, | |
| "grad_norm": 9.017184493131936e-05, | |
| "learning_rate": 8.753937899416727e-05, | |
| "loss": 0.5315, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.9202869129787522, | |
| "eval_accuracy": 0.9229832572298325, | |
| "eval_loss": 0.3090192675590515, | |
| "eval_runtime": 585.6276, | |
| "eval_samples_per_second": 5.609, | |
| "eval_steps_per_second": 5.609, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.9216402760860739, | |
| "grad_norm": 21.234371185302734, | |
| "learning_rate": 8.74873234656908e-05, | |
| "loss": 0.3668, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 0.9229936391933956, | |
| "grad_norm": 4.857106685638428, | |
| "learning_rate": 8.743517497674896e-05, | |
| "loss": 0.3368, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 0.9243470023007173, | |
| "grad_norm": 0.0022790771909058094, | |
| "learning_rate": 8.738293365665872e-05, | |
| "loss": 0.0965, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 0.925700365408039, | |
| "grad_norm": 25.66297149658203, | |
| "learning_rate": 8.73305996349673e-05, | |
| "loss": 0.5797, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 0.9270537285153607, | |
| "grad_norm": 0.9596951603889465, | |
| "learning_rate": 8.727817304145176e-05, | |
| "loss": 0.2861, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.9284070916226823, | |
| "grad_norm": 0.6504172086715698, | |
| "learning_rate": 8.722565400611872e-05, | |
| "loss": 0.1645, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 0.929760454730004, | |
| "grad_norm": 0.5617231130599976, | |
| "learning_rate": 8.717304265920402e-05, | |
| "loss": 0.5579, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 0.9311138178373257, | |
| "grad_norm": 8.67705249786377, | |
| "learning_rate": 8.71203391311725e-05, | |
| "loss": 0.6897, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.9324671809446474, | |
| "grad_norm": 8.441317558288574, | |
| "learning_rate": 8.70675435527175e-05, | |
| "loss": 0.1798, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 0.9338205440519691, | |
| "grad_norm": 11.730077743530273, | |
| "learning_rate": 8.70146560547606e-05, | |
| "loss": 0.1287, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.9338205440519691, | |
| "eval_accuracy": 0.9187214611872146, | |
| "eval_loss": 0.22067341208457947, | |
| "eval_runtime": 585.0326, | |
| "eval_samples_per_second": 5.615, | |
| "eval_steps_per_second": 5.615, | |
| "step": 6900 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 22167, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |