| { | |
| "best_metric": 0.8666666666666667, | |
| "best_model_checkpoint": "CTMAE2_CS_V7_4/checkpoint-1404", | |
| "epoch": 49.01367741935484, | |
| "eval_steps": 500, | |
| "global_step": 7750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0012903225806451613, | |
| "grad_norm": 2.9930078983306885, | |
| "learning_rate": 1.2903225806451614e-07, | |
| "loss": 0.6825, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0025806451612903226, | |
| "grad_norm": 3.1541030406951904, | |
| "learning_rate": 2.580645161290323e-07, | |
| "loss": 0.6758, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.003870967741935484, | |
| "grad_norm": 4.803792476654053, | |
| "learning_rate": 3.870967741935484e-07, | |
| "loss": 0.6704, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.005161290322580645, | |
| "grad_norm": 1.6744341850280762, | |
| "learning_rate": 5.161290322580646e-07, | |
| "loss": 0.661, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0064516129032258064, | |
| "grad_norm": 2.0169718265533447, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 0.6543, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.007741935483870968, | |
| "grad_norm": 2.0901169776916504, | |
| "learning_rate": 7.741935483870968e-07, | |
| "loss": 0.6622, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.00903225806451613, | |
| "grad_norm": 5.307899475097656, | |
| "learning_rate": 9.032258064516129e-07, | |
| "loss": 0.665, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01032258064516129, | |
| "grad_norm": 4.631078720092773, | |
| "learning_rate": 1.0322580645161291e-06, | |
| "loss": 0.6854, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.011612903225806452, | |
| "grad_norm": 2.43985652923584, | |
| "learning_rate": 1.1612903225806454e-06, | |
| "loss": 0.6704, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.012903225806451613, | |
| "grad_norm": 2.5190656185150146, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 0.606, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.014193548387096775, | |
| "grad_norm": 13.296693801879883, | |
| "learning_rate": 1.4193548387096776e-06, | |
| "loss": 0.5469, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.015483870967741935, | |
| "grad_norm": 22.31141471862793, | |
| "learning_rate": 1.5483870967741937e-06, | |
| "loss": 0.6627, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.016774193548387096, | |
| "grad_norm": 5.563461780548096, | |
| "learning_rate": 1.67741935483871e-06, | |
| "loss": 0.6442, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.01806451612903226, | |
| "grad_norm": 6.656688213348389, | |
| "learning_rate": 1.8064516129032258e-06, | |
| "loss": 0.5698, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.01935483870967742, | |
| "grad_norm": 7.344876289367676, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 0.6652, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.020129032258064516, | |
| "eval_accuracy": 0.4666666666666667, | |
| "eval_loss": 0.7404347062110901, | |
| "eval_runtime": 14.531, | |
| "eval_samples_per_second": 3.097, | |
| "eval_steps_per_second": 0.619, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0005161290322582, | |
| "grad_norm": 12.970778465270996, | |
| "learning_rate": 2.0645161290322582e-06, | |
| "loss": 0.6031, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0018064516129033, | |
| "grad_norm": 5.201364994049072, | |
| "learning_rate": 2.1935483870967745e-06, | |
| "loss": 0.6682, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0030967741935484, | |
| "grad_norm": 7.8400139808654785, | |
| "learning_rate": 2.3225806451612907e-06, | |
| "loss": 0.6292, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.0043870967741935, | |
| "grad_norm": 6.222132205963135, | |
| "learning_rate": 2.4516129032258066e-06, | |
| "loss": 0.6417, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0056774193548388, | |
| "grad_norm": 6.516747951507568, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 0.6335, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0069677419354839, | |
| "grad_norm": 4.24305534362793, | |
| "learning_rate": 2.709677419354839e-06, | |
| "loss": 0.5944, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.008258064516129, | |
| "grad_norm": 4.02984094619751, | |
| "learning_rate": 2.8387096774193553e-06, | |
| "loss": 0.6735, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.0095483870967743, | |
| "grad_norm": 3.8313159942626953, | |
| "learning_rate": 2.967741935483871e-06, | |
| "loss": 0.6809, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0108387096774194, | |
| "grad_norm": 7.125197410583496, | |
| "learning_rate": 3.0967741935483874e-06, | |
| "loss": 0.5828, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0121290322580645, | |
| "grad_norm": 4.8547163009643555, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.7288, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0134193548387096, | |
| "grad_norm": 5.060990333557129, | |
| "learning_rate": 3.35483870967742e-06, | |
| "loss": 0.5894, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.014709677419355, | |
| "grad_norm": 3.9383673667907715, | |
| "learning_rate": 3.4838709677419357e-06, | |
| "loss": 0.6024, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 5.0263214111328125, | |
| "learning_rate": 3.6129032258064515e-06, | |
| "loss": 0.6688, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0172903225806451, | |
| "grad_norm": 9.285834312438965, | |
| "learning_rate": 3.741935483870968e-06, | |
| "loss": 0.5496, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.0185806451612902, | |
| "grad_norm": 14.92795181274414, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 0.5025, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0198709677419355, | |
| "grad_norm": 13.292335510253906, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.5679, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0201290322580645, | |
| "eval_accuracy": 0.4666666666666667, | |
| "eval_loss": 0.6957474946975708, | |
| "eval_runtime": 12.7815, | |
| "eval_samples_per_second": 3.521, | |
| "eval_steps_per_second": 0.704, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.0010322580645163, | |
| "grad_norm": 7.391489028930664, | |
| "learning_rate": 4.1290322580645165e-06, | |
| "loss": 0.4973, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.002322580645161, | |
| "grad_norm": 14.326224327087402, | |
| "learning_rate": 4.258064516129032e-06, | |
| "loss": 0.6073, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.0036129032258065, | |
| "grad_norm": 6.367290019989014, | |
| "learning_rate": 4.387096774193549e-06, | |
| "loss": 0.537, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.0049032258064514, | |
| "grad_norm": 24.50838851928711, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 0.5098, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.0061935483870967, | |
| "grad_norm": 6.082309246063232, | |
| "learning_rate": 4.6451612903225815e-06, | |
| "loss": 0.584, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.007483870967742, | |
| "grad_norm": 32.96268081665039, | |
| "learning_rate": 4.774193548387097e-06, | |
| "loss": 0.6034, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.008774193548387, | |
| "grad_norm": 9.968070983886719, | |
| "learning_rate": 4.903225806451613e-06, | |
| "loss": 0.5638, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.0100645161290323, | |
| "grad_norm": 13.696681022644043, | |
| "learning_rate": 5.032258064516129e-06, | |
| "loss": 0.5094, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.0113548387096776, | |
| "grad_norm": 8.534320831298828, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 0.5637, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0126451612903224, | |
| "grad_norm": 10.044099807739258, | |
| "learning_rate": 5.290322580645162e-06, | |
| "loss": 0.6784, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0139354838709678, | |
| "grad_norm": 47.13435363769531, | |
| "learning_rate": 5.419354838709678e-06, | |
| "loss": 0.5562, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.015225806451613, | |
| "grad_norm": 15.391922950744629, | |
| "learning_rate": 5.548387096774194e-06, | |
| "loss": 0.4974, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.016516129032258, | |
| "grad_norm": 37.72578048706055, | |
| "learning_rate": 5.677419354838711e-06, | |
| "loss": 0.4648, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.0178064516129033, | |
| "grad_norm": 10.507546424865723, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 0.505, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.0190967741935486, | |
| "grad_norm": 6.1668829917907715, | |
| "learning_rate": 5.935483870967742e-06, | |
| "loss": 0.4455, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.0201290322580645, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 0.47290292382240295, | |
| "eval_runtime": 12.3627, | |
| "eval_samples_per_second": 3.64, | |
| "eval_steps_per_second": 0.728, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 3.000258064516129, | |
| "grad_norm": 7.66517448425293, | |
| "learning_rate": 6.064516129032259e-06, | |
| "loss": 0.4085, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.0015483870967743, | |
| "grad_norm": 15.525689125061035, | |
| "learning_rate": 6.193548387096775e-06, | |
| "loss": 0.5533, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.002838709677419, | |
| "grad_norm": 29.974485397338867, | |
| "learning_rate": 6.3225806451612906e-06, | |
| "loss": 0.578, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.0041290322580645, | |
| "grad_norm": 5.562129020690918, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.6254, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.00541935483871, | |
| "grad_norm": 3.9507057666778564, | |
| "learning_rate": 6.580645161290323e-06, | |
| "loss": 0.4409, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.0067096774193547, | |
| "grad_norm": 31.996253967285156, | |
| "learning_rate": 6.70967741935484e-06, | |
| "loss": 0.3241, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 66.41522216796875, | |
| "learning_rate": 6.838709677419355e-06, | |
| "loss": 0.7365, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.0092903225806453, | |
| "grad_norm": 7.483922481536865, | |
| "learning_rate": 6.967741935483871e-06, | |
| "loss": 0.7, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.01058064516129, | |
| "grad_norm": 2.7322657108306885, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.5528, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.0118709677419355, | |
| "grad_norm": 9.521190643310547, | |
| "learning_rate": 7.225806451612903e-06, | |
| "loss": 0.6523, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 3.013161290322581, | |
| "grad_norm": 3.5681607723236084, | |
| "learning_rate": 7.35483870967742e-06, | |
| "loss": 0.5665, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 3.0144516129032257, | |
| "grad_norm": 28.108043670654297, | |
| "learning_rate": 7.483870967741936e-06, | |
| "loss": 0.6268, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.015741935483871, | |
| "grad_norm": 8.838102340698242, | |
| "learning_rate": 7.612903225806451e-06, | |
| "loss": 0.4694, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.017032258064516, | |
| "grad_norm": 21.64659309387207, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.524, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.0183225806451612, | |
| "grad_norm": 5.254949569702148, | |
| "learning_rate": 7.870967741935484e-06, | |
| "loss": 0.4841, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.0196129032258066, | |
| "grad_norm": 13.017827987670898, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.479, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.0201290322580645, | |
| "eval_accuracy": 0.6, | |
| "eval_loss": 0.8237701654434204, | |
| "eval_runtime": 11.5741, | |
| "eval_samples_per_second": 3.888, | |
| "eval_steps_per_second": 0.778, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 4.000774193548387, | |
| "grad_norm": 7.159151554107666, | |
| "learning_rate": 8.129032258064517e-06, | |
| "loss": 0.3456, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.002064516129033, | |
| "grad_norm": 9.60805892944336, | |
| "learning_rate": 8.258064516129033e-06, | |
| "loss": 0.7947, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.003354838709678, | |
| "grad_norm": 8.285265922546387, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.4209, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.004645161290322, | |
| "grad_norm": 12.667649269104004, | |
| "learning_rate": 8.516129032258065e-06, | |
| "loss": 0.6283, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 4.005935483870967, | |
| "grad_norm": 9.29390811920166, | |
| "learning_rate": 8.64516129032258e-06, | |
| "loss": 0.3608, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 4.007225806451613, | |
| "grad_norm": 34.86391067504883, | |
| "learning_rate": 8.774193548387098e-06, | |
| "loss": 0.4912, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 4.008516129032258, | |
| "grad_norm": 3.0919666290283203, | |
| "learning_rate": 8.903225806451614e-06, | |
| "loss": 0.5047, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 4.009806451612903, | |
| "grad_norm": 3.6202268600463867, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.3924, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.011096774193549, | |
| "grad_norm": 23.536909103393555, | |
| "learning_rate": 9.161290322580645e-06, | |
| "loss": 0.5627, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 4.0123870967741935, | |
| "grad_norm": 12.18332290649414, | |
| "learning_rate": 9.290322580645163e-06, | |
| "loss": 0.4785, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 4.013677419354838, | |
| "grad_norm": 14.957155227661133, | |
| "learning_rate": 9.419354838709677e-06, | |
| "loss": 0.4945, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 4.014967741935484, | |
| "grad_norm": 9.993709564208984, | |
| "learning_rate": 9.548387096774195e-06, | |
| "loss": 0.5742, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 4.016258064516129, | |
| "grad_norm": 10.588777542114258, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.4755, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.017548387096774, | |
| "grad_norm": 24.638578414916992, | |
| "learning_rate": 9.806451612903226e-06, | |
| "loss": 0.5454, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 4.01883870967742, | |
| "grad_norm": 8.419483184814453, | |
| "learning_rate": 9.935483870967742e-06, | |
| "loss": 0.4759, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 4.0201290322580645, | |
| "grad_norm": 0.7586236596107483, | |
| "learning_rate": 9.992831541218639e-06, | |
| "loss": 0.4476, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 4.0201290322580645, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 0.5751302242279053, | |
| "eval_runtime": 12.6336, | |
| "eval_samples_per_second": 3.562, | |
| "eval_steps_per_second": 0.712, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 5.001290322580645, | |
| "grad_norm": 9.112713813781738, | |
| "learning_rate": 9.978494623655915e-06, | |
| "loss": 0.4095, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 5.002580645161291, | |
| "grad_norm": 14.822427749633789, | |
| "learning_rate": 9.96415770609319e-06, | |
| "loss": 0.4537, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.0038709677419355, | |
| "grad_norm": 16.80013084411621, | |
| "learning_rate": 9.949820788530466e-06, | |
| "loss": 0.5184, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 5.00516129032258, | |
| "grad_norm": 15.491483688354492, | |
| "learning_rate": 9.935483870967742e-06, | |
| "loss": 0.4734, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 5.006451612903226, | |
| "grad_norm": 12.78894329071045, | |
| "learning_rate": 9.921146953405018e-06, | |
| "loss": 0.6542, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 5.007741935483871, | |
| "grad_norm": 20.72500228881836, | |
| "learning_rate": 9.906810035842294e-06, | |
| "loss": 0.5551, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 5.009032258064516, | |
| "grad_norm": 14.697334289550781, | |
| "learning_rate": 9.89247311827957e-06, | |
| "loss": 0.4696, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.010322580645162, | |
| "grad_norm": 6.555180072784424, | |
| "learning_rate": 9.878136200716847e-06, | |
| "loss": 0.4804, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 5.0116129032258065, | |
| "grad_norm": 7.139140605926514, | |
| "learning_rate": 9.863799283154123e-06, | |
| "loss": 0.4613, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 5.012903225806451, | |
| "grad_norm": 3.5184733867645264, | |
| "learning_rate": 9.8494623655914e-06, | |
| "loss": 0.5416, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 5.014193548387097, | |
| "grad_norm": 6.236945152282715, | |
| "learning_rate": 9.835125448028676e-06, | |
| "loss": 0.5338, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 5.015483870967742, | |
| "grad_norm": 12.424110412597656, | |
| "learning_rate": 9.820788530465952e-06, | |
| "loss": 0.5935, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.016774193548387, | |
| "grad_norm": 20.812471389770508, | |
| "learning_rate": 9.806451612903226e-06, | |
| "loss": 0.6243, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 5.018064516129032, | |
| "grad_norm": 10.67922592163086, | |
| "learning_rate": 9.792114695340502e-06, | |
| "loss": 0.5427, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 5.019354838709678, | |
| "grad_norm": 11.50655746459961, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 0.5195, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 5.0201290322580645, | |
| "eval_accuracy": 0.5333333333333333, | |
| "eval_loss": 0.8028479814529419, | |
| "eval_runtime": 12.3443, | |
| "eval_samples_per_second": 3.645, | |
| "eval_steps_per_second": 0.729, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 6.000516129032258, | |
| "grad_norm": 5.780803203582764, | |
| "learning_rate": 9.763440860215055e-06, | |
| "loss": 0.5746, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 6.001806451612903, | |
| "grad_norm": 4.691468238830566, | |
| "learning_rate": 9.749103942652331e-06, | |
| "loss": 0.5012, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.003096774193549, | |
| "grad_norm": 6.017818927764893, | |
| "learning_rate": 9.734767025089607e-06, | |
| "loss": 0.5094, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 6.0043870967741935, | |
| "grad_norm": 9.175492286682129, | |
| "learning_rate": 9.720430107526882e-06, | |
| "loss": 0.4829, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 6.005677419354838, | |
| "grad_norm": 18.013885498046875, | |
| "learning_rate": 9.706093189964158e-06, | |
| "loss": 0.5643, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 6.006967741935484, | |
| "grad_norm": 9.79544734954834, | |
| "learning_rate": 9.691756272401434e-06, | |
| "loss": 0.5024, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 6.008258064516129, | |
| "grad_norm": 6.72033166885376, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.4552, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.009548387096774, | |
| "grad_norm": 36.5327033996582, | |
| "learning_rate": 9.663082437275987e-06, | |
| "loss": 0.54, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 6.01083870967742, | |
| "grad_norm": 2.145115613937378, | |
| "learning_rate": 9.648745519713263e-06, | |
| "loss": 0.3285, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 6.0121290322580645, | |
| "grad_norm": 15.595919609069824, | |
| "learning_rate": 9.634408602150539e-06, | |
| "loss": 0.5151, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 6.013419354838709, | |
| "grad_norm": 15.70621395111084, | |
| "learning_rate": 9.620071684587814e-06, | |
| "loss": 0.4021, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 6.014709677419355, | |
| "grad_norm": 54.31192398071289, | |
| "learning_rate": 9.60573476702509e-06, | |
| "loss": 0.7821, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.016, | |
| "grad_norm": 9.83252239227295, | |
| "learning_rate": 9.591397849462366e-06, | |
| "loss": 0.3528, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 6.017290322580645, | |
| "grad_norm": 6.726585388183594, | |
| "learning_rate": 9.577060931899642e-06, | |
| "loss": 0.3144, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 6.018580645161291, | |
| "grad_norm": 18.89128875732422, | |
| "learning_rate": 9.562724014336918e-06, | |
| "loss": 0.7871, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 6.0198709677419355, | |
| "grad_norm": 12.1157865524292, | |
| "learning_rate": 9.548387096774195e-06, | |
| "loss": 0.4507, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 6.0201290322580645, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 0.5427281260490417, | |
| "eval_runtime": 12.4587, | |
| "eval_samples_per_second": 3.612, | |
| "eval_steps_per_second": 0.722, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 7.001032258064516, | |
| "grad_norm": 25.068649291992188, | |
| "learning_rate": 9.53405017921147e-06, | |
| "loss": 0.4748, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.002322580645162, | |
| "grad_norm": 1.43007230758667, | |
| "learning_rate": 9.519713261648747e-06, | |
| "loss": 0.4112, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 7.0036129032258065, | |
| "grad_norm": 11.848410606384277, | |
| "learning_rate": 9.505376344086023e-06, | |
| "loss": 0.7485, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 7.004903225806451, | |
| "grad_norm": 4.169070243835449, | |
| "learning_rate": 9.491039426523298e-06, | |
| "loss": 0.5412, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 7.006193548387097, | |
| "grad_norm": 13.245758056640625, | |
| "learning_rate": 9.476702508960574e-06, | |
| "loss": 0.4974, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 7.007483870967742, | |
| "grad_norm": 9.883331298828125, | |
| "learning_rate": 9.46236559139785e-06, | |
| "loss": 0.4081, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.008774193548387, | |
| "grad_norm": 44.25878143310547, | |
| "learning_rate": 9.448028673835126e-06, | |
| "loss": 0.5144, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 7.010064516129033, | |
| "grad_norm": 19.449512481689453, | |
| "learning_rate": 9.433691756272403e-06, | |
| "loss": 0.3876, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 7.011354838709678, | |
| "grad_norm": 12.099861145019531, | |
| "learning_rate": 9.419354838709677e-06, | |
| "loss": 0.4465, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 7.0126451612903224, | |
| "grad_norm": 7.420779705047607, | |
| "learning_rate": 9.405017921146953e-06, | |
| "loss": 0.5317, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 7.013935483870967, | |
| "grad_norm": 17.574254989624023, | |
| "learning_rate": 9.39068100358423e-06, | |
| "loss": 0.3704, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.015225806451613, | |
| "grad_norm": 68.1356201171875, | |
| "learning_rate": 9.376344086021506e-06, | |
| "loss": 0.3728, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 7.016516129032258, | |
| "grad_norm": 27.221202850341797, | |
| "learning_rate": 9.362007168458782e-06, | |
| "loss": 0.3151, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 7.017806451612903, | |
| "grad_norm": 31.54282569885254, | |
| "learning_rate": 9.347670250896058e-06, | |
| "loss": 0.5659, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 7.019096774193549, | |
| "grad_norm": 10.042457580566406, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.8518, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 7.0201290322580645, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 0.5461336970329285, | |
| "eval_runtime": 12.7639, | |
| "eval_samples_per_second": 3.526, | |
| "eval_steps_per_second": 0.705, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 8.00025806451613, | |
| "grad_norm": 29.309894561767578, | |
| "learning_rate": 9.31899641577061e-06, | |
| "loss": 0.3766, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.001548387096774, | |
| "grad_norm": 12.317828178405762, | |
| "learning_rate": 9.304659498207887e-06, | |
| "loss": 0.3815, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 8.00283870967742, | |
| "grad_norm": 0.46714508533477783, | |
| "learning_rate": 9.290322580645163e-06, | |
| "loss": 0.3796, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 8.004129032258065, | |
| "grad_norm": 6.1579694747924805, | |
| "learning_rate": 9.27598566308244e-06, | |
| "loss": 0.5415, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 8.00541935483871, | |
| "grad_norm": 3.1224958896636963, | |
| "learning_rate": 9.261648745519714e-06, | |
| "loss": 0.6147, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 8.006709677419355, | |
| "grad_norm": 5.903903007507324, | |
| "learning_rate": 9.24731182795699e-06, | |
| "loss": 0.5985, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.008, | |
| "grad_norm": 8.758337020874023, | |
| "learning_rate": 9.232974910394266e-06, | |
| "loss": 0.3555, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 8.009290322580645, | |
| "grad_norm": 32.05942153930664, | |
| "learning_rate": 9.218637992831542e-06, | |
| "loss": 0.874, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 8.01058064516129, | |
| "grad_norm": 5.0790205001831055, | |
| "learning_rate": 9.204301075268819e-06, | |
| "loss": 0.4734, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 8.011870967741935, | |
| "grad_norm": 3.0711236000061035, | |
| "learning_rate": 9.189964157706093e-06, | |
| "loss": 0.3428, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 8.01316129032258, | |
| "grad_norm": 41.974037170410156, | |
| "learning_rate": 9.17562724014337e-06, | |
| "loss": 0.4555, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 8.014451612903226, | |
| "grad_norm": 12.20572280883789, | |
| "learning_rate": 9.161290322580645e-06, | |
| "loss": 0.4709, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 8.01574193548387, | |
| "grad_norm": 20.7106876373291, | |
| "learning_rate": 9.146953405017922e-06, | |
| "loss": 0.574, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 8.017032258064516, | |
| "grad_norm": 9.659872055053711, | |
| "learning_rate": 9.132616487455198e-06, | |
| "loss": 0.454, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 8.018322580645162, | |
| "grad_norm": 10.134319305419922, | |
| "learning_rate": 9.118279569892474e-06, | |
| "loss": 0.2553, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 8.019612903225806, | |
| "grad_norm": 2.39729905128479, | |
| "learning_rate": 9.10394265232975e-06, | |
| "loss": 0.5845, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 8.020129032258065, | |
| "eval_accuracy": 0.8666666666666667, | |
| "eval_loss": 0.33630362153053284, | |
| "eval_runtime": 12.4402, | |
| "eval_samples_per_second": 3.617, | |
| "eval_steps_per_second": 0.723, | |
| "step": 1404 | |
| }, | |
| { | |
| "epoch": 9.000774193548388, | |
| "grad_norm": 8.21097183227539, | |
| "learning_rate": 9.089605734767026e-06, | |
| "loss": 0.384, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 9.002064516129032, | |
| "grad_norm": 8.226821899414062, | |
| "learning_rate": 9.075268817204301e-06, | |
| "loss": 0.3429, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 9.003354838709678, | |
| "grad_norm": 27.480941772460938, | |
| "learning_rate": 9.060931899641577e-06, | |
| "loss": 0.3982, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 9.004645161290323, | |
| "grad_norm": 5.669180870056152, | |
| "learning_rate": 9.046594982078853e-06, | |
| "loss": 0.527, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 9.005935483870967, | |
| "grad_norm": 3.3830082416534424, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.3796, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.007225806451613, | |
| "grad_norm": 10.208439826965332, | |
| "learning_rate": 9.017921146953406e-06, | |
| "loss": 0.6444, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 9.008516129032259, | |
| "grad_norm": 13.923842430114746, | |
| "learning_rate": 9.003584229390682e-06, | |
| "loss": 0.6296, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 9.009806451612903, | |
| "grad_norm": 2.5853426456451416, | |
| "learning_rate": 8.989247311827958e-06, | |
| "loss": 0.4328, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 9.011096774193549, | |
| "grad_norm": 7.062355995178223, | |
| "learning_rate": 8.974910394265234e-06, | |
| "loss": 0.314, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 9.012387096774194, | |
| "grad_norm": 14.065876960754395, | |
| "learning_rate": 8.96057347670251e-06, | |
| "loss": 0.3748, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.013677419354838, | |
| "grad_norm": 28.12191390991211, | |
| "learning_rate": 8.946236559139785e-06, | |
| "loss": 0.6978, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 9.014967741935484, | |
| "grad_norm": 15.321383476257324, | |
| "learning_rate": 8.931899641577061e-06, | |
| "loss": 0.3152, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 9.01625806451613, | |
| "grad_norm": 28.618267059326172, | |
| "learning_rate": 8.917562724014338e-06, | |
| "loss": 0.4733, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 9.017548387096774, | |
| "grad_norm": 23.342119216918945, | |
| "learning_rate": 8.903225806451614e-06, | |
| "loss": 0.4027, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 9.01883870967742, | |
| "grad_norm": 9.924079895019531, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.4406, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 9.020129032258065, | |
| "grad_norm": 2.8093783855438232, | |
| "learning_rate": 8.874551971326165e-06, | |
| "loss": 0.3814, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 9.020129032258065, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 0.48341792821884155, | |
| "eval_runtime": 12.4062, | |
| "eval_samples_per_second": 3.627, | |
| "eval_steps_per_second": 0.725, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 10.001290322580646, | |
| "grad_norm": 12.093438148498535, | |
| "learning_rate": 8.86021505376344e-06, | |
| "loss": 0.3282, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 10.00258064516129, | |
| "grad_norm": 31.386138916015625, | |
| "learning_rate": 8.845878136200717e-06, | |
| "loss": 0.2942, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 10.003870967741936, | |
| "grad_norm": 40.8341064453125, | |
| "learning_rate": 8.831541218637993e-06, | |
| "loss": 0.3487, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 10.005161290322581, | |
| "grad_norm": 17.18274688720703, | |
| "learning_rate": 8.81720430107527e-06, | |
| "loss": 0.4068, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.006451612903225, | |
| "grad_norm": 12.916933059692383, | |
| "learning_rate": 8.802867383512546e-06, | |
| "loss": 0.5222, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 10.007741935483871, | |
| "grad_norm": 18.985637664794922, | |
| "learning_rate": 8.788530465949822e-06, | |
| "loss": 0.4755, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 10.009032258064517, | |
| "grad_norm": 12.837810516357422, | |
| "learning_rate": 8.774193548387098e-06, | |
| "loss": 0.5596, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 10.01032258064516, | |
| "grad_norm": 22.213176727294922, | |
| "learning_rate": 8.759856630824374e-06, | |
| "loss": 0.3784, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 10.011612903225807, | |
| "grad_norm": 17.21725845336914, | |
| "learning_rate": 8.74551971326165e-06, | |
| "loss": 0.4079, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 10.012903225806452, | |
| "grad_norm": 6.838178634643555, | |
| "learning_rate": 8.731182795698927e-06, | |
| "loss": 0.3469, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 10.014193548387096, | |
| "grad_norm": 16.972368240356445, | |
| "learning_rate": 8.716845878136203e-06, | |
| "loss": 0.4364, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 10.015483870967742, | |
| "grad_norm": 9.705157279968262, | |
| "learning_rate": 8.702508960573477e-06, | |
| "loss": 0.6421, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 10.016774193548388, | |
| "grad_norm": 39.67078399658203, | |
| "learning_rate": 8.688172043010754e-06, | |
| "loss": 0.3624, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 10.018064516129032, | |
| "grad_norm": 8.396621704101562, | |
| "learning_rate": 8.67383512544803e-06, | |
| "loss": 0.3446, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 10.019354838709678, | |
| "grad_norm": 2.1214261054992676, | |
| "learning_rate": 8.659498207885306e-06, | |
| "loss": 0.3419, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 10.020129032258065, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 0.9275798201560974, | |
| "eval_runtime": 12.6693, | |
| "eval_samples_per_second": 3.552, | |
| "eval_steps_per_second": 0.71, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 11.000516129032258, | |
| "grad_norm": 5.564540863037109, | |
| "learning_rate": 8.64516129032258e-06, | |
| "loss": 0.3787, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 11.001806451612904, | |
| "grad_norm": 19.6820125579834, | |
| "learning_rate": 8.630824372759857e-06, | |
| "loss": 0.5847, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 11.003096774193548, | |
| "grad_norm": 7.581094264984131, | |
| "learning_rate": 8.616487455197133e-06, | |
| "loss": 0.4978, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 11.004387096774193, | |
| "grad_norm": 6.024228572845459, | |
| "learning_rate": 8.602150537634409e-06, | |
| "loss": 0.3745, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 11.00567741935484, | |
| "grad_norm": 33.71268081665039, | |
| "learning_rate": 8.587813620071685e-06, | |
| "loss": 0.379, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 11.006967741935483, | |
| "grad_norm": 13.256340026855469, | |
| "learning_rate": 8.573476702508961e-06, | |
| "loss": 0.3595, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 11.008258064516129, | |
| "grad_norm": 29.387556076049805, | |
| "learning_rate": 8.559139784946238e-06, | |
| "loss": 0.5523, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 11.009548387096775, | |
| "grad_norm": 11.03555679321289, | |
| "learning_rate": 8.544802867383514e-06, | |
| "loss": 0.3211, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 11.010838709677419, | |
| "grad_norm": 60.02790451049805, | |
| "learning_rate": 8.530465949820788e-06, | |
| "loss": 0.3525, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 11.012129032258064, | |
| "grad_norm": 29.51386260986328, | |
| "learning_rate": 8.516129032258065e-06, | |
| "loss": 0.6549, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 11.01341935483871, | |
| "grad_norm": 26.95381736755371, | |
| "learning_rate": 8.50179211469534e-06, | |
| "loss": 0.4983, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 11.014709677419354, | |
| "grad_norm": 10.231369972229004, | |
| "learning_rate": 8.487455197132617e-06, | |
| "loss": 0.355, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 11.016, | |
| "grad_norm": 48.44913101196289, | |
| "learning_rate": 8.473118279569893e-06, | |
| "loss": 0.4358, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 11.017290322580646, | |
| "grad_norm": 28.62180519104004, | |
| "learning_rate": 8.45878136200717e-06, | |
| "loss": 0.3035, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 11.01858064516129, | |
| "grad_norm": 52.0974006652832, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 0.2934, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 11.019870967741936, | |
| "grad_norm": 35.910186767578125, | |
| "learning_rate": 8.430107526881722e-06, | |
| "loss": 0.2682, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 11.020129032258065, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.0908188819885254, | |
| "eval_runtime": 12.7702, | |
| "eval_samples_per_second": 3.524, | |
| "eval_steps_per_second": 0.705, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 12.001032258064516, | |
| "grad_norm": 5.6883978843688965, | |
| "learning_rate": 8.415770609318998e-06, | |
| "loss": 0.4105, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 12.002322580645162, | |
| "grad_norm": 2.781935691833496, | |
| "learning_rate": 8.401433691756273e-06, | |
| "loss": 0.2933, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 12.003612903225806, | |
| "grad_norm": 49.51652526855469, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.4624, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 12.004903225806451, | |
| "grad_norm": 2.9524998664855957, | |
| "learning_rate": 8.372759856630825e-06, | |
| "loss": 0.3423, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 12.006193548387097, | |
| "grad_norm": 27.094207763671875, | |
| "learning_rate": 8.358422939068101e-06, | |
| "loss": 0.4242, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 12.007483870967741, | |
| "grad_norm": 30.611806869506836, | |
| "learning_rate": 8.344086021505376e-06, | |
| "loss": 0.3531, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 12.008774193548387, | |
| "grad_norm": 5.029760360717773, | |
| "learning_rate": 8.329749103942652e-06, | |
| "loss": 0.4756, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 12.010064516129033, | |
| "grad_norm": 15.222090721130371, | |
| "learning_rate": 8.315412186379928e-06, | |
| "loss": 0.5881, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 12.011354838709677, | |
| "grad_norm": 0.34555259346961975, | |
| "learning_rate": 8.301075268817204e-06, | |
| "loss": 0.2341, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 12.012645161290322, | |
| "grad_norm": 0.8221472501754761, | |
| "learning_rate": 8.28673835125448e-06, | |
| "loss": 0.2678, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 12.013935483870968, | |
| "grad_norm": 22.073829650878906, | |
| "learning_rate": 8.272401433691757e-06, | |
| "loss": 0.5473, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 12.015225806451612, | |
| "grad_norm": 42.95087432861328, | |
| "learning_rate": 8.258064516129033e-06, | |
| "loss": 0.4656, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 12.016516129032258, | |
| "grad_norm": 3.4676356315612793, | |
| "learning_rate": 8.24372759856631e-06, | |
| "loss": 0.3188, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.017806451612904, | |
| "grad_norm": 24.06205177307129, | |
| "learning_rate": 8.229390681003585e-06, | |
| "loss": 0.6528, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 12.019096774193548, | |
| "grad_norm": 10.741403579711914, | |
| "learning_rate": 8.215053763440862e-06, | |
| "loss": 0.272, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 12.020129032258065, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 0.7781123518943787, | |
| "eval_runtime": 12.3726, | |
| "eval_samples_per_second": 3.637, | |
| "eval_steps_per_second": 0.727, | |
| "step": 2028 | |
| }, | |
| { | |
| "epoch": 13.00025806451613, | |
| "grad_norm": 17.455461502075195, | |
| "learning_rate": 8.200716845878138e-06, | |
| "loss": 0.3592, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 13.001548387096774, | |
| "grad_norm": 16.679277420043945, | |
| "learning_rate": 8.186379928315414e-06, | |
| "loss": 0.2775, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 13.00283870967742, | |
| "grad_norm": 20.2856502532959, | |
| "learning_rate": 8.172043010752689e-06, | |
| "loss": 0.3932, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 13.004129032258065, | |
| "grad_norm": 1.3590325117111206, | |
| "learning_rate": 8.157706093189965e-06, | |
| "loss": 0.2778, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 13.00541935483871, | |
| "grad_norm": 60.31137466430664, | |
| "learning_rate": 8.143369175627241e-06, | |
| "loss": 0.5134, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 13.006709677419355, | |
| "grad_norm": 10.152859687805176, | |
| "learning_rate": 8.129032258064517e-06, | |
| "loss": 0.3191, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 13.008, | |
| "grad_norm": 18.664304733276367, | |
| "learning_rate": 8.114695340501793e-06, | |
| "loss": 0.5233, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 13.009290322580645, | |
| "grad_norm": 17.228160858154297, | |
| "learning_rate": 8.100358422939068e-06, | |
| "loss": 0.3784, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 13.01058064516129, | |
| "grad_norm": 27.762767791748047, | |
| "learning_rate": 8.086021505376344e-06, | |
| "loss": 0.4604, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 13.011870967741935, | |
| "grad_norm": 1.8264316320419312, | |
| "learning_rate": 8.07168458781362e-06, | |
| "loss": 0.4964, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 13.01316129032258, | |
| "grad_norm": 20.46810531616211, | |
| "learning_rate": 8.057347670250897e-06, | |
| "loss": 0.3672, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 13.014451612903226, | |
| "grad_norm": 43.744483947753906, | |
| "learning_rate": 8.043010752688173e-06, | |
| "loss": 0.2465, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 13.01574193548387, | |
| "grad_norm": 12.962421417236328, | |
| "learning_rate": 8.028673835125449e-06, | |
| "loss": 0.3561, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 13.017032258064516, | |
| "grad_norm": 6.151007652282715, | |
| "learning_rate": 8.014336917562725e-06, | |
| "loss": 0.3438, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 13.018322580645162, | |
| "grad_norm": 66.533203125, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4317, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 13.019612903225806, | |
| "grad_norm": 10.530814170837402, | |
| "learning_rate": 7.985663082437278e-06, | |
| "loss": 0.313, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 13.020129032258065, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 1.4004600048065186, | |
| "eval_runtime": 12.4747, | |
| "eval_samples_per_second": 3.607, | |
| "eval_steps_per_second": 0.721, | |
| "step": 2184 | |
| }, | |
| { | |
| "epoch": 14.000774193548388, | |
| "grad_norm": 5.235982418060303, | |
| "learning_rate": 7.971326164874552e-06, | |
| "loss": 0.3872, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 14.002064516129032, | |
| "grad_norm": 25.094335556030273, | |
| "learning_rate": 7.956989247311828e-06, | |
| "loss": 0.4856, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 14.003354838709678, | |
| "grad_norm": 27.851547241210938, | |
| "learning_rate": 7.942652329749104e-06, | |
| "loss": 0.5354, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 14.004645161290323, | |
| "grad_norm": 7.955545425415039, | |
| "learning_rate": 7.92831541218638e-06, | |
| "loss": 0.4407, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 14.005935483870967, | |
| "grad_norm": 13.76785945892334, | |
| "learning_rate": 7.913978494623657e-06, | |
| "loss": 0.2369, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 14.007225806451613, | |
| "grad_norm": 11.216278076171875, | |
| "learning_rate": 7.899641577060933e-06, | |
| "loss": 0.4398, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 14.008516129032259, | |
| "grad_norm": 23.19121551513672, | |
| "learning_rate": 7.88530465949821e-06, | |
| "loss": 0.234, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 14.009806451612903, | |
| "grad_norm": 19.719554901123047, | |
| "learning_rate": 7.870967741935484e-06, | |
| "loss": 0.4499, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 14.011096774193549, | |
| "grad_norm": 1.4593003988265991, | |
| "learning_rate": 7.85663082437276e-06, | |
| "loss": 0.5409, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 14.012387096774194, | |
| "grad_norm": 10.19871711730957, | |
| "learning_rate": 7.842293906810036e-06, | |
| "loss": 0.5856, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 14.013677419354838, | |
| "grad_norm": 4.29229736328125, | |
| "learning_rate": 7.827956989247312e-06, | |
| "loss": 0.4059, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 14.014967741935484, | |
| "grad_norm": 9.68092155456543, | |
| "learning_rate": 7.813620071684589e-06, | |
| "loss": 0.5861, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 14.01625806451613, | |
| "grad_norm": 0.605811357498169, | |
| "learning_rate": 7.799283154121865e-06, | |
| "loss": 0.3349, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 14.017548387096774, | |
| "grad_norm": 22.93714141845703, | |
| "learning_rate": 7.78494623655914e-06, | |
| "loss": 0.2628, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 14.01883870967742, | |
| "grad_norm": 2.228055715560913, | |
| "learning_rate": 7.770609318996416e-06, | |
| "loss": 0.3993, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 14.020129032258065, | |
| "grad_norm": 3.1606695652008057, | |
| "learning_rate": 7.756272401433692e-06, | |
| "loss": 0.2476, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 14.020129032258065, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 0.9078776836395264, | |
| "eval_runtime": 12.7936, | |
| "eval_samples_per_second": 3.517, | |
| "eval_steps_per_second": 0.703, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 15.001290322580646, | |
| "grad_norm": 8.388017654418945, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.2463, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 15.00258064516129, | |
| "grad_norm": 1.4157217741012573, | |
| "learning_rate": 7.727598566308244e-06, | |
| "loss": 0.1762, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 15.003870967741936, | |
| "grad_norm": 73.77959442138672, | |
| "learning_rate": 7.71326164874552e-06, | |
| "loss": 0.4477, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 15.005161290322581, | |
| "grad_norm": 129.7151336669922, | |
| "learning_rate": 7.698924731182797e-06, | |
| "loss": 0.4238, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 15.006451612903225, | |
| "grad_norm": 54.64049530029297, | |
| "learning_rate": 7.684587813620073e-06, | |
| "loss": 0.3693, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 15.007741935483871, | |
| "grad_norm": 1.2686078548431396, | |
| "learning_rate": 7.670250896057349e-06, | |
| "loss": 0.5296, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 15.009032258064517, | |
| "grad_norm": 0.11383429169654846, | |
| "learning_rate": 7.655913978494625e-06, | |
| "loss": 0.3847, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 15.01032258064516, | |
| "grad_norm": 49.59299850463867, | |
| "learning_rate": 7.641577060931901e-06, | |
| "loss": 0.3088, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 15.011612903225807, | |
| "grad_norm": 32.66481399536133, | |
| "learning_rate": 7.627240143369177e-06, | |
| "loss": 0.6174, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 15.012903225806452, | |
| "grad_norm": 5.530159950256348, | |
| "learning_rate": 7.612903225806451e-06, | |
| "loss": 0.5322, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 15.014193548387096, | |
| "grad_norm": 21.824308395385742, | |
| "learning_rate": 7.5985663082437275e-06, | |
| "loss": 0.2274, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 15.015483870967742, | |
| "grad_norm": 2.1449851989746094, | |
| "learning_rate": 7.584229390681004e-06, | |
| "loss": 0.3631, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 15.016774193548388, | |
| "grad_norm": 31.965057373046875, | |
| "learning_rate": 7.56989247311828e-06, | |
| "loss": 0.7662, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 15.018064516129032, | |
| "grad_norm": 20.789730072021484, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 0.7048, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 15.019354838709678, | |
| "grad_norm": 21.57069969177246, | |
| "learning_rate": 7.541218637992832e-06, | |
| "loss": 0.6739, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 15.020129032258065, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 0.7557557225227356, | |
| "eval_runtime": 11.99, | |
| "eval_samples_per_second": 3.753, | |
| "eval_steps_per_second": 0.751, | |
| "step": 2496 | |
| }, | |
| { | |
| "epoch": 16.00051612903226, | |
| "grad_norm": 12.548541069030762, | |
| "learning_rate": 7.526881720430108e-06, | |
| "loss": 0.4202, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 16.001806451612904, | |
| "grad_norm": 19.834489822387695, | |
| "learning_rate": 7.512544802867384e-06, | |
| "loss": 0.3242, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 16.003096774193548, | |
| "grad_norm": 28.00609016418457, | |
| "learning_rate": 7.49820788530466e-06, | |
| "loss": 0.3601, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 16.004387096774195, | |
| "grad_norm": 29.896514892578125, | |
| "learning_rate": 7.483870967741936e-06, | |
| "loss": 0.3028, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 16.00567741935484, | |
| "grad_norm": 29.4525203704834, | |
| "learning_rate": 7.4695340501792126e-06, | |
| "loss": 0.7224, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 16.006967741935483, | |
| "grad_norm": 74.91488647460938, | |
| "learning_rate": 7.455197132616489e-06, | |
| "loss": 0.3392, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 16.00825806451613, | |
| "grad_norm": 25.67245864868164, | |
| "learning_rate": 7.440860215053764e-06, | |
| "loss": 0.4847, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 16.009548387096775, | |
| "grad_norm": 31.815969467163086, | |
| "learning_rate": 7.4265232974910395e-06, | |
| "loss": 0.4092, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 16.01083870967742, | |
| "grad_norm": 0.19238120317459106, | |
| "learning_rate": 7.412186379928316e-06, | |
| "loss": 0.2497, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 16.012129032258066, | |
| "grad_norm": 6.9821038246154785, | |
| "learning_rate": 7.397849462365592e-06, | |
| "loss": 0.411, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 16.01341935483871, | |
| "grad_norm": 47.721553802490234, | |
| "learning_rate": 7.383512544802868e-06, | |
| "loss": 0.3704, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 16.014709677419354, | |
| "grad_norm": 64.65721130371094, | |
| "learning_rate": 7.3691756272401435e-06, | |
| "loss": 0.3476, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 16.016, | |
| "grad_norm": 12.756269454956055, | |
| "learning_rate": 7.35483870967742e-06, | |
| "loss": 0.296, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 16.017290322580646, | |
| "grad_norm": 83.34323120117188, | |
| "learning_rate": 7.340501792114696e-06, | |
| "loss": 0.3619, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 16.01858064516129, | |
| "grad_norm": 40.10869598388672, | |
| "learning_rate": 7.326164874551972e-06, | |
| "loss": 0.728, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 16.019870967741934, | |
| "grad_norm": 24.635774612426758, | |
| "learning_rate": 7.311827956989248e-06, | |
| "loss": 0.3332, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 16.020129032258065, | |
| "eval_accuracy": 0.8666666666666667, | |
| "eval_loss": 0.34080299735069275, | |
| "eval_runtime": 12.4593, | |
| "eval_samples_per_second": 3.612, | |
| "eval_steps_per_second": 0.722, | |
| "step": 2652 | |
| }, | |
| { | |
| "epoch": 17.001032258064516, | |
| "grad_norm": 36.969512939453125, | |
| "learning_rate": 7.2974910394265245e-06, | |
| "loss": 0.2571, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 17.00232258064516, | |
| "grad_norm": 26.82135581970215, | |
| "learning_rate": 7.2831541218638e-06, | |
| "loss": 0.3187, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 17.003612903225807, | |
| "grad_norm": 61.53876876831055, | |
| "learning_rate": 7.268817204301076e-06, | |
| "loss": 0.3182, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 17.00490322580645, | |
| "grad_norm": 65.02365112304688, | |
| "learning_rate": 7.254480286738352e-06, | |
| "loss": 0.4658, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 17.006193548387095, | |
| "grad_norm": 16.802270889282227, | |
| "learning_rate": 7.240143369175628e-06, | |
| "loss": 0.1815, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 17.007483870967743, | |
| "grad_norm": 8.099735260009766, | |
| "learning_rate": 7.225806451612903e-06, | |
| "loss": 0.1619, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 17.008774193548387, | |
| "grad_norm": 4.278258800506592, | |
| "learning_rate": 7.211469534050179e-06, | |
| "loss": 0.4259, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 17.01006451612903, | |
| "grad_norm": 6.792086601257324, | |
| "learning_rate": 7.1971326164874554e-06, | |
| "loss": 0.3659, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 17.01135483870968, | |
| "grad_norm": 67.52592468261719, | |
| "learning_rate": 7.182795698924732e-06, | |
| "loss": 0.4207, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 17.012645161290322, | |
| "grad_norm": 27.389123916625977, | |
| "learning_rate": 7.168458781362008e-06, | |
| "loss": 0.1438, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 17.013935483870966, | |
| "grad_norm": 0.19431276619434357, | |
| "learning_rate": 7.154121863799284e-06, | |
| "loss": 0.5633, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 17.015225806451614, | |
| "grad_norm": 14.940716743469238, | |
| "learning_rate": 7.139784946236559e-06, | |
| "loss": 0.2484, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 17.016516129032258, | |
| "grad_norm": 44.51100540161133, | |
| "learning_rate": 7.125448028673836e-06, | |
| "loss": 0.2306, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 17.017806451612902, | |
| "grad_norm": 52.78997802734375, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 0.4458, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 17.01909677419355, | |
| "grad_norm": 2.4495134353637695, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.4414, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 17.020129032258065, | |
| "eval_accuracy": 0.8666666666666667, | |
| "eval_loss": 0.45585525035858154, | |
| "eval_runtime": 12.3846, | |
| "eval_samples_per_second": 3.634, | |
| "eval_steps_per_second": 0.727, | |
| "step": 2808 | |
| }, | |
| { | |
| "epoch": 18.000258064516128, | |
| "grad_norm": 51.98246765136719, | |
| "learning_rate": 7.082437275985664e-06, | |
| "loss": 0.2975, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 18.001548387096776, | |
| "grad_norm": 0.20406349003314972, | |
| "learning_rate": 7.0681003584229404e-06, | |
| "loss": 0.1485, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 18.00283870967742, | |
| "grad_norm": 41.591121673583984, | |
| "learning_rate": 7.053763440860215e-06, | |
| "loss": 0.5301, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 18.004129032258064, | |
| "grad_norm": 3.681441068649292, | |
| "learning_rate": 7.039426523297491e-06, | |
| "loss": 0.2969, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 18.00541935483871, | |
| "grad_norm": 0.1328764259815216, | |
| "learning_rate": 7.025089605734767e-06, | |
| "loss": 0.5148, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 18.006709677419355, | |
| "grad_norm": 58.09383010864258, | |
| "learning_rate": 7.010752688172044e-06, | |
| "loss": 0.5326, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 18.008, | |
| "grad_norm": 64.18909454345703, | |
| "learning_rate": 6.99641577060932e-06, | |
| "loss": 0.4416, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 18.009290322580647, | |
| "grad_norm": 23.151090621948242, | |
| "learning_rate": 6.982078853046595e-06, | |
| "loss": 0.3543, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 18.01058064516129, | |
| "grad_norm": 26.064599990844727, | |
| "learning_rate": 6.967741935483871e-06, | |
| "loss": 0.3632, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 18.011870967741935, | |
| "grad_norm": 47.92241668701172, | |
| "learning_rate": 6.9534050179211476e-06, | |
| "loss": 0.2792, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 18.013161290322582, | |
| "grad_norm": 2.298640012741089, | |
| "learning_rate": 6.939068100358424e-06, | |
| "loss": 0.117, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 18.014451612903226, | |
| "grad_norm": 0.6145700216293335, | |
| "learning_rate": 6.9247311827957e-06, | |
| "loss": 0.5346, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 18.01574193548387, | |
| "grad_norm": 0.37437283992767334, | |
| "learning_rate": 6.910394265232976e-06, | |
| "loss": 0.2575, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 18.017032258064518, | |
| "grad_norm": 0.5951812863349915, | |
| "learning_rate": 6.8960573476702516e-06, | |
| "loss": 0.4508, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 18.01832258064516, | |
| "grad_norm": 0.2956177592277527, | |
| "learning_rate": 6.881720430107528e-06, | |
| "loss": 0.292, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 18.019612903225806, | |
| "grad_norm": 38.594276428222656, | |
| "learning_rate": 6.867383512544803e-06, | |
| "loss": 0.5799, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 18.020129032258065, | |
| "eval_accuracy": 0.8444444444444444, | |
| "eval_loss": 0.3860296607017517, | |
| "eval_runtime": 14.1806, | |
| "eval_samples_per_second": 3.173, | |
| "eval_steps_per_second": 0.635, | |
| "step": 2964 | |
| }, | |
| { | |
| "epoch": 19.000774193548388, | |
| "grad_norm": 0.9255644679069519, | |
| "learning_rate": 6.853046594982079e-06, | |
| "loss": 0.2206, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 19.002064516129032, | |
| "grad_norm": 21.96172332763672, | |
| "learning_rate": 6.838709677419355e-06, | |
| "loss": 0.2664, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 19.003354838709676, | |
| "grad_norm": 47.029502868652344, | |
| "learning_rate": 6.824372759856631e-06, | |
| "loss": 0.1402, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 19.004645161290323, | |
| "grad_norm": 0.06740839034318924, | |
| "learning_rate": 6.810035842293907e-06, | |
| "loss": 0.304, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 19.005935483870967, | |
| "grad_norm": 48.14612579345703, | |
| "learning_rate": 6.795698924731183e-06, | |
| "loss": 0.6275, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 19.00722580645161, | |
| "grad_norm": 0.8753212690353394, | |
| "learning_rate": 6.7813620071684595e-06, | |
| "loss": 0.3483, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 19.00851612903226, | |
| "grad_norm": 61.9237174987793, | |
| "learning_rate": 6.767025089605736e-06, | |
| "loss": 0.4497, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 19.009806451612903, | |
| "grad_norm": 13.719686508178711, | |
| "learning_rate": 6.752688172043012e-06, | |
| "loss": 0.4342, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 19.011096774193547, | |
| "grad_norm": 9.002520561218262, | |
| "learning_rate": 6.738351254480287e-06, | |
| "loss": 0.1794, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 19.012387096774194, | |
| "grad_norm": 5.730069160461426, | |
| "learning_rate": 6.7240143369175635e-06, | |
| "loss": 0.4597, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 19.01367741935484, | |
| "grad_norm": 6.476169586181641, | |
| "learning_rate": 6.70967741935484e-06, | |
| "loss": 0.1593, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 19.014967741935482, | |
| "grad_norm": 22.35500144958496, | |
| "learning_rate": 6.695340501792115e-06, | |
| "loss": 0.5272, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 19.01625806451613, | |
| "grad_norm": 2.3751094341278076, | |
| "learning_rate": 6.6810035842293904e-06, | |
| "loss": 0.2401, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 19.017548387096774, | |
| "grad_norm": 63.143310546875, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.5057, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 19.018838709677418, | |
| "grad_norm": 17.578876495361328, | |
| "learning_rate": 6.652329749103943e-06, | |
| "loss": 0.2632, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 19.020129032258065, | |
| "grad_norm": 4.3501176834106445, | |
| "learning_rate": 6.637992831541219e-06, | |
| "loss": 0.3393, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 19.020129032258065, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 1.2055385112762451, | |
| "eval_runtime": 12.3892, | |
| "eval_samples_per_second": 3.632, | |
| "eval_steps_per_second": 0.726, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 20.001290322580644, | |
| "grad_norm": 18.48643684387207, | |
| "learning_rate": 6.623655913978495e-06, | |
| "loss": 0.2538, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 20.00258064516129, | |
| "grad_norm": 0.6086944341659546, | |
| "learning_rate": 6.6093189964157715e-06, | |
| "loss": 0.1029, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 20.003870967741936, | |
| "grad_norm": 91.40310668945312, | |
| "learning_rate": 6.594982078853047e-06, | |
| "loss": 0.226, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 20.00516129032258, | |
| "grad_norm": 0.6372993588447571, | |
| "learning_rate": 6.580645161290323e-06, | |
| "loss": 0.4993, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 20.006451612903227, | |
| "grad_norm": 69.5063247680664, | |
| "learning_rate": 6.566308243727599e-06, | |
| "loss": 0.552, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 20.00774193548387, | |
| "grad_norm": 21.01985740661621, | |
| "learning_rate": 6.5519713261648755e-06, | |
| "loss": 0.275, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 20.009032258064515, | |
| "grad_norm": 6.748633861541748, | |
| "learning_rate": 6.537634408602152e-06, | |
| "loss": 0.3271, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 20.010322580645163, | |
| "grad_norm": 0.1616794317960739, | |
| "learning_rate": 6.523297491039428e-06, | |
| "loss": 0.3492, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 20.011612903225807, | |
| "grad_norm": 10.018768310546875, | |
| "learning_rate": 6.508960573476702e-06, | |
| "loss": 0.7704, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 20.01290322580645, | |
| "grad_norm": 8.020637512207031, | |
| "learning_rate": 6.494623655913979e-06, | |
| "loss": 0.3829, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 20.014193548387098, | |
| "grad_norm": 0.08563937246799469, | |
| "learning_rate": 6.480286738351255e-06, | |
| "loss": 0.4351, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 20.015483870967742, | |
| "grad_norm": 65.92079162597656, | |
| "learning_rate": 6.465949820788531e-06, | |
| "loss": 0.4459, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 20.016774193548386, | |
| "grad_norm": 50.6903190612793, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.3612, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 20.018064516129034, | |
| "grad_norm": 85.24955749511719, | |
| "learning_rate": 6.437275985663083e-06, | |
| "loss": 0.3458, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 20.019354838709678, | |
| "grad_norm": 1.5930240154266357, | |
| "learning_rate": 6.422939068100359e-06, | |
| "loss": 0.1268, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 20.020129032258065, | |
| "eval_accuracy": 0.8444444444444444, | |
| "eval_loss": 0.5241544842720032, | |
| "eval_runtime": 12.4133, | |
| "eval_samples_per_second": 3.625, | |
| "eval_steps_per_second": 0.725, | |
| "step": 3276 | |
| }, | |
| { | |
| "epoch": 21.00051612903226, | |
| "grad_norm": 6.929024696350098, | |
| "learning_rate": 6.408602150537635e-06, | |
| "loss": 0.1939, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 21.001806451612904, | |
| "grad_norm": 0.387129545211792, | |
| "learning_rate": 6.394265232974911e-06, | |
| "loss": 0.2652, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 21.003096774193548, | |
| "grad_norm": 19.081331253051758, | |
| "learning_rate": 6.379928315412187e-06, | |
| "loss": 0.3289, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 21.004387096774195, | |
| "grad_norm": 0.08913157880306244, | |
| "learning_rate": 6.365591397849464e-06, | |
| "loss": 0.2781, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 21.00567741935484, | |
| "grad_norm": 5.859015464782715, | |
| "learning_rate": 6.351254480286739e-06, | |
| "loss": 0.2153, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 21.006967741935483, | |
| "grad_norm": 0.22481068968772888, | |
| "learning_rate": 6.336917562724015e-06, | |
| "loss": 0.196, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 21.00825806451613, | |
| "grad_norm": 9.10342025756836, | |
| "learning_rate": 6.3225806451612906e-06, | |
| "loss": 0.3705, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 21.009548387096775, | |
| "grad_norm": 114.85432434082031, | |
| "learning_rate": 6.308243727598567e-06, | |
| "loss": 0.338, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 21.01083870967742, | |
| "grad_norm": 0.7558663487434387, | |
| "learning_rate": 6.293906810035842e-06, | |
| "loss": 0.5103, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 21.012129032258066, | |
| "grad_norm": 36.05562210083008, | |
| "learning_rate": 6.279569892473118e-06, | |
| "loss": 0.469, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 21.01341935483871, | |
| "grad_norm": 10.676998138427734, | |
| "learning_rate": 6.2652329749103945e-06, | |
| "loss": 0.2335, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 21.014709677419354, | |
| "grad_norm": 3.4675915241241455, | |
| "learning_rate": 6.250896057347671e-06, | |
| "loss": 0.2347, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 21.016, | |
| "grad_norm": 27.237436294555664, | |
| "learning_rate": 6.236559139784947e-06, | |
| "loss": 0.349, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 21.017290322580646, | |
| "grad_norm": 0.49276021122932434, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 0.2839, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 21.01858064516129, | |
| "grad_norm": 0.3784695565700531, | |
| "learning_rate": 6.2078853046594985e-06, | |
| "loss": 0.1199, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 21.019870967741934, | |
| "grad_norm": 3.458200693130493, | |
| "learning_rate": 6.193548387096775e-06, | |
| "loss": 0.1138, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 21.020129032258065, | |
| "eval_accuracy": 0.8444444444444444, | |
| "eval_loss": 0.46266043186187744, | |
| "eval_runtime": 12.4947, | |
| "eval_samples_per_second": 3.602, | |
| "eval_steps_per_second": 0.72, | |
| "step": 3432 | |
| }, | |
| { | |
| "epoch": 22.001032258064516, | |
| "grad_norm": 0.38675570487976074, | |
| "learning_rate": 6.179211469534051e-06, | |
| "loss": 0.1098, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 22.00232258064516, | |
| "grad_norm": 1.0825914144515991, | |
| "learning_rate": 6.164874551971327e-06, | |
| "loss": 0.1732, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 22.003612903225807, | |
| "grad_norm": 0.0654192566871643, | |
| "learning_rate": 6.150537634408603e-06, | |
| "loss": 0.2489, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 22.00490322580645, | |
| "grad_norm": 70.72102355957031, | |
| "learning_rate": 6.136200716845878e-06, | |
| "loss": 0.7633, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 22.006193548387095, | |
| "grad_norm": 81.68818664550781, | |
| "learning_rate": 6.121863799283154e-06, | |
| "loss": 0.2802, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 22.007483870967743, | |
| "grad_norm": 69.21981048583984, | |
| "learning_rate": 6.10752688172043e-06, | |
| "loss": 0.438, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 22.008774193548387, | |
| "grad_norm": 44.38734817504883, | |
| "learning_rate": 6.0931899641577065e-06, | |
| "loss": 0.4938, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 22.01006451612903, | |
| "grad_norm": 0.6449613571166992, | |
| "learning_rate": 6.078853046594983e-06, | |
| "loss": 0.2029, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 22.01135483870968, | |
| "grad_norm": 0.20172294974327087, | |
| "learning_rate": 6.064516129032259e-06, | |
| "loss": 0.1766, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 22.012645161290322, | |
| "grad_norm": 72.92123413085938, | |
| "learning_rate": 6.050179211469534e-06, | |
| "loss": 0.9753, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 22.013935483870966, | |
| "grad_norm": 2.54245924949646, | |
| "learning_rate": 6.0358422939068105e-06, | |
| "loss": 0.379, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 22.015225806451614, | |
| "grad_norm": 0.32834023237228394, | |
| "learning_rate": 6.021505376344087e-06, | |
| "loss": 0.1858, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 22.016516129032258, | |
| "grad_norm": 0.2698739767074585, | |
| "learning_rate": 6.007168458781363e-06, | |
| "loss": 0.3366, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 22.017806451612902, | |
| "grad_norm": 23.720914840698242, | |
| "learning_rate": 5.992831541218639e-06, | |
| "loss": 0.3285, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 22.01909677419355, | |
| "grad_norm": 33.760955810546875, | |
| "learning_rate": 5.978494623655915e-06, | |
| "loss": 0.6119, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 22.020129032258065, | |
| "eval_accuracy": 0.7555555555555555, | |
| "eval_loss": 0.8547287583351135, | |
| "eval_runtime": 30.0345, | |
| "eval_samples_per_second": 1.498, | |
| "eval_steps_per_second": 0.3, | |
| "step": 3588 | |
| }, | |
| { | |
| "epoch": 23.000258064516128, | |
| "grad_norm": 2.298129081726074, | |
| "learning_rate": 5.964157706093191e-06, | |
| "loss": 0.1481, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 23.001548387096776, | |
| "grad_norm": 0.36253607273101807, | |
| "learning_rate": 5.949820788530466e-06, | |
| "loss": 0.38, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 23.00283870967742, | |
| "grad_norm": 72.27323913574219, | |
| "learning_rate": 5.935483870967742e-06, | |
| "loss": 0.307, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 23.004129032258064, | |
| "grad_norm": 0.09664549678564072, | |
| "learning_rate": 5.9211469534050184e-06, | |
| "loss": 0.2919, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 23.00541935483871, | |
| "grad_norm": 0.9734894037246704, | |
| "learning_rate": 5.906810035842294e-06, | |
| "loss": 0.3973, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 23.006709677419355, | |
| "grad_norm": 44.79798889160156, | |
| "learning_rate": 5.89247311827957e-06, | |
| "loss": 0.4191, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 23.008, | |
| "grad_norm": 2.0593411922454834, | |
| "learning_rate": 5.878136200716846e-06, | |
| "loss": 0.1328, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 23.009290322580647, | |
| "grad_norm": 75.23008728027344, | |
| "learning_rate": 5.8637992831541224e-06, | |
| "loss": 0.455, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 23.01058064516129, | |
| "grad_norm": 1.4643235206604004, | |
| "learning_rate": 5.849462365591399e-06, | |
| "loss": 0.0429, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 23.011870967741935, | |
| "grad_norm": 57.34547805786133, | |
| "learning_rate": 5.835125448028675e-06, | |
| "loss": 0.4096, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 23.013161290322582, | |
| "grad_norm": 1.0468956232070923, | |
| "learning_rate": 5.82078853046595e-06, | |
| "loss": 0.5174, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 23.014451612903226, | |
| "grad_norm": 0.7758573293685913, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 0.1903, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 23.01574193548387, | |
| "grad_norm": 116.5160903930664, | |
| "learning_rate": 5.792114695340503e-06, | |
| "loss": 0.296, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 23.017032258064518, | |
| "grad_norm": 100.15150451660156, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 0.6699, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 23.01832258064516, | |
| "grad_norm": 128.29710388183594, | |
| "learning_rate": 5.763440860215054e-06, | |
| "loss": 0.3131, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 23.019612903225806, | |
| "grad_norm": 31.59364891052246, | |
| "learning_rate": 5.7491039426523296e-06, | |
| "loss": 0.1607, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 23.020129032258065, | |
| "eval_accuracy": 0.8222222222222222, | |
| "eval_loss": 0.76637202501297, | |
| "eval_runtime": 12.4118, | |
| "eval_samples_per_second": 3.626, | |
| "eval_steps_per_second": 0.725, | |
| "step": 3744 | |
| }, | |
| { | |
| "epoch": 24.000774193548388, | |
| "grad_norm": 0.016302738338708878, | |
| "learning_rate": 5.734767025089606e-06, | |
| "loss": 0.2076, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 24.002064516129032, | |
| "grad_norm": 23.496017456054688, | |
| "learning_rate": 5.720430107526882e-06, | |
| "loss": 0.2031, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 24.003354838709676, | |
| "grad_norm": 0.20120538771152496, | |
| "learning_rate": 5.706093189964158e-06, | |
| "loss": 0.3169, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 24.004645161290323, | |
| "grad_norm": 0.810869038105011, | |
| "learning_rate": 5.691756272401434e-06, | |
| "loss": 0.2218, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 24.005935483870967, | |
| "grad_norm": 102.47581481933594, | |
| "learning_rate": 5.677419354838711e-06, | |
| "loss": 0.4929, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 24.00722580645161, | |
| "grad_norm": 0.4209055006504059, | |
| "learning_rate": 5.663082437275986e-06, | |
| "loss": 0.3321, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 24.00851612903226, | |
| "grad_norm": 0.12239155173301697, | |
| "learning_rate": 5.648745519713262e-06, | |
| "loss": 0.1542, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 24.009806451612903, | |
| "grad_norm": 0.6548942923545837, | |
| "learning_rate": 5.634408602150538e-06, | |
| "loss": 0.1785, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 24.011096774193547, | |
| "grad_norm": 0.06878427416086197, | |
| "learning_rate": 5.620071684587815e-06, | |
| "loss": 0.3115, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 24.012387096774194, | |
| "grad_norm": 0.3307665288448334, | |
| "learning_rate": 5.605734767025091e-06, | |
| "loss": 0.35, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 24.01367741935484, | |
| "grad_norm": 0.5849574208259583, | |
| "learning_rate": 5.591397849462365e-06, | |
| "loss": 0.1148, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 24.014967741935482, | |
| "grad_norm": 0.4524058401584625, | |
| "learning_rate": 5.5770609318996415e-06, | |
| "loss": 0.3095, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 24.01625806451613, | |
| "grad_norm": 51.53236770629883, | |
| "learning_rate": 5.562724014336918e-06, | |
| "loss": 0.2, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 24.017548387096774, | |
| "grad_norm": 75.92852020263672, | |
| "learning_rate": 5.548387096774194e-06, | |
| "loss": 0.2467, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 24.018838709677418, | |
| "grad_norm": 138.65054321289062, | |
| "learning_rate": 5.53405017921147e-06, | |
| "loss": 0.554, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 24.020129032258065, | |
| "grad_norm": 0.004257502965629101, | |
| "learning_rate": 5.5197132616487455e-06, | |
| "loss": 0.1289, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 24.020129032258065, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.7741568088531494, | |
| "eval_runtime": 12.7403, | |
| "eval_samples_per_second": 3.532, | |
| "eval_steps_per_second": 0.706, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 25.001290322580644, | |
| "grad_norm": 0.04135989397764206, | |
| "learning_rate": 5.505376344086022e-06, | |
| "loss": 0.1466, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 25.00258064516129, | |
| "grad_norm": 0.10639182478189468, | |
| "learning_rate": 5.491039426523298e-06, | |
| "loss": 0.1246, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 25.003870967741936, | |
| "grad_norm": 1.0761823654174805, | |
| "learning_rate": 5.476702508960574e-06, | |
| "loss": 0.6441, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 25.00516129032258, | |
| "grad_norm": 105.61331939697266, | |
| "learning_rate": 5.46236559139785e-06, | |
| "loss": 0.3217, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 25.006451612903227, | |
| "grad_norm": 69.04462432861328, | |
| "learning_rate": 5.4480286738351265e-06, | |
| "loss": 0.4453, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 25.00774193548387, | |
| "grad_norm": 7.438420295715332, | |
| "learning_rate": 5.433691756272402e-06, | |
| "loss": 0.0321, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 25.009032258064515, | |
| "grad_norm": 0.010565834119915962, | |
| "learning_rate": 5.419354838709678e-06, | |
| "loss": 0.2326, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 25.010322580645163, | |
| "grad_norm": 0.2766633927822113, | |
| "learning_rate": 5.4050179211469535e-06, | |
| "loss": 0.1925, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 25.011612903225807, | |
| "grad_norm": 0.09161556512117386, | |
| "learning_rate": 5.39068100358423e-06, | |
| "loss": 0.2467, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 25.01290322580645, | |
| "grad_norm": 0.11548233032226562, | |
| "learning_rate": 5.376344086021506e-06, | |
| "loss": 0.164, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 25.014193548387098, | |
| "grad_norm": 122.91968536376953, | |
| "learning_rate": 5.362007168458781e-06, | |
| "loss": 0.3615, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 25.015483870967742, | |
| "grad_norm": 54.03779983520508, | |
| "learning_rate": 5.3476702508960574e-06, | |
| "loss": 0.1533, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 25.016774193548386, | |
| "grad_norm": 0.053860463201999664, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.4189, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 25.018064516129034, | |
| "grad_norm": 4.591695785522461, | |
| "learning_rate": 5.31899641577061e-06, | |
| "loss": 0.5156, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 25.019354838709678, | |
| "grad_norm": 7.412461280822754, | |
| "learning_rate": 5.304659498207886e-06, | |
| "loss": 0.3687, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 25.020129032258065, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.7859394550323486, | |
| "eval_runtime": 12.3775, | |
| "eval_samples_per_second": 3.636, | |
| "eval_steps_per_second": 0.727, | |
| "step": 4056 | |
| }, | |
| { | |
| "epoch": 26.00051612903226, | |
| "grad_norm": 0.34613585472106934, | |
| "learning_rate": 5.290322580645162e-06, | |
| "loss": 0.081, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 26.001806451612904, | |
| "grad_norm": 32.229827880859375, | |
| "learning_rate": 5.275985663082438e-06, | |
| "loss": 0.2012, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 26.003096774193548, | |
| "grad_norm": 0.054103899747133255, | |
| "learning_rate": 5.261648745519714e-06, | |
| "loss": 0.2473, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 26.004387096774195, | |
| "grad_norm": 10.594367980957031, | |
| "learning_rate": 5.24731182795699e-06, | |
| "loss": 0.4263, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 26.00567741935484, | |
| "grad_norm": 36.54576110839844, | |
| "learning_rate": 5.232974910394266e-06, | |
| "loss": 0.5038, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 26.006967741935483, | |
| "grad_norm": 1.2245818376541138, | |
| "learning_rate": 5.218637992831541e-06, | |
| "loss": 0.1015, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 26.00825806451613, | |
| "grad_norm": 0.22598019242286682, | |
| "learning_rate": 5.204301075268817e-06, | |
| "loss": 0.1273, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 26.009548387096775, | |
| "grad_norm": 9.139720916748047, | |
| "learning_rate": 5.189964157706093e-06, | |
| "loss": 0.3239, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 26.01083870967742, | |
| "grad_norm": 32.884239196777344, | |
| "learning_rate": 5.175627240143369e-06, | |
| "loss": 0.6147, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 26.012129032258066, | |
| "grad_norm": 27.457473754882812, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 0.18, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 26.01341935483871, | |
| "grad_norm": 2.117966413497925, | |
| "learning_rate": 5.146953405017922e-06, | |
| "loss": 0.0768, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 26.014709677419354, | |
| "grad_norm": 0.1101021021604538, | |
| "learning_rate": 5.132616487455197e-06, | |
| "loss": 0.1728, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 26.016, | |
| "grad_norm": 0.09035364538431168, | |
| "learning_rate": 5.118279569892473e-06, | |
| "loss": 0.0833, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 26.017290322580646, | |
| "grad_norm": 31.332761764526367, | |
| "learning_rate": 5.10394265232975e-06, | |
| "loss": 0.4058, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 26.01858064516129, | |
| "grad_norm": 0.013916853815317154, | |
| "learning_rate": 5.089605734767026e-06, | |
| "loss": 0.6245, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 26.019870967741934, | |
| "grad_norm": 0.17179369926452637, | |
| "learning_rate": 5.075268817204302e-06, | |
| "loss": 0.1596, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 26.020129032258065, | |
| "eval_accuracy": 0.8222222222222222, | |
| "eval_loss": 0.678537130355835, | |
| "eval_runtime": 12.393, | |
| "eval_samples_per_second": 3.631, | |
| "eval_steps_per_second": 0.726, | |
| "step": 4212 | |
| }, | |
| { | |
| "epoch": 27.001032258064516, | |
| "grad_norm": 0.7308743596076965, | |
| "learning_rate": 5.060931899641578e-06, | |
| "loss": 0.2694, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 27.00232258064516, | |
| "grad_norm": 3.456517219543457, | |
| "learning_rate": 5.0465949820788544e-06, | |
| "loss": 0.1401, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 27.003612903225807, | |
| "grad_norm": 102.73038482666016, | |
| "learning_rate": 5.032258064516129e-06, | |
| "loss": 0.4411, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 27.00490322580645, | |
| "grad_norm": 0.05057060718536377, | |
| "learning_rate": 5.017921146953405e-06, | |
| "loss": 0.274, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 27.006193548387095, | |
| "grad_norm": 128.64810180664062, | |
| "learning_rate": 5.003584229390681e-06, | |
| "loss": 0.2204, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 27.007483870967743, | |
| "grad_norm": 122.1602783203125, | |
| "learning_rate": 4.9892473118279576e-06, | |
| "loss": 0.1641, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 27.008774193548387, | |
| "grad_norm": 55.78334045410156, | |
| "learning_rate": 4.974910394265233e-06, | |
| "loss": 0.4618, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 27.01006451612903, | |
| "grad_norm": 9.603472709655762, | |
| "learning_rate": 4.960573476702509e-06, | |
| "loss": 0.5115, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 27.01135483870968, | |
| "grad_norm": 29.618284225463867, | |
| "learning_rate": 4.946236559139785e-06, | |
| "loss": 0.4004, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 27.012645161290322, | |
| "grad_norm": 0.04179683327674866, | |
| "learning_rate": 4.9318996415770615e-06, | |
| "loss": 0.1981, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 27.013935483870966, | |
| "grad_norm": 54.665374755859375, | |
| "learning_rate": 4.917562724014338e-06, | |
| "loss": 0.4391, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 27.015225806451614, | |
| "grad_norm": 0.1683672070503235, | |
| "learning_rate": 4.903225806451613e-06, | |
| "loss": 0.258, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 27.016516129032258, | |
| "grad_norm": 2.3406288623809814, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 0.2624, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 27.017806451612902, | |
| "grad_norm": 4.3582444190979, | |
| "learning_rate": 4.8745519713261655e-06, | |
| "loss": 0.2358, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 27.01909677419355, | |
| "grad_norm": 37.09184265136719, | |
| "learning_rate": 4.860215053763441e-06, | |
| "loss": 0.1539, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 27.020129032258065, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.7017784714698792, | |
| "eval_runtime": 12.4767, | |
| "eval_samples_per_second": 3.607, | |
| "eval_steps_per_second": 0.721, | |
| "step": 4368 | |
| }, | |
| { | |
| "epoch": 28.000258064516128, | |
| "grad_norm": 0.015434903092682362, | |
| "learning_rate": 4.845878136200717e-06, | |
| "loss": 0.4581, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 28.001548387096776, | |
| "grad_norm": 0.4754023849964142, | |
| "learning_rate": 4.831541218637993e-06, | |
| "loss": 0.1722, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 28.00283870967742, | |
| "grad_norm": 0.06658592075109482, | |
| "learning_rate": 4.8172043010752695e-06, | |
| "loss": 0.0679, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 28.004129032258064, | |
| "grad_norm": 0.14160186052322388, | |
| "learning_rate": 4.802867383512545e-06, | |
| "loss": 0.1153, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 28.00541935483871, | |
| "grad_norm": 0.10064123570919037, | |
| "learning_rate": 4.788530465949821e-06, | |
| "loss": 0.4708, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 28.006709677419355, | |
| "grad_norm": 68.63744354248047, | |
| "learning_rate": 4.774193548387097e-06, | |
| "loss": 0.308, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 28.008, | |
| "grad_norm": 0.04068288952112198, | |
| "learning_rate": 4.7598566308243735e-06, | |
| "loss": 0.0934, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 28.009290322580647, | |
| "grad_norm": 60.843116760253906, | |
| "learning_rate": 4.745519713261649e-06, | |
| "loss": 0.2206, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 28.01058064516129, | |
| "grad_norm": 0.11511409282684326, | |
| "learning_rate": 4.731182795698925e-06, | |
| "loss": 0.1969, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 28.011870967741935, | |
| "grad_norm": 0.05606415867805481, | |
| "learning_rate": 4.716845878136201e-06, | |
| "loss": 0.14, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 28.013161290322582, | |
| "grad_norm": 0.06901150941848755, | |
| "learning_rate": 4.702508960573477e-06, | |
| "loss": 0.5648, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 28.014451612903226, | |
| "grad_norm": 9.159394264221191, | |
| "learning_rate": 4.688172043010753e-06, | |
| "loss": 0.1338, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 28.01574193548387, | |
| "grad_norm": 0.26266661286354065, | |
| "learning_rate": 4.673835125448029e-06, | |
| "loss": 0.347, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 28.017032258064518, | |
| "grad_norm": 0.8105466961860657, | |
| "learning_rate": 4.659498207885305e-06, | |
| "loss": 0.357, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 28.01832258064516, | |
| "grad_norm": 35.22748947143555, | |
| "learning_rate": 4.6451612903225815e-06, | |
| "loss": 0.346, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 28.019612903225806, | |
| "grad_norm": 0.025991085916757584, | |
| "learning_rate": 4.630824372759857e-06, | |
| "loss": 0.0724, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 28.020129032258065, | |
| "eval_accuracy": 0.8666666666666667, | |
| "eval_loss": 0.5836610794067383, | |
| "eval_runtime": 11.0937, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.811, | |
| "step": 4524 | |
| }, | |
| { | |
| "epoch": 29.000774193548388, | |
| "grad_norm": 22.36833953857422, | |
| "learning_rate": 4.616487455197133e-06, | |
| "loss": 0.2497, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 29.002064516129032, | |
| "grad_norm": 10.795868873596191, | |
| "learning_rate": 4.602150537634409e-06, | |
| "loss": 0.1859, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 29.003354838709676, | |
| "grad_norm": 0.03331925719976425, | |
| "learning_rate": 4.587813620071685e-06, | |
| "loss": 0.2687, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 29.004645161290323, | |
| "grad_norm": 0.1016969308257103, | |
| "learning_rate": 4.573476702508961e-06, | |
| "loss": 0.3733, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 29.005935483870967, | |
| "grad_norm": 0.6122087240219116, | |
| "learning_rate": 4.559139784946237e-06, | |
| "loss": 0.2503, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 29.00722580645161, | |
| "grad_norm": 45.6820068359375, | |
| "learning_rate": 4.544802867383513e-06, | |
| "loss": 0.229, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 29.00851612903226, | |
| "grad_norm": 0.10590338706970215, | |
| "learning_rate": 4.530465949820789e-06, | |
| "loss": 0.4161, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 29.009806451612903, | |
| "grad_norm": 0.6490383148193359, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 0.2941, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 29.011096774193547, | |
| "grad_norm": 0.4290194511413574, | |
| "learning_rate": 4.501792114695341e-06, | |
| "loss": 0.4443, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 29.012387096774194, | |
| "grad_norm": 0.13596759736537933, | |
| "learning_rate": 4.487455197132617e-06, | |
| "loss": 0.0077, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 29.01367741935484, | |
| "grad_norm": 59.526424407958984, | |
| "learning_rate": 4.473118279569893e-06, | |
| "loss": 0.3228, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 29.014967741935482, | |
| "grad_norm": 27.774105072021484, | |
| "learning_rate": 4.458781362007169e-06, | |
| "loss": 0.1542, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 29.01625806451613, | |
| "grad_norm": 0.11487451940774918, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.1157, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 29.017548387096774, | |
| "grad_norm": 16.48320960998535, | |
| "learning_rate": 4.43010752688172e-06, | |
| "loss": 0.5424, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 29.018838709677418, | |
| "grad_norm": 0.364417165517807, | |
| "learning_rate": 4.4157706093189966e-06, | |
| "loss": 0.1977, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 29.020129032258065, | |
| "grad_norm": 3.548923969268799, | |
| "learning_rate": 4.401433691756273e-06, | |
| "loss": 0.1998, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 29.020129032258065, | |
| "eval_accuracy": 0.8222222222222222, | |
| "eval_loss": 0.9271312355995178, | |
| "eval_runtime": 11.1996, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.804, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 30.001290322580644, | |
| "grad_norm": 72.71692657470703, | |
| "learning_rate": 4.387096774193549e-06, | |
| "loss": 0.4276, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 30.00258064516129, | |
| "grad_norm": 1.4667835235595703, | |
| "learning_rate": 4.372759856630825e-06, | |
| "loss": 0.0907, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 30.003870967741936, | |
| "grad_norm": 0.09754376858472824, | |
| "learning_rate": 4.358422939068101e-06, | |
| "loss": 0.2188, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 30.00516129032258, | |
| "grad_norm": 0.08105052262544632, | |
| "learning_rate": 4.344086021505377e-06, | |
| "loss": 0.009, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 30.006451612903227, | |
| "grad_norm": 160.77835083007812, | |
| "learning_rate": 4.329749103942653e-06, | |
| "loss": 0.2415, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 30.00774193548387, | |
| "grad_norm": 0.08130495995283127, | |
| "learning_rate": 4.315412186379928e-06, | |
| "loss": 0.0029, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 30.009032258064515, | |
| "grad_norm": 44.8267936706543, | |
| "learning_rate": 4.3010752688172045e-06, | |
| "loss": 0.4865, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 30.010322580645163, | |
| "grad_norm": 0.062065478414297104, | |
| "learning_rate": 4.286738351254481e-06, | |
| "loss": 0.1237, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 30.011612903225807, | |
| "grad_norm": 94.81366729736328, | |
| "learning_rate": 4.272401433691757e-06, | |
| "loss": 0.1576, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 30.01290322580645, | |
| "grad_norm": 0.23950950801372528, | |
| "learning_rate": 4.258064516129032e-06, | |
| "loss": 0.3001, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 30.014193548387098, | |
| "grad_norm": 1.1783959865570068, | |
| "learning_rate": 4.2437275985663085e-06, | |
| "loss": 0.0383, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 30.015483870967742, | |
| "grad_norm": 74.16380310058594, | |
| "learning_rate": 4.229390681003585e-06, | |
| "loss": 0.2856, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 30.016774193548386, | |
| "grad_norm": 0.0347282700240612, | |
| "learning_rate": 4.215053763440861e-06, | |
| "loss": 0.5245, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 30.018064516129034, | |
| "grad_norm": 2.53531551361084, | |
| "learning_rate": 4.200716845878136e-06, | |
| "loss": 0.0145, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 30.019354838709678, | |
| "grad_norm": 25.381006240844727, | |
| "learning_rate": 4.1863799283154125e-06, | |
| "loss": 0.2286, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 30.020129032258065, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 1.7056816816329956, | |
| "eval_runtime": 11.1225, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.809, | |
| "step": 4836 | |
| }, | |
| { | |
| "epoch": 31.00051612903226, | |
| "grad_norm": 25.31055450439453, | |
| "learning_rate": 4.172043010752688e-06, | |
| "loss": 0.2005, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 31.001806451612904, | |
| "grad_norm": 0.07410001754760742, | |
| "learning_rate": 4.157706093189964e-06, | |
| "loss": 0.1723, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 31.003096774193548, | |
| "grad_norm": 81.27484893798828, | |
| "learning_rate": 4.14336917562724e-06, | |
| "loss": 0.1116, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 31.004387096774195, | |
| "grad_norm": 139.99208068847656, | |
| "learning_rate": 4.1290322580645165e-06, | |
| "loss": 0.0662, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 31.00567741935484, | |
| "grad_norm": 236.75772094726562, | |
| "learning_rate": 4.114695340501793e-06, | |
| "loss": 0.0752, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 31.006967741935483, | |
| "grad_norm": 0.2467179298400879, | |
| "learning_rate": 4.100358422939069e-06, | |
| "loss": 0.3826, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 31.00825806451613, | |
| "grad_norm": 0.08100493252277374, | |
| "learning_rate": 4.086021505376344e-06, | |
| "loss": 0.1579, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 31.009548387096775, | |
| "grad_norm": 15.623895645141602, | |
| "learning_rate": 4.0716845878136205e-06, | |
| "loss": 0.3839, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 31.01083870967742, | |
| "grad_norm": 2.5538089275360107, | |
| "learning_rate": 4.057347670250897e-06, | |
| "loss": 0.2829, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 31.012129032258066, | |
| "grad_norm": 0.021700290963053703, | |
| "learning_rate": 4.043010752688172e-06, | |
| "loss": 0.2271, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 31.01341935483871, | |
| "grad_norm": 95.48535919189453, | |
| "learning_rate": 4.028673835125448e-06, | |
| "loss": 0.2686, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 31.014709677419354, | |
| "grad_norm": 0.05281487852334976, | |
| "learning_rate": 4.0143369175627245e-06, | |
| "loss": 0.421, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 31.016, | |
| "grad_norm": 0.03669271618127823, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.707, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 31.017290322580646, | |
| "grad_norm": 122.49234771728516, | |
| "learning_rate": 3.985663082437276e-06, | |
| "loss": 0.3185, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 31.01858064516129, | |
| "grad_norm": 27.71544075012207, | |
| "learning_rate": 3.971326164874552e-06, | |
| "loss": 0.2691, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 31.019870967741934, | |
| "grad_norm": 129.3389892578125, | |
| "learning_rate": 3.9569892473118284e-06, | |
| "loss": 0.192, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 31.020129032258065, | |
| "eval_accuracy": 0.8222222222222222, | |
| "eval_loss": 0.7721913456916809, | |
| "eval_runtime": 14.2262, | |
| "eval_samples_per_second": 3.163, | |
| "eval_steps_per_second": 0.633, | |
| "step": 4992 | |
| }, | |
| { | |
| "epoch": 32.00103225806452, | |
| "grad_norm": 84.79925537109375, | |
| "learning_rate": 3.942652329749105e-06, | |
| "loss": 0.1914, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 32.00232258064516, | |
| "grad_norm": 0.014622915536165237, | |
| "learning_rate": 3.92831541218638e-06, | |
| "loss": 0.0137, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 32.00361290322581, | |
| "grad_norm": 0.06631313264369965, | |
| "learning_rate": 3.913978494623656e-06, | |
| "loss": 0.3702, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 32.00490322580645, | |
| "grad_norm": 0.016402188688516617, | |
| "learning_rate": 3.8996415770609324e-06, | |
| "loss": 0.2934, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 32.006193548387095, | |
| "grad_norm": 0.13503485918045044, | |
| "learning_rate": 3.885304659498208e-06, | |
| "loss": 0.3808, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 32.00748387096774, | |
| "grad_norm": 0.7589678764343262, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 0.1125, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 32.00877419354839, | |
| "grad_norm": 186.81280517578125, | |
| "learning_rate": 3.85663082437276e-06, | |
| "loss": 0.2812, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 32.010064516129034, | |
| "grad_norm": 0.06653529405593872, | |
| "learning_rate": 3.842293906810036e-06, | |
| "loss": 0.1013, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 32.01135483870968, | |
| "grad_norm": 0.0339493602514267, | |
| "learning_rate": 3.827956989247313e-06, | |
| "loss": 0.2587, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 32.01264516129032, | |
| "grad_norm": 0.053266000002622604, | |
| "learning_rate": 3.8136200716845884e-06, | |
| "loss": 0.1226, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 32.013935483870966, | |
| "grad_norm": 163.95733642578125, | |
| "learning_rate": 3.7992831541218638e-06, | |
| "loss": 0.0879, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 32.01522580645161, | |
| "grad_norm": 0.04351264238357544, | |
| "learning_rate": 3.78494623655914e-06, | |
| "loss": 0.0121, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 32.01651612903226, | |
| "grad_norm": 164.31834411621094, | |
| "learning_rate": 3.770609318996416e-06, | |
| "loss": 0.0753, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 32.017806451612906, | |
| "grad_norm": 142.49822998046875, | |
| "learning_rate": 3.756272401433692e-06, | |
| "loss": 0.1366, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 32.01909677419355, | |
| "grad_norm": 14.688565254211426, | |
| "learning_rate": 3.741935483870968e-06, | |
| "loss": 0.1149, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 32.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.9803315997123718, | |
| "eval_runtime": 11.786, | |
| "eval_samples_per_second": 3.818, | |
| "eval_steps_per_second": 0.764, | |
| "step": 5148 | |
| }, | |
| { | |
| "epoch": 33.00025806451613, | |
| "grad_norm": 0.11231986433267593, | |
| "learning_rate": 3.7275985663082444e-06, | |
| "loss": 0.0025, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 33.001548387096776, | |
| "grad_norm": 0.01792258769273758, | |
| "learning_rate": 3.7132616487455197e-06, | |
| "loss": 0.111, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 33.00283870967742, | |
| "grad_norm": 274.67547607421875, | |
| "learning_rate": 3.698924731182796e-06, | |
| "loss": 0.3681, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 33.004129032258064, | |
| "grad_norm": 169.23826599121094, | |
| "learning_rate": 3.6845878136200717e-06, | |
| "loss": 0.275, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 33.00541935483871, | |
| "grad_norm": 6.449133396148682, | |
| "learning_rate": 3.670250896057348e-06, | |
| "loss": 0.3115, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 33.00670967741935, | |
| "grad_norm": 0.006865628529340029, | |
| "learning_rate": 3.655913978494624e-06, | |
| "loss": 0.2667, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 33.008, | |
| "grad_norm": 39.192447662353516, | |
| "learning_rate": 3.6415770609319e-06, | |
| "loss": 0.2327, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 33.00929032258065, | |
| "grad_norm": 0.21899880468845367, | |
| "learning_rate": 3.627240143369176e-06, | |
| "loss": 0.3446, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 33.01058064516129, | |
| "grad_norm": 0.03009025938808918, | |
| "learning_rate": 3.6129032258064515e-06, | |
| "loss": 0.6423, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 33.011870967741935, | |
| "grad_norm": 0.21091091632843018, | |
| "learning_rate": 3.5985663082437277e-06, | |
| "loss": 0.1147, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 33.01316129032258, | |
| "grad_norm": 0.05113637074828148, | |
| "learning_rate": 3.584229390681004e-06, | |
| "loss": 0.2681, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 33.01445161290322, | |
| "grad_norm": 36.46299362182617, | |
| "learning_rate": 3.5698924731182797e-06, | |
| "loss": 0.2061, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 33.015741935483874, | |
| "grad_norm": 0.16068463027477264, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 0.1026, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 33.01703225806452, | |
| "grad_norm": 0.040871601551771164, | |
| "learning_rate": 3.541218637992832e-06, | |
| "loss": 0.0533, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 33.01832258064516, | |
| "grad_norm": 0.10143303126096725, | |
| "learning_rate": 3.5268817204301075e-06, | |
| "loss": 0.0612, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 33.019612903225806, | |
| "grad_norm": 72.79322052001953, | |
| "learning_rate": 3.5125448028673837e-06, | |
| "loss": 0.1723, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 33.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.9822748899459839, | |
| "eval_runtime": 11.7583, | |
| "eval_samples_per_second": 3.827, | |
| "eval_steps_per_second": 0.765, | |
| "step": 5304 | |
| }, | |
| { | |
| "epoch": 34.00077419354839, | |
| "grad_norm": 0.8634727597236633, | |
| "learning_rate": 3.49820788530466e-06, | |
| "loss": 0.1172, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 34.00206451612903, | |
| "grad_norm": 40.82673645019531, | |
| "learning_rate": 3.4838709677419357e-06, | |
| "loss": 0.2339, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 34.003354838709676, | |
| "grad_norm": 183.44642639160156, | |
| "learning_rate": 3.469534050179212e-06, | |
| "loss": 0.0282, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 34.00464516129032, | |
| "grad_norm": 54.94170379638672, | |
| "learning_rate": 3.455197132616488e-06, | |
| "loss": 0.2871, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 34.00593548387097, | |
| "grad_norm": 0.04819169640541077, | |
| "learning_rate": 3.440860215053764e-06, | |
| "loss": 0.0932, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 34.007225806451615, | |
| "grad_norm": 0.02546178176999092, | |
| "learning_rate": 3.4265232974910397e-06, | |
| "loss": 0.1141, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 34.00851612903226, | |
| "grad_norm": 0.23394398391246796, | |
| "learning_rate": 3.4121863799283155e-06, | |
| "loss": 0.1246, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 34.0098064516129, | |
| "grad_norm": 0.01693810150027275, | |
| "learning_rate": 3.3978494623655917e-06, | |
| "loss": 0.092, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 34.01109677419355, | |
| "grad_norm": 0.020266568288207054, | |
| "learning_rate": 3.383512544802868e-06, | |
| "loss": 0.1385, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 34.01238709677419, | |
| "grad_norm": 6.0975022315979, | |
| "learning_rate": 3.3691756272401437e-06, | |
| "loss": 0.2113, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 34.01367741935484, | |
| "grad_norm": 0.029755596071481705, | |
| "learning_rate": 3.35483870967742e-06, | |
| "loss": 0.1239, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 34.014967741935486, | |
| "grad_norm": 4.003732681274414, | |
| "learning_rate": 3.3405017921146952e-06, | |
| "loss": 0.2932, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 34.01625806451613, | |
| "grad_norm": 0.44571009278297424, | |
| "learning_rate": 3.3261648745519714e-06, | |
| "loss": 0.1222, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 34.017548387096774, | |
| "grad_norm": 118.21929931640625, | |
| "learning_rate": 3.3118279569892476e-06, | |
| "loss": 0.2565, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 34.01883870967742, | |
| "grad_norm": 0.022239074110984802, | |
| "learning_rate": 3.2974910394265234e-06, | |
| "loss": 0.1699, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 34.02012903225806, | |
| "grad_norm": 3.5563700199127197, | |
| "learning_rate": 3.2831541218637996e-06, | |
| "loss": 0.0624, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 34.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.1380023956298828, | |
| "eval_runtime": 13.1323, | |
| "eval_samples_per_second": 3.427, | |
| "eval_steps_per_second": 0.685, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 35.001290322580644, | |
| "grad_norm": 0.035845834761857986, | |
| "learning_rate": 3.268817204301076e-06, | |
| "loss": 0.1927, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 35.00258064516129, | |
| "grad_norm": 144.52903747558594, | |
| "learning_rate": 3.254480286738351e-06, | |
| "loss": 0.1464, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 35.00387096774193, | |
| "grad_norm": 0.028881637379527092, | |
| "learning_rate": 3.2401433691756274e-06, | |
| "loss": 0.009, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 35.00516129032258, | |
| "grad_norm": 0.4867962598800659, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.0611, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 35.00645161290323, | |
| "grad_norm": 9.892499923706055, | |
| "learning_rate": 3.2114695340501794e-06, | |
| "loss": 0.2048, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 35.00774193548387, | |
| "grad_norm": 0.03587242588400841, | |
| "learning_rate": 3.1971326164874556e-06, | |
| "loss": 0.0087, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 35.009032258064515, | |
| "grad_norm": 0.3965907692909241, | |
| "learning_rate": 3.182795698924732e-06, | |
| "loss": 0.3048, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 35.01032258064516, | |
| "grad_norm": 0.026851637288928032, | |
| "learning_rate": 3.1684587813620076e-06, | |
| "loss": 0.1183, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 35.0116129032258, | |
| "grad_norm": 172.9999542236328, | |
| "learning_rate": 3.1541218637992834e-06, | |
| "loss": 0.4033, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 35.012903225806454, | |
| "grad_norm": 0.010580616071820259, | |
| "learning_rate": 3.139784946236559e-06, | |
| "loss": 0.0341, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 35.0141935483871, | |
| "grad_norm": 0.18182462453842163, | |
| "learning_rate": 3.1254480286738354e-06, | |
| "loss": 0.5813, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 35.01548387096774, | |
| "grad_norm": 0.018430812284350395, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 0.0615, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 35.016774193548386, | |
| "grad_norm": 0.27715563774108887, | |
| "learning_rate": 3.0967741935483874e-06, | |
| "loss": 0.1651, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 35.01806451612903, | |
| "grad_norm": 0.05246179178357124, | |
| "learning_rate": 3.0824372759856636e-06, | |
| "loss": 0.3838, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 35.019354838709674, | |
| "grad_norm": 48.31368637084961, | |
| "learning_rate": 3.068100358422939e-06, | |
| "loss": 0.3264, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 35.02012903225806, | |
| "eval_accuracy": 0.8222222222222222, | |
| "eval_loss": 0.9570736289024353, | |
| "eval_runtime": 16.3873, | |
| "eval_samples_per_second": 2.746, | |
| "eval_steps_per_second": 0.549, | |
| "step": 5616 | |
| }, | |
| { | |
| "epoch": 36.000516129032256, | |
| "grad_norm": 0.9066537618637085, | |
| "learning_rate": 3.053763440860215e-06, | |
| "loss": 0.229, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 36.0018064516129, | |
| "grad_norm": 0.897160530090332, | |
| "learning_rate": 3.0394265232974914e-06, | |
| "loss": 0.174, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 36.00309677419355, | |
| "grad_norm": 153.85252380371094, | |
| "learning_rate": 3.025089605734767e-06, | |
| "loss": 0.0998, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 36.004387096774195, | |
| "grad_norm": 0.013687233440577984, | |
| "learning_rate": 3.0107526881720433e-06, | |
| "loss": 0.1984, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 36.00567741935484, | |
| "grad_norm": 0.03366924449801445, | |
| "learning_rate": 2.9964157706093196e-06, | |
| "loss": 0.0131, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 36.00696774193548, | |
| "grad_norm": 0.02471003495156765, | |
| "learning_rate": 2.9820788530465953e-06, | |
| "loss": 0.1701, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 36.00825806451613, | |
| "grad_norm": 0.15641465783119202, | |
| "learning_rate": 2.967741935483871e-06, | |
| "loss": 0.3584, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 36.00954838709677, | |
| "grad_norm": 4.849608898162842, | |
| "learning_rate": 2.953405017921147e-06, | |
| "loss": 0.4201, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 36.01083870967742, | |
| "grad_norm": 58.6971321105957, | |
| "learning_rate": 2.939068100358423e-06, | |
| "loss": 0.1367, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 36.012129032258066, | |
| "grad_norm": 0.08391012996435165, | |
| "learning_rate": 2.9247311827956993e-06, | |
| "loss": 0.1407, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 36.01341935483871, | |
| "grad_norm": 0.44904467463493347, | |
| "learning_rate": 2.910394265232975e-06, | |
| "loss": 0.0042, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 36.014709677419354, | |
| "grad_norm": 0.01743984967470169, | |
| "learning_rate": 2.8960573476702513e-06, | |
| "loss": 0.2206, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 36.016, | |
| "grad_norm": 3.4085276126861572, | |
| "learning_rate": 2.881720430107527e-06, | |
| "loss": 0.6049, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 36.01729032258064, | |
| "grad_norm": 0.0050637186504900455, | |
| "learning_rate": 2.867383512544803e-06, | |
| "loss": 0.0878, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 36.01858064516129, | |
| "grad_norm": 0.02858421951532364, | |
| "learning_rate": 2.853046594982079e-06, | |
| "loss": 0.2625, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 36.01987096774194, | |
| "grad_norm": 212.74087524414062, | |
| "learning_rate": 2.8387096774193553e-06, | |
| "loss": 0.0631, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 36.02012903225806, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 1.4243476390838623, | |
| "eval_runtime": 12.515, | |
| "eval_samples_per_second": 3.596, | |
| "eval_steps_per_second": 0.719, | |
| "step": 5772 | |
| }, | |
| { | |
| "epoch": 37.00103225806452, | |
| "grad_norm": 0.14299911260604858, | |
| "learning_rate": 2.824372759856631e-06, | |
| "loss": 0.0316, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 37.00232258064516, | |
| "grad_norm": 0.05638037621974945, | |
| "learning_rate": 2.8100358422939073e-06, | |
| "loss": 0.3119, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 37.00361290322581, | |
| "grad_norm": 0.014610113576054573, | |
| "learning_rate": 2.7956989247311827e-06, | |
| "loss": 0.1079, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 37.00490322580645, | |
| "grad_norm": 32.98031234741211, | |
| "learning_rate": 2.781362007168459e-06, | |
| "loss": 0.1343, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 37.006193548387095, | |
| "grad_norm": 0.011130129918456078, | |
| "learning_rate": 2.767025089605735e-06, | |
| "loss": 0.0007, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 37.00748387096774, | |
| "grad_norm": 153.53753662109375, | |
| "learning_rate": 2.752688172043011e-06, | |
| "loss": 0.3429, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 37.00877419354839, | |
| "grad_norm": 49.70185852050781, | |
| "learning_rate": 2.738351254480287e-06, | |
| "loss": 0.311, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 37.010064516129034, | |
| "grad_norm": 17.71489143371582, | |
| "learning_rate": 2.7240143369175633e-06, | |
| "loss": 0.1211, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 37.01135483870968, | |
| "grad_norm": 0.038397543132305145, | |
| "learning_rate": 2.709677419354839e-06, | |
| "loss": 0.1386, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 37.01264516129032, | |
| "grad_norm": 0.3241770565509796, | |
| "learning_rate": 2.695340501792115e-06, | |
| "loss": 0.252, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 37.013935483870966, | |
| "grad_norm": 44.44465255737305, | |
| "learning_rate": 2.6810035842293906e-06, | |
| "loss": 0.3877, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 37.01522580645161, | |
| "grad_norm": 7.57636833190918, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.1813, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 37.01651612903226, | |
| "grad_norm": 88.82908630371094, | |
| "learning_rate": 2.652329749103943e-06, | |
| "loss": 0.1592, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 37.017806451612906, | |
| "grad_norm": 158.9471893310547, | |
| "learning_rate": 2.637992831541219e-06, | |
| "loss": 0.0847, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 37.01909677419355, | |
| "grad_norm": 0.9167165160179138, | |
| "learning_rate": 2.623655913978495e-06, | |
| "loss": 0.1043, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 37.02012903225806, | |
| "eval_accuracy": 0.8444444444444444, | |
| "eval_loss": 0.8813353776931763, | |
| "eval_runtime": 11.1988, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.804, | |
| "step": 5928 | |
| }, | |
| { | |
| "epoch": 38.00025806451613, | |
| "grad_norm": 82.5613021850586, | |
| "learning_rate": 2.6093189964157704e-06, | |
| "loss": 0.0705, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 38.001548387096776, | |
| "grad_norm": 7.3600969314575195, | |
| "learning_rate": 2.5949820788530466e-06, | |
| "loss": 0.0927, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 38.00283870967742, | |
| "grad_norm": 18.823772430419922, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 0.0693, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 38.004129032258064, | |
| "grad_norm": 0.07058124989271164, | |
| "learning_rate": 2.5663082437275986e-06, | |
| "loss": 0.2642, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 38.00541935483871, | |
| "grad_norm": 0.3029346466064453, | |
| "learning_rate": 2.551971326164875e-06, | |
| "loss": 0.0295, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 38.00670967741935, | |
| "grad_norm": 25.69213104248047, | |
| "learning_rate": 2.537634408602151e-06, | |
| "loss": 0.2919, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 38.008, | |
| "grad_norm": 32.14357376098633, | |
| "learning_rate": 2.5232974910394272e-06, | |
| "loss": 0.2199, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 38.00929032258065, | |
| "grad_norm": 1.4096826314926147, | |
| "learning_rate": 2.5089605734767026e-06, | |
| "loss": 0.1615, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 38.01058064516129, | |
| "grad_norm": 0.020581666380167007, | |
| "learning_rate": 2.4946236559139788e-06, | |
| "loss": 0.0028, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 38.011870967741935, | |
| "grad_norm": 0.001719414140097797, | |
| "learning_rate": 2.4802867383512546e-06, | |
| "loss": 0.1343, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 38.01316129032258, | |
| "grad_norm": 68.00126647949219, | |
| "learning_rate": 2.4659498207885308e-06, | |
| "loss": 0.2686, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 38.01445161290322, | |
| "grad_norm": 12.022477149963379, | |
| "learning_rate": 2.4516129032258066e-06, | |
| "loss": 0.1364, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 38.015741935483874, | |
| "grad_norm": 0.012526830658316612, | |
| "learning_rate": 2.4372759856630828e-06, | |
| "loss": 0.0016, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 38.01703225806452, | |
| "grad_norm": 172.25982666015625, | |
| "learning_rate": 2.4229390681003586e-06, | |
| "loss": 0.2391, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 38.01832258064516, | |
| "grad_norm": 0.1778022050857544, | |
| "learning_rate": 2.4086021505376348e-06, | |
| "loss": 0.1831, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 38.019612903225806, | |
| "grad_norm": 51.6696891784668, | |
| "learning_rate": 2.3942652329749105e-06, | |
| "loss": 0.1337, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 38.02012903225806, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.617709994316101, | |
| "eval_runtime": 11.5473, | |
| "eval_samples_per_second": 3.897, | |
| "eval_steps_per_second": 0.779, | |
| "step": 6084 | |
| }, | |
| { | |
| "epoch": 39.00077419354839, | |
| "grad_norm": 0.3095431625843048, | |
| "learning_rate": 2.3799283154121868e-06, | |
| "loss": 0.1136, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 39.00206451612903, | |
| "grad_norm": 0.7548017501831055, | |
| "learning_rate": 2.3655913978494625e-06, | |
| "loss": 0.028, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 39.003354838709676, | |
| "grad_norm": 227.6057586669922, | |
| "learning_rate": 2.3512544802867383e-06, | |
| "loss": 0.0666, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 39.00464516129032, | |
| "grad_norm": 0.030988622456789017, | |
| "learning_rate": 2.3369175627240145e-06, | |
| "loss": 0.0552, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 39.00593548387097, | |
| "grad_norm": 78.96713256835938, | |
| "learning_rate": 2.3225806451612907e-06, | |
| "loss": 0.2064, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 39.007225806451615, | |
| "grad_norm": 0.07105328142642975, | |
| "learning_rate": 2.3082437275985665e-06, | |
| "loss": 0.1224, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 39.00851612903226, | |
| "grad_norm": 0.031410399824380875, | |
| "learning_rate": 2.2939068100358423e-06, | |
| "loss": 0.0425, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 39.0098064516129, | |
| "grad_norm": 0.37001341581344604, | |
| "learning_rate": 2.2795698924731185e-06, | |
| "loss": 0.1499, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 39.01109677419355, | |
| "grad_norm": 0.38967448472976685, | |
| "learning_rate": 2.2652329749103943e-06, | |
| "loss": 0.0029, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 39.01238709677419, | |
| "grad_norm": 11.713730812072754, | |
| "learning_rate": 2.2508960573476705e-06, | |
| "loss": 0.3785, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 39.01367741935484, | |
| "grad_norm": 182.08062744140625, | |
| "learning_rate": 2.2365591397849463e-06, | |
| "loss": 0.3781, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 39.014967741935486, | |
| "grad_norm": 0.0234206672757864, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.0018, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 39.01625806451613, | |
| "grad_norm": 114.47459411621094, | |
| "learning_rate": 2.2078853046594983e-06, | |
| "loss": 0.1431, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 39.017548387096774, | |
| "grad_norm": 0.33611953258514404, | |
| "learning_rate": 2.1935483870967745e-06, | |
| "loss": 0.0072, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 39.01883870967742, | |
| "grad_norm": 0.006744945887476206, | |
| "learning_rate": 2.1792114695340507e-06, | |
| "loss": 0.1151, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 39.02012903225806, | |
| "grad_norm": 0.40540820360183716, | |
| "learning_rate": 2.1648745519713265e-06, | |
| "loss": 0.271, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 39.02012903225806, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 1.1880711317062378, | |
| "eval_runtime": 12.1531, | |
| "eval_samples_per_second": 3.703, | |
| "eval_steps_per_second": 0.741, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 40.001290322580644, | |
| "grad_norm": 0.022850753739476204, | |
| "learning_rate": 2.1505376344086023e-06, | |
| "loss": 0.0875, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 40.00258064516129, | |
| "grad_norm": 4.465057373046875, | |
| "learning_rate": 2.1362007168458785e-06, | |
| "loss": 0.1549, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 40.00387096774193, | |
| "grad_norm": 4.138294696807861, | |
| "learning_rate": 2.1218637992831543e-06, | |
| "loss": 0.1441, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 40.00516129032258, | |
| "grad_norm": 0.024796660989522934, | |
| "learning_rate": 2.1075268817204305e-06, | |
| "loss": 0.0527, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 40.00645161290323, | |
| "grad_norm": 153.99562072753906, | |
| "learning_rate": 2.0931899641577063e-06, | |
| "loss": 0.1709, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 40.00774193548387, | |
| "grad_norm": 0.013526755385100842, | |
| "learning_rate": 2.078853046594982e-06, | |
| "loss": 0.1901, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 40.009032258064515, | |
| "grad_norm": 21.57200050354004, | |
| "learning_rate": 2.0645161290322582e-06, | |
| "loss": 0.188, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 40.01032258064516, | |
| "grad_norm": 0.3606230318546295, | |
| "learning_rate": 2.0501792114695345e-06, | |
| "loss": 0.0031, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 40.0116129032258, | |
| "grad_norm": 0.007888257503509521, | |
| "learning_rate": 2.0358422939068102e-06, | |
| "loss": 0.0067, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 40.012903225806454, | |
| "grad_norm": 0.0844966471195221, | |
| "learning_rate": 2.021505376344086e-06, | |
| "loss": 0.1609, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 40.0141935483871, | |
| "grad_norm": 0.013470242731273174, | |
| "learning_rate": 2.0071684587813622e-06, | |
| "loss": 0.0686, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 40.01548387096774, | |
| "grad_norm": 0.0016926495591178536, | |
| "learning_rate": 1.992831541218638e-06, | |
| "loss": 0.0821, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 40.016774193548386, | |
| "grad_norm": 0.031637683510780334, | |
| "learning_rate": 1.9784946236559142e-06, | |
| "loss": 0.0009, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 40.01806451612903, | |
| "grad_norm": 2.9717295169830322, | |
| "learning_rate": 1.96415770609319e-06, | |
| "loss": 0.2078, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 40.019354838709674, | |
| "grad_norm": 0.03920386731624603, | |
| "learning_rate": 1.9498207885304662e-06, | |
| "loss": 0.3917, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 40.02012903225806, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 1.2264965772628784, | |
| "eval_runtime": 12.3989, | |
| "eval_samples_per_second": 3.629, | |
| "eval_steps_per_second": 0.726, | |
| "step": 6396 | |
| }, | |
| { | |
| "epoch": 41.000516129032256, | |
| "grad_norm": 67.1982421875, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 0.1175, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 41.0018064516129, | |
| "grad_norm": 0.11867904663085938, | |
| "learning_rate": 1.921146953405018e-06, | |
| "loss": 0.0074, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 41.00309677419355, | |
| "grad_norm": 0.13042046129703522, | |
| "learning_rate": 1.9068100358422942e-06, | |
| "loss": 0.1299, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 41.004387096774195, | |
| "grad_norm": 0.012479264289140701, | |
| "learning_rate": 1.89247311827957e-06, | |
| "loss": 0.1325, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 41.00567741935484, | |
| "grad_norm": 0.026974236592650414, | |
| "learning_rate": 1.878136200716846e-06, | |
| "loss": 0.0036, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 41.00696774193548, | |
| "grad_norm": 0.053469158709049225, | |
| "learning_rate": 1.8637992831541222e-06, | |
| "loss": 0.1131, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 41.00825806451613, | |
| "grad_norm": 287.1971130371094, | |
| "learning_rate": 1.849462365591398e-06, | |
| "loss": 0.0701, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 41.00954838709677, | |
| "grad_norm": 0.022596845403313637, | |
| "learning_rate": 1.835125448028674e-06, | |
| "loss": 0.2626, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 41.01083870967742, | |
| "grad_norm": 0.021237093955278397, | |
| "learning_rate": 1.82078853046595e-06, | |
| "loss": 0.0135, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 41.012129032258066, | |
| "grad_norm": 0.005842030048370361, | |
| "learning_rate": 1.8064516129032258e-06, | |
| "loss": 0.0867, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 41.01341935483871, | |
| "grad_norm": 1.0644536018371582, | |
| "learning_rate": 1.792114695340502e-06, | |
| "loss": 0.0682, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 41.014709677419354, | |
| "grad_norm": 0.014805124141275883, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 0.2025, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 41.016, | |
| "grad_norm": 0.02180282399058342, | |
| "learning_rate": 1.7634408602150537e-06, | |
| "loss": 0.1783, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 41.01729032258064, | |
| "grad_norm": 1.1481382846832275, | |
| "learning_rate": 1.74910394265233e-06, | |
| "loss": 0.1882, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 41.01858064516129, | |
| "grad_norm": 0.42159610986709595, | |
| "learning_rate": 1.734767025089606e-06, | |
| "loss": 0.0795, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 41.01987096774194, | |
| "grad_norm": 4.157188415527344, | |
| "learning_rate": 1.720430107526882e-06, | |
| "loss": 0.3894, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 41.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.1420141458511353, | |
| "eval_runtime": 11.134, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.808, | |
| "step": 6552 | |
| }, | |
| { | |
| "epoch": 42.00103225806452, | |
| "grad_norm": 0.11696040630340576, | |
| "learning_rate": 1.7060931899641577e-06, | |
| "loss": 0.2078, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 42.00232258064516, | |
| "grad_norm": 0.047910962253808975, | |
| "learning_rate": 1.691756272401434e-06, | |
| "loss": 0.1006, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 42.00361290322581, | |
| "grad_norm": 0.5936770439147949, | |
| "learning_rate": 1.67741935483871e-06, | |
| "loss": 0.0044, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 42.00490322580645, | |
| "grad_norm": 0.041977811604738235, | |
| "learning_rate": 1.6630824372759857e-06, | |
| "loss": 0.233, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 42.006193548387095, | |
| "grad_norm": 0.015213453210890293, | |
| "learning_rate": 1.6487455197132617e-06, | |
| "loss": 0.0006, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 42.00748387096774, | |
| "grad_norm": 232.71258544921875, | |
| "learning_rate": 1.634408602150538e-06, | |
| "loss": 0.2646, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 42.00877419354839, | |
| "grad_norm": 0.07088331133127213, | |
| "learning_rate": 1.6200716845878137e-06, | |
| "loss": 0.0009, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 42.010064516129034, | |
| "grad_norm": 0.04214366897940636, | |
| "learning_rate": 1.6057347670250897e-06, | |
| "loss": 0.0004, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 42.01135483870968, | |
| "grad_norm": 0.11403486132621765, | |
| "learning_rate": 1.591397849462366e-06, | |
| "loss": 0.103, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 42.01264516129032, | |
| "grad_norm": 0.018779393285512924, | |
| "learning_rate": 1.5770609318996417e-06, | |
| "loss": 0.1607, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 42.013935483870966, | |
| "grad_norm": 0.06166236102581024, | |
| "learning_rate": 1.5627240143369177e-06, | |
| "loss": 0.1148, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 42.01522580645161, | |
| "grad_norm": 0.7400681376457214, | |
| "learning_rate": 1.5483870967741937e-06, | |
| "loss": 0.1983, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 42.01651612903226, | |
| "grad_norm": 274.43035888671875, | |
| "learning_rate": 1.5340501792114695e-06, | |
| "loss": 0.2049, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 42.017806451612906, | |
| "grad_norm": 0.05334322154521942, | |
| "learning_rate": 1.5197132616487457e-06, | |
| "loss": 0.0006, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 42.01909677419355, | |
| "grad_norm": 29.896432876586914, | |
| "learning_rate": 1.5053763440860217e-06, | |
| "loss": 0.0459, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 42.02012903225806, | |
| "eval_accuracy": 0.8222222222222222, | |
| "eval_loss": 1.0436769723892212, | |
| "eval_runtime": 11.58, | |
| "eval_samples_per_second": 3.886, | |
| "eval_steps_per_second": 0.777, | |
| "step": 6708 | |
| }, | |
| { | |
| "epoch": 43.00025806451613, | |
| "grad_norm": 0.10032159090042114, | |
| "learning_rate": 1.4910394265232977e-06, | |
| "loss": 0.0611, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 43.001548387096776, | |
| "grad_norm": 34.7204704284668, | |
| "learning_rate": 1.4767025089605735e-06, | |
| "loss": 0.1306, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 43.00283870967742, | |
| "grad_norm": 0.20480941236019135, | |
| "learning_rate": 1.4623655913978497e-06, | |
| "loss": 0.0008, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 43.004129032258064, | |
| "grad_norm": 0.015655193477869034, | |
| "learning_rate": 1.4480286738351257e-06, | |
| "loss": 0.1466, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 43.00541935483871, | |
| "grad_norm": 0.014934848062694073, | |
| "learning_rate": 1.4336917562724014e-06, | |
| "loss": 0.0424, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 43.00670967741935, | |
| "grad_norm": 2.6653757095336914, | |
| "learning_rate": 1.4193548387096776e-06, | |
| "loss": 0.0008, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 43.008, | |
| "grad_norm": 0.05319586396217346, | |
| "learning_rate": 1.4050179211469536e-06, | |
| "loss": 0.1862, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 43.00929032258065, | |
| "grad_norm": 79.42096710205078, | |
| "learning_rate": 1.3906810035842294e-06, | |
| "loss": 0.1399, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 43.01058064516129, | |
| "grad_norm": 117.84407043457031, | |
| "learning_rate": 1.3763440860215054e-06, | |
| "loss": 0.0399, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 43.011870967741935, | |
| "grad_norm": 0.003360538277775049, | |
| "learning_rate": 1.3620071684587816e-06, | |
| "loss": 0.2066, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 43.01316129032258, | |
| "grad_norm": 0.1315963715314865, | |
| "learning_rate": 1.3476702508960574e-06, | |
| "loss": 0.1055, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 43.01445161290322, | |
| "grad_norm": 0.24959410727024078, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.0099, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 43.015741935483874, | |
| "grad_norm": 0.30342984199523926, | |
| "learning_rate": 1.3189964157706094e-06, | |
| "loss": 0.0794, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 43.01703225806452, | |
| "grad_norm": 0.025268543511629105, | |
| "learning_rate": 1.3046594982078852e-06, | |
| "loss": 0.1073, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 43.01832258064516, | |
| "grad_norm": 0.026141099631786346, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 0.1541, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 43.019612903225806, | |
| "grad_norm": 0.04978553205728531, | |
| "learning_rate": 1.2759856630824374e-06, | |
| "loss": 0.1631, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 43.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.2079893350601196, | |
| "eval_runtime": 11.2397, | |
| "eval_samples_per_second": 4.004, | |
| "eval_steps_per_second": 0.801, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 44.00077419354839, | |
| "grad_norm": 0.03287717327475548, | |
| "learning_rate": 1.2616487455197136e-06, | |
| "loss": 0.0339, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 44.00206451612903, | |
| "grad_norm": 0.04084869101643562, | |
| "learning_rate": 1.2473118279569894e-06, | |
| "loss": 0.001, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 44.003354838709676, | |
| "grad_norm": 0.024442797526717186, | |
| "learning_rate": 1.2329749103942654e-06, | |
| "loss": 0.0084, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 44.00464516129032, | |
| "grad_norm": 0.048785168677568436, | |
| "learning_rate": 1.2186379928315414e-06, | |
| "loss": 0.0061, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 44.00593548387097, | |
| "grad_norm": 0.0352037213742733, | |
| "learning_rate": 1.2043010752688174e-06, | |
| "loss": 0.0004, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 44.007225806451615, | |
| "grad_norm": 0.00967736542224884, | |
| "learning_rate": 1.1899641577060934e-06, | |
| "loss": 0.3168, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 44.00851612903226, | |
| "grad_norm": 0.2255047857761383, | |
| "learning_rate": 1.1756272401433692e-06, | |
| "loss": 0.0177, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 44.0098064516129, | |
| "grad_norm": 0.006142869126051664, | |
| "learning_rate": 1.1612903225806454e-06, | |
| "loss": 0.0006, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 44.01109677419355, | |
| "grad_norm": 131.3245086669922, | |
| "learning_rate": 1.1469534050179212e-06, | |
| "loss": 0.2193, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 44.01238709677419, | |
| "grad_norm": 127.88922119140625, | |
| "learning_rate": 1.1326164874551971e-06, | |
| "loss": 0.0235, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 44.01367741935484, | |
| "grad_norm": 0.01740800403058529, | |
| "learning_rate": 1.1182795698924731e-06, | |
| "loss": 0.0432, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 44.014967741935486, | |
| "grad_norm": 3.726701259613037, | |
| "learning_rate": 1.1039426523297491e-06, | |
| "loss": 0.1276, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 44.01625806451613, | |
| "grad_norm": 0.021030189469456673, | |
| "learning_rate": 1.0896057347670253e-06, | |
| "loss": 0.0004, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 44.017548387096774, | |
| "grad_norm": 198.07481384277344, | |
| "learning_rate": 1.0752688172043011e-06, | |
| "loss": 0.2345, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 44.01883870967742, | |
| "grad_norm": 31.0013370513916, | |
| "learning_rate": 1.0609318996415771e-06, | |
| "loss": 0.0067, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 44.02012903225806, | |
| "grad_norm": 6.965691089630127, | |
| "learning_rate": 1.0465949820788531e-06, | |
| "loss": 0.0029, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 44.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.19989812374115, | |
| "eval_runtime": 11.1203, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.809, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 45.001290322580644, | |
| "grad_norm": 0.0045283399522304535, | |
| "learning_rate": 1.0322580645161291e-06, | |
| "loss": 0.3109, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 45.00258064516129, | |
| "grad_norm": 0.012233160436153412, | |
| "learning_rate": 1.0179211469534051e-06, | |
| "loss": 0.1216, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 45.00387096774193, | |
| "grad_norm": 88.97980499267578, | |
| "learning_rate": 1.0035842293906811e-06, | |
| "loss": 0.0103, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 45.00516129032258, | |
| "grad_norm": 0.037353161722421646, | |
| "learning_rate": 9.892473118279571e-07, | |
| "loss": 0.0005, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 45.00645161290323, | |
| "grad_norm": 0.012419509701430798, | |
| "learning_rate": 9.749103942652331e-07, | |
| "loss": 0.0004, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 45.00774193548387, | |
| "grad_norm": 0.027250424027442932, | |
| "learning_rate": 9.60573476702509e-07, | |
| "loss": 0.0029, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 45.009032258064515, | |
| "grad_norm": 0.030230935662984848, | |
| "learning_rate": 9.46236559139785e-07, | |
| "loss": 0.1261, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 45.01032258064516, | |
| "grad_norm": 0.6563981175422668, | |
| "learning_rate": 9.318996415770611e-07, | |
| "loss": 0.0004, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 45.0116129032258, | |
| "grad_norm": 0.014341190457344055, | |
| "learning_rate": 9.17562724014337e-07, | |
| "loss": 0.0415, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 45.012903225806454, | |
| "grad_norm": 16.875062942504883, | |
| "learning_rate": 9.032258064516129e-07, | |
| "loss": 0.0046, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 45.0141935483871, | |
| "grad_norm": 0.07907381653785706, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 0.0004, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 45.01548387096774, | |
| "grad_norm": 0.002687829779461026, | |
| "learning_rate": 8.74551971326165e-07, | |
| "loss": 0.2535, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 45.016774193548386, | |
| "grad_norm": 0.017619280144572258, | |
| "learning_rate": 8.60215053763441e-07, | |
| "loss": 0.3106, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 45.01806451612903, | |
| "grad_norm": 0.0871070846915245, | |
| "learning_rate": 8.45878136200717e-07, | |
| "loss": 0.0004, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 45.019354838709674, | |
| "grad_norm": 115.92878723144531, | |
| "learning_rate": 8.315412186379929e-07, | |
| "loss": 0.1862, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 45.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.246695637702942, | |
| "eval_runtime": 12.7879, | |
| "eval_samples_per_second": 3.519, | |
| "eval_steps_per_second": 0.704, | |
| "step": 7176 | |
| }, | |
| { | |
| "epoch": 46.000516129032256, | |
| "grad_norm": 0.07080364227294922, | |
| "learning_rate": 8.17204301075269e-07, | |
| "loss": 0.0934, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 46.0018064516129, | |
| "grad_norm": 238.50723266601562, | |
| "learning_rate": 8.028673835125448e-07, | |
| "loss": 0.195, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 46.00309677419355, | |
| "grad_norm": 0.12389879673719406, | |
| "learning_rate": 7.885304659498208e-07, | |
| "loss": 0.2498, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 46.004387096774195, | |
| "grad_norm": 0.09104171395301819, | |
| "learning_rate": 7.741935483870968e-07, | |
| "loss": 0.1334, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 46.00567741935484, | |
| "grad_norm": 0.5613530278205872, | |
| "learning_rate": 7.598566308243728e-07, | |
| "loss": 0.1509, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 46.00696774193548, | |
| "grad_norm": 3.0590741634368896, | |
| "learning_rate": 7.455197132616488e-07, | |
| "loss": 0.0008, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 46.00825806451613, | |
| "grad_norm": 31.719104766845703, | |
| "learning_rate": 7.311827956989248e-07, | |
| "loss": 0.2307, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 46.00954838709677, | |
| "grad_norm": 0.07125529646873474, | |
| "learning_rate": 7.168458781362007e-07, | |
| "loss": 0.0008, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 46.01083870967742, | |
| "grad_norm": 0.0011962425196543336, | |
| "learning_rate": 7.025089605734768e-07, | |
| "loss": 0.1139, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 46.012129032258066, | |
| "grad_norm": 0.03515659272670746, | |
| "learning_rate": 6.881720430107527e-07, | |
| "loss": 0.163, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 46.01341935483871, | |
| "grad_norm": 126.44549560546875, | |
| "learning_rate": 6.738351254480287e-07, | |
| "loss": 0.1118, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 46.014709677419354, | |
| "grad_norm": 0.04712965339422226, | |
| "learning_rate": 6.594982078853047e-07, | |
| "loss": 0.1168, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 46.016, | |
| "grad_norm": 2.0991053581237793, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 0.1808, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 46.01729032258064, | |
| "grad_norm": 0.026332564651966095, | |
| "learning_rate": 6.308243727598568e-07, | |
| "loss": 0.0008, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 46.01858064516129, | |
| "grad_norm": 0.022915642708539963, | |
| "learning_rate": 6.164874551971327e-07, | |
| "loss": 0.0567, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 46.01987096774194, | |
| "grad_norm": 0.028331952169537544, | |
| "learning_rate": 6.021505376344087e-07, | |
| "loss": 0.0266, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 46.02012903225806, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 1.3164331912994385, | |
| "eval_runtime": 12.3588, | |
| "eval_samples_per_second": 3.641, | |
| "eval_steps_per_second": 0.728, | |
| "step": 7332 | |
| }, | |
| { | |
| "epoch": 47.00103225806452, | |
| "grad_norm": 0.03460884466767311, | |
| "learning_rate": 5.878136200716846e-07, | |
| "loss": 0.005, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 47.00232258064516, | |
| "grad_norm": 0.01784307137131691, | |
| "learning_rate": 5.734767025089606e-07, | |
| "loss": 0.0004, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 47.00361290322581, | |
| "grad_norm": 0.022959072142839432, | |
| "learning_rate": 5.591397849462366e-07, | |
| "loss": 0.0071, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 47.00490322580645, | |
| "grad_norm": 0.01595255918800831, | |
| "learning_rate": 5.448028673835127e-07, | |
| "loss": 0.0008, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 47.006193548387095, | |
| "grad_norm": 0.033321935683488846, | |
| "learning_rate": 5.304659498207886e-07, | |
| "loss": 0.0004, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 47.00748387096774, | |
| "grad_norm": 243.83221435546875, | |
| "learning_rate": 5.161290322580646e-07, | |
| "loss": 0.1434, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 47.00877419354839, | |
| "grad_norm": 0.36720991134643555, | |
| "learning_rate": 5.017921146953406e-07, | |
| "loss": 0.0004, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 47.010064516129034, | |
| "grad_norm": 0.01275189220905304, | |
| "learning_rate": 4.874551971326166e-07, | |
| "loss": 0.1149, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 47.01135483870968, | |
| "grad_norm": 0.013302352279424667, | |
| "learning_rate": 4.731182795698925e-07, | |
| "loss": 0.0031, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 47.01264516129032, | |
| "grad_norm": 0.005342531017959118, | |
| "learning_rate": 4.587813620071685e-07, | |
| "loss": 0.3822, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 47.013935483870966, | |
| "grad_norm": 0.029405364766716957, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 0.0012, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 47.01522580645161, | |
| "grad_norm": 0.00923657976090908, | |
| "learning_rate": 4.301075268817205e-07, | |
| "loss": 0.1319, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 47.01651612903226, | |
| "grad_norm": 0.02844962663948536, | |
| "learning_rate": 4.1577060931899643e-07, | |
| "loss": 0.1112, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 47.017806451612906, | |
| "grad_norm": 311.14520263671875, | |
| "learning_rate": 4.014336917562724e-07, | |
| "loss": 0.156, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 47.01909677419355, | |
| "grad_norm": 6.184717178344727, | |
| "learning_rate": 3.870967741935484e-07, | |
| "loss": 0.0013, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 47.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.2166378498077393, | |
| "eval_runtime": 11.1295, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.809, | |
| "step": 7488 | |
| }, | |
| { | |
| "epoch": 48.00025806451613, | |
| "grad_norm": 0.06741901487112045, | |
| "learning_rate": 3.727598566308244e-07, | |
| "loss": 0.0005, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 48.001548387096776, | |
| "grad_norm": 33.09029769897461, | |
| "learning_rate": 3.5842293906810036e-07, | |
| "loss": 0.0031, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 48.00283870967742, | |
| "grad_norm": 0.006431705318391323, | |
| "learning_rate": 3.4408602150537636e-07, | |
| "loss": 0.0003, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 48.004129032258064, | |
| "grad_norm": 0.009529894217848778, | |
| "learning_rate": 3.2974910394265235e-07, | |
| "loss": 0.0447, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 48.00541935483871, | |
| "grad_norm": 0.012879936024546623, | |
| "learning_rate": 3.154121863799284e-07, | |
| "loss": 0.0774, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 48.00670967741935, | |
| "grad_norm": 268.919921875, | |
| "learning_rate": 3.0107526881720434e-07, | |
| "loss": 0.2787, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 48.008, | |
| "grad_norm": 0.022163085639476776, | |
| "learning_rate": 2.867383512544803e-07, | |
| "loss": 0.0003, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 48.00929032258065, | |
| "grad_norm": 102.45469665527344, | |
| "learning_rate": 2.7240143369175634e-07, | |
| "loss": 0.1184, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 48.01058064516129, | |
| "grad_norm": 0.017046399414539337, | |
| "learning_rate": 2.580645161290323e-07, | |
| "loss": 0.0643, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 48.011870967741935, | |
| "grad_norm": 0.008543446660041809, | |
| "learning_rate": 2.437275985663083e-07, | |
| "loss": 0.0007, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 48.01316129032258, | |
| "grad_norm": 0.012028884142637253, | |
| "learning_rate": 2.2939068100358425e-07, | |
| "loss": 0.1149, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 48.01445161290322, | |
| "grad_norm": 0.05450389161705971, | |
| "learning_rate": 2.1505376344086024e-07, | |
| "loss": 0.0013, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 48.015741935483874, | |
| "grad_norm": 0.0074774823151528835, | |
| "learning_rate": 2.007168458781362e-07, | |
| "loss": 0.0009, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 48.01703225806452, | |
| "grad_norm": 0.00835146103054285, | |
| "learning_rate": 1.863799283154122e-07, | |
| "loss": 0.0003, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 48.01832258064516, | |
| "grad_norm": 0.014417874626815319, | |
| "learning_rate": 1.7204301075268818e-07, | |
| "loss": 0.1268, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 48.019612903225806, | |
| "grad_norm": 0.024401506409049034, | |
| "learning_rate": 1.577060931899642e-07, | |
| "loss": 0.0007, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 48.02012903225806, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.2551714181900024, | |
| "eval_runtime": 12.1327, | |
| "eval_samples_per_second": 3.709, | |
| "eval_steps_per_second": 0.742, | |
| "step": 7644 | |
| }, | |
| { | |
| "epoch": 49.00077419354839, | |
| "grad_norm": 0.07428643852472305, | |
| "learning_rate": 1.4336917562724014e-07, | |
| "loss": 0.2398, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 49.00206451612903, | |
| "grad_norm": 0.006770737003535032, | |
| "learning_rate": 1.2903225806451614e-07, | |
| "loss": 0.0004, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 49.003354838709676, | |
| "grad_norm": 0.008373833261430264, | |
| "learning_rate": 1.1469534050179212e-07, | |
| "loss": 0.2454, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 49.00464516129032, | |
| "grad_norm": 0.022648228332400322, | |
| "learning_rate": 1.003584229390681e-07, | |
| "loss": 0.0004, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 49.00593548387097, | |
| "grad_norm": 0.02756650373339653, | |
| "learning_rate": 8.602150537634409e-08, | |
| "loss": 0.0004, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 49.007225806451615, | |
| "grad_norm": 0.0037723516579717398, | |
| "learning_rate": 7.168458781362007e-08, | |
| "loss": 0.0004, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 49.00851612903226, | |
| "grad_norm": 0.05359743908047676, | |
| "learning_rate": 5.734767025089606e-08, | |
| "loss": 0.0006, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 49.0098064516129, | |
| "grad_norm": 0.029262199997901917, | |
| "learning_rate": 4.3010752688172045e-08, | |
| "loss": 0.0006, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 49.01109677419355, | |
| "grad_norm": 0.019295159727334976, | |
| "learning_rate": 2.867383512544803e-08, | |
| "loss": 0.0926, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 49.01238709677419, | |
| "grad_norm": 22.579675674438477, | |
| "learning_rate": 1.4336917562724015e-08, | |
| "loss": 0.1257, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 49.01367741935484, | |
| "grad_norm": 116.33914947509766, | |
| "learning_rate": 0.0, | |
| "loss": 0.0089, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 49.01367741935484, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 1.2637289762496948, | |
| "eval_runtime": 13.8945, | |
| "eval_samples_per_second": 3.239, | |
| "eval_steps_per_second": 0.648, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 49.01367741935484, | |
| "step": 7750, | |
| "total_flos": 1.6950678884940114e+20, | |
| "train_loss": 0.30147023142159224, | |
| "train_runtime": 21493.7197, | |
| "train_samples_per_second": 1.803, | |
| "train_steps_per_second": 0.361 | |
| }, | |
| { | |
| "epoch": 49.01367741935484, | |
| "eval_accuracy": 0.8666666666666667, | |
| "eval_loss": 0.3363037407398224, | |
| "eval_runtime": 12.5631, | |
| "eval_samples_per_second": 3.582, | |
| "eval_steps_per_second": 0.716, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 49.01367741935484, | |
| "eval_accuracy": 0.8666666666666667, | |
| "eval_loss": 0.3363037109375, | |
| "eval_runtime": 12.6357, | |
| "eval_samples_per_second": 3.561, | |
| "eval_steps_per_second": 0.712, | |
| "step": 7750 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 7750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.6950678884940114e+20, | |
| "train_batch_size": 5, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |