| { | |
| "best_metric": 0.5490196078431373, | |
| "best_model_checkpoint": "SwinLarge/checkpoint-946", | |
| "epoch": 9.984152139461173, | |
| "eval_steps": 500, | |
| "global_step": 3150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03169572107765452, | |
| "grad_norm": 8.620658874511719, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "loss": 2.6801, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06339144215530904, | |
| "grad_norm": 10.387493133544922, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "loss": 2.5071, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09508716323296355, | |
| "grad_norm": 11.921625137329102, | |
| "learning_rate": 4.7619047619047615e-06, | |
| "loss": 2.3099, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12678288431061807, | |
| "grad_norm": 7.416236877441406, | |
| "learning_rate": 6.349206349206349e-06, | |
| "loss": 2.0759, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15847860538827258, | |
| "grad_norm": 10.799355506896973, | |
| "learning_rate": 7.936507936507936e-06, | |
| "loss": 1.9247, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1901743264659271, | |
| "grad_norm": 13.448376655578613, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 1.7113, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2218700475435816, | |
| "grad_norm": 10.010295867919922, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 1.8997, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25356576862123614, | |
| "grad_norm": 7.308131217956543, | |
| "learning_rate": 1.2698412698412699e-05, | |
| "loss": 1.8577, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.28526148969889065, | |
| "grad_norm": 7.589808464050293, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 1.5864, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.31695721077654515, | |
| "grad_norm": 32.497920989990234, | |
| "learning_rate": 1.5873015873015872e-05, | |
| "loss": 1.5627, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3486529318541997, | |
| "grad_norm": 8.728975296020508, | |
| "learning_rate": 1.746031746031746e-05, | |
| "loss": 2.0147, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3803486529318542, | |
| "grad_norm": 11.442461013793945, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 1.8042, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4120443740095087, | |
| "grad_norm": 7.592460632324219, | |
| "learning_rate": 2.0634920634920636e-05, | |
| "loss": 1.5589, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4437400950871632, | |
| "grad_norm": 8.4674711227417, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 1.8606, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4754358161648177, | |
| "grad_norm": 5.960061073303223, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 1.5667, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5071315372424723, | |
| "grad_norm": 6.8911333084106445, | |
| "learning_rate": 2.5396825396825397e-05, | |
| "loss": 1.6166, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5388272583201268, | |
| "grad_norm": 5.806614875793457, | |
| "learning_rate": 2.6984126984126984e-05, | |
| "loss": 1.5221, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5705229793977813, | |
| "grad_norm": 13.649744033813477, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 1.6278, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6022187004754358, | |
| "grad_norm": 6.668085098266602, | |
| "learning_rate": 3.0158730158730158e-05, | |
| "loss": 1.7091, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6339144215530903, | |
| "grad_norm": 7.859160900115967, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 1.5582, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6656101426307448, | |
| "grad_norm": 4.802083492279053, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.4414, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6973058637083994, | |
| "grad_norm": 15.605195045471191, | |
| "learning_rate": 3.492063492063492e-05, | |
| "loss": 1.7171, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7290015847860539, | |
| "grad_norm": 7.946794033050537, | |
| "learning_rate": 3.650793650793651e-05, | |
| "loss": 1.738, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7606973058637084, | |
| "grad_norm": 7.9303364753723145, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 1.5497, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7923930269413629, | |
| "grad_norm": 6.65696382522583, | |
| "learning_rate": 3.968253968253968e-05, | |
| "loss": 1.7273, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8240887480190174, | |
| "grad_norm": 4.192816734313965, | |
| "learning_rate": 4.126984126984127e-05, | |
| "loss": 1.8298, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8557844690966719, | |
| "grad_norm": 6.851734161376953, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 1.728, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8874801901743264, | |
| "grad_norm": 8.173200607299805, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 1.5258, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.919175911251981, | |
| "grad_norm": 5.898636817932129, | |
| "learning_rate": 4.603174603174603e-05, | |
| "loss": 1.817, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9508716323296355, | |
| "grad_norm": 7.8518452644348145, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 1.6367, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9825673534072901, | |
| "grad_norm": 4.014355659484863, | |
| "learning_rate": 4.9206349206349204e-05, | |
| "loss": 1.8318, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9984152139461173, | |
| "eval_accuracy": 0.5436720142602496, | |
| "eval_loss": 1.7651361227035522, | |
| "eval_runtime": 149.0995, | |
| "eval_samples_per_second": 3.763, | |
| "eval_steps_per_second": 0.946, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0142630744849446, | |
| "grad_norm": 5.769894599914551, | |
| "learning_rate": 4.991181657848325e-05, | |
| "loss": 1.5269, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.045958795562599, | |
| "grad_norm": 4.770431041717529, | |
| "learning_rate": 4.973544973544973e-05, | |
| "loss": 1.71, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0776545166402536, | |
| "grad_norm": 3.420660972595215, | |
| "learning_rate": 4.955908289241622e-05, | |
| "loss": 1.856, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.109350237717908, | |
| "grad_norm": 4.487371444702148, | |
| "learning_rate": 4.938271604938271e-05, | |
| "loss": 1.5407, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1410459587955626, | |
| "grad_norm": 5.092081546783447, | |
| "learning_rate": 4.9206349206349204e-05, | |
| "loss": 1.6359, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.172741679873217, | |
| "grad_norm": 3.577042818069458, | |
| "learning_rate": 4.9029982363315695e-05, | |
| "loss": 1.6652, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2044374009508716, | |
| "grad_norm": 3.7257213592529297, | |
| "learning_rate": 4.8853615520282185e-05, | |
| "loss": 1.5741, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.236133122028526, | |
| "grad_norm": 5.686126232147217, | |
| "learning_rate": 4.8677248677248676e-05, | |
| "loss": 1.6713, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2678288431061806, | |
| "grad_norm": 4.3698577880859375, | |
| "learning_rate": 4.850088183421517e-05, | |
| "loss": 1.7232, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.299524564183835, | |
| "grad_norm": 6.864264965057373, | |
| "learning_rate": 4.832451499118166e-05, | |
| "loss": 1.6318, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3312202852614896, | |
| "grad_norm": 3.44631290435791, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 1.5783, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3629160063391441, | |
| "grad_norm": 11.141667366027832, | |
| "learning_rate": 4.797178130511464e-05, | |
| "loss": 1.5381, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.3946117274167986, | |
| "grad_norm": 4.2202959060668945, | |
| "learning_rate": 4.779541446208113e-05, | |
| "loss": 1.5854, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4263074484944531, | |
| "grad_norm": 4.3586039543151855, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 1.7901, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4580031695721076, | |
| "grad_norm": 4.2103352546691895, | |
| "learning_rate": 4.744268077601411e-05, | |
| "loss": 1.4986, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.4896988906497624, | |
| "grad_norm": 5.0563225746154785, | |
| "learning_rate": 4.72663139329806e-05, | |
| "loss": 1.5424, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5213946117274166, | |
| "grad_norm": 4.409386157989502, | |
| "learning_rate": 4.708994708994709e-05, | |
| "loss": 1.4963, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5530903328050714, | |
| "grad_norm": 6.303650856018066, | |
| "learning_rate": 4.691358024691358e-05, | |
| "loss": 1.5943, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.5847860538827259, | |
| "grad_norm": 4.192474842071533, | |
| "learning_rate": 4.673721340388007e-05, | |
| "loss": 1.6115, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6164817749603804, | |
| "grad_norm": 4.521521091461182, | |
| "learning_rate": 4.656084656084656e-05, | |
| "loss": 1.5737, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6481774960380349, | |
| "grad_norm": 8.733231544494629, | |
| "learning_rate": 4.638447971781305e-05, | |
| "loss": 1.5708, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.6798732171156894, | |
| "grad_norm": 5.1407575607299805, | |
| "learning_rate": 4.620811287477954e-05, | |
| "loss": 1.5751, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.7115689381933439, | |
| "grad_norm": 4.215906143188477, | |
| "learning_rate": 4.603174603174603e-05, | |
| "loss": 1.5678, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.7432646592709984, | |
| "grad_norm": 7.169848918914795, | |
| "learning_rate": 4.585537918871252e-05, | |
| "loss": 1.6263, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7749603803486529, | |
| "grad_norm": 5.950140953063965, | |
| "learning_rate": 4.567901234567901e-05, | |
| "loss": 1.5155, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8066561014263076, | |
| "grad_norm": 6.070963382720947, | |
| "learning_rate": 4.55026455026455e-05, | |
| "loss": 1.6662, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8383518225039621, | |
| "grad_norm": 5.032849311828613, | |
| "learning_rate": 4.532627865961199e-05, | |
| "loss": 1.6791, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.8700475435816166, | |
| "grad_norm": 6.853456020355225, | |
| "learning_rate": 4.5149911816578484e-05, | |
| "loss": 1.6334, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9017432646592711, | |
| "grad_norm": 4.782254219055176, | |
| "learning_rate": 4.4973544973544974e-05, | |
| "loss": 1.6209, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9334389857369256, | |
| "grad_norm": 7.97931432723999, | |
| "learning_rate": 4.4797178130511465e-05, | |
| "loss": 1.8784, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.9651347068145801, | |
| "grad_norm": 3.8500170707702637, | |
| "learning_rate": 4.4620811287477956e-05, | |
| "loss": 1.721, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.9968304278922346, | |
| "grad_norm": 3.8997135162353516, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 1.6067, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5454545454545454, | |
| "eval_loss": 1.6392648220062256, | |
| "eval_runtime": 147.9577, | |
| "eval_samples_per_second": 3.792, | |
| "eval_steps_per_second": 0.953, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.028526148969889, | |
| "grad_norm": 4.185525894165039, | |
| "learning_rate": 4.426807760141094e-05, | |
| "loss": 1.4786, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.0602218700475436, | |
| "grad_norm": 5.7630391120910645, | |
| "learning_rate": 4.409171075837743e-05, | |
| "loss": 1.3732, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.091917591125198, | |
| "grad_norm": 6.025218963623047, | |
| "learning_rate": 4.391534391534391e-05, | |
| "loss": 1.5661, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.1236133122028527, | |
| "grad_norm": 8.51380443572998, | |
| "learning_rate": 4.37389770723104e-05, | |
| "loss": 1.5385, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.155309033280507, | |
| "grad_norm": 7.913331985473633, | |
| "learning_rate": 4.3562610229276893e-05, | |
| "loss": 1.3816, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.1870047543581617, | |
| "grad_norm": 10.290911674499512, | |
| "learning_rate": 4.3386243386243384e-05, | |
| "loss": 1.5289, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.218700475435816, | |
| "grad_norm": 12.427474021911621, | |
| "learning_rate": 4.3209876543209875e-05, | |
| "loss": 1.3822, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.2503961965134707, | |
| "grad_norm": 8.04323959350586, | |
| "learning_rate": 4.3033509700176366e-05, | |
| "loss": 1.699, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.282091917591125, | |
| "grad_norm": 6.580540180206299, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 1.488, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3137876386687797, | |
| "grad_norm": 6.697164058685303, | |
| "learning_rate": 4.268077601410935e-05, | |
| "loss": 1.6108, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.345483359746434, | |
| "grad_norm": 8.002148628234863, | |
| "learning_rate": 4.250440917107584e-05, | |
| "loss": 1.4629, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.3771790808240887, | |
| "grad_norm": 6.550644874572754, | |
| "learning_rate": 4.232804232804233e-05, | |
| "loss": 1.3738, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.408874801901743, | |
| "grad_norm": 10.595471382141113, | |
| "learning_rate": 4.215167548500882e-05, | |
| "loss": 1.4171, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.4405705229793977, | |
| "grad_norm": 8.15519905090332, | |
| "learning_rate": 4.197530864197531e-05, | |
| "loss": 1.5467, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.472266244057052, | |
| "grad_norm": 5.568257808685303, | |
| "learning_rate": 4.17989417989418e-05, | |
| "loss": 1.3498, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.5039619651347067, | |
| "grad_norm": 5.846574783325195, | |
| "learning_rate": 4.162257495590829e-05, | |
| "loss": 1.4658, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.535657686212361, | |
| "grad_norm": 5.970884323120117, | |
| "learning_rate": 4.144620811287478e-05, | |
| "loss": 1.5717, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.5673534072900157, | |
| "grad_norm": 10.46259593963623, | |
| "learning_rate": 4.126984126984127e-05, | |
| "loss": 1.5871, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.59904912836767, | |
| "grad_norm": 6.0144267082214355, | |
| "learning_rate": 4.109347442680776e-05, | |
| "loss": 1.5136, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.6307448494453247, | |
| "grad_norm": 7.098198890686035, | |
| "learning_rate": 4.091710758377425e-05, | |
| "loss": 1.5325, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.662440570522979, | |
| "grad_norm": 6.445565700531006, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 1.4732, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.6941362916006337, | |
| "grad_norm": 7.529149532318115, | |
| "learning_rate": 4.056437389770723e-05, | |
| "loss": 1.4529, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.7258320126782882, | |
| "grad_norm": 5.447256088256836, | |
| "learning_rate": 4.038800705467372e-05, | |
| "loss": 1.3855, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.7575277337559427, | |
| "grad_norm": 5.000948429107666, | |
| "learning_rate": 4.021164021164021e-05, | |
| "loss": 1.248, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.7892234548335972, | |
| "grad_norm": 5.39021110534668, | |
| "learning_rate": 4.00352733686067e-05, | |
| "loss": 1.7784, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.8209191759112517, | |
| "grad_norm": 5.756132125854492, | |
| "learning_rate": 3.985890652557319e-05, | |
| "loss": 1.6598, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.8526148969889062, | |
| "grad_norm": 5.659241199493408, | |
| "learning_rate": 3.968253968253968e-05, | |
| "loss": 1.4138, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.8843106180665607, | |
| "grad_norm": 7.678089141845703, | |
| "learning_rate": 3.950617283950617e-05, | |
| "loss": 1.3823, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.9160063391442153, | |
| "grad_norm": 7.0097479820251465, | |
| "learning_rate": 3.9329805996472664e-05, | |
| "loss": 1.6495, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.94770206022187, | |
| "grad_norm": 7.138586521148682, | |
| "learning_rate": 3.9153439153439155e-05, | |
| "loss": 1.4128, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.9793977812995247, | |
| "grad_norm": 8.242341041564941, | |
| "learning_rate": 3.8977072310405645e-05, | |
| "loss": 1.406, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.9984152139461173, | |
| "eval_accuracy": 0.5490196078431373, | |
| "eval_loss": 1.6471669673919678, | |
| "eval_runtime": 147.6692, | |
| "eval_samples_per_second": 3.799, | |
| "eval_steps_per_second": 0.955, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.011093502377179, | |
| "grad_norm": 8.209331512451172, | |
| "learning_rate": 3.8800705467372136e-05, | |
| "loss": 1.4082, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.0427892234548337, | |
| "grad_norm": 6.750119209289551, | |
| "learning_rate": 3.862433862433863e-05, | |
| "loss": 1.1927, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.074484944532488, | |
| "grad_norm": 18.73226547241211, | |
| "learning_rate": 3.844797178130512e-05, | |
| "loss": 1.2399, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.1061806656101427, | |
| "grad_norm": 19.268539428710938, | |
| "learning_rate": 3.82716049382716e-05, | |
| "loss": 1.1511, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.1378763866877972, | |
| "grad_norm": 13.193881034851074, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 1.3139, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.1695721077654517, | |
| "grad_norm": 9.010955810546875, | |
| "learning_rate": 3.791887125220458e-05, | |
| "loss": 1.2482, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2012678288431062, | |
| "grad_norm": 6.472105026245117, | |
| "learning_rate": 3.7742504409171074e-05, | |
| "loss": 1.0941, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.2329635499207607, | |
| "grad_norm": 14.444457054138184, | |
| "learning_rate": 3.7566137566137564e-05, | |
| "loss": 1.1845, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.2646592709984152, | |
| "grad_norm": 9.034687995910645, | |
| "learning_rate": 3.7389770723104055e-05, | |
| "loss": 1.2469, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.2963549920760697, | |
| "grad_norm": 10.975589752197266, | |
| "learning_rate": 3.7213403880070546e-05, | |
| "loss": 1.3324, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.3280507131537242, | |
| "grad_norm": 9.528573036193848, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.0736, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.3597464342313788, | |
| "grad_norm": 9.448025703430176, | |
| "learning_rate": 3.686067019400353e-05, | |
| "loss": 1.2401, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.3914421553090333, | |
| "grad_norm": 7.891082286834717, | |
| "learning_rate": 3.668430335097002e-05, | |
| "loss": 1.3604, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.4231378763866878, | |
| "grad_norm": 12.8311185836792, | |
| "learning_rate": 3.650793650793651e-05, | |
| "loss": 1.2031, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.4548335974643423, | |
| "grad_norm": 10.008551597595215, | |
| "learning_rate": 3.6331569664903e-05, | |
| "loss": 1.1598, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.4865293185419968, | |
| "grad_norm": 10.480392456054688, | |
| "learning_rate": 3.615520282186949e-05, | |
| "loss": 1.3213, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.5182250396196513, | |
| "grad_norm": 10.108732223510742, | |
| "learning_rate": 3.597883597883598e-05, | |
| "loss": 1.3552, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.5499207606973058, | |
| "grad_norm": 9.28955364227295, | |
| "learning_rate": 3.580246913580247e-05, | |
| "loss": 1.1861, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.5816164817749603, | |
| "grad_norm": 9.186285972595215, | |
| "learning_rate": 3.562610229276896e-05, | |
| "loss": 1.1303, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.613312202852615, | |
| "grad_norm": 9.3871488571167, | |
| "learning_rate": 3.5449735449735446e-05, | |
| "loss": 1.2186, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.6450079239302693, | |
| "grad_norm": 14.419343948364258, | |
| "learning_rate": 3.527336860670194e-05, | |
| "loss": 1.3413, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.676703645007924, | |
| "grad_norm": 8.462416648864746, | |
| "learning_rate": 3.509700176366843e-05, | |
| "loss": 1.2544, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.7083993660855783, | |
| "grad_norm": 6.36189079284668, | |
| "learning_rate": 3.492063492063492e-05, | |
| "loss": 1.0951, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.740095087163233, | |
| "grad_norm": 16.61732292175293, | |
| "learning_rate": 3.474426807760141e-05, | |
| "loss": 1.1044, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.7717908082408877, | |
| "grad_norm": 9.842411041259766, | |
| "learning_rate": 3.45679012345679e-05, | |
| "loss": 1.2549, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.8034865293185423, | |
| "grad_norm": 16.95412826538086, | |
| "learning_rate": 3.439153439153439e-05, | |
| "loss": 1.2285, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8351822503961968, | |
| "grad_norm": 16.060531616210938, | |
| "learning_rate": 3.421516754850088e-05, | |
| "loss": 1.3635, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.8668779714738513, | |
| "grad_norm": 12.322091102600098, | |
| "learning_rate": 3.403880070546737e-05, | |
| "loss": 1.2121, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.8985736925515058, | |
| "grad_norm": 10.384512901306152, | |
| "learning_rate": 3.386243386243386e-05, | |
| "loss": 1.321, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.9302694136291603, | |
| "grad_norm": 10.984822273254395, | |
| "learning_rate": 3.3686067019400353e-05, | |
| "loss": 1.1268, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.9619651347068148, | |
| "grad_norm": 12.061741828918457, | |
| "learning_rate": 3.3509700176366844e-05, | |
| "loss": 1.1616, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.9936608557844693, | |
| "grad_norm": 13.65400505065918, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.3983, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5454545454545454, | |
| "eval_loss": 1.7344454526901245, | |
| "eval_runtime": 147.3715, | |
| "eval_samples_per_second": 3.807, | |
| "eval_steps_per_second": 0.957, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 4.025356576862124, | |
| "grad_norm": 14.69643497467041, | |
| "learning_rate": 3.3156966490299826e-05, | |
| "loss": 0.8817, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.057052297939778, | |
| "grad_norm": 16.983091354370117, | |
| "learning_rate": 3.2980599647266316e-05, | |
| "loss": 0.9822, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.088748019017433, | |
| "grad_norm": 7.896317481994629, | |
| "learning_rate": 3.280423280423281e-05, | |
| "loss": 0.9038, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.120443740095087, | |
| "grad_norm": 19.647830963134766, | |
| "learning_rate": 3.262786596119929e-05, | |
| "loss": 0.9003, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.152139461172742, | |
| "grad_norm": 14.798089981079102, | |
| "learning_rate": 3.245149911816578e-05, | |
| "loss": 0.9009, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.183835182250396, | |
| "grad_norm": 21.833358764648438, | |
| "learning_rate": 3.227513227513227e-05, | |
| "loss": 0.832, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.215530903328051, | |
| "grad_norm": 16.340986251831055, | |
| "learning_rate": 3.209876543209876e-05, | |
| "loss": 0.7249, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.247226624405705, | |
| "grad_norm": 13.776719093322754, | |
| "learning_rate": 3.1922398589065254e-05, | |
| "loss": 0.782, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.27892234548336, | |
| "grad_norm": 22.321552276611328, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 0.7917, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.310618066561014, | |
| "grad_norm": 13.473883628845215, | |
| "learning_rate": 3.1569664902998235e-05, | |
| "loss": 0.7287, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.342313787638669, | |
| "grad_norm": 21.802818298339844, | |
| "learning_rate": 3.1393298059964726e-05, | |
| "loss": 0.7062, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.374009508716323, | |
| "grad_norm": 16.524831771850586, | |
| "learning_rate": 3.121693121693122e-05, | |
| "loss": 0.8148, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.405705229793978, | |
| "grad_norm": 16.520061492919922, | |
| "learning_rate": 3.104056437389771e-05, | |
| "loss": 0.8808, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.437400950871632, | |
| "grad_norm": 22.931140899658203, | |
| "learning_rate": 3.08641975308642e-05, | |
| "loss": 0.7967, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.469096671949287, | |
| "grad_norm": 14.37611198425293, | |
| "learning_rate": 3.068783068783069e-05, | |
| "loss": 0.7436, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.500792393026941, | |
| "grad_norm": 15.375365257263184, | |
| "learning_rate": 3.0511463844797176e-05, | |
| "loss": 0.8335, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.532488114104596, | |
| "grad_norm": 13.653996467590332, | |
| "learning_rate": 3.0335097001763667e-05, | |
| "loss": 0.8232, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.56418383518225, | |
| "grad_norm": 28.006961822509766, | |
| "learning_rate": 3.0158730158730158e-05, | |
| "loss": 0.7759, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.595879556259905, | |
| "grad_norm": 7.853826522827148, | |
| "learning_rate": 2.998236331569665e-05, | |
| "loss": 0.898, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.627575277337559, | |
| "grad_norm": 14.124467849731445, | |
| "learning_rate": 2.980599647266314e-05, | |
| "loss": 0.9063, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.659270998415214, | |
| "grad_norm": 16.662132263183594, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.6931, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.690966719492868, | |
| "grad_norm": 17.079029083251953, | |
| "learning_rate": 2.945326278659612e-05, | |
| "loss": 0.676, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.722662440570523, | |
| "grad_norm": 15.269068717956543, | |
| "learning_rate": 2.927689594356261e-05, | |
| "loss": 0.8302, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.754358161648177, | |
| "grad_norm": 11.715784072875977, | |
| "learning_rate": 2.91005291005291e-05, | |
| "loss": 0.8037, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.786053882725832, | |
| "grad_norm": 15.774462699890137, | |
| "learning_rate": 2.892416225749559e-05, | |
| "loss": 0.8735, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.817749603803486, | |
| "grad_norm": 18.308950424194336, | |
| "learning_rate": 2.874779541446208e-05, | |
| "loss": 0.6236, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.849445324881141, | |
| "grad_norm": 21.661277770996094, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.8501, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 4.881141045958795, | |
| "grad_norm": 10.561766624450684, | |
| "learning_rate": 2.839506172839506e-05, | |
| "loss": 0.8987, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 4.91283676703645, | |
| "grad_norm": 17.965679168701172, | |
| "learning_rate": 2.8218694885361552e-05, | |
| "loss": 0.8309, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 4.944532488114104, | |
| "grad_norm": 15.322041511535645, | |
| "learning_rate": 2.8042328042328043e-05, | |
| "loss": 0.6263, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 4.976228209191759, | |
| "grad_norm": 8.305255889892578, | |
| "learning_rate": 2.7865961199294534e-05, | |
| "loss": 0.7272, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 4.998415213946117, | |
| "eval_accuracy": 0.5258467023172906, | |
| "eval_loss": 2.12831974029541, | |
| "eval_runtime": 147.5846, | |
| "eval_samples_per_second": 3.801, | |
| "eval_steps_per_second": 0.955, | |
| "step": 1577 | |
| }, | |
| { | |
| "epoch": 5.007923930269413, | |
| "grad_norm": 12.602760314941406, | |
| "learning_rate": 2.768959435626102e-05, | |
| "loss": 0.7548, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 5.039619651347068, | |
| "grad_norm": 10.254925727844238, | |
| "learning_rate": 2.7513227513227512e-05, | |
| "loss": 0.4662, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.071315372424722, | |
| "grad_norm": 6.985943794250488, | |
| "learning_rate": 2.7336860670194003e-05, | |
| "loss": 0.3352, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.103011093502377, | |
| "grad_norm": 5.529629230499268, | |
| "learning_rate": 2.7160493827160493e-05, | |
| "loss": 0.4111, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 5.134706814580031, | |
| "grad_norm": 26.3822021484375, | |
| "learning_rate": 2.6984126984126984e-05, | |
| "loss": 0.445, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 5.166402535657686, | |
| "grad_norm": 15.873011589050293, | |
| "learning_rate": 2.6807760141093475e-05, | |
| "loss": 0.4571, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 5.19809825673534, | |
| "grad_norm": 15.90845012664795, | |
| "learning_rate": 2.6631393298059965e-05, | |
| "loss": 0.6197, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.229793977812995, | |
| "grad_norm": 21.822021484375, | |
| "learning_rate": 2.6455026455026456e-05, | |
| "loss": 0.416, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 5.261489698890649, | |
| "grad_norm": 16.9212646484375, | |
| "learning_rate": 2.6278659611992943e-05, | |
| "loss": 0.3351, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 5.293185419968304, | |
| "grad_norm": 18.473758697509766, | |
| "learning_rate": 2.6102292768959434e-05, | |
| "loss": 0.4751, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 5.324881141045958, | |
| "grad_norm": 19.178316116333008, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.535, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 5.356576862123613, | |
| "grad_norm": 11.751172065734863, | |
| "learning_rate": 2.5749559082892416e-05, | |
| "loss": 0.6542, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 5.3882725832012675, | |
| "grad_norm": 16.899391174316406, | |
| "learning_rate": 2.5573192239858906e-05, | |
| "loss": 0.4733, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.419968304278922, | |
| "grad_norm": 19.778823852539062, | |
| "learning_rate": 2.5396825396825397e-05, | |
| "loss": 0.448, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 5.4516640253565765, | |
| "grad_norm": 30.306079864501953, | |
| "learning_rate": 2.5220458553791888e-05, | |
| "loss": 0.6095, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 5.483359746434231, | |
| "grad_norm": 7.676305770874023, | |
| "learning_rate": 2.504409171075838e-05, | |
| "loss": 0.3816, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 5.5150554675118855, | |
| "grad_norm": 12.745634078979492, | |
| "learning_rate": 2.4867724867724866e-05, | |
| "loss": 0.5836, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 5.546751188589541, | |
| "grad_norm": 19.767845153808594, | |
| "learning_rate": 2.4691358024691357e-05, | |
| "loss": 0.4719, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 5.5784469096671945, | |
| "grad_norm": 22.378528594970703, | |
| "learning_rate": 2.4514991181657847e-05, | |
| "loss": 0.4079, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 5.61014263074485, | |
| "grad_norm": 16.106239318847656, | |
| "learning_rate": 2.4338624338624338e-05, | |
| "loss": 0.4083, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 5.6418383518225035, | |
| "grad_norm": 11.803277015686035, | |
| "learning_rate": 2.416225749559083e-05, | |
| "loss": 0.5222, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 5.673534072900159, | |
| "grad_norm": 20.511564254760742, | |
| "learning_rate": 2.398589065255732e-05, | |
| "loss": 0.4857, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 5.705229793977813, | |
| "grad_norm": 30.313129425048828, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 0.5033, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.736925515055468, | |
| "grad_norm": 17.92380142211914, | |
| "learning_rate": 2.36331569664903e-05, | |
| "loss": 0.3853, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 5.768621236133122, | |
| "grad_norm": 11.325889587402344, | |
| "learning_rate": 2.345679012345679e-05, | |
| "loss": 0.4485, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 5.800316957210777, | |
| "grad_norm": 30.107561111450195, | |
| "learning_rate": 2.328042328042328e-05, | |
| "loss": 0.4691, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 5.832012678288431, | |
| "grad_norm": 7.061257362365723, | |
| "learning_rate": 2.310405643738977e-05, | |
| "loss": 0.5587, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 5.863708399366086, | |
| "grad_norm": 21.1574764251709, | |
| "learning_rate": 2.292768959435626e-05, | |
| "loss": 0.3672, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 5.89540412044374, | |
| "grad_norm": 21.53982162475586, | |
| "learning_rate": 2.275132275132275e-05, | |
| "loss": 0.4423, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 5.927099841521395, | |
| "grad_norm": 9.077971458435059, | |
| "learning_rate": 2.2574955908289242e-05, | |
| "loss": 0.2509, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 5.958795562599049, | |
| "grad_norm": 21.368032455444336, | |
| "learning_rate": 2.2398589065255733e-05, | |
| "loss": 0.4079, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 5.990491283676704, | |
| "grad_norm": 13.913603782653809, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.3975, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5133689839572193, | |
| "eval_loss": 2.522934913635254, | |
| "eval_runtime": 147.3195, | |
| "eval_samples_per_second": 3.808, | |
| "eval_steps_per_second": 0.957, | |
| "step": 1893 | |
| }, | |
| { | |
| "epoch": 6.022187004754358, | |
| "grad_norm": 3.1691958904266357, | |
| "learning_rate": 2.2045855379188714e-05, | |
| "loss": 0.2654, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.053882725832013, | |
| "grad_norm": 9.656085014343262, | |
| "learning_rate": 2.18694885361552e-05, | |
| "loss": 0.3378, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 6.085578446909667, | |
| "grad_norm": 6.1951375007629395, | |
| "learning_rate": 2.1693121693121692e-05, | |
| "loss": 0.2532, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 6.117274167987322, | |
| "grad_norm": 17.4242000579834, | |
| "learning_rate": 2.1516754850088183e-05, | |
| "loss": 0.3027, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 6.148969889064976, | |
| "grad_norm": 27.61707305908203, | |
| "learning_rate": 2.1340388007054674e-05, | |
| "loss": 0.2637, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 6.180665610142631, | |
| "grad_norm": 10.691701889038086, | |
| "learning_rate": 2.1164021164021164e-05, | |
| "loss": 0.2931, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 6.212361331220285, | |
| "grad_norm": 35.979496002197266, | |
| "learning_rate": 2.0987654320987655e-05, | |
| "loss": 0.2336, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 6.24405705229794, | |
| "grad_norm": 21.92683219909668, | |
| "learning_rate": 2.0811287477954146e-05, | |
| "loss": 0.3806, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 6.2757527733755945, | |
| "grad_norm": 8.734770774841309, | |
| "learning_rate": 2.0634920634920636e-05, | |
| "loss": 0.2853, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 6.307448494453249, | |
| "grad_norm": 13.567136764526367, | |
| "learning_rate": 2.0458553791887124e-05, | |
| "loss": 0.2756, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 6.3391442155309035, | |
| "grad_norm": 14.175586700439453, | |
| "learning_rate": 2.0282186948853614e-05, | |
| "loss": 0.3459, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.370839936608558, | |
| "grad_norm": 6.7237749099731445, | |
| "learning_rate": 2.0105820105820105e-05, | |
| "loss": 0.13, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 6.4025356576862125, | |
| "grad_norm": 3.6393444538116455, | |
| "learning_rate": 1.9929453262786596e-05, | |
| "loss": 0.1789, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 6.434231378763867, | |
| "grad_norm": 4.694659233093262, | |
| "learning_rate": 1.9753086419753087e-05, | |
| "loss": 0.2228, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 6.4659270998415215, | |
| "grad_norm": 9.12720775604248, | |
| "learning_rate": 1.9576719576719577e-05, | |
| "loss": 0.2706, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 6.497622820919176, | |
| "grad_norm": 18.74335479736328, | |
| "learning_rate": 1.9400352733686068e-05, | |
| "loss": 0.2674, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.5293185419968305, | |
| "grad_norm": 7.866713523864746, | |
| "learning_rate": 1.922398589065256e-05, | |
| "loss": 0.3087, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 6.561014263074485, | |
| "grad_norm": 4.733059406280518, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 0.2031, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 6.5927099841521395, | |
| "grad_norm": 15.15184497833252, | |
| "learning_rate": 1.8871252204585537e-05, | |
| "loss": 0.18, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 6.624405705229794, | |
| "grad_norm": 19.79001235961914, | |
| "learning_rate": 1.8694885361552028e-05, | |
| "loss": 0.2417, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 6.6561014263074485, | |
| "grad_norm": 13.751955032348633, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.2938, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.687797147385103, | |
| "grad_norm": 15.962373733520508, | |
| "learning_rate": 1.834215167548501e-05, | |
| "loss": 0.1829, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 6.7194928684627575, | |
| "grad_norm": 15.361172676086426, | |
| "learning_rate": 1.81657848324515e-05, | |
| "loss": 0.1573, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 6.751188589540412, | |
| "grad_norm": 3.690701961517334, | |
| "learning_rate": 1.798941798941799e-05, | |
| "loss": 0.3055, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 6.7828843106180665, | |
| "grad_norm": 10.492719650268555, | |
| "learning_rate": 1.781305114638448e-05, | |
| "loss": 0.2244, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 6.814580031695721, | |
| "grad_norm": 13.18823528289795, | |
| "learning_rate": 1.763668430335097e-05, | |
| "loss": 0.2039, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 6.8462757527733755, | |
| "grad_norm": 5.904299259185791, | |
| "learning_rate": 1.746031746031746e-05, | |
| "loss": 0.2987, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 6.87797147385103, | |
| "grad_norm": 25.814712524414062, | |
| "learning_rate": 1.728395061728395e-05, | |
| "loss": 0.2196, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 6.9096671949286845, | |
| "grad_norm": 17.685895919799805, | |
| "learning_rate": 1.710758377425044e-05, | |
| "loss": 0.2306, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 6.941362916006339, | |
| "grad_norm": 34.82439422607422, | |
| "learning_rate": 1.693121693121693e-05, | |
| "loss": 0.2668, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 6.9730586370839935, | |
| "grad_norm": 6.761794567108154, | |
| "learning_rate": 1.6754850088183422e-05, | |
| "loss": 0.2648, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 6.998415213946117, | |
| "eval_accuracy": 0.5080213903743316, | |
| "eval_loss": 3.0332891941070557, | |
| "eval_runtime": 147.5737, | |
| "eval_samples_per_second": 3.801, | |
| "eval_steps_per_second": 0.955, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 7.004754358161648, | |
| "grad_norm": 18.470195770263672, | |
| "learning_rate": 1.6578483245149913e-05, | |
| "loss": 0.1705, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 7.0364500792393025, | |
| "grad_norm": 13.634916305541992, | |
| "learning_rate": 1.6402116402116404e-05, | |
| "loss": 0.1345, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 7.068145800316957, | |
| "grad_norm": 7.778036594390869, | |
| "learning_rate": 1.622574955908289e-05, | |
| "loss": 0.0799, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 7.0998415213946116, | |
| "grad_norm": 15.194972038269043, | |
| "learning_rate": 1.604938271604938e-05, | |
| "loss": 0.1687, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 7.131537242472266, | |
| "grad_norm": 7.493090629577637, | |
| "learning_rate": 1.5873015873015872e-05, | |
| "loss": 0.0908, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 7.163232963549921, | |
| "grad_norm": 1.1790108680725098, | |
| "learning_rate": 1.5696649029982363e-05, | |
| "loss": 0.1564, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 7.194928684627575, | |
| "grad_norm": 20.04016876220703, | |
| "learning_rate": 1.5520282186948854e-05, | |
| "loss": 0.166, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 7.22662440570523, | |
| "grad_norm": 3.1165976524353027, | |
| "learning_rate": 1.5343915343915344e-05, | |
| "loss": 0.1283, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 7.258320126782884, | |
| "grad_norm": 23.948923110961914, | |
| "learning_rate": 1.5167548500881834e-05, | |
| "loss": 0.176, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 7.290015847860539, | |
| "grad_norm": 3.0111184120178223, | |
| "learning_rate": 1.4991181657848324e-05, | |
| "loss": 0.1272, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.321711568938193, | |
| "grad_norm": 11.723183631896973, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.1757, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 7.353407290015848, | |
| "grad_norm": 35.94911193847656, | |
| "learning_rate": 1.4638447971781306e-05, | |
| "loss": 0.1217, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 7.385103011093502, | |
| "grad_norm": 11.777710914611816, | |
| "learning_rate": 1.4462081128747795e-05, | |
| "loss": 0.1587, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 7.416798732171157, | |
| "grad_norm": 31.9948673248291, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.116, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 7.448494453248811, | |
| "grad_norm": 9.983025550842285, | |
| "learning_rate": 1.4109347442680776e-05, | |
| "loss": 0.1015, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 7.480190174326466, | |
| "grad_norm": 13.77480411529541, | |
| "learning_rate": 1.3932980599647267e-05, | |
| "loss": 0.1539, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 7.51188589540412, | |
| "grad_norm": 18.11697769165039, | |
| "learning_rate": 1.3756613756613756e-05, | |
| "loss": 0.1695, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 7.543581616481775, | |
| "grad_norm": 0.5398086905479431, | |
| "learning_rate": 1.3580246913580247e-05, | |
| "loss": 0.0802, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 7.575277337559429, | |
| "grad_norm": 22.803916931152344, | |
| "learning_rate": 1.3403880070546737e-05, | |
| "loss": 0.1697, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 7.606973058637084, | |
| "grad_norm": 24.52759552001953, | |
| "learning_rate": 1.3227513227513228e-05, | |
| "loss": 0.1614, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.638668779714738, | |
| "grad_norm": 14.2177734375, | |
| "learning_rate": 1.3051146384479717e-05, | |
| "loss": 0.1197, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 7.6703645007923935, | |
| "grad_norm": 3.4550652503967285, | |
| "learning_rate": 1.2874779541446208e-05, | |
| "loss": 0.0921, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 7.702060221870047, | |
| "grad_norm": 2.2358646392822266, | |
| "learning_rate": 1.2698412698412699e-05, | |
| "loss": 0.1467, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 7.7337559429477025, | |
| "grad_norm": 24.631845474243164, | |
| "learning_rate": 1.252204585537919e-05, | |
| "loss": 0.2073, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 7.765451664025356, | |
| "grad_norm": 1.8313311338424683, | |
| "learning_rate": 1.2345679012345678e-05, | |
| "loss": 0.0878, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 7.7971473851030115, | |
| "grad_norm": 5.266528606414795, | |
| "learning_rate": 1.2169312169312169e-05, | |
| "loss": 0.1029, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.828843106180665, | |
| "grad_norm": 5.7267632484436035, | |
| "learning_rate": 1.199294532627866e-05, | |
| "loss": 0.1302, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 7.8605388272583205, | |
| "grad_norm": 24.080692291259766, | |
| "learning_rate": 1.181657848324515e-05, | |
| "loss": 0.1879, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 7.892234548335974, | |
| "grad_norm": 29.952129364013672, | |
| "learning_rate": 1.164021164021164e-05, | |
| "loss": 0.1314, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 7.9239302694136295, | |
| "grad_norm": 10.011947631835938, | |
| "learning_rate": 1.146384479717813e-05, | |
| "loss": 0.1335, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 7.955625990491284, | |
| "grad_norm": 4.504798412322998, | |
| "learning_rate": 1.1287477954144621e-05, | |
| "loss": 0.1474, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 7.9873217115689386, | |
| "grad_norm": 13.406722068786621, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1232, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5240641711229946, | |
| "eval_loss": 3.4625513553619385, | |
| "eval_runtime": 147.6895, | |
| "eval_samples_per_second": 3.799, | |
| "eval_steps_per_second": 0.955, | |
| "step": 2524 | |
| }, | |
| { | |
| "epoch": 8.019017432646592, | |
| "grad_norm": 2.257606267929077, | |
| "learning_rate": 1.09347442680776e-05, | |
| "loss": 0.1443, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 8.050713153724248, | |
| "grad_norm": 4.254097938537598, | |
| "learning_rate": 1.0758377425044091e-05, | |
| "loss": 0.0859, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 8.082408874801901, | |
| "grad_norm": 16.948497772216797, | |
| "learning_rate": 1.0582010582010582e-05, | |
| "loss": 0.0859, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 8.114104595879557, | |
| "grad_norm": 10.674163818359375, | |
| "learning_rate": 1.0405643738977073e-05, | |
| "loss": 0.0403, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 8.14580031695721, | |
| "grad_norm": 31.626522064208984, | |
| "learning_rate": 1.0229276895943562e-05, | |
| "loss": 0.1069, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 8.177496038034866, | |
| "grad_norm": 14.414571762084961, | |
| "learning_rate": 1.0052910052910053e-05, | |
| "loss": 0.0578, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 8.20919175911252, | |
| "grad_norm": 12.272441864013672, | |
| "learning_rate": 9.876543209876543e-06, | |
| "loss": 0.1106, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 8.240887480190175, | |
| "grad_norm": 39.95854187011719, | |
| "learning_rate": 9.700176366843034e-06, | |
| "loss": 0.1382, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.272583201267828, | |
| "grad_norm": 5.969349384307861, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 0.0664, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 8.304278922345484, | |
| "grad_norm": 6.0112128257751465, | |
| "learning_rate": 9.347442680776014e-06, | |
| "loss": 0.066, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 8.335974643423137, | |
| "grad_norm": 9.331475257873535, | |
| "learning_rate": 9.171075837742504e-06, | |
| "loss": 0.0728, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 8.367670364500793, | |
| "grad_norm": 1.4397279024124146, | |
| "learning_rate": 8.994708994708995e-06, | |
| "loss": 0.0607, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 8.399366085578446, | |
| "grad_norm": 16.279735565185547, | |
| "learning_rate": 8.818342151675484e-06, | |
| "loss": 0.1049, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 8.431061806656102, | |
| "grad_norm": 28.912487030029297, | |
| "learning_rate": 8.641975308641975e-06, | |
| "loss": 0.1594, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 8.462757527733755, | |
| "grad_norm": 7.044099807739258, | |
| "learning_rate": 8.465608465608466e-06, | |
| "loss": 0.0905, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 8.49445324881141, | |
| "grad_norm": 1.4151721000671387, | |
| "learning_rate": 8.289241622574956e-06, | |
| "loss": 0.1646, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 8.526148969889064, | |
| "grad_norm": 8.555072784423828, | |
| "learning_rate": 8.112874779541445e-06, | |
| "loss": 0.0751, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 8.55784469096672, | |
| "grad_norm": 10.666037559509277, | |
| "learning_rate": 7.936507936507936e-06, | |
| "loss": 0.0631, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.589540412044373, | |
| "grad_norm": 21.815340042114258, | |
| "learning_rate": 7.760141093474427e-06, | |
| "loss": 0.0646, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 8.621236133122029, | |
| "grad_norm": 2.177539587020874, | |
| "learning_rate": 7.583774250440917e-06, | |
| "loss": 0.1986, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 8.652931854199682, | |
| "grad_norm": 3.416982889175415, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.0619, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 8.684627575277338, | |
| "grad_norm": 7.183048248291016, | |
| "learning_rate": 7.231040564373897e-06, | |
| "loss": 0.1483, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 8.716323296354991, | |
| "grad_norm": 27.301572799682617, | |
| "learning_rate": 7.054673721340388e-06, | |
| "loss": 0.0735, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 8.748019017432647, | |
| "grad_norm": 4.7548651695251465, | |
| "learning_rate": 6.878306878306878e-06, | |
| "loss": 0.075, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 8.7797147385103, | |
| "grad_norm": 19.81043815612793, | |
| "learning_rate": 6.701940035273369e-06, | |
| "loss": 0.0404, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 8.811410459587956, | |
| "grad_norm": 3.3048360347747803, | |
| "learning_rate": 6.5255731922398585e-06, | |
| "loss": 0.084, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 8.843106180665611, | |
| "grad_norm": 22.454687118530273, | |
| "learning_rate": 6.349206349206349e-06, | |
| "loss": 0.0535, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 8.874801901743265, | |
| "grad_norm": 10.355389595031738, | |
| "learning_rate": 6.172839506172839e-06, | |
| "loss": 0.101, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 8.906497622820918, | |
| "grad_norm": 1.525553822517395, | |
| "learning_rate": 5.99647266313933e-06, | |
| "loss": 0.0479, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 8.938193343898574, | |
| "grad_norm": 2.0415096282958984, | |
| "learning_rate": 5.82010582010582e-06, | |
| "loss": 0.0523, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 8.969889064976229, | |
| "grad_norm": 4.97711706161499, | |
| "learning_rate": 5.6437389770723105e-06, | |
| "loss": 0.0873, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 8.998415213946117, | |
| "eval_accuracy": 0.5026737967914439, | |
| "eval_loss": 3.6218764781951904, | |
| "eval_runtime": 146.8598, | |
| "eval_samples_per_second": 3.82, | |
| "eval_steps_per_second": 0.96, | |
| "step": 2839 | |
| }, | |
| { | |
| "epoch": 9.001584786053883, | |
| "grad_norm": 2.5172929763793945, | |
| "learning_rate": 5.4673721340388e-06, | |
| "loss": 0.042, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 9.033280507131538, | |
| "grad_norm": 19.426034927368164, | |
| "learning_rate": 5.291005291005291e-06, | |
| "loss": 0.0626, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 9.064976228209192, | |
| "grad_norm": 5.145648956298828, | |
| "learning_rate": 5.114638447971781e-06, | |
| "loss": 0.0465, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 9.096671949286847, | |
| "grad_norm": 5.119009971618652, | |
| "learning_rate": 4.938271604938272e-06, | |
| "loss": 0.0241, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 9.1283676703645, | |
| "grad_norm": 3.799389362335205, | |
| "learning_rate": 4.7619047619047615e-06, | |
| "loss": 0.1072, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 9.160063391442156, | |
| "grad_norm": 10.141913414001465, | |
| "learning_rate": 4.585537918871252e-06, | |
| "loss": 0.0517, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 9.19175911251981, | |
| "grad_norm": 26.335498809814453, | |
| "learning_rate": 4.409171075837742e-06, | |
| "loss": 0.0353, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 9.223454833597465, | |
| "grad_norm": 72.42816162109375, | |
| "learning_rate": 4.232804232804233e-06, | |
| "loss": 0.1109, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 9.255150554675119, | |
| "grad_norm": 6.7033538818359375, | |
| "learning_rate": 4.056437389770723e-06, | |
| "loss": 0.0249, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 9.286846275752774, | |
| "grad_norm": 0.21264280378818512, | |
| "learning_rate": 3.8800705467372134e-06, | |
| "loss": 0.0571, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 9.318541996830428, | |
| "grad_norm": 0.8612923622131348, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0355, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 9.350237717908083, | |
| "grad_norm": 0.4571894407272339, | |
| "learning_rate": 3.527336860670194e-06, | |
| "loss": 0.0297, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 9.381933438985737, | |
| "grad_norm": 0.8260334730148315, | |
| "learning_rate": 3.3509700176366843e-06, | |
| "loss": 0.0477, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 9.413629160063392, | |
| "grad_norm": 15.636239051818848, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "loss": 0.0422, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 9.445324881141046, | |
| "grad_norm": 20.841358184814453, | |
| "learning_rate": 2.998236331569665e-06, | |
| "loss": 0.0369, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 9.477020602218701, | |
| "grad_norm": 1.6072678565979004, | |
| "learning_rate": 2.8218694885361552e-06, | |
| "loss": 0.0503, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 9.508716323296355, | |
| "grad_norm": 1.6855651140213013, | |
| "learning_rate": 2.6455026455026455e-06, | |
| "loss": 0.0472, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.54041204437401, | |
| "grad_norm": 4.713263988494873, | |
| "learning_rate": 2.469135802469136e-06, | |
| "loss": 0.1097, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 9.572107765451664, | |
| "grad_norm": 2.9448232650756836, | |
| "learning_rate": 2.292768959435626e-06, | |
| "loss": 0.0553, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 9.60380348652932, | |
| "grad_norm": 18.76097869873047, | |
| "learning_rate": 2.1164021164021164e-06, | |
| "loss": 0.0342, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 9.635499207606973, | |
| "grad_norm": 6.058113098144531, | |
| "learning_rate": 1.9400352733686067e-06, | |
| "loss": 0.041, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 9.667194928684628, | |
| "grad_norm": 32.21320343017578, | |
| "learning_rate": 1.763668430335097e-06, | |
| "loss": 0.0386, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 9.698890649762282, | |
| "grad_norm": 0.48055747151374817, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "loss": 0.0454, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 9.730586370839937, | |
| "grad_norm": 7.426022529602051, | |
| "learning_rate": 1.4109347442680776e-06, | |
| "loss": 0.1654, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 9.76228209191759, | |
| "grad_norm": 7.643261909484863, | |
| "learning_rate": 1.234567901234568e-06, | |
| "loss": 0.0294, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 9.793977812995246, | |
| "grad_norm": 3.1924824714660645, | |
| "learning_rate": 1.0582010582010582e-06, | |
| "loss": 0.1084, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 9.8256735340729, | |
| "grad_norm": 31.949647903442383, | |
| "learning_rate": 8.818342151675485e-07, | |
| "loss": 0.0582, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 9.857369255150555, | |
| "grad_norm": 41.80002212524414, | |
| "learning_rate": 7.054673721340388e-07, | |
| "loss": 0.0771, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 9.889064976228209, | |
| "grad_norm": 1.5549315214157104, | |
| "learning_rate": 5.291005291005291e-07, | |
| "loss": 0.0501, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 9.920760697305864, | |
| "grad_norm": 14.312984466552734, | |
| "learning_rate": 3.527336860670194e-07, | |
| "loss": 0.0516, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 9.952456418383518, | |
| "grad_norm": 3.448463201522827, | |
| "learning_rate": 1.763668430335097e-07, | |
| "loss": 0.0653, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 9.984152139461173, | |
| "grad_norm": 5.9251389503479, | |
| "learning_rate": 0.0, | |
| "loss": 0.0554, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 9.984152139461173, | |
| "eval_accuracy": 0.49376114081996436, | |
| "eval_loss": 3.771087884902954, | |
| "eval_runtime": 147.6096, | |
| "eval_samples_per_second": 3.801, | |
| "eval_steps_per_second": 0.955, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 9.984152139461173, | |
| "step": 3150, | |
| "total_flos": 2.6101205954547646e+19, | |
| "train_loss": 0.7937991834822156, | |
| "train_runtime": 16586.4479, | |
| "train_samples_per_second": 3.042, | |
| "train_steps_per_second": 0.19 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 2.6101205954547646e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |