| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999929740743343, | |
| "eval_steps": 100, | |
| "global_step": 1779, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 30.728321433017737, | |
| "learning_rate": 2.9915682967959526e-07, | |
| "loss": 0.7843, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 36.77551634614557, | |
| "learning_rate": 2.9831365935919053e-07, | |
| "loss": 0.7146, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 35.07871745362434, | |
| "learning_rate": 2.9747048903878585e-07, | |
| "loss": 0.6938, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 22.884080905805792, | |
| "learning_rate": 2.9662731871838107e-07, | |
| "loss": 0.6587, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 20.00517219363236, | |
| "learning_rate": 2.957841483979764e-07, | |
| "loss": 0.653, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 18.36397890422735, | |
| "learning_rate": 2.9494097807757167e-07, | |
| "loss": 0.6644, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 25.380353671575364, | |
| "learning_rate": 2.9409780775716694e-07, | |
| "loss": 0.6591, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 22.96117971805064, | |
| "learning_rate": 2.932546374367622e-07, | |
| "loss": 0.6649, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 18.570972155544734, | |
| "learning_rate": 2.924114671163575e-07, | |
| "loss": 0.6461, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 31.463777178706607, | |
| "learning_rate": 2.915682967959528e-07, | |
| "loss": 0.6402, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_accuracy": 0.71716621253406, | |
| "eval_loss": 0.6716727614402771, | |
| "eval_runtime": 81.6137, | |
| "eval_samples_per_second": 22.484, | |
| "eval_steps_per_second": 0.711, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 16.52215127003667, | |
| "learning_rate": 2.90725126475548e-07, | |
| "loss": 0.6428, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 19.497709198702577, | |
| "learning_rate": 2.8988195615514335e-07, | |
| "loss": 0.6215, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 18.350616329827186, | |
| "learning_rate": 2.890387858347386e-07, | |
| "loss": 0.6408, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 21.31149136639458, | |
| "learning_rate": 2.881956155143339e-07, | |
| "loss": 0.6435, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 18.652897776189004, | |
| "learning_rate": 2.8735244519392916e-07, | |
| "loss": 0.6429, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 15.905534553134705, | |
| "learning_rate": 2.8650927487352443e-07, | |
| "loss": 0.6406, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 16.811536357450365, | |
| "learning_rate": 2.856661045531197e-07, | |
| "loss": 0.6433, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 18.105305932357698, | |
| "learning_rate": 2.84822934232715e-07, | |
| "loss": 0.6305, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 17.99993276337699, | |
| "learning_rate": 2.839797639123103e-07, | |
| "loss": 0.6339, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 19.99711675781162, | |
| "learning_rate": 2.8313659359190557e-07, | |
| "loss": 0.6293, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_accuracy": 0.740599455040872, | |
| "eval_loss": 0.6711810231208801, | |
| "eval_runtime": 81.6564, | |
| "eval_samples_per_second": 22.472, | |
| "eval_steps_per_second": 0.71, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 16.424824162884995, | |
| "learning_rate": 2.8229342327150084e-07, | |
| "loss": 0.6053, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 27.512486316457252, | |
| "learning_rate": 2.814502529510961e-07, | |
| "loss": 0.6109, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 22.320535945754912, | |
| "learning_rate": 2.806070826306914e-07, | |
| "loss": 0.6358, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 13.83846516782303, | |
| "learning_rate": 2.7976391231028666e-07, | |
| "loss": 0.6261, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 16.27927043935017, | |
| "learning_rate": 2.7892074198988193e-07, | |
| "loss": 0.6395, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 21.72712116315618, | |
| "learning_rate": 2.780775716694772e-07, | |
| "loss": 0.6273, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 19.022167698615288, | |
| "learning_rate": 2.7723440134907247e-07, | |
| "loss": 0.6206, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 16.64363952350825, | |
| "learning_rate": 2.763912310286678e-07, | |
| "loss": 0.6304, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 20.294764983153055, | |
| "learning_rate": 2.7554806070826307e-07, | |
| "loss": 0.6225, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 11.467691935838706, | |
| "learning_rate": 2.7470489038785834e-07, | |
| "loss": 0.6297, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_accuracy": 0.7444141689373297, | |
| "eval_loss": 0.658985435962677, | |
| "eval_runtime": 81.6887, | |
| "eval_samples_per_second": 22.463, | |
| "eval_steps_per_second": 0.71, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 21.910401668214213, | |
| "learning_rate": 2.738617200674536e-07, | |
| "loss": 0.6182, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 24.007473629091503, | |
| "learning_rate": 2.730185497470489e-07, | |
| "loss": 0.6038, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 21.290646545312395, | |
| "learning_rate": 2.7217537942664415e-07, | |
| "loss": 0.6195, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 23.68987516909537, | |
| "learning_rate": 2.713322091062394e-07, | |
| "loss": 0.6308, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 30.80192565983983, | |
| "learning_rate": 2.704890387858347e-07, | |
| "loss": 0.6341, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 21.8133450286374, | |
| "learning_rate": 2.6964586846543e-07, | |
| "loss": 0.6106, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 20.299330239483883, | |
| "learning_rate": 2.688026981450253e-07, | |
| "loss": 0.64, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 20.648521192709566, | |
| "learning_rate": 2.6795952782462056e-07, | |
| "loss": 0.6169, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 30.447858291866996, | |
| "learning_rate": 2.6711635750421584e-07, | |
| "loss": 0.6416, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 15.933419590444712, | |
| "learning_rate": 2.662731871838111e-07, | |
| "loss": 0.6112, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "eval_accuracy": 0.7667574931880109, | |
| "eval_loss": 0.6638922095298767, | |
| "eval_runtime": 81.6313, | |
| "eval_samples_per_second": 22.479, | |
| "eval_steps_per_second": 0.711, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 23.46830344261005, | |
| "learning_rate": 2.654300168634064e-07, | |
| "loss": 0.6277, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 27.935995699348798, | |
| "learning_rate": 2.6458684654300165e-07, | |
| "loss": 0.6221, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 17.973218257874134, | |
| "learning_rate": 2.63743676222597e-07, | |
| "loss": 0.6477, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 18.376434311461416, | |
| "learning_rate": 2.629005059021922e-07, | |
| "loss": 0.6309, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 18.526800694282237, | |
| "learning_rate": 2.620573355817875e-07, | |
| "loss": 0.6316, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 15.15241254088161, | |
| "learning_rate": 2.612141652613828e-07, | |
| "loss": 0.6201, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 14.410194781468793, | |
| "learning_rate": 2.6037099494097806e-07, | |
| "loss": 0.6218, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 40.740369918242074, | |
| "learning_rate": 2.5952782462057333e-07, | |
| "loss": 0.609, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 17.442343289782034, | |
| "learning_rate": 2.586846543001686e-07, | |
| "loss": 0.6031, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 14.525640800311784, | |
| "learning_rate": 2.5784148397976393e-07, | |
| "loss": 0.6065, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_accuracy": 0.7623978201634878, | |
| "eval_loss": 0.6653993725776672, | |
| "eval_runtime": 81.6511, | |
| "eval_samples_per_second": 22.474, | |
| "eval_steps_per_second": 0.71, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 16.37698400765353, | |
| "learning_rate": 2.5699831365935915e-07, | |
| "loss": 0.6044, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 18.46191451753819, | |
| "learning_rate": 2.5615514333895447e-07, | |
| "loss": 0.6026, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 16.60110193637955, | |
| "learning_rate": 2.5531197301854974e-07, | |
| "loss": 0.6117, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 12.90641142874894, | |
| "learning_rate": 2.54468802698145e-07, | |
| "loss": 0.6218, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 12.782297037944778, | |
| "learning_rate": 2.536256323777403e-07, | |
| "loss": 0.6094, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 13.132450106152808, | |
| "learning_rate": 2.5278246205733556e-07, | |
| "loss": 0.6065, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 15.505668610650245, | |
| "learning_rate": 2.519392917369309e-07, | |
| "loss": 0.6141, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 10.877242291946278, | |
| "learning_rate": 2.510961214165261e-07, | |
| "loss": 0.6132, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 16.260577656983788, | |
| "learning_rate": 2.502529510961214e-07, | |
| "loss": 0.618, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 17.74551723827062, | |
| "learning_rate": 2.494097807757167e-07, | |
| "loss": 0.6011, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "eval_accuracy": 0.7656675749318801, | |
| "eval_loss": 0.6725944876670837, | |
| "eval_runtime": 81.6007, | |
| "eval_samples_per_second": 22.488, | |
| "eval_steps_per_second": 0.711, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 13.489006632275235, | |
| "learning_rate": 2.4856661045531197e-07, | |
| "loss": 0.5972, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 11.884271769605707, | |
| "learning_rate": 2.4772344013490724e-07, | |
| "loss": 0.62, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 13.261788153937319, | |
| "learning_rate": 2.468802698145025e-07, | |
| "loss": 0.6167, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 14.172098159046573, | |
| "learning_rate": 2.4603709949409783e-07, | |
| "loss": 0.6153, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 21.378776702080216, | |
| "learning_rate": 2.4519392917369305e-07, | |
| "loss": 0.619, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 12.09294701146547, | |
| "learning_rate": 2.443507588532884e-07, | |
| "loss": 0.6282, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 14.012823513290494, | |
| "learning_rate": 2.4350758853288365e-07, | |
| "loss": 0.6329, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 12.10834170031786, | |
| "learning_rate": 2.426644182124789e-07, | |
| "loss": 0.6302, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 18.577060775724657, | |
| "learning_rate": 2.418212478920742e-07, | |
| "loss": 0.6311, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 13.914957035063992, | |
| "learning_rate": 2.4097807757166946e-07, | |
| "loss": 0.6188, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_accuracy": 0.7716621253405994, | |
| "eval_loss": 0.6559094190597534, | |
| "eval_runtime": 81.6864, | |
| "eval_samples_per_second": 22.464, | |
| "eval_steps_per_second": 0.71, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 18.235947830015988, | |
| "learning_rate": 2.4013490725126473e-07, | |
| "loss": 0.6333, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 18.663534626617114, | |
| "learning_rate": 2.3929173693086e-07, | |
| "loss": 0.6277, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 15.048399094387358, | |
| "learning_rate": 2.3844856661045533e-07, | |
| "loss": 0.6209, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 14.114525067135927, | |
| "learning_rate": 2.3760539629005057e-07, | |
| "loss": 0.6192, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 19.517752622102307, | |
| "learning_rate": 2.3676222596964587e-07, | |
| "loss": 0.6016, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 11.012158965376576, | |
| "learning_rate": 2.3591905564924112e-07, | |
| "loss": 0.619, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 12.289099272322433, | |
| "learning_rate": 2.3507588532883641e-07, | |
| "loss": 0.625, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 11.949697745076817, | |
| "learning_rate": 2.3423271500843169e-07, | |
| "loss": 0.6147, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 20.873971042525024, | |
| "learning_rate": 2.3338954468802698e-07, | |
| "loss": 0.6059, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 23.749605034362084, | |
| "learning_rate": 2.3254637436762223e-07, | |
| "loss": 0.5964, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "eval_accuracy": 0.7623978201634878, | |
| "eval_loss": 0.6648799180984497, | |
| "eval_runtime": 81.6698, | |
| "eval_samples_per_second": 22.469, | |
| "eval_steps_per_second": 0.71, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 17.378412987521198, | |
| "learning_rate": 2.3170320404721753e-07, | |
| "loss": 0.5999, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 14.09219874944554, | |
| "learning_rate": 2.3086003372681282e-07, | |
| "loss": 0.6123, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 15.751831833203504, | |
| "learning_rate": 2.3001686340640807e-07, | |
| "loss": 0.6173, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 10.58472297190868, | |
| "learning_rate": 2.2917369308600337e-07, | |
| "loss": 0.6107, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 16.766833417242065, | |
| "learning_rate": 2.2833052276559864e-07, | |
| "loss": 0.5977, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 10.191079428211454, | |
| "learning_rate": 2.2748735244519394e-07, | |
| "loss": 0.6249, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 10.892435339015812, | |
| "learning_rate": 2.2664418212478918e-07, | |
| "loss": 0.6071, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 14.857284393951572, | |
| "learning_rate": 2.2580101180438448e-07, | |
| "loss": 0.6277, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 16.280040840533015, | |
| "learning_rate": 2.2495784148397975e-07, | |
| "loss": 0.6117, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 15.402943403313595, | |
| "learning_rate": 2.2411467116357502e-07, | |
| "loss": 0.6263, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_accuracy": 0.7596730245231608, | |
| "eval_loss": 0.6583617329597473, | |
| "eval_runtime": 81.6999, | |
| "eval_samples_per_second": 22.46, | |
| "eval_steps_per_second": 0.71, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 16.324857723183708, | |
| "learning_rate": 2.2327150084317032e-07, | |
| "loss": 0.6189, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 10.98662853747513, | |
| "learning_rate": 2.224283305227656e-07, | |
| "loss": 0.6097, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 17.67105325082076, | |
| "learning_rate": 2.215851602023609e-07, | |
| "loss": 0.6122, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 20.41264564438468, | |
| "learning_rate": 2.2074198988195613e-07, | |
| "loss": 0.6133, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 17.79087613152158, | |
| "learning_rate": 2.1989881956155143e-07, | |
| "loss": 0.6211, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 14.606633495853552, | |
| "learning_rate": 2.1905564924114668e-07, | |
| "loss": 0.6127, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 15.658044929958105, | |
| "learning_rate": 2.1821247892074197e-07, | |
| "loss": 0.613, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 10.939022037806527, | |
| "learning_rate": 2.1736930860033725e-07, | |
| "loss": 0.611, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 17.074626515502477, | |
| "learning_rate": 2.1652613827993254e-07, | |
| "loss": 0.6235, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 9.873729304084089, | |
| "learning_rate": 2.1568296795952782e-07, | |
| "loss": 0.6173, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "eval_accuracy": 0.7787465940054495, | |
| "eval_loss": 0.6586682796478271, | |
| "eval_runtime": 81.6576, | |
| "eval_samples_per_second": 22.472, | |
| "eval_steps_per_second": 0.71, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 17.977983053245335, | |
| "learning_rate": 2.148397976391231e-07, | |
| "loss": 0.6085, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 22.720038161057953, | |
| "learning_rate": 2.1399662731871838e-07, | |
| "loss": 0.6077, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 14.934140128371691, | |
| "learning_rate": 2.1315345699831363e-07, | |
| "loss": 0.5936, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 17.025807021499876, | |
| "learning_rate": 2.1231028667790893e-07, | |
| "loss": 0.6077, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 13.012674899332776, | |
| "learning_rate": 2.114671163575042e-07, | |
| "loss": 0.6143, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 12.997561334592964, | |
| "learning_rate": 2.106239460370995e-07, | |
| "loss": 0.6079, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 12.442902629191648, | |
| "learning_rate": 2.0978077571669474e-07, | |
| "loss": 0.6027, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 16.45898926071221, | |
| "learning_rate": 2.0893760539629004e-07, | |
| "loss": 0.6027, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 21.037037529928906, | |
| "learning_rate": 2.0809443507588534e-07, | |
| "loss": 0.6048, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 9.922931220178954, | |
| "learning_rate": 2.0725126475548058e-07, | |
| "loss": 0.6133, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "eval_accuracy": 0.7754768392370572, | |
| "eval_loss": 0.6589598655700684, | |
| "eval_runtime": 81.5819, | |
| "eval_samples_per_second": 22.493, | |
| "eval_steps_per_second": 0.711, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 16.07724841249849, | |
| "learning_rate": 2.0640809443507588e-07, | |
| "loss": 0.6202, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 12.964684643299604, | |
| "learning_rate": 2.0556492411467115e-07, | |
| "loss": 0.6114, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 11.819826038425457, | |
| "learning_rate": 2.0472175379426645e-07, | |
| "loss": 0.6042, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 10.985202176776713, | |
| "learning_rate": 2.038785834738617e-07, | |
| "loss": 0.5989, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 15.914126285773381, | |
| "learning_rate": 2.03035413153457e-07, | |
| "loss": 0.6111, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 20.198502391005317, | |
| "learning_rate": 2.0219224283305226e-07, | |
| "loss": 0.6193, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 16.375353496861376, | |
| "learning_rate": 2.0134907251264754e-07, | |
| "loss": 0.6124, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 14.424027706264562, | |
| "learning_rate": 2.0050590219224283e-07, | |
| "loss": 0.6095, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 17.448118657270804, | |
| "learning_rate": 1.996627318718381e-07, | |
| "loss": 0.594, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 15.835051950720354, | |
| "learning_rate": 1.988195615514334e-07, | |
| "loss": 0.5902, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "eval_accuracy": 0.7673024523160763, | |
| "eval_loss": 0.6717323660850525, | |
| "eval_runtime": 81.1264, | |
| "eval_samples_per_second": 22.619, | |
| "eval_steps_per_second": 0.715, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 10.104802089341579, | |
| "learning_rate": 1.9797639123102865e-07, | |
| "loss": 0.6083, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 16.88055370044505, | |
| "learning_rate": 1.9713322091062395e-07, | |
| "loss": 0.6208, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 13.25377180665782, | |
| "learning_rate": 1.962900505902192e-07, | |
| "loss": 0.6195, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 12.03529826522156, | |
| "learning_rate": 1.954468802698145e-07, | |
| "loss": 0.6013, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 21.966805810887724, | |
| "learning_rate": 1.9460370994940976e-07, | |
| "loss": 0.5955, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 25.150403183144306, | |
| "learning_rate": 1.9376053962900506e-07, | |
| "loss": 0.6123, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 28.00330027046741, | |
| "learning_rate": 1.9291736930860033e-07, | |
| "loss": 0.6242, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 10.71380749264463, | |
| "learning_rate": 1.920741989881956e-07, | |
| "loss": 0.6288, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 14.346837709212238, | |
| "learning_rate": 1.912310286677909e-07, | |
| "loss": 0.6212, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 11.295633948457253, | |
| "learning_rate": 1.9038785834738614e-07, | |
| "loss": 0.6027, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "eval_accuracy": 0.7787465940054495, | |
| "eval_loss": 0.6629257202148438, | |
| "eval_runtime": 81.1053, | |
| "eval_samples_per_second": 22.625, | |
| "eval_steps_per_second": 0.715, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 11.81270801886205, | |
| "learning_rate": 1.8954468802698144e-07, | |
| "loss": 0.6127, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 11.458987921580261, | |
| "learning_rate": 1.887015177065767e-07, | |
| "loss": 0.6027, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 14.114816619174277, | |
| "learning_rate": 1.87858347386172e-07, | |
| "loss": 0.617, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 9.445103600368194, | |
| "learning_rate": 1.8701517706576726e-07, | |
| "loss": 0.6016, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 16.378432231814056, | |
| "learning_rate": 1.8617200674536255e-07, | |
| "loss": 0.607, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 18.105605603170623, | |
| "learning_rate": 1.8532883642495785e-07, | |
| "loss": 0.6059, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 25.98754135538099, | |
| "learning_rate": 1.844856661045531e-07, | |
| "loss": 0.6079, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 13.387903869057888, | |
| "learning_rate": 1.836424957841484e-07, | |
| "loss": 0.6248, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 15.072500301887933, | |
| "learning_rate": 1.8279932546374367e-07, | |
| "loss": 0.6059, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 28.195360534555086, | |
| "learning_rate": 1.8195615514333896e-07, | |
| "loss": 0.6094, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "eval_accuracy": 0.7825613079019074, | |
| "eval_loss": 0.6670619249343872, | |
| "eval_runtime": 81.1088, | |
| "eval_samples_per_second": 22.624, | |
| "eval_steps_per_second": 0.715, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 16.78097475585066, | |
| "learning_rate": 1.811129848229342e-07, | |
| "loss": 0.6026, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 9.27565083160915, | |
| "learning_rate": 1.802698145025295e-07, | |
| "loss": 0.6061, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 15.151917102254139, | |
| "learning_rate": 1.7942664418212478e-07, | |
| "loss": 0.6121, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 15.628107133180718, | |
| "learning_rate": 1.7858347386172005e-07, | |
| "loss": 0.6136, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 11.41747344754936, | |
| "learning_rate": 1.7774030354131535e-07, | |
| "loss": 0.6013, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 17.94793616613114, | |
| "learning_rate": 1.7689713322091062e-07, | |
| "loss": 0.6094, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 16.847404718639655, | |
| "learning_rate": 1.7605396290050592e-07, | |
| "loss": 0.6006, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 19.943006958068334, | |
| "learning_rate": 1.7521079258010116e-07, | |
| "loss": 0.6203, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 17.544121262898, | |
| "learning_rate": 1.7436762225969646e-07, | |
| "loss": 0.6118, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 16.348862133230483, | |
| "learning_rate": 1.735244519392917e-07, | |
| "loss": 0.606, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_accuracy": 0.7771117166212534, | |
| "eval_loss": 0.6631556749343872, | |
| "eval_runtime": 81.1021, | |
| "eval_samples_per_second": 22.626, | |
| "eval_steps_per_second": 0.715, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 15.065882473200872, | |
| "learning_rate": 1.72681281618887e-07, | |
| "loss": 0.6105, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 10.334423134734958, | |
| "learning_rate": 1.718381112984823e-07, | |
| "loss": 0.603, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 13.630372232978868, | |
| "learning_rate": 1.7099494097807757e-07, | |
| "loss": 0.6177, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 21.60021911475766, | |
| "learning_rate": 1.7015177065767284e-07, | |
| "loss": 0.6138, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 27.660710692587415, | |
| "learning_rate": 1.6930860033726811e-07, | |
| "loss": 0.6058, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 15.6224127860944, | |
| "learning_rate": 1.684654300168634e-07, | |
| "loss": 0.6275, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 9.768689869942213, | |
| "learning_rate": 1.6762225969645866e-07, | |
| "loss": 0.6181, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 29.86529261497021, | |
| "learning_rate": 1.6677908937605395e-07, | |
| "loss": 0.6175, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 12.971455009975848, | |
| "learning_rate": 1.6593591905564923e-07, | |
| "loss": 0.6177, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 15.387843128502691, | |
| "learning_rate": 1.6509274873524452e-07, | |
| "loss": 0.6119, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_accuracy": 0.7754768392370572, | |
| "eval_loss": 0.6606820225715637, | |
| "eval_runtime": 81.096, | |
| "eval_samples_per_second": 22.627, | |
| "eval_steps_per_second": 0.715, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 14.683073977357795, | |
| "learning_rate": 1.642495784148398e-07, | |
| "loss": 0.6132, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 20.101674008953353, | |
| "learning_rate": 1.6340640809443507e-07, | |
| "loss": 0.6129, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 9.565771214700746, | |
| "learning_rate": 1.6256323777403036e-07, | |
| "loss": 0.6032, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 19.63652279943903, | |
| "learning_rate": 1.617200674536256e-07, | |
| "loss": 0.6044, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 11.798398386935885, | |
| "learning_rate": 1.608768971332209e-07, | |
| "loss": 0.6104, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 8.45308349951647, | |
| "learning_rate": 1.6003372681281618e-07, | |
| "loss": 0.5946, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 6.748385507096111, | |
| "learning_rate": 1.5919055649241148e-07, | |
| "loss": 0.6124, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 11.792848227458215, | |
| "learning_rate": 1.5834738617200672e-07, | |
| "loss": 0.6118, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 15.049508549188333, | |
| "learning_rate": 1.5750421585160202e-07, | |
| "loss": 0.6085, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 10.01303081102771, | |
| "learning_rate": 1.5666104553119732e-07, | |
| "loss": 0.5992, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_accuracy": 0.7798365122615804, | |
| "eval_loss": 0.6598241925239563, | |
| "eval_runtime": 81.0778, | |
| "eval_samples_per_second": 22.633, | |
| "eval_steps_per_second": 0.715, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 13.816105839597155, | |
| "learning_rate": 1.5581787521079256e-07, | |
| "loss": 0.614, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 15.916990016500973, | |
| "learning_rate": 1.5497470489038786e-07, | |
| "loss": 0.6071, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 14.639137907706433, | |
| "learning_rate": 1.5413153456998313e-07, | |
| "loss": 0.6069, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 15.234658834038223, | |
| "learning_rate": 1.5328836424957843e-07, | |
| "loss": 0.5919, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 13.325766641648226, | |
| "learning_rate": 1.5244519392917367e-07, | |
| "loss": 0.6179, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 9.90933302447297, | |
| "learning_rate": 1.5160202360876897e-07, | |
| "loss": 0.6017, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 18.624449453795865, | |
| "learning_rate": 1.5075885328836422e-07, | |
| "loss": 0.6055, | |
| "step": 1770 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3558, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |