| { | |
| "best_metric": 0.9832942485809326, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/google_t5/t5_small_patent/checkpoint-1400", | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 2346, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 2.1583292484283447, | |
| "learning_rate": 0.0004978687127024723, | |
| "loss": 2.0566, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.600330114364624, | |
| "learning_rate": 0.0004957374254049446, | |
| "loss": 1.9478, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.930787444114685, | |
| "learning_rate": 0.0004936061381074169, | |
| "loss": 1.8597, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.5029382705688477, | |
| "learning_rate": 0.0004914748508098892, | |
| "loss": 1.5716, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.0470213890075684, | |
| "learning_rate": 0.0004893435635123615, | |
| "loss": 1.4814, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_accuracy": 0.492, | |
| "eval_f1_macro": 0.34138249695786205, | |
| "eval_f1_micro": 0.492, | |
| "eval_loss": 1.4703689813613892, | |
| "eval_runtime": 4.5651, | |
| "eval_samples_per_second": 1095.26, | |
| "eval_steps_per_second": 34.391, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.6178905963897705, | |
| "learning_rate": 0.00048721227621483377, | |
| "loss": 1.3366, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.6837921142578125, | |
| "learning_rate": 0.00048508098891730605, | |
| "loss": 1.3933, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.290954828262329, | |
| "learning_rate": 0.0004829497016197784, | |
| "loss": 1.3195, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.3305504322052, | |
| "learning_rate": 0.00048081841432225065, | |
| "loss": 1.3749, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.9327062368392944, | |
| "learning_rate": 0.0004786871270247229, | |
| "loss": 1.3003, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_accuracy": 0.5512, | |
| "eval_f1_macro": 0.3876411779838728, | |
| "eval_f1_micro": 0.5512, | |
| "eval_loss": 1.2652499675750732, | |
| "eval_runtime": 4.5124, | |
| "eval_samples_per_second": 1108.067, | |
| "eval_steps_per_second": 34.793, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.8468806743621826, | |
| "learning_rate": 0.00047655583972719526, | |
| "loss": 1.3399, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.5669147968292236, | |
| "learning_rate": 0.00047442455242966753, | |
| "loss": 1.2677, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.0333309173583984, | |
| "learning_rate": 0.0004722932651321398, | |
| "loss": 1.264, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.6710681915283203, | |
| "learning_rate": 0.00047016197783461214, | |
| "loss": 1.2956, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.8155871629714966, | |
| "learning_rate": 0.0004680306905370844, | |
| "loss": 1.2291, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_accuracy": 0.563, | |
| "eval_f1_macro": 0.3999749460001978, | |
| "eval_f1_micro": 0.563, | |
| "eval_loss": 1.2304304838180542, | |
| "eval_runtime": 4.58, | |
| "eval_samples_per_second": 1091.714, | |
| "eval_steps_per_second": 34.28, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.5728886127471924, | |
| "learning_rate": 0.0004658994032395567, | |
| "loss": 1.3538, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.9312357902526855, | |
| "learning_rate": 0.000463768115942029, | |
| "loss": 1.3509, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.0137836933135986, | |
| "learning_rate": 0.0004616368286445013, | |
| "loss": 1.2469, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.5608091354370117, | |
| "learning_rate": 0.00045950554134697357, | |
| "loss": 1.2728, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.511885404586792, | |
| "learning_rate": 0.0004573742540494459, | |
| "loss": 1.142, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_accuracy": 0.5894, | |
| "eval_f1_macro": 0.45857413022807464, | |
| "eval_f1_micro": 0.5894, | |
| "eval_loss": 1.164367914199829, | |
| "eval_runtime": 4.5868, | |
| "eval_samples_per_second": 1090.093, | |
| "eval_steps_per_second": 34.229, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 2.0801782608032227, | |
| "learning_rate": 0.0004552429667519182, | |
| "loss": 1.1444, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.706130266189575, | |
| "learning_rate": 0.00045311167945439045, | |
| "loss": 1.2788, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.2432401180267334, | |
| "learning_rate": 0.0004509803921568628, | |
| "loss": 1.2085, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.283546209335327, | |
| "learning_rate": 0.00044884910485933505, | |
| "loss": 1.1881, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.2058613300323486, | |
| "learning_rate": 0.00044671781756180733, | |
| "loss": 1.0581, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.603, | |
| "eval_f1_macro": 0.4563218777873921, | |
| "eval_f1_micro": 0.603, | |
| "eval_loss": 1.1396372318267822, | |
| "eval_runtime": 4.6256, | |
| "eval_samples_per_second": 1080.947, | |
| "eval_steps_per_second": 33.942, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.8488962650299072, | |
| "learning_rate": 0.00044458653026427966, | |
| "loss": 1.1765, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.4596290588378906, | |
| "learning_rate": 0.00044245524296675193, | |
| "loss": 1.1927, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 2.053030014038086, | |
| "learning_rate": 0.0004403239556692242, | |
| "loss": 1.0985, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.664118766784668, | |
| "learning_rate": 0.00043819266837169654, | |
| "loss": 1.1782, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 2.8193089962005615, | |
| "learning_rate": 0.0004360613810741688, | |
| "loss": 1.2415, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_accuracy": 0.613, | |
| "eval_f1_macro": 0.4936808407307906, | |
| "eval_f1_micro": 0.613, | |
| "eval_loss": 1.1215277910232544, | |
| "eval_runtime": 4.6437, | |
| "eval_samples_per_second": 1076.718, | |
| "eval_steps_per_second": 33.809, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.5075747966766357, | |
| "learning_rate": 0.0004339300937766411, | |
| "loss": 1.1684, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 2.044328212738037, | |
| "learning_rate": 0.0004317988064791134, | |
| "loss": 1.0177, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.432016611099243, | |
| "learning_rate": 0.0004296675191815857, | |
| "loss": 1.2053, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.8879696130752563, | |
| "learning_rate": 0.00042753623188405797, | |
| "loss": 1.1831, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.592384099960327, | |
| "learning_rate": 0.0004254049445865303, | |
| "loss": 1.1336, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "eval_accuracy": 0.6172, | |
| "eval_f1_macro": 0.5291747385702272, | |
| "eval_f1_micro": 0.6172, | |
| "eval_loss": 1.1041539907455444, | |
| "eval_runtime": 4.6545, | |
| "eval_samples_per_second": 1074.222, | |
| "eval_steps_per_second": 33.731, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 2.692321538925171, | |
| "learning_rate": 0.0004232736572890026, | |
| "loss": 1.1718, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.972010612487793, | |
| "learning_rate": 0.00042114236999147485, | |
| "loss": 1.2379, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 2.035459041595459, | |
| "learning_rate": 0.0004190110826939472, | |
| "loss": 1.219, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 3.327613115310669, | |
| "learning_rate": 0.00041687979539641946, | |
| "loss": 1.2285, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 2.2528626918792725, | |
| "learning_rate": 0.00041474850809889173, | |
| "loss": 1.045, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_accuracy": 0.624, | |
| "eval_f1_macro": 0.5271110605500514, | |
| "eval_f1_micro": 0.624, | |
| "eval_loss": 1.092409610748291, | |
| "eval_runtime": 4.6526, | |
| "eval_samples_per_second": 1074.679, | |
| "eval_steps_per_second": 33.745, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 3.0680856704711914, | |
| "learning_rate": 0.00041261722080136406, | |
| "loss": 1.1587, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 2.2462947368621826, | |
| "learning_rate": 0.00041048593350383634, | |
| "loss": 1.1927, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 2.637155771255493, | |
| "learning_rate": 0.0004083546462063086, | |
| "loss": 1.1396, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.410139799118042, | |
| "learning_rate": 0.00040622335890878094, | |
| "loss": 1.1035, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 2.298185110092163, | |
| "learning_rate": 0.0004040920716112532, | |
| "loss": 1.1204, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "eval_accuracy": 0.6184, | |
| "eval_f1_macro": 0.5146174720519366, | |
| "eval_f1_micro": 0.6184, | |
| "eval_loss": 1.089658260345459, | |
| "eval_runtime": 4.6519, | |
| "eval_samples_per_second": 1074.824, | |
| "eval_steps_per_second": 33.749, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.443023920059204, | |
| "learning_rate": 0.0004019607843137255, | |
| "loss": 1.1065, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.6598963737487793, | |
| "learning_rate": 0.0003998294970161978, | |
| "loss": 1.143, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.6134822368621826, | |
| "learning_rate": 0.0003976982097186701, | |
| "loss": 1.1093, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 2.171005964279175, | |
| "learning_rate": 0.00039556692242114237, | |
| "loss": 1.2507, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.519437551498413, | |
| "learning_rate": 0.0003934356351236147, | |
| "loss": 1.0691, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.6236, | |
| "eval_f1_macro": 0.51689845984142, | |
| "eval_f1_micro": 0.6236, | |
| "eval_loss": 1.0826703310012817, | |
| "eval_runtime": 4.6344, | |
| "eval_samples_per_second": 1078.897, | |
| "eval_steps_per_second": 33.877, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 3.738205671310425, | |
| "learning_rate": 0.000391304347826087, | |
| "loss": 1.1863, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 2.0040698051452637, | |
| "learning_rate": 0.00038917306052855925, | |
| "loss": 1.1366, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 2.250598907470703, | |
| "learning_rate": 0.0003870417732310316, | |
| "loss": 1.1372, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.850501537322998, | |
| "learning_rate": 0.00038491048593350386, | |
| "loss": 1.0724, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.0993905067443848, | |
| "learning_rate": 0.00038277919863597613, | |
| "loss": 0.9782, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_accuracy": 0.6258, | |
| "eval_f1_macro": 0.5302739353593817, | |
| "eval_f1_micro": 0.6258, | |
| "eval_loss": 1.0663546323776245, | |
| "eval_runtime": 4.6409, | |
| "eval_samples_per_second": 1077.372, | |
| "eval_steps_per_second": 33.829, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.872110605239868, | |
| "learning_rate": 0.00038064791133844846, | |
| "loss": 1.103, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 2.1900711059570312, | |
| "learning_rate": 0.00037851662404092074, | |
| "loss": 1.1209, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 2.3909690380096436, | |
| "learning_rate": 0.000376385336743393, | |
| "loss": 1.0616, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.6566083431243896, | |
| "learning_rate": 0.00037425404944586534, | |
| "loss": 1.0967, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 2.543503761291504, | |
| "learning_rate": 0.0003721227621483376, | |
| "loss": 1.081, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_accuracy": 0.638, | |
| "eval_f1_macro": 0.5580934642391258, | |
| "eval_f1_micro": 0.638, | |
| "eval_loss": 1.0547568798065186, | |
| "eval_runtime": 4.6359, | |
| "eval_samples_per_second": 1078.528, | |
| "eval_steps_per_second": 33.866, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 2.8259775638580322, | |
| "learning_rate": 0.0003699914748508099, | |
| "loss": 1.1526, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.438499093055725, | |
| "learning_rate": 0.0003678601875532822, | |
| "loss": 1.0818, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 2.1894028186798096, | |
| "learning_rate": 0.0003657289002557545, | |
| "loss": 1.1406, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 2.273362159729004, | |
| "learning_rate": 0.0003635976129582268, | |
| "loss": 1.0805, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.018456220626831, | |
| "learning_rate": 0.0003614663256606991, | |
| "loss": 1.1033, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_accuracy": 0.6398, | |
| "eval_f1_macro": 0.5593218396680832, | |
| "eval_f1_micro": 0.6398, | |
| "eval_loss": 1.0300291776657104, | |
| "eval_runtime": 4.6355, | |
| "eval_samples_per_second": 1078.633, | |
| "eval_steps_per_second": 33.869, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.58931303024292, | |
| "learning_rate": 0.0003593350383631714, | |
| "loss": 1.0768, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 2.5274384021759033, | |
| "learning_rate": 0.00035720375106564365, | |
| "loss": 1.1398, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.7806419134140015, | |
| "learning_rate": 0.000355072463768116, | |
| "loss": 1.1315, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.8781118392944336, | |
| "learning_rate": 0.00035294117647058826, | |
| "loss": 1.2041, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.604295253753662, | |
| "learning_rate": 0.00035080988917306053, | |
| "loss": 1.0946, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_accuracy": 0.632, | |
| "eval_f1_macro": 0.5544640143725398, | |
| "eval_f1_micro": 0.632, | |
| "eval_loss": 1.0619752407073975, | |
| "eval_runtime": 4.6258, | |
| "eval_samples_per_second": 1080.883, | |
| "eval_steps_per_second": 33.94, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.7683440446853638, | |
| "learning_rate": 0.00034867860187553286, | |
| "loss": 1.1201, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.7039743661880493, | |
| "learning_rate": 0.00034654731457800514, | |
| "loss": 1.1598, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 2.3733344078063965, | |
| "learning_rate": 0.0003444160272804774, | |
| "loss": 1.0286, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 3.2192821502685547, | |
| "learning_rate": 0.00034228473998294974, | |
| "loss": 1.1397, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.8821942806243896, | |
| "learning_rate": 0.000340153452685422, | |
| "loss": 1.0261, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_accuracy": 0.6422, | |
| "eval_f1_macro": 0.5648413230502044, | |
| "eval_f1_micro": 0.6422, | |
| "eval_loss": 1.0328495502471924, | |
| "eval_runtime": 4.6303, | |
| "eval_samples_per_second": 1079.842, | |
| "eval_steps_per_second": 33.907, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 3.0460636615753174, | |
| "learning_rate": 0.0003380221653878943, | |
| "loss": 1.0396, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.969018816947937, | |
| "learning_rate": 0.0003358908780903666, | |
| "loss": 1.1069, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.0290212631225586, | |
| "learning_rate": 0.0003337595907928389, | |
| "loss": 1.0933, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 2.099224328994751, | |
| "learning_rate": 0.0003316283034953112, | |
| "loss": 0.9485, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.5207663774490356, | |
| "learning_rate": 0.0003294970161977835, | |
| "loss": 0.9153, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.6438, | |
| "eval_f1_macro": 0.5706285939519296, | |
| "eval_f1_micro": 0.6438, | |
| "eval_loss": 1.0378421545028687, | |
| "eval_runtime": 4.6333, | |
| "eval_samples_per_second": 1079.143, | |
| "eval_steps_per_second": 33.885, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.9443259239196777, | |
| "learning_rate": 0.0003273657289002558, | |
| "loss": 1.0276, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.0874528884887695, | |
| "learning_rate": 0.00032523444160272806, | |
| "loss": 0.8604, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 2.344442367553711, | |
| "learning_rate": 0.0003231031543052004, | |
| "loss": 0.9834, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 2.2925143241882324, | |
| "learning_rate": 0.0003209718670076726, | |
| "loss": 0.9142, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 2.234575033187866, | |
| "learning_rate": 0.0003188405797101449, | |
| "loss": 0.9678, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "eval_accuracy": 0.6402, | |
| "eval_f1_macro": 0.5482744671072567, | |
| "eval_f1_micro": 0.6402, | |
| "eval_loss": 1.0520384311676025, | |
| "eval_runtime": 4.6291, | |
| "eval_samples_per_second": 1080.114, | |
| "eval_steps_per_second": 33.916, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 2.2082581520080566, | |
| "learning_rate": 0.0003167092924126172, | |
| "loss": 0.9128, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 2.396441698074341, | |
| "learning_rate": 0.0003145780051150895, | |
| "loss": 0.9842, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 2.544588565826416, | |
| "learning_rate": 0.00031244671781756176, | |
| "loss": 0.899, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 3.3907265663146973, | |
| "learning_rate": 0.0003103154305200341, | |
| "loss": 1.0218, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 2.5349113941192627, | |
| "learning_rate": 0.00030818414322250637, | |
| "loss": 0.9619, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_accuracy": 0.6408, | |
| "eval_f1_macro": 0.5593284960739974, | |
| "eval_f1_micro": 0.6408, | |
| "eval_loss": 1.0482654571533203, | |
| "eval_runtime": 4.5768, | |
| "eval_samples_per_second": 1092.472, | |
| "eval_steps_per_second": 34.304, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 2.689525842666626, | |
| "learning_rate": 0.00030605285592497864, | |
| "loss": 0.9623, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 2.4175920486450195, | |
| "learning_rate": 0.00030392156862745097, | |
| "loss": 0.9547, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 2.392862319946289, | |
| "learning_rate": 0.00030179028132992325, | |
| "loss": 0.9661, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.7447049617767334, | |
| "learning_rate": 0.0002996589940323955, | |
| "loss": 0.9644, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 2.1388115882873535, | |
| "learning_rate": 0.00029752770673486785, | |
| "loss": 0.9972, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "eval_accuracy": 0.6496, | |
| "eval_f1_macro": 0.5684800364514564, | |
| "eval_f1_micro": 0.6496, | |
| "eval_loss": 1.0254615545272827, | |
| "eval_runtime": 4.6237, | |
| "eval_samples_per_second": 1081.385, | |
| "eval_steps_per_second": 33.955, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 2.1175084114074707, | |
| "learning_rate": 0.00029539641943734013, | |
| "loss": 0.8829, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 2.6198832988739014, | |
| "learning_rate": 0.0002932651321398124, | |
| "loss": 1.0612, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.1012232303619385, | |
| "learning_rate": 0.00029113384484228473, | |
| "loss": 0.9605, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 2.3467049598693848, | |
| "learning_rate": 0.000289002557544757, | |
| "loss": 0.9891, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 3.222963809967041, | |
| "learning_rate": 0.0002868712702472293, | |
| "loss": 1.027, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.645, | |
| "eval_f1_macro": 0.5742246375864342, | |
| "eval_f1_micro": 0.645, | |
| "eval_loss": 1.0296003818511963, | |
| "eval_runtime": 4.632, | |
| "eval_samples_per_second": 1079.448, | |
| "eval_steps_per_second": 33.895, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 2.7652339935302734, | |
| "learning_rate": 0.0002847399829497016, | |
| "loss": 1.0417, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.1574389934539795, | |
| "learning_rate": 0.0002826086956521739, | |
| "loss": 0.9338, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 2.7982587814331055, | |
| "learning_rate": 0.00028047740835464616, | |
| "loss": 0.8798, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 3.097017526626587, | |
| "learning_rate": 0.0002783461210571185, | |
| "loss": 0.9041, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 2.7650091648101807, | |
| "learning_rate": 0.00027621483375959077, | |
| "loss": 0.8248, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "eval_accuracy": 0.655, | |
| "eval_f1_macro": 0.5811778167583621, | |
| "eval_f1_micro": 0.655, | |
| "eval_loss": 1.0331394672393799, | |
| "eval_runtime": 4.5784, | |
| "eval_samples_per_second": 1092.075, | |
| "eval_steps_per_second": 34.291, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 2.5823092460632324, | |
| "learning_rate": 0.0002740835464620631, | |
| "loss": 1.1103, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 2.0402040481567383, | |
| "learning_rate": 0.0002719522591645354, | |
| "loss": 1.0841, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 3.023350238800049, | |
| "learning_rate": 0.00026982097186700765, | |
| "loss": 0.971, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 1.5279247760772705, | |
| "learning_rate": 0.00026768968456948, | |
| "loss": 0.9861, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 2.5153307914733887, | |
| "learning_rate": 0.00026555839727195225, | |
| "loss": 0.9405, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "eval_accuracy": 0.6502, | |
| "eval_f1_macro": 0.5719450847979869, | |
| "eval_f1_micro": 0.6502, | |
| "eval_loss": 1.020789623260498, | |
| "eval_runtime": 4.5771, | |
| "eval_samples_per_second": 1092.404, | |
| "eval_steps_per_second": 34.301, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 2.871000289916992, | |
| "learning_rate": 0.00026342710997442453, | |
| "loss": 0.9694, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 2.492405414581299, | |
| "learning_rate": 0.00026129582267689686, | |
| "loss": 0.8878, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 2.770477056503296, | |
| "learning_rate": 0.00025916453537936913, | |
| "loss": 0.9456, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 7.3541340827941895, | |
| "learning_rate": 0.0002570332480818414, | |
| "loss": 0.9401, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 2.690429449081421, | |
| "learning_rate": 0.00025490196078431374, | |
| "loss": 0.9735, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_accuracy": 0.6388, | |
| "eval_f1_macro": 0.5743608505759401, | |
| "eval_f1_micro": 0.6388, | |
| "eval_loss": 1.0388753414154053, | |
| "eval_runtime": 4.6105, | |
| "eval_samples_per_second": 1084.485, | |
| "eval_steps_per_second": 34.053, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 3.4364089965820312, | |
| "learning_rate": 0.000252770673486786, | |
| "loss": 1.124, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.5237576961517334, | |
| "learning_rate": 0.0002506393861892583, | |
| "loss": 1.0429, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 2.4886770248413086, | |
| "learning_rate": 0.0002485080988917306, | |
| "loss": 0.9049, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.1807761192321777, | |
| "learning_rate": 0.0002463768115942029, | |
| "loss": 0.9292, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 2.830251455307007, | |
| "learning_rate": 0.0002442455242966752, | |
| "loss": 0.9566, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "eval_accuracy": 0.658, | |
| "eval_f1_macro": 0.5749866859043913, | |
| "eval_f1_micro": 0.658, | |
| "eval_loss": 0.9963122010231018, | |
| "eval_runtime": 4.6173, | |
| "eval_samples_per_second": 1082.894, | |
| "eval_steps_per_second": 34.003, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 2.474280834197998, | |
| "learning_rate": 0.0002421142369991475, | |
| "loss": 0.8845, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.4428656101226807, | |
| "learning_rate": 0.0002399829497016198, | |
| "loss": 1.0546, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 2.1537678241729736, | |
| "learning_rate": 0.00023785166240409208, | |
| "loss": 0.9547, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.9569634199142456, | |
| "learning_rate": 0.00023572037510656438, | |
| "loss": 0.8683, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.2532451152801514, | |
| "learning_rate": 0.00023358908780903668, | |
| "loss": 0.9423, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 0.6496, | |
| "eval_f1_macro": 0.5831534303823815, | |
| "eval_f1_micro": 0.6496, | |
| "eval_loss": 0.9966222643852234, | |
| "eval_runtime": 4.6211, | |
| "eval_samples_per_second": 1081.985, | |
| "eval_steps_per_second": 33.974, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 2.022592306137085, | |
| "learning_rate": 0.00023145780051150893, | |
| "loss": 0.8889, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 3.371534585952759, | |
| "learning_rate": 0.00022932651321398123, | |
| "loss": 0.999, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 2.3057377338409424, | |
| "learning_rate": 0.00022719522591645354, | |
| "loss": 0.9226, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.5777747631073, | |
| "learning_rate": 0.0002250639386189258, | |
| "loss": 0.8813, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 2.6509506702423096, | |
| "learning_rate": 0.00022293265132139811, | |
| "loss": 0.9248, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "eval_accuracy": 0.6558, | |
| "eval_f1_macro": 0.5856739858943232, | |
| "eval_f1_micro": 0.6558, | |
| "eval_loss": 0.9953044652938843, | |
| "eval_runtime": 4.6144, | |
| "eval_samples_per_second": 1083.57, | |
| "eval_steps_per_second": 34.024, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.321070671081543, | |
| "learning_rate": 0.00022080136402387042, | |
| "loss": 0.9733, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 3.0813984870910645, | |
| "learning_rate": 0.0002186700767263427, | |
| "loss": 0.9548, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.4469900131225586, | |
| "learning_rate": 0.000216538789428815, | |
| "loss": 0.9282, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 2.527189016342163, | |
| "learning_rate": 0.0002144075021312873, | |
| "loss": 0.972, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.339242935180664, | |
| "learning_rate": 0.00021227621483375957, | |
| "loss": 1.008, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "eval_accuracy": 0.6588, | |
| "eval_f1_macro": 0.5809236785268852, | |
| "eval_f1_micro": 0.6588, | |
| "eval_loss": 0.9940443634986877, | |
| "eval_runtime": 4.6228, | |
| "eval_samples_per_second": 1081.603, | |
| "eval_steps_per_second": 33.962, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 2.1901655197143555, | |
| "learning_rate": 0.00021014492753623187, | |
| "loss": 0.9911, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.4824706315994263, | |
| "learning_rate": 0.00020801364023870418, | |
| "loss": 0.9618, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.87557852268219, | |
| "learning_rate": 0.00020588235294117645, | |
| "loss": 0.9933, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 2.4000425338745117, | |
| "learning_rate": 0.00020375106564364876, | |
| "loss": 0.9995, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 2.9583659172058105, | |
| "learning_rate": 0.00020161977834612106, | |
| "loss": 0.9098, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_accuracy": 0.657, | |
| "eval_f1_macro": 0.5821992906460801, | |
| "eval_f1_micro": 0.657, | |
| "eval_loss": 0.9832942485809326, | |
| "eval_runtime": 4.6161, | |
| "eval_samples_per_second": 1083.16, | |
| "eval_steps_per_second": 34.011, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.1622211933135986, | |
| "learning_rate": 0.00019948849104859333, | |
| "loss": 1.0026, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 2.2868545055389404, | |
| "learning_rate": 0.00019735720375106564, | |
| "loss": 0.9563, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 1.8215919733047485, | |
| "learning_rate": 0.00019522591645353794, | |
| "loss": 0.8711, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 2.21563458442688, | |
| "learning_rate": 0.0001930946291560102, | |
| "loss": 0.8976, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 2.023301839828491, | |
| "learning_rate": 0.00019096334185848252, | |
| "loss": 0.8679, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_accuracy": 0.6644, | |
| "eval_f1_macro": 0.589941204224004, | |
| "eval_f1_micro": 0.6644, | |
| "eval_loss": 0.9842157959938049, | |
| "eval_runtime": 4.6173, | |
| "eval_samples_per_second": 1082.894, | |
| "eval_steps_per_second": 34.003, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 3.6565377712249756, | |
| "learning_rate": 0.00018883205456095482, | |
| "loss": 0.9224, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 2.6100504398345947, | |
| "learning_rate": 0.0001867007672634271, | |
| "loss": 0.9561, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 2.7163302898406982, | |
| "learning_rate": 0.0001845694799658994, | |
| "loss": 0.9514, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 2.600562572479248, | |
| "learning_rate": 0.0001824381926683717, | |
| "loss": 0.9922, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 2.355883836746216, | |
| "learning_rate": 0.000180306905370844, | |
| "loss": 1.1342, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_accuracy": 0.6526, | |
| "eval_f1_macro": 0.5761908672451729, | |
| "eval_f1_micro": 0.6526, | |
| "eval_loss": 0.9932846426963806, | |
| "eval_runtime": 4.6151, | |
| "eval_samples_per_second": 1083.409, | |
| "eval_steps_per_second": 34.019, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 2.483553409576416, | |
| "learning_rate": 0.00017817561807331628, | |
| "loss": 0.8869, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 2.1051151752471924, | |
| "learning_rate": 0.00017604433077578858, | |
| "loss": 0.8881, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 2.524996757507324, | |
| "learning_rate": 0.00017391304347826088, | |
| "loss": 0.9509, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 2.025562286376953, | |
| "learning_rate": 0.00017178175618073316, | |
| "loss": 0.9632, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 2.144996404647827, | |
| "learning_rate": 0.00016965046888320546, | |
| "loss": 0.9157, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "eval_accuracy": 0.6626, | |
| "eval_f1_macro": 0.5923777456303617, | |
| "eval_f1_micro": 0.6626, | |
| "eval_loss": 0.986947774887085, | |
| "eval_runtime": 4.6155, | |
| "eval_samples_per_second": 1083.31, | |
| "eval_steps_per_second": 34.016, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 2.602703809738159, | |
| "learning_rate": 0.00016751918158567776, | |
| "loss": 0.9573, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.3346447944641113, | |
| "learning_rate": 0.00016538789428815004, | |
| "loss": 0.781, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 1.8888440132141113, | |
| "learning_rate": 0.00016325660699062234, | |
| "loss": 0.9018, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 2.854541063308716, | |
| "learning_rate": 0.00016112531969309464, | |
| "loss": 0.833, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 2.8383681774139404, | |
| "learning_rate": 0.00015899403239556692, | |
| "loss": 0.8084, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_accuracy": 0.6654, | |
| "eval_f1_macro": 0.5892559746460942, | |
| "eval_f1_micro": 0.6654, | |
| "eval_loss": 0.9908738136291504, | |
| "eval_runtime": 4.6175, | |
| "eval_samples_per_second": 1082.838, | |
| "eval_steps_per_second": 34.001, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 2.7356762886047363, | |
| "learning_rate": 0.00015686274509803922, | |
| "loss": 0.8014, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 1.9319524765014648, | |
| "learning_rate": 0.00015473145780051152, | |
| "loss": 0.9392, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 2.2316906452178955, | |
| "learning_rate": 0.0001526001705029838, | |
| "loss": 0.8349, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.030834674835205, | |
| "learning_rate": 0.0001504688832054561, | |
| "loss": 0.804, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 2.651904821395874, | |
| "learning_rate": 0.0001483375959079284, | |
| "loss": 0.7373, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "eval_accuracy": 0.6622, | |
| "eval_f1_macro": 0.5964545001055274, | |
| "eval_f1_micro": 0.6622, | |
| "eval_loss": 0.9894343614578247, | |
| "eval_runtime": 4.6227, | |
| "eval_samples_per_second": 1081.625, | |
| "eval_steps_per_second": 33.963, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 2.3340699672698975, | |
| "learning_rate": 0.00014620630861040068, | |
| "loss": 0.79, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 1.6943384408950806, | |
| "learning_rate": 0.00014407502131287298, | |
| "loss": 0.8374, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 2.2380564212799072, | |
| "learning_rate": 0.00014194373401534528, | |
| "loss": 0.8359, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 2.5444910526275635, | |
| "learning_rate": 0.00013981244671781756, | |
| "loss": 0.73, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 2.747690200805664, | |
| "learning_rate": 0.00013768115942028986, | |
| "loss": 0.9081, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_accuracy": 0.6614, | |
| "eval_f1_macro": 0.5879935284170169, | |
| "eval_f1_micro": 0.6614, | |
| "eval_loss": 0.999671459197998, | |
| "eval_runtime": 4.6154, | |
| "eval_samples_per_second": 1083.329, | |
| "eval_steps_per_second": 34.017, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 2.335947275161743, | |
| "learning_rate": 0.00013554987212276216, | |
| "loss": 0.8419, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.619969129562378, | |
| "learning_rate": 0.00013341858482523444, | |
| "loss": 0.8277, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 2.147700071334839, | |
| "learning_rate": 0.00013128729752770674, | |
| "loss": 0.7419, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 3.1116886138916016, | |
| "learning_rate": 0.00012915601023017904, | |
| "loss": 0.8837, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.3517539501190186, | |
| "learning_rate": 0.00012702472293265132, | |
| "loss": 0.8064, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_accuracy": 0.659, | |
| "eval_f1_macro": 0.5918523615308761, | |
| "eval_f1_micro": 0.659, | |
| "eval_loss": 0.9998078346252441, | |
| "eval_runtime": 4.6138, | |
| "eval_samples_per_second": 1083.711, | |
| "eval_steps_per_second": 34.029, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 2.4132673740386963, | |
| "learning_rate": 0.00012489343563512362, | |
| "loss": 0.7836, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 2.801809787750244, | |
| "learning_rate": 0.00012276214833759592, | |
| "loss": 0.9516, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 2.2415406703948975, | |
| "learning_rate": 0.00012063086104006821, | |
| "loss": 0.7053, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 2.9133055210113525, | |
| "learning_rate": 0.0001184995737425405, | |
| "loss": 0.8583, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 2.091689109802246, | |
| "learning_rate": 0.00011636828644501279, | |
| "loss": 0.8519, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "eval_accuracy": 0.6584, | |
| "eval_f1_macro": 0.5880330853167925, | |
| "eval_f1_micro": 0.6584, | |
| "eval_loss": 1.0030735731124878, | |
| "eval_runtime": 4.6181, | |
| "eval_samples_per_second": 1082.688, | |
| "eval_steps_per_second": 33.996, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 2.7555630207061768, | |
| "learning_rate": 0.0001142369991474851, | |
| "loss": 0.8504, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 2.736027717590332, | |
| "learning_rate": 0.00011210571184995738, | |
| "loss": 0.7777, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 2.5624210834503174, | |
| "learning_rate": 0.00010997442455242967, | |
| "loss": 0.8027, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 3.113119602203369, | |
| "learning_rate": 0.00010784313725490197, | |
| "loss": 0.7215, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 2.8675148487091064, | |
| "learning_rate": 0.00010571184995737426, | |
| "loss": 0.8711, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "eval_accuracy": 0.6666, | |
| "eval_f1_macro": 0.59808390789386, | |
| "eval_f1_micro": 0.6666, | |
| "eval_loss": 0.9974775314331055, | |
| "eval_runtime": 4.6223, | |
| "eval_samples_per_second": 1081.724, | |
| "eval_steps_per_second": 33.966, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 3.3633134365081787, | |
| "learning_rate": 0.00010358056265984655, | |
| "loss": 0.872, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 2.622450828552246, | |
| "learning_rate": 0.00010144927536231885, | |
| "loss": 0.8464, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.2080349922180176, | |
| "learning_rate": 9.931798806479114e-05, | |
| "loss": 0.7183, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 2.735830068588257, | |
| "learning_rate": 9.718670076726342e-05, | |
| "loss": 0.7964, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 2.720543384552002, | |
| "learning_rate": 9.505541346973572e-05, | |
| "loss": 0.7617, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "eval_accuracy": 0.6584, | |
| "eval_f1_macro": 0.5848770420558893, | |
| "eval_f1_micro": 0.6584, | |
| "eval_loss": 1.014381766319275, | |
| "eval_runtime": 4.6129, | |
| "eval_samples_per_second": 1083.915, | |
| "eval_steps_per_second": 34.035, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 2.777467966079712, | |
| "learning_rate": 9.292412617220801e-05, | |
| "loss": 0.7938, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 2.8683433532714844, | |
| "learning_rate": 9.07928388746803e-05, | |
| "loss": 0.7492, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 2.9691097736358643, | |
| "learning_rate": 8.86615515771526e-05, | |
| "loss": 0.8453, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 3.0169878005981445, | |
| "learning_rate": 8.653026427962489e-05, | |
| "loss": 0.7834, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 2.8592581748962402, | |
| "learning_rate": 8.439897698209718e-05, | |
| "loss": 0.717, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "eval_accuracy": 0.6622, | |
| "eval_f1_macro": 0.5902745243208923, | |
| "eval_f1_micro": 0.6622, | |
| "eval_loss": 1.010201334953308, | |
| "eval_runtime": 4.606, | |
| "eval_samples_per_second": 1085.541, | |
| "eval_steps_per_second": 34.086, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 2.6895742416381836, | |
| "learning_rate": 8.226768968456948e-05, | |
| "loss": 0.7258, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 2.743197441101074, | |
| "learning_rate": 8.013640238704177e-05, | |
| "loss": 0.8436, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 3.289165735244751, | |
| "learning_rate": 7.800511508951406e-05, | |
| "loss": 0.8923, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 2.217371940612793, | |
| "learning_rate": 7.587382779198636e-05, | |
| "loss": 0.7365, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 3.4667258262634277, | |
| "learning_rate": 7.374254049445865e-05, | |
| "loss": 0.857, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.6622, | |
| "eval_f1_macro": 0.5923250147204171, | |
| "eval_f1_micro": 0.6622, | |
| "eval_loss": 1.0059385299682617, | |
| "eval_runtime": 4.5385, | |
| "eval_samples_per_second": 1101.676, | |
| "eval_steps_per_second": 34.593, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 2.3217034339904785, | |
| "learning_rate": 7.161125319693094e-05, | |
| "loss": 0.8445, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 2.4780850410461426, | |
| "learning_rate": 6.947996589940324e-05, | |
| "loss": 0.7369, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 2.254239559173584, | |
| "learning_rate": 6.734867860187553e-05, | |
| "loss": 0.7862, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 1.7889450788497925, | |
| "learning_rate": 6.521739130434782e-05, | |
| "loss": 0.7451, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 3.1137242317199707, | |
| "learning_rate": 6.308610400682012e-05, | |
| "loss": 0.8623, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "eval_accuracy": 0.664, | |
| "eval_f1_macro": 0.5971048543279723, | |
| "eval_f1_micro": 0.664, | |
| "eval_loss": 1.0025016069412231, | |
| "eval_runtime": 4.6042, | |
| "eval_samples_per_second": 1085.954, | |
| "eval_steps_per_second": 34.099, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 2.7454991340637207, | |
| "learning_rate": 6.095481670929241e-05, | |
| "loss": 0.8, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 2.9460010528564453, | |
| "learning_rate": 5.882352941176471e-05, | |
| "loss": 0.7284, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 3.019240379333496, | |
| "learning_rate": 5.6692242114237e-05, | |
| "loss": 0.7608, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 2.111807346343994, | |
| "learning_rate": 5.456095481670929e-05, | |
| "loss": 0.829, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 3.3570430278778076, | |
| "learning_rate": 5.242966751918159e-05, | |
| "loss": 0.782, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "eval_accuracy": 0.6644, | |
| "eval_f1_macro": 0.5985316673414206, | |
| "eval_f1_micro": 0.6644, | |
| "eval_loss": 1.0012755393981934, | |
| "eval_runtime": 4.5602, | |
| "eval_samples_per_second": 1096.451, | |
| "eval_steps_per_second": 34.429, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 2.958252191543579, | |
| "learning_rate": 5.0298380221653884e-05, | |
| "loss": 0.7804, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 2.9975624084472656, | |
| "learning_rate": 4.816709292412617e-05, | |
| "loss": 0.7807, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 2.5771234035491943, | |
| "learning_rate": 4.603580562659847e-05, | |
| "loss": 0.7609, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 2.9194953441619873, | |
| "learning_rate": 4.3904518329070764e-05, | |
| "loss": 0.7783, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 3.0004827976226807, | |
| "learning_rate": 4.177323103154305e-05, | |
| "loss": 0.8018, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "eval_accuracy": 0.6652, | |
| "eval_f1_macro": 0.5984558375770885, | |
| "eval_f1_micro": 0.6652, | |
| "eval_loss": 1.0043922662734985, | |
| "eval_runtime": 4.5535, | |
| "eval_samples_per_second": 1098.053, | |
| "eval_steps_per_second": 34.479, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 1.9689658880233765, | |
| "learning_rate": 3.964194373401535e-05, | |
| "loss": 0.8252, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 2.3247179985046387, | |
| "learning_rate": 3.7510656436487644e-05, | |
| "loss": 0.8301, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 3.333657741546631, | |
| "learning_rate": 3.537936913895993e-05, | |
| "loss": 0.7161, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 3.242521047592163, | |
| "learning_rate": 3.324808184143223e-05, | |
| "loss": 0.796, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.6882381439208984, | |
| "learning_rate": 3.111679454390452e-05, | |
| "loss": 0.7901, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "eval_accuracy": 0.6678, | |
| "eval_f1_macro": 0.6030435999289231, | |
| "eval_f1_micro": 0.6678, | |
| "eval_loss": 0.9987305402755737, | |
| "eval_runtime": 4.6038, | |
| "eval_samples_per_second": 1086.055, | |
| "eval_steps_per_second": 34.102, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 2.786713123321533, | |
| "learning_rate": 2.8985507246376814e-05, | |
| "loss": 0.7368, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 2.7163522243499756, | |
| "learning_rate": 2.6854219948849106e-05, | |
| "loss": 0.7976, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 2.269670248031616, | |
| "learning_rate": 2.4722932651321398e-05, | |
| "loss": 0.7612, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 2.7400472164154053, | |
| "learning_rate": 2.2591645353793694e-05, | |
| "loss": 0.8234, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 4.061866760253906, | |
| "learning_rate": 2.0460358056265986e-05, | |
| "loss": 0.8835, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_accuracy": 0.6644, | |
| "eval_f1_macro": 0.5985551818802296, | |
| "eval_f1_micro": 0.6644, | |
| "eval_loss": 1.0015017986297607, | |
| "eval_runtime": 4.6024, | |
| "eval_samples_per_second": 1086.396, | |
| "eval_steps_per_second": 34.113, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 3.1923465728759766, | |
| "learning_rate": 1.8329070758738275e-05, | |
| "loss": 0.8002, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 2.8874006271362305, | |
| "learning_rate": 1.619778346121057e-05, | |
| "loss": 0.9052, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 2.9610722064971924, | |
| "learning_rate": 1.4066496163682863e-05, | |
| "loss": 0.7826, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 2.6525473594665527, | |
| "learning_rate": 1.1935208866155157e-05, | |
| "loss": 0.8008, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 2.8157360553741455, | |
| "learning_rate": 9.803921568627451e-06, | |
| "loss": 0.8679, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "eval_accuracy": 0.6636, | |
| "eval_f1_macro": 0.5960754763007041, | |
| "eval_f1_micro": 0.6636, | |
| "eval_loss": 0.9994349479675293, | |
| "eval_runtime": 4.5976, | |
| "eval_samples_per_second": 1087.518, | |
| "eval_steps_per_second": 34.148, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 2.7772059440612793, | |
| "learning_rate": 7.672634271099745e-06, | |
| "loss": 0.7572, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 2.3767988681793213, | |
| "learning_rate": 5.541346973572038e-06, | |
| "loss": 0.7519, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 3.354130506515503, | |
| "learning_rate": 3.4100596760443308e-06, | |
| "loss": 0.798, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 2.4790210723876953, | |
| "learning_rate": 1.2787723785166241e-06, | |
| "loss": 0.732, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 2346, | |
| "total_flos": 2555504607363072.0, | |
| "train_loss": 0.9882089091688776, | |
| "train_runtime": 435.725, | |
| "train_samples_per_second": 172.127, | |
| "train_steps_per_second": 5.384 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2346, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 2555504607363072.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |