| { | |
| "best_metric": 0.5568962693214417, | |
| "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-18500", | |
| "epoch": 2.3640661938534278, | |
| "eval_steps": 500, | |
| "global_step": 21000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.056287290329843524, | |
| "grad_norm": 150.1642303466797, | |
| "learning_rate": 4.929640887087696e-07, | |
| "loss": 0.3486, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.056287290329843524, | |
| "eval_loss": 1.1265727281570435, | |
| "eval_runtime": 123.7806, | |
| "eval_samples_per_second": 15.948, | |
| "eval_steps_per_second": 1.995, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11257458065968705, | |
| "grad_norm": 366.6913146972656, | |
| "learning_rate": 4.859281774175391e-07, | |
| "loss": 0.2733, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.11257458065968705, | |
| "eval_loss": 0.9742079377174377, | |
| "eval_runtime": 125.8047, | |
| "eval_samples_per_second": 15.691, | |
| "eval_steps_per_second": 1.963, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16886187098953057, | |
| "grad_norm": 555.5016479492188, | |
| "learning_rate": 4.788922661263087e-07, | |
| "loss": 0.1851, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.16886187098953057, | |
| "eval_loss": 0.9162159562110901, | |
| "eval_runtime": 126.7606, | |
| "eval_samples_per_second": 15.573, | |
| "eval_steps_per_second": 1.949, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2251491613193741, | |
| "grad_norm": 0.423260897397995, | |
| "learning_rate": 4.7185635483507824e-07, | |
| "loss": 0.1973, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2251491613193741, | |
| "eval_loss": 0.8716973662376404, | |
| "eval_runtime": 127.6955, | |
| "eval_samples_per_second": 15.459, | |
| "eval_steps_per_second": 1.934, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2814364516492176, | |
| "grad_norm": 6.840139389038086, | |
| "learning_rate": 4.6482044354384774e-07, | |
| "loss": 0.1881, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.2814364516492176, | |
| "eval_loss": 0.830047070980072, | |
| "eval_runtime": 128.0771, | |
| "eval_samples_per_second": 15.413, | |
| "eval_steps_per_second": 1.929, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.33772374197906113, | |
| "grad_norm": 729.8377685546875, | |
| "learning_rate": 4.577845322526173e-07, | |
| "loss": 0.1695, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.33772374197906113, | |
| "eval_loss": 0.8265627026557922, | |
| "eval_runtime": 128.052, | |
| "eval_samples_per_second": 15.416, | |
| "eval_steps_per_second": 1.929, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.39401103230890466, | |
| "grad_norm": 7.96999938756926e-06, | |
| "learning_rate": 4.507486209613869e-07, | |
| "loss": 0.155, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.39401103230890466, | |
| "eval_loss": 0.7721038460731506, | |
| "eval_runtime": 128.1871, | |
| "eval_samples_per_second": 15.399, | |
| "eval_steps_per_second": 1.927, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4502983226387482, | |
| "grad_norm": 0.006555848754942417, | |
| "learning_rate": 4.4371270967015645e-07, | |
| "loss": 0.1216, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.4502983226387482, | |
| "eval_loss": 0.7490401864051819, | |
| "eval_runtime": 127.5516, | |
| "eval_samples_per_second": 15.476, | |
| "eval_steps_per_second": 1.936, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5065856129685917, | |
| "grad_norm": 0.0004515685432124883, | |
| "learning_rate": 4.36676798378926e-07, | |
| "loss": 0.1531, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5065856129685917, | |
| "eval_loss": 0.7533182501792908, | |
| "eval_runtime": 127.4797, | |
| "eval_samples_per_second": 15.485, | |
| "eval_steps_per_second": 1.938, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5628729032984352, | |
| "grad_norm": 4.4633176003117114e-05, | |
| "learning_rate": 4.2964088708769556e-07, | |
| "loss": 0.1154, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5628729032984352, | |
| "eval_loss": 0.747530460357666, | |
| "eval_runtime": 127.8675, | |
| "eval_samples_per_second": 15.438, | |
| "eval_steps_per_second": 1.932, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6191601936282788, | |
| "grad_norm": 4.158839702606201, | |
| "learning_rate": 4.2260497579646517e-07, | |
| "loss": 0.1407, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6191601936282788, | |
| "eval_loss": 0.724854588508606, | |
| "eval_runtime": 125.1629, | |
| "eval_samples_per_second": 15.771, | |
| "eval_steps_per_second": 1.973, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6754474839581223, | |
| "grad_norm": 0.748630166053772, | |
| "learning_rate": 4.155690645052347e-07, | |
| "loss": 0.1447, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6754474839581223, | |
| "eval_loss": 0.7091771960258484, | |
| "eval_runtime": 125.2549, | |
| "eval_samples_per_second": 15.76, | |
| "eval_steps_per_second": 1.972, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7317347742879657, | |
| "grad_norm": 2.8120924980612472e-05, | |
| "learning_rate": 4.0853315321400427e-07, | |
| "loss": 0.1146, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7317347742879657, | |
| "eval_loss": 0.7105826735496521, | |
| "eval_runtime": 127.7786, | |
| "eval_samples_per_second": 15.449, | |
| "eval_steps_per_second": 1.933, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7880220646178093, | |
| "grad_norm": 1.4741635823156685e-07, | |
| "learning_rate": 4.014972419227738e-07, | |
| "loss": 0.171, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7880220646178093, | |
| "eval_loss": 0.6978534460067749, | |
| "eval_runtime": 125.2567, | |
| "eval_samples_per_second": 15.76, | |
| "eval_steps_per_second": 1.972, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.8443093549476528, | |
| "grad_norm": 0.002975167240947485, | |
| "learning_rate": 3.944613306315434e-07, | |
| "loss": 0.1584, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8443093549476528, | |
| "eval_loss": 0.6923494935035706, | |
| "eval_runtime": 125.3384, | |
| "eval_samples_per_second": 15.749, | |
| "eval_steps_per_second": 1.971, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.9005966452774964, | |
| "grad_norm": 9.225498797604814e-05, | |
| "learning_rate": 3.8742541934031293e-07, | |
| "loss": 0.1384, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9005966452774964, | |
| "eval_loss": 0.6872902512550354, | |
| "eval_runtime": 128.114, | |
| "eval_samples_per_second": 15.408, | |
| "eval_steps_per_second": 1.928, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9568839356073399, | |
| "grad_norm": 1.9474915902151224e-08, | |
| "learning_rate": 3.803895080490825e-07, | |
| "loss": 0.1243, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9568839356073399, | |
| "eval_loss": 0.6788680553436279, | |
| "eval_runtime": 125.0525, | |
| "eval_samples_per_second": 15.785, | |
| "eval_steps_per_second": 1.975, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.0131712259371835, | |
| "grad_norm": 5.165647506713867, | |
| "learning_rate": 3.7335359675785204e-07, | |
| "loss": 0.1102, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0131712259371835, | |
| "eval_loss": 0.6788864135742188, | |
| "eval_runtime": 124.7777, | |
| "eval_samples_per_second": 15.82, | |
| "eval_steps_per_second": 1.98, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.069458516267027, | |
| "grad_norm": 0.0005840375670231879, | |
| "learning_rate": 3.6631768546662164e-07, | |
| "loss": 0.0523, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.069458516267027, | |
| "eval_loss": 0.6628187894821167, | |
| "eval_runtime": 127.8018, | |
| "eval_samples_per_second": 15.446, | |
| "eval_steps_per_second": 1.933, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.1257458065968704, | |
| "grad_norm": 0.001191094284877181, | |
| "learning_rate": 3.592817741753912e-07, | |
| "loss": 0.0617, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1257458065968704, | |
| "eval_loss": 0.6507942080497742, | |
| "eval_runtime": 125.484, | |
| "eval_samples_per_second": 15.731, | |
| "eval_steps_per_second": 1.968, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1820330969267139, | |
| "grad_norm": 0.0003092484548687935, | |
| "learning_rate": 3.5224586288416075e-07, | |
| "loss": 0.0598, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.1820330969267139, | |
| "eval_loss": 0.6480408906936646, | |
| "eval_runtime": 125.7661, | |
| "eval_samples_per_second": 15.696, | |
| "eval_steps_per_second": 1.964, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.2383203872565574, | |
| "grad_norm": 5.508670710696606e-07, | |
| "learning_rate": 3.452099515929303e-07, | |
| "loss": 0.0782, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.2383203872565574, | |
| "eval_loss": 0.6403237581253052, | |
| "eval_runtime": 128.3947, | |
| "eval_samples_per_second": 15.374, | |
| "eval_steps_per_second": 1.924, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.294607677586401, | |
| "grad_norm": 0.1470593810081482, | |
| "learning_rate": 3.3817404030169986e-07, | |
| "loss": 0.0616, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.294607677586401, | |
| "eval_loss": 0.6376460790634155, | |
| "eval_runtime": 125.5391, | |
| "eval_samples_per_second": 15.724, | |
| "eval_steps_per_second": 1.968, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.3508949679162445, | |
| "grad_norm": 1.8021532014245167e-05, | |
| "learning_rate": 3.3113812901046946e-07, | |
| "loss": 0.0723, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3508949679162445, | |
| "eval_loss": 0.6378474831581116, | |
| "eval_runtime": 125.5914, | |
| "eval_samples_per_second": 15.718, | |
| "eval_steps_per_second": 1.967, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.407182258246088, | |
| "grad_norm": 0.00013483635848388076, | |
| "learning_rate": 3.2410221771923896e-07, | |
| "loss": 0.0459, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.407182258246088, | |
| "eval_loss": 0.6378260850906372, | |
| "eval_runtime": 128.2725, | |
| "eval_samples_per_second": 15.389, | |
| "eval_steps_per_second": 1.926, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.4634695485759315, | |
| "grad_norm": 4.757418707868055e-07, | |
| "learning_rate": 3.170663064280085e-07, | |
| "loss": 0.0571, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.4634695485759315, | |
| "eval_loss": 0.6366899609565735, | |
| "eval_runtime": 125.6986, | |
| "eval_samples_per_second": 15.704, | |
| "eval_steps_per_second": 1.965, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.5197568389057752, | |
| "grad_norm": 0.00028420978924259543, | |
| "learning_rate": 3.1003039513677807e-07, | |
| "loss": 0.0499, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.5197568389057752, | |
| "eval_loss": 0.6286123394966125, | |
| "eval_runtime": 127.9165, | |
| "eval_samples_per_second": 15.432, | |
| "eval_steps_per_second": 1.931, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.5760441292356187, | |
| "grad_norm": 2.1999912291903456e-07, | |
| "learning_rate": 3.0299448384554767e-07, | |
| "loss": 0.0679, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5760441292356187, | |
| "eval_loss": 0.634914755821228, | |
| "eval_runtime": 125.0396, | |
| "eval_samples_per_second": 15.787, | |
| "eval_steps_per_second": 1.975, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.6323314195654621, | |
| "grad_norm": 0.00023505109129473567, | |
| "learning_rate": 2.959585725543172e-07, | |
| "loss": 0.0554, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.6323314195654621, | |
| "eval_loss": 0.623322069644928, | |
| "eval_runtime": 125.2429, | |
| "eval_samples_per_second": 15.761, | |
| "eval_steps_per_second": 1.972, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.6886187098953056, | |
| "grad_norm": 11.346132278442383, | |
| "learning_rate": 2.889226612630868e-07, | |
| "loss": 0.0617, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.6886187098953056, | |
| "eval_loss": 0.6115825772285461, | |
| "eval_runtime": 127.8813, | |
| "eval_samples_per_second": 15.436, | |
| "eval_steps_per_second": 1.931, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.744906000225149, | |
| "grad_norm": 308.43804931640625, | |
| "learning_rate": 2.8188674997185633e-07, | |
| "loss": 0.0561, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.744906000225149, | |
| "eval_loss": 0.5951005220413208, | |
| "eval_runtime": 125.2491, | |
| "eval_samples_per_second": 15.761, | |
| "eval_steps_per_second": 1.972, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.8011932905549926, | |
| "grad_norm": 6.91558055265773e-09, | |
| "learning_rate": 2.7485083868062594e-07, | |
| "loss": 0.0672, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.8011932905549926, | |
| "eval_loss": 0.5809568166732788, | |
| "eval_runtime": 125.2557, | |
| "eval_samples_per_second": 15.76, | |
| "eval_steps_per_second": 1.972, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.8574805808848363, | |
| "grad_norm": 2.2250075915053458e-07, | |
| "learning_rate": 2.678149273893955e-07, | |
| "loss": 0.0812, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.8574805808848363, | |
| "eval_loss": 0.5691242218017578, | |
| "eval_runtime": 128.3032, | |
| "eval_samples_per_second": 15.385, | |
| "eval_steps_per_second": 1.925, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.9137678712146797, | |
| "grad_norm": 1.1724663972854614, | |
| "learning_rate": 2.6077901609816504e-07, | |
| "loss": 0.0614, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.9137678712146797, | |
| "eval_loss": 0.5643230080604553, | |
| "eval_runtime": 125.7191, | |
| "eval_samples_per_second": 15.702, | |
| "eval_steps_per_second": 1.965, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.9700551615445232, | |
| "grad_norm": 0.00030226208036765456, | |
| "learning_rate": 2.5374310480693454e-07, | |
| "loss": 0.0462, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.9700551615445232, | |
| "eval_loss": 0.5657238364219666, | |
| "eval_runtime": 124.9923, | |
| "eval_samples_per_second": 15.793, | |
| "eval_steps_per_second": 1.976, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.026342451874367, | |
| "grad_norm": 1.949754704355655e-07, | |
| "learning_rate": 2.4670719351570415e-07, | |
| "loss": 0.0565, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.026342451874367, | |
| "eval_loss": 0.5607953667640686, | |
| "eval_runtime": 127.7187, | |
| "eval_samples_per_second": 15.456, | |
| "eval_steps_per_second": 1.934, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0826297422042104, | |
| "grad_norm": 14.2212553024292, | |
| "learning_rate": 2.396712822244737e-07, | |
| "loss": 0.0549, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.0826297422042104, | |
| "eval_loss": 0.5568962693214417, | |
| "eval_runtime": 125.0503, | |
| "eval_samples_per_second": 15.786, | |
| "eval_steps_per_second": 1.975, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.138917032534054, | |
| "grad_norm": 1.2045973539352417, | |
| "learning_rate": 2.3263537093324326e-07, | |
| "loss": 0.0245, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.138917032534054, | |
| "eval_loss": 0.5595241785049438, | |
| "eval_runtime": 124.8527, | |
| "eval_samples_per_second": 15.811, | |
| "eval_steps_per_second": 1.978, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.1952043228638973, | |
| "grad_norm": 8.770056410867255e-06, | |
| "learning_rate": 2.2559945964201284e-07, | |
| "loss": 0.0428, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.1952043228638973, | |
| "eval_loss": 0.5740877985954285, | |
| "eval_runtime": 127.9211, | |
| "eval_samples_per_second": 15.431, | |
| "eval_steps_per_second": 1.931, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.251491613193741, | |
| "grad_norm": 6.243853079013206e-08, | |
| "learning_rate": 2.185635483507824e-07, | |
| "loss": 0.0301, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.251491613193741, | |
| "eval_loss": 0.5761672854423523, | |
| "eval_runtime": 125.2829, | |
| "eval_samples_per_second": 15.756, | |
| "eval_steps_per_second": 1.972, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.3077789035235843, | |
| "grad_norm": 2.299124268745345e-08, | |
| "learning_rate": 2.1152763705955197e-07, | |
| "loss": 0.04, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.3077789035235843, | |
| "eval_loss": 0.5712103843688965, | |
| "eval_runtime": 125.3797, | |
| "eval_samples_per_second": 15.744, | |
| "eval_steps_per_second": 1.97, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.3640661938534278, | |
| "grad_norm": 1.971009623957798e-06, | |
| "learning_rate": 2.044917257683215e-07, | |
| "loss": 0.0349, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.3640661938534278, | |
| "eval_loss": 0.5690667629241943, | |
| "eval_runtime": 125.4733, | |
| "eval_samples_per_second": 15.732, | |
| "eval_steps_per_second": 1.969, | |
| "step": 21000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 35532, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7555285134811260.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |