| { | |
| "best_metric": 0.09835900366306305, | |
| "best_model_checkpoint": "./vit-base-beans-demo-v5/checkpoint-500", | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 567, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01763668430335097, | |
| "grad_norm": 0.15796910226345062, | |
| "learning_rate": 0.00019647266313932982, | |
| "loss": 0.1281, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03527336860670194, | |
| "grad_norm": 0.291827917098999, | |
| "learning_rate": 0.0001929453262786596, | |
| "loss": 0.1139, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05291005291005291, | |
| "grad_norm": 0.489213764667511, | |
| "learning_rate": 0.00018941798941798943, | |
| "loss": 0.1035, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07054673721340388, | |
| "grad_norm": 0.11527382582426071, | |
| "learning_rate": 0.00018589065255731924, | |
| "loss": 0.1207, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08818342151675485, | |
| "grad_norm": 0.15595707297325134, | |
| "learning_rate": 0.00018236331569664903, | |
| "loss": 0.1186, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10582010582010581, | |
| "grad_norm": 0.7361170053482056, | |
| "learning_rate": 0.00017883597883597884, | |
| "loss": 0.1058, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 0.289522647857666, | |
| "learning_rate": 0.00017530864197530866, | |
| "loss": 0.0913, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.14109347442680775, | |
| "grad_norm": 0.0984591692686081, | |
| "learning_rate": 0.00017178130511463847, | |
| "loss": 0.0928, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 0.08082891255617142, | |
| "learning_rate": 0.00016825396825396826, | |
| "loss": 0.0829, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1763668430335097, | |
| "grad_norm": 0.5304596424102783, | |
| "learning_rate": 0.00016472663139329807, | |
| "loss": 0.1192, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1763668430335097, | |
| "eval_accuracy": 0.9692460317460317, | |
| "eval_loss": 0.1157514676451683, | |
| "eval_runtime": 12.4102, | |
| "eval_samples_per_second": 81.224, | |
| "eval_steps_per_second": 10.153, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19400352733686066, | |
| "grad_norm": 0.2734580636024475, | |
| "learning_rate": 0.0001611992945326279, | |
| "loss": 0.0381, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.21164021164021163, | |
| "grad_norm": 1.1482329368591309, | |
| "learning_rate": 0.00015767195767195767, | |
| "loss": 0.1579, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2292768959435626, | |
| "grad_norm": 0.29367437958717346, | |
| "learning_rate": 0.0001541446208112875, | |
| "loss": 0.1296, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 0.4296620190143585, | |
| "learning_rate": 0.0001506172839506173, | |
| "loss": 0.0876, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.26455026455026454, | |
| "grad_norm": 0.430569589138031, | |
| "learning_rate": 0.0001470899470899471, | |
| "loss": 0.1431, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2821869488536155, | |
| "grad_norm": 0.17311625182628632, | |
| "learning_rate": 0.0001435626102292769, | |
| "loss": 0.1194, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2998236331569665, | |
| "grad_norm": 0.21777546405792236, | |
| "learning_rate": 0.00014003527336860672, | |
| "loss": 0.1202, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 2.8384833335876465, | |
| "learning_rate": 0.0001365079365079365, | |
| "loss": 0.1352, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3350970017636684, | |
| "grad_norm": 0.3133362829685211, | |
| "learning_rate": 0.00013298059964726632, | |
| "loss": 0.0935, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3527336860670194, | |
| "grad_norm": 0.27794110774993896, | |
| "learning_rate": 0.00012945326278659614, | |
| "loss": 0.0734, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3527336860670194, | |
| "eval_accuracy": 0.9702380952380952, | |
| "eval_loss": 0.1267656832933426, | |
| "eval_runtime": 12.6893, | |
| "eval_samples_per_second": 79.437, | |
| "eval_steps_per_second": 9.93, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 0.12122131884098053, | |
| "learning_rate": 0.00012592592592592592, | |
| "loss": 0.1548, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3880070546737213, | |
| "grad_norm": 0.14606191217899323, | |
| "learning_rate": 0.00012239858906525574, | |
| "loss": 0.0166, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4056437389770723, | |
| "grad_norm": 0.118919737637043, | |
| "learning_rate": 0.00011887125220458555, | |
| "loss": 0.1295, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.42328042328042326, | |
| "grad_norm": 0.14326384663581848, | |
| "learning_rate": 0.00011534391534391535, | |
| "loss": 0.1287, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4409171075837742, | |
| "grad_norm": 1.4116628170013428, | |
| "learning_rate": 0.00011181657848324515, | |
| "loss": 0.148, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4585537918871252, | |
| "grad_norm": 0.18267543613910675, | |
| "learning_rate": 0.00010828924162257497, | |
| "loss": 0.1161, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 0.6114774346351624, | |
| "learning_rate": 0.00010476190476190477, | |
| "loss": 0.0778, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 0.10120945423841476, | |
| "learning_rate": 0.00010123456790123458, | |
| "loss": 0.026, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5114638447971781, | |
| "grad_norm": 0.5115650296211243, | |
| "learning_rate": 9.770723104056437e-05, | |
| "loss": 0.0964, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5291005291005291, | |
| "grad_norm": 0.2577248513698578, | |
| "learning_rate": 9.417989417989419e-05, | |
| "loss": 0.0701, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5291005291005291, | |
| "eval_accuracy": 0.9672619047619048, | |
| "eval_loss": 0.1057305857539177, | |
| "eval_runtime": 12.4383, | |
| "eval_samples_per_second": 81.04, | |
| "eval_steps_per_second": 10.13, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.54673721340388, | |
| "grad_norm": 1.4628558158874512, | |
| "learning_rate": 9.065255731922399e-05, | |
| "loss": 0.1241, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.564373897707231, | |
| "grad_norm": 0.4355175793170929, | |
| "learning_rate": 8.712522045855379e-05, | |
| "loss": 0.125, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.582010582010582, | |
| "grad_norm": 0.09273388981819153, | |
| "learning_rate": 8.35978835978836e-05, | |
| "loss": 0.0728, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.599647266313933, | |
| "grad_norm": 0.777564525604248, | |
| "learning_rate": 8.00705467372134e-05, | |
| "loss": 0.0974, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 0.8731770515441895, | |
| "learning_rate": 7.65432098765432e-05, | |
| "loss": 0.1852, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 0.18054471909999847, | |
| "learning_rate": 7.301587301587302e-05, | |
| "loss": 0.0722, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6525573192239859, | |
| "grad_norm": 0.10277850925922394, | |
| "learning_rate": 6.948853615520282e-05, | |
| "loss": 0.0705, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6701940035273368, | |
| "grad_norm": 0.2326202243566513, | |
| "learning_rate": 6.596119929453263e-05, | |
| "loss": 0.0813, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6878306878306878, | |
| "grad_norm": 0.4453868865966797, | |
| "learning_rate": 6.243386243386243e-05, | |
| "loss": 0.0443, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7054673721340388, | |
| "grad_norm": 0.18326041102409363, | |
| "learning_rate": 5.890652557319224e-05, | |
| "loss": 0.1107, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7054673721340388, | |
| "eval_accuracy": 0.9722222222222222, | |
| "eval_loss": 0.10807543247938156, | |
| "eval_runtime": 12.4927, | |
| "eval_samples_per_second": 80.687, | |
| "eval_steps_per_second": 10.086, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7231040564373897, | |
| "grad_norm": 0.08458270132541656, | |
| "learning_rate": 5.537918871252204e-05, | |
| "loss": 0.0699, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 0.3042624294757843, | |
| "learning_rate": 5.185185185185185e-05, | |
| "loss": 0.0649, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7583774250440917, | |
| "grad_norm": 0.1482134908437729, | |
| "learning_rate": 4.832451499118166e-05, | |
| "loss": 0.0652, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7760141093474426, | |
| "grad_norm": 0.08534322679042816, | |
| "learning_rate": 4.4797178130511465e-05, | |
| "loss": 0.0572, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 0.09549910575151443, | |
| "learning_rate": 4.126984126984127e-05, | |
| "loss": 0.0527, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8112874779541446, | |
| "grad_norm": 1.4032962322235107, | |
| "learning_rate": 3.7742504409171074e-05, | |
| "loss": 0.0643, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8289241622574955, | |
| "grad_norm": 0.5806692242622375, | |
| "learning_rate": 3.421516754850088e-05, | |
| "loss": 0.0699, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8465608465608465, | |
| "grad_norm": 0.07148485630750656, | |
| "learning_rate": 3.068783068783069e-05, | |
| "loss": 0.1102, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 0.6352106928825378, | |
| "learning_rate": 2.7160493827160493e-05, | |
| "loss": 0.0574, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8818342151675485, | |
| "grad_norm": 0.8136057257652283, | |
| "learning_rate": 2.36331569664903e-05, | |
| "loss": 0.0413, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8818342151675485, | |
| "eval_accuracy": 0.9761904761904762, | |
| "eval_loss": 0.09835900366306305, | |
| "eval_runtime": 12.4974, | |
| "eval_samples_per_second": 80.657, | |
| "eval_steps_per_second": 10.082, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8994708994708994, | |
| "grad_norm": 0.11122062802314758, | |
| "learning_rate": 2.0105820105820105e-05, | |
| "loss": 0.0287, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9171075837742504, | |
| "grad_norm": 0.08222879469394684, | |
| "learning_rate": 1.6578483245149913e-05, | |
| "loss": 0.0793, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9347442680776014, | |
| "grad_norm": 0.08997409790754318, | |
| "learning_rate": 1.3051146384479717e-05, | |
| "loss": 0.0523, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.14071638882160187, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 0.0828, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.9700176366843033, | |
| "grad_norm": 0.0937449261546135, | |
| "learning_rate": 5.99647266313933e-06, | |
| "loss": 0.1201, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 2.3273768424987793, | |
| "learning_rate": 2.469135802469136e-06, | |
| "loss": 0.1994, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 567, | |
| "total_flos": 7.023873938666619e+17, | |
| "train_loss": 0.09499678115575612, | |
| "train_runtime": 238.2789, | |
| "train_samples_per_second": 38.039, | |
| "train_steps_per_second": 2.38 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 567, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.023873938666619e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |