| { | |
| "best_metric": 1.239951252937317, | |
| "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l18-l/checkpoint-17000", | |
| "epoch": 1.8131399317406145, | |
| "eval_steps": 500, | |
| "global_step": 17000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05332764505119454, | |
| "grad_norm": 186.70851135253906, | |
| "learning_rate": 9.866680887372013e-09, | |
| "loss": 0.4804, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.05332764505119454, | |
| "eval_loss": 1.4967491626739502, | |
| "eval_runtime": 60.9128, | |
| "eval_samples_per_second": 16.204, | |
| "eval_steps_per_second": 2.036, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.10665529010238908, | |
| "grad_norm": 86.50623321533203, | |
| "learning_rate": 9.733361774744027e-09, | |
| "loss": 0.4928, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.10665529010238908, | |
| "eval_loss": 1.482089877128601, | |
| "eval_runtime": 62.7665, | |
| "eval_samples_per_second": 15.725, | |
| "eval_steps_per_second": 1.976, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1599829351535836, | |
| "grad_norm": 603.98974609375, | |
| "learning_rate": 9.600042662116041e-09, | |
| "loss": 0.4325, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1599829351535836, | |
| "eval_loss": 1.4686362743377686, | |
| "eval_runtime": 61.6915, | |
| "eval_samples_per_second": 15.999, | |
| "eval_steps_per_second": 2.01, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21331058020477817, | |
| "grad_norm": 13.387365341186523, | |
| "learning_rate": 9.466723549488055e-09, | |
| "loss": 0.3964, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21331058020477817, | |
| "eval_loss": 1.4561632871627808, | |
| "eval_runtime": 61.8331, | |
| "eval_samples_per_second": 15.962, | |
| "eval_steps_per_second": 2.005, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2666382252559727, | |
| "grad_norm": 322.0577392578125, | |
| "learning_rate": 9.333404436860068e-09, | |
| "loss": 0.4294, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.2666382252559727, | |
| "eval_loss": 1.445016622543335, | |
| "eval_runtime": 63.1527, | |
| "eval_samples_per_second": 15.629, | |
| "eval_steps_per_second": 1.963, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3199658703071672, | |
| "grad_norm": 0.010774992406368256, | |
| "learning_rate": 9.200085324232083e-09, | |
| "loss": 0.3954, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3199658703071672, | |
| "eval_loss": 1.4342260360717773, | |
| "eval_runtime": 63.3562, | |
| "eval_samples_per_second": 15.579, | |
| "eval_steps_per_second": 1.957, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.37329351535836175, | |
| "grad_norm": 42.48988723754883, | |
| "learning_rate": 9.066766211604096e-09, | |
| "loss": 0.4116, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.37329351535836175, | |
| "eval_loss": 1.4254305362701416, | |
| "eval_runtime": 63.1723, | |
| "eval_samples_per_second": 15.624, | |
| "eval_steps_per_second": 1.963, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.42662116040955633, | |
| "grad_norm": 6.53631067276001, | |
| "learning_rate": 8.933447098976109e-09, | |
| "loss": 0.3366, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.42662116040955633, | |
| "eval_loss": 1.4154412746429443, | |
| "eval_runtime": 63.1863, | |
| "eval_samples_per_second": 15.62, | |
| "eval_steps_per_second": 1.962, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.47994880546075086, | |
| "grad_norm": 219.85145568847656, | |
| "learning_rate": 8.800127986348124e-09, | |
| "loss": 0.3941, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.47994880546075086, | |
| "eval_loss": 1.406192421913147, | |
| "eval_runtime": 63.3732, | |
| "eval_samples_per_second": 15.574, | |
| "eval_steps_per_second": 1.957, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5332764505119454, | |
| "grad_norm": 0.5011767148971558, | |
| "learning_rate": 8.666808873720137e-09, | |
| "loss": 0.4351, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5332764505119454, | |
| "eval_loss": 1.3958221673965454, | |
| "eval_runtime": 61.9375, | |
| "eval_samples_per_second": 15.935, | |
| "eval_steps_per_second": 2.002, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.58660409556314, | |
| "grad_norm": 1.465696096420288, | |
| "learning_rate": 8.53348976109215e-09, | |
| "loss": 0.4269, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.58660409556314, | |
| "eval_loss": 1.3861993551254272, | |
| "eval_runtime": 62.145, | |
| "eval_samples_per_second": 15.882, | |
| "eval_steps_per_second": 1.995, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6399317406143344, | |
| "grad_norm": 514.3063354492188, | |
| "learning_rate": 8.400170648464164e-09, | |
| "loss": 0.3764, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6399317406143344, | |
| "eval_loss": 1.377150058746338, | |
| "eval_runtime": 63.2598, | |
| "eval_samples_per_second": 15.602, | |
| "eval_steps_per_second": 1.96, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.693259385665529, | |
| "grad_norm": 267.90301513671875, | |
| "learning_rate": 8.266851535836177e-09, | |
| "loss": 0.3354, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.693259385665529, | |
| "eval_loss": 1.3687273263931274, | |
| "eval_runtime": 61.8197, | |
| "eval_samples_per_second": 15.966, | |
| "eval_steps_per_second": 2.006, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7465870307167235, | |
| "grad_norm": 0.4357345700263977, | |
| "learning_rate": 8.133532423208192e-09, | |
| "loss": 0.3666, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7465870307167235, | |
| "eval_loss": 1.360427975654602, | |
| "eval_runtime": 61.8202, | |
| "eval_samples_per_second": 15.966, | |
| "eval_steps_per_second": 2.006, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7999146757679181, | |
| "grad_norm": 151.71505737304688, | |
| "learning_rate": 8.000213310580205e-09, | |
| "loss": 0.3991, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7999146757679181, | |
| "eval_loss": 1.351379156112671, | |
| "eval_runtime": 62.9768, | |
| "eval_samples_per_second": 15.672, | |
| "eval_steps_per_second": 1.969, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8532423208191127, | |
| "grad_norm": 467.88653564453125, | |
| "learning_rate": 7.86689419795222e-09, | |
| "loss": 0.4028, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8532423208191127, | |
| "eval_loss": 1.3436161279678345, | |
| "eval_runtime": 62.1794, | |
| "eval_samples_per_second": 15.873, | |
| "eval_steps_per_second": 1.994, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9065699658703071, | |
| "grad_norm": 186.75767517089844, | |
| "learning_rate": 7.733575085324233e-09, | |
| "loss": 0.3412, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9065699658703071, | |
| "eval_loss": 1.3365025520324707, | |
| "eval_runtime": 62.3614, | |
| "eval_samples_per_second": 15.827, | |
| "eval_steps_per_second": 1.988, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9598976109215017, | |
| "grad_norm": 62.3108024597168, | |
| "learning_rate": 7.600255972696246e-09, | |
| "loss": 0.3685, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.9598976109215017, | |
| "eval_loss": 1.3287090063095093, | |
| "eval_runtime": 63.3854, | |
| "eval_samples_per_second": 15.571, | |
| "eval_steps_per_second": 1.956, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0132252559726962, | |
| "grad_norm": 537.1036987304688, | |
| "learning_rate": 7.46693686006826e-09, | |
| "loss": 0.3502, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0132252559726962, | |
| "eval_loss": 1.3230267763137817, | |
| "eval_runtime": 61.9741, | |
| "eval_samples_per_second": 15.926, | |
| "eval_steps_per_second": 2.001, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0665529010238908, | |
| "grad_norm": 433.3678283691406, | |
| "learning_rate": 7.333617747440273e-09, | |
| "loss": 0.3423, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0665529010238908, | |
| "eval_loss": 1.3153482675552368, | |
| "eval_runtime": 63.108, | |
| "eval_samples_per_second": 15.64, | |
| "eval_steps_per_second": 1.965, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1198805460750854, | |
| "grad_norm": 16.9348087310791, | |
| "learning_rate": 7.2002986348122865e-09, | |
| "loss": 0.3562, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.1198805460750854, | |
| "eval_loss": 1.307684063911438, | |
| "eval_runtime": 62.1177, | |
| "eval_samples_per_second": 15.889, | |
| "eval_steps_per_second": 1.996, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.17320819112628, | |
| "grad_norm": 333.85516357421875, | |
| "learning_rate": 7.0669795221843005e-09, | |
| "loss": 0.34, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.17320819112628, | |
| "eval_loss": 1.3010845184326172, | |
| "eval_runtime": 63.2514, | |
| "eval_samples_per_second": 15.604, | |
| "eval_steps_per_second": 1.96, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.2265358361774745, | |
| "grad_norm": 0.8008983731269836, | |
| "learning_rate": 6.933660409556314e-09, | |
| "loss": 0.3681, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.2265358361774745, | |
| "eval_loss": 1.2945556640625, | |
| "eval_runtime": 62.1026, | |
| "eval_samples_per_second": 15.893, | |
| "eval_steps_per_second": 1.997, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.2798634812286689, | |
| "grad_norm": 61.41884994506836, | |
| "learning_rate": 6.800341296928328e-09, | |
| "loss": 0.304, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.2798634812286689, | |
| "eval_loss": 1.2891267538070679, | |
| "eval_runtime": 63.0412, | |
| "eval_samples_per_second": 15.656, | |
| "eval_steps_per_second": 1.967, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3331911262798635, | |
| "grad_norm": 0.1409968137741089, | |
| "learning_rate": 6.667022184300341e-09, | |
| "loss": 0.3365, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.3331911262798635, | |
| "eval_loss": 1.2835596799850464, | |
| "eval_runtime": 62.1735, | |
| "eval_samples_per_second": 15.875, | |
| "eval_steps_per_second": 1.994, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.386518771331058, | |
| "grad_norm": 12.641498565673828, | |
| "learning_rate": 6.533703071672356e-09, | |
| "loss": 0.323, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.386518771331058, | |
| "eval_loss": 1.2784309387207031, | |
| "eval_runtime": 63.1295, | |
| "eval_samples_per_second": 15.635, | |
| "eval_steps_per_second": 1.964, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.4398464163822526, | |
| "grad_norm": 44.88418960571289, | |
| "learning_rate": 6.400383959044369e-09, | |
| "loss": 0.3823, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.4398464163822526, | |
| "eval_loss": 1.2729544639587402, | |
| "eval_runtime": 62.2234, | |
| "eval_samples_per_second": 15.862, | |
| "eval_steps_per_second": 1.993, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.493174061433447, | |
| "grad_norm": 331.7489013671875, | |
| "learning_rate": 6.267064846416382e-09, | |
| "loss": 0.2957, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.493174061433447, | |
| "eval_loss": 1.267383098602295, | |
| "eval_runtime": 63.223, | |
| "eval_samples_per_second": 15.611, | |
| "eval_steps_per_second": 1.961, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5465017064846416, | |
| "grad_norm": 144.97119140625, | |
| "learning_rate": 6.133745733788396e-09, | |
| "loss": 0.3459, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.5465017064846416, | |
| "eval_loss": 1.26187002658844, | |
| "eval_runtime": 62.0237, | |
| "eval_samples_per_second": 15.913, | |
| "eval_steps_per_second": 1.999, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.5998293515358362, | |
| "grad_norm": 193.1491241455078, | |
| "learning_rate": 6.0004266211604095e-09, | |
| "loss": 0.2956, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.5998293515358362, | |
| "eval_loss": 1.2574553489685059, | |
| "eval_runtime": 63.3494, | |
| "eval_samples_per_second": 15.58, | |
| "eval_steps_per_second": 1.957, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.6531569965870307, | |
| "grad_norm": 229.95928955078125, | |
| "learning_rate": 5.8671075085324235e-09, | |
| "loss": 0.3341, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.6531569965870307, | |
| "eval_loss": 1.2525171041488647, | |
| "eval_runtime": 62.229, | |
| "eval_samples_per_second": 15.861, | |
| "eval_steps_per_second": 1.993, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.7064846416382253, | |
| "grad_norm": 340.9511413574219, | |
| "learning_rate": 5.733788395904437e-09, | |
| "loss": 0.3303, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7064846416382253, | |
| "eval_loss": 1.2486214637756348, | |
| "eval_runtime": 61.8176, | |
| "eval_samples_per_second": 15.966, | |
| "eval_steps_per_second": 2.006, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.75981228668942, | |
| "grad_norm": 343.5415954589844, | |
| "learning_rate": 5.600469283276451e-09, | |
| "loss": 0.3257, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.75981228668942, | |
| "eval_loss": 1.2440201044082642, | |
| "eval_runtime": 62.3742, | |
| "eval_samples_per_second": 15.824, | |
| "eval_steps_per_second": 1.988, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.8131399317406145, | |
| "grad_norm": 340.19482421875, | |
| "learning_rate": 5.467150170648464e-09, | |
| "loss": 0.3302, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8131399317406145, | |
| "eval_loss": 1.239951252937317, | |
| "eval_runtime": 63.4481, | |
| "eval_samples_per_second": 15.556, | |
| "eval_steps_per_second": 1.954, | |
| "step": 17000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 37504, | |
| "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": 6116611488122520.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |