diff --git a/ood-viclickbait2025-xlmr-source-source-04-s44-20260901T140530Z-6863e5c6/trainer/checkpoint-720/trainer_state.json b/ood-viclickbait2025-xlmr-source-source-04-s44-20260901T140530Z-6863e5c6/trainer/checkpoint-720/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..eb4a669588fb210b9ca6ab9c66e052c6efda1c0d --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-04-s44-20260901T140530Z-6863e5c6/trainer/checkpoint-720/trainer_state.json @@ -0,0 +1,338 @@ +{ + "best_global_step": 480, + "best_metric": 0.8749119097956307, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-viclickbait2025-xlmr-source-source-04-s44-20260901T140530Z-6863e5c6/trainer/checkpoint-480", + "epoch": 9.0, + "eval_steps": 500, + "global_step": 720, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.3125, + "grad_norm": 36.113670349121094, + "learning_rate": 3e-06, + "loss": 2.9318, + "step": 25 + }, + { + "epoch": 0.625, + "grad_norm": 120.8009262084961, + "learning_rate": 6.125000000000001e-06, + "loss": 2.6839, + "step": 50 + }, + { + "epoch": 0.9375, + "grad_norm": 27.118284225463867, + "learning_rate": 9.250000000000001e-06, + "loss": 2.4275, + "step": 75 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.676056338028169, + "eval_loss": 0.5310995578765869, + "eval_macro_f1": 0.40336134453781514, + "eval_runtime": 0.6358, + "eval_samples_per_second": 446.665, + "eval_steps_per_second": 28.31, + "eval_weighted_f1": 0.5453899869807078, + "step": 80 + }, + { + "epoch": 1.25, + "grad_norm": 33.2912712097168, + "learning_rate": 1.2375000000000001e-05, + "loss": 2.1637, + "step": 100 + }, + { + "epoch": 1.5625, + "grad_norm": 93.63400268554688, + "learning_rate": 1.55e-05, + "loss": 2.1012, + "step": 125 + }, + { + "epoch": 1.875, + "grad_norm": 59.89738845825195, + "learning_rate": 1.8625000000000002e-05, + "loss": 2.019, + "step": 150 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.8345070422535211, + "eval_loss": 0.39797243475914, + "eval_macro_f1": 0.8198528915581348, + "eval_runtime": 0.639, + "eval_samples_per_second": 444.427, + "eval_steps_per_second": 28.168, + "eval_weighted_f1": 0.837944435626513, + "step": 160 + }, + { + "epoch": 2.1875, + "grad_norm": 48.018943786621094, + "learning_rate": 1.9805555555555557e-05, + "loss": 1.8596, + "step": 175 + }, + { + "epoch": 2.5, + "grad_norm": 97.55175018310547, + "learning_rate": 1.9458333333333333e-05, + "loss": 1.7207, + "step": 200 + }, + { + "epoch": 2.8125, + "grad_norm": 115.1563491821289, + "learning_rate": 1.9111111111111113e-05, + "loss": 1.6247, + "step": 225 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.8908450704225352, + "eval_loss": 0.2916692793369293, + "eval_macro_f1": 0.8743129612974145, + "eval_runtime": 0.6489, + "eval_samples_per_second": 437.665, + "eval_steps_per_second": 27.739, + "eval_weighted_f1": 0.8903635526810268, + "step": 240 + }, + { + "epoch": 3.125, + "grad_norm": 30.12644386291504, + "learning_rate": 1.876388888888889e-05, + "loss": 1.4328, + "step": 250 + }, + { + "epoch": 3.4375, + "grad_norm": 69.03034210205078, + "learning_rate": 1.8416666666666666e-05, + "loss": 1.1091, + "step": 275 + }, + { + "epoch": 3.75, + "grad_norm": 72.46318054199219, + "learning_rate": 1.8069444444444446e-05, + "loss": 1.4007, + "step": 300 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8591549295774648, + "eval_loss": 0.32971417903900146, + "eval_macro_f1": 0.8409854423292273, + "eval_runtime": 0.6502, + "eval_samples_per_second": 436.762, + "eval_steps_per_second": 27.682, + "eval_weighted_f1": 0.8599119915461413, + "step": 320 + }, + { + "epoch": 4.0625, + "grad_norm": 28.667325973510742, + "learning_rate": 1.7722222222222222e-05, + "loss": 1.2847, + "step": 325 + }, + { + "epoch": 4.375, + "grad_norm": 44.14274597167969, + "learning_rate": 1.7375000000000002e-05, + "loss": 0.8504, + "step": 350 + }, + { + "epoch": 4.6875, + "grad_norm": 68.7794189453125, + "learning_rate": 1.702777777777778e-05, + "loss": 1.1538, + "step": 375 + }, + { + "epoch": 5.0, + "grad_norm": 70.4148178100586, + "learning_rate": 1.668055555555556e-05, + "loss": 1.0259, + "step": 400 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8802816901408451, + "eval_loss": 0.32677745819091797, + "eval_macro_f1": 0.8648376259798433, + "eval_runtime": 0.6628, + "eval_samples_per_second": 428.517, + "eval_steps_per_second": 27.16, + "eval_weighted_f1": 0.8809251928142202, + "step": 400 + }, + { + "epoch": 5.3125, + "grad_norm": 61.19955062866211, + "learning_rate": 1.6333333333333335e-05, + "loss": 0.7135, + "step": 425 + }, + { + "epoch": 5.625, + "grad_norm": 22.916311264038086, + "learning_rate": 1.5986111111111115e-05, + "loss": 0.7698, + "step": 450 + }, + { + "epoch": 5.9375, + "grad_norm": 21.670488357543945, + "learning_rate": 1.563888888888889e-05, + "loss": 0.5699, + "step": 475 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8943661971830986, + "eval_loss": 0.382527619600296, + "eval_macro_f1": 0.8749119097956307, + "eval_runtime": 0.6631, + "eval_samples_per_second": 428.31, + "eval_steps_per_second": 27.146, + "eval_weighted_f1": 0.892281809248727, + "step": 480 + }, + { + "epoch": 6.25, + "grad_norm": 108.23480224609375, + "learning_rate": 1.5291666666666668e-05, + "loss": 0.6613, + "step": 500 + }, + { + "epoch": 6.5625, + "grad_norm": 176.19325256347656, + "learning_rate": 1.4944444444444446e-05, + "loss": 0.4539, + "step": 525 + }, + { + "epoch": 6.875, + "grad_norm": 49.169189453125, + "learning_rate": 1.4597222222222223e-05, + "loss": 0.3864, + "step": 550 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8838028169014085, + "eval_loss": 0.37622955441474915, + "eval_macro_f1": 0.8684576192681797, + "eval_runtime": 0.6716, + "eval_samples_per_second": 422.896, + "eval_steps_per_second": 26.803, + "eval_weighted_f1": 0.8842774106426424, + "step": 560 + }, + { + "epoch": 7.1875, + "grad_norm": 55.59708786010742, + "learning_rate": 1.425e-05, + "loss": 0.4167, + "step": 575 + }, + { + "epoch": 7.5, + "grad_norm": 123.0556869506836, + "learning_rate": 1.3902777777777779e-05, + "loss": 0.3474, + "step": 600 + }, + { + "epoch": 7.8125, + "grad_norm": 126.29974365234375, + "learning_rate": 1.3555555555555557e-05, + "loss": 0.5173, + "step": 625 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8591549295774648, + "eval_loss": 0.5031241774559021, + "eval_macro_f1": 0.8441957428132543, + "eval_runtime": 0.6342, + "eval_samples_per_second": 447.795, + "eval_steps_per_second": 28.381, + "eval_weighted_f1": 0.8611948186816752, + "step": 640 + }, + { + "epoch": 8.125, + "grad_norm": 1.4242525100708008, + "learning_rate": 1.3208333333333334e-05, + "loss": 0.3358, + "step": 650 + }, + { + "epoch": 8.4375, + "grad_norm": 91.3457260131836, + "learning_rate": 1.2861111111111114e-05, + "loss": 0.4218, + "step": 675 + }, + { + "epoch": 8.75, + "grad_norm": 113.89402770996094, + "learning_rate": 1.251388888888889e-05, + "loss": 0.2973, + "step": 700 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8732394366197183, + "eval_loss": 0.7847744822502136, + "eval_macro_f1": 0.8544667767465695, + "eval_runtime": 0.6354, + "eval_samples_per_second": 446.983, + "eval_steps_per_second": 28.33, + "eval_weighted_f1": 0.8728713452496566, + "step": 720 + } + ], + "logging_steps": 25, + "max_steps": 1600, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/special_tokens_map.json b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/tokenizer_config.json b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/trainer_state.json b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..bfcb6003623c03f1ed7965633d8d9571e217c318 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350/trainer_state.json @@ -0,0 +1,586 @@ +{ + "best_global_step": 1350, + "best_metric": 0.8314726209463051, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350", + "epoch": 15.0, + "eval_steps": 500, + "global_step": 1350, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.2801120448179272, + "grad_norm": 21.286026000976562, + "learning_rate": 2.666666666666667e-06, + "loss": 2.9303, + "step": 25 + }, + { + "epoch": 0.5602240896358543, + "grad_norm": 18.28496742248535, + "learning_rate": 5.444444444444445e-06, + "loss": 2.7658, + "step": 50 + }, + { + "epoch": 0.8403361344537815, + "grad_norm": 23.925424575805664, + "learning_rate": 8.222222222222222e-06, + "loss": 2.427, + "step": 75 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.7665615141955836, + "eval_loss": 0.5131620168685913, + "eval_macro_f1": 0.7311835350201688, + "eval_runtime": 0.7127, + "eval_samples_per_second": 444.81, + "eval_steps_per_second": 28.064, + "eval_weighted_f1": 0.759793552962026, + "step": 90 + }, + { + "epoch": 1.112044817927171, + "grad_norm": 42.43369674682617, + "learning_rate": 1.1000000000000001e-05, + "loss": 2.151, + "step": 100 + }, + { + "epoch": 1.392156862745098, + "grad_norm": 75.90241241455078, + "learning_rate": 1.377777777777778e-05, + "loss": 1.8566, + "step": 125 + }, + { + "epoch": 1.6722689075630253, + "grad_norm": 39.13203048706055, + "learning_rate": 1.6555555555555556e-05, + "loss": 2.0029, + "step": 150 + }, + { + "epoch": 1.9523809523809523, + "grad_norm": 113.73194122314453, + "learning_rate": 1.9333333333333333e-05, + "loss": 1.7688, + "step": 175 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.8107255520504731, + "eval_loss": 0.535222053527832, + "eval_macro_f1": 0.7784456248252726, + "eval_runtime": 0.6496, + "eval_samples_per_second": 488.029, + "eval_steps_per_second": 30.791, + "eval_weighted_f1": 0.803255816824807, + "step": 180 + }, + { + "epoch": 2.224089635854342, + "grad_norm": 58.54833984375, + "learning_rate": 1.9765432098765434e-05, + "loss": 1.5445, + "step": 200 + }, + { + "epoch": 2.504201680672269, + "grad_norm": 121.798095703125, + "learning_rate": 1.9456790123456793e-05, + "loss": 1.6574, + "step": 225 + }, + { + "epoch": 2.784313725490196, + "grad_norm": 34.078330993652344, + "learning_rate": 1.9148148148148148e-05, + "loss": 1.455, + "step": 250 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.8107255520504731, + "eval_loss": 0.5058051943778992, + "eval_macro_f1": 0.803933696190005, + "eval_runtime": 0.6798, + "eval_samples_per_second": 466.328, + "eval_steps_per_second": 29.421, + "eval_weighted_f1": 0.8146395028853194, + "step": 270 + }, + { + "epoch": 3.0560224089635852, + "grad_norm": 53.804298400878906, + "learning_rate": 1.8839506172839507e-05, + "loss": 1.3164, + "step": 275 + }, + { + "epoch": 3.3361344537815127, + "grad_norm": 46.20854949951172, + "learning_rate": 1.8530864197530866e-05, + "loss": 1.3639, + "step": 300 + }, + { + "epoch": 3.61624649859944, + "grad_norm": 45.22013473510742, + "learning_rate": 1.8222222222222224e-05, + "loss": 1.1226, + "step": 325 + }, + { + "epoch": 3.896358543417367, + "grad_norm": 48.710594177246094, + "learning_rate": 1.7913580246913583e-05, + "loss": 1.1293, + "step": 350 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8170347003154574, + "eval_loss": 0.3787948191165924, + "eval_macro_f1": 0.7998040069686412, + "eval_runtime": 0.7137, + "eval_samples_per_second": 444.141, + "eval_steps_per_second": 28.022, + "eval_weighted_f1": 0.8170347003154574, + "step": 360 + }, + { + "epoch": 4.168067226890757, + "grad_norm": 83.72383880615234, + "learning_rate": 1.760493827160494e-05, + "loss": 0.8385, + "step": 375 + }, + { + "epoch": 4.448179271708684, + "grad_norm": 62.41585922241211, + "learning_rate": 1.7296296296296297e-05, + "loss": 0.9109, + "step": 400 + }, + { + "epoch": 4.728291316526611, + "grad_norm": 37.59474563598633, + "learning_rate": 1.6987654320987656e-05, + "loss": 0.9558, + "step": 425 + }, + { + "epoch": 5.0, + "grad_norm": 4.466841697692871, + "learning_rate": 1.6679012345679015e-05, + "loss": 0.8452, + "step": 450 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8138801261829653, + "eval_loss": 0.561353325843811, + "eval_macro_f1": 0.8031532527127867, + "eval_runtime": 0.6452, + "eval_samples_per_second": 491.315, + "eval_steps_per_second": 30.998, + "eval_weighted_f1": 0.8166343234253083, + "step": 450 + }, + { + "epoch": 5.280112044817927, + "grad_norm": 46.44953918457031, + "learning_rate": 1.6370370370370374e-05, + "loss": 0.6156, + "step": 475 + }, + { + "epoch": 5.560224089635854, + "grad_norm": 44.48462677001953, + "learning_rate": 1.606172839506173e-05, + "loss": 0.615, + "step": 500 + }, + { + "epoch": 5.840336134453781, + "grad_norm": 77.2662353515625, + "learning_rate": 1.5753086419753088e-05, + "loss": 0.7461, + "step": 525 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8422712933753943, + "eval_loss": 0.6332255601882935, + "eval_macro_f1": 0.8163979241960894, + "eval_runtime": 0.666, + "eval_samples_per_second": 476.01, + "eval_steps_per_second": 30.032, + "eval_weighted_f1": 0.8366182883446218, + "step": 540 + }, + { + "epoch": 6.1120448179271705, + "grad_norm": 5.867457389831543, + "learning_rate": 1.5444444444444446e-05, + "loss": 0.6319, + "step": 550 + }, + { + "epoch": 6.392156862745098, + "grad_norm": 3.1487138271331787, + "learning_rate": 1.5135802469135803e-05, + "loss": 0.3849, + "step": 575 + }, + { + "epoch": 6.6722689075630255, + "grad_norm": 39.26576232910156, + "learning_rate": 1.4827160493827162e-05, + "loss": 0.5494, + "step": 600 + }, + { + "epoch": 6.9523809523809526, + "grad_norm": 3.9887986183166504, + "learning_rate": 1.4518518518518521e-05, + "loss": 0.6336, + "step": 625 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8391167192429022, + "eval_loss": 0.7031358480453491, + "eval_macro_f1": 0.8197116188819378, + "eval_runtime": 0.6708, + "eval_samples_per_second": 472.594, + "eval_steps_per_second": 29.817, + "eval_weighted_f1": 0.8370642567047232, + "step": 630 + }, + { + "epoch": 7.224089635854342, + "grad_norm": 11.982047080993652, + "learning_rate": 1.4209876543209878e-05, + "loss": 0.3573, + "step": 650 + }, + { + "epoch": 7.504201680672269, + "grad_norm": 40.24326705932617, + "learning_rate": 1.3901234567901237e-05, + "loss": 0.3807, + "step": 675 + }, + { + "epoch": 7.784313725490196, + "grad_norm": 1.3944097757339478, + "learning_rate": 1.3592592592592592e-05, + "loss": 0.4012, + "step": 700 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8170347003154574, + "eval_loss": 1.148803472518921, + "eval_macro_f1": 0.805562605752961, + "eval_runtime": 0.665, + "eval_samples_per_second": 476.714, + "eval_steps_per_second": 30.077, + "eval_weighted_f1": 0.8194185121725995, + "step": 720 + }, + { + "epoch": 8.056022408963585, + "grad_norm": 212.5960235595703, + "learning_rate": 1.3283950617283951e-05, + "loss": 0.4906, + "step": 725 + }, + { + "epoch": 8.336134453781513, + "grad_norm": 318.5392761230469, + "learning_rate": 1.297530864197531e-05, + "loss": 0.3477, + "step": 750 + }, + { + "epoch": 8.61624649859944, + "grad_norm": 46.817447662353516, + "learning_rate": 1.2666666666666667e-05, + "loss": 0.3973, + "step": 775 + }, + { + "epoch": 8.896358543417367, + "grad_norm": 1.5035258531570435, + "learning_rate": 1.2358024691358026e-05, + "loss": 0.2958, + "step": 800 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8454258675078864, + "eval_loss": 0.9726073741912842, + "eval_macro_f1": 0.8290917093029653, + "eval_runtime": 0.6293, + "eval_samples_per_second": 503.774, + "eval_steps_per_second": 31.784, + "eval_weighted_f1": 0.8445924920892679, + "step": 810 + }, + { + "epoch": 9.168067226890756, + "grad_norm": 0.4589764177799225, + "learning_rate": 1.2049382716049384e-05, + "loss": 0.1497, + "step": 825 + }, + { + "epoch": 9.448179271708684, + "grad_norm": 8.860508918762207, + "learning_rate": 1.1740740740740741e-05, + "loss": 0.243, + "step": 850 + }, + { + "epoch": 9.72829131652661, + "grad_norm": 130.12161254882812, + "learning_rate": 1.14320987654321e-05, + "loss": 0.2929, + "step": 875 + }, + { + "epoch": 10.0, + "grad_norm": 0.031857386231422424, + "learning_rate": 1.1123456790123457e-05, + "loss": 0.2966, + "step": 900 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8422712933753943, + "eval_loss": 1.1647114753723145, + "eval_macro_f1": 0.8259771629336847, + "eval_runtime": 0.6312, + "eval_samples_per_second": 502.21, + "eval_steps_per_second": 31.685, + "eval_weighted_f1": 0.8415993704705815, + "step": 900 + }, + { + "epoch": 10.280112044817928, + "grad_norm": 1010.2288208007812, + "learning_rate": 1.0814814814814816e-05, + "loss": 0.1748, + "step": 925 + }, + { + "epoch": 10.560224089635854, + "grad_norm": 0.4934237003326416, + "learning_rate": 1.0506172839506175e-05, + "loss": 0.2865, + "step": 950 + }, + { + "epoch": 10.840336134453782, + "grad_norm": 6.401635646820068, + "learning_rate": 1.0197530864197532e-05, + "loss": 0.2486, + "step": 975 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.8359621451104101, + "eval_loss": 1.4146658182144165, + "eval_macro_f1": 0.8219255034137067, + "eval_runtime": 0.6608, + "eval_samples_per_second": 479.755, + "eval_steps_per_second": 30.268, + "eval_weighted_f1": 0.8365930054113856, + "step": 990 + }, + { + "epoch": 11.11204481792717, + "grad_norm": 438.1229248046875, + "learning_rate": 9.88888888888889e-06, + "loss": 0.1684, + "step": 1000 + }, + { + "epoch": 11.392156862745098, + "grad_norm": 416.82550048828125, + "learning_rate": 9.580246913580248e-06, + "loss": 0.1978, + "step": 1025 + }, + { + "epoch": 11.672268907563025, + "grad_norm": 204.48353576660156, + "learning_rate": 9.271604938271606e-06, + "loss": 0.2504, + "step": 1050 + }, + { + "epoch": 11.952380952380953, + "grad_norm": 3.435577869415283, + "learning_rate": 8.962962962962963e-06, + "loss": 0.2117, + "step": 1075 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8485804416403786, + "eval_loss": 1.2211965322494507, + "eval_macro_f1": 0.8299204149154968, + "eval_runtime": 0.6933, + "eval_samples_per_second": 457.219, + "eval_steps_per_second": 28.847, + "eval_weighted_f1": 0.8464478671575351, + "step": 1080 + }, + { + "epoch": 12.224089635854341, + "grad_norm": 0.07730665057897568, + "learning_rate": 8.654320987654322e-06, + "loss": 0.1066, + "step": 1100 + }, + { + "epoch": 12.504201680672269, + "grad_norm": 0.006811291445046663, + "learning_rate": 8.34567901234568e-06, + "loss": 0.1689, + "step": 1125 + }, + { + "epoch": 12.784313725490197, + "grad_norm": 26.207834243774414, + "learning_rate": 8.037037037037038e-06, + "loss": 0.102, + "step": 1150 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.8454258675078864, + "eval_loss": 1.3052610158920288, + "eval_macro_f1": 0.8290917093029653, + "eval_runtime": 0.6859, + "eval_samples_per_second": 462.174, + "eval_steps_per_second": 29.159, + "eval_weighted_f1": 0.8445924920892679, + "step": 1170 + }, + { + "epoch": 13.056022408963585, + "grad_norm": 161.05384826660156, + "learning_rate": 7.728395061728395e-06, + "loss": 0.199, + "step": 1175 + }, + { + "epoch": 13.336134453781513, + "grad_norm": 158.7040557861328, + "learning_rate": 7.419753086419754e-06, + "loss": 0.1567, + "step": 1200 + }, + { + "epoch": 13.61624649859944, + "grad_norm": 0.009637095965445042, + "learning_rate": 7.111111111111112e-06, + "loss": 0.0224, + "step": 1225 + }, + { + "epoch": 13.896358543417367, + "grad_norm": 4.374759197235107, + "learning_rate": 6.80246913580247e-06, + "loss": 0.1211, + "step": 1250 + }, + { + "epoch": 14.0, + "eval_accuracy": 0.8391167192429022, + "eval_loss": 1.3063347339630127, + "eval_macro_f1": 0.8205361602930565, + "eval_runtime": 0.6592, + "eval_samples_per_second": 480.858, + "eval_steps_per_second": 30.338, + "eval_weighted_f1": 0.8374772581590924, + "step": 1260 + }, + { + "epoch": 14.168067226890756, + "grad_norm": 0.043752290308475494, + "learning_rate": 6.493827160493828e-06, + "loss": 0.1534, + "step": 1275 + }, + { + "epoch": 14.448179271708684, + "grad_norm": 0.2848133146762848, + "learning_rate": 6.1851851851851856e-06, + "loss": 0.1551, + "step": 1300 + }, + { + "epoch": 14.72829131652661, + "grad_norm": 0.20860472321510315, + "learning_rate": 5.876543209876544e-06, + "loss": 0.0696, + "step": 1325 + }, + { + "epoch": 15.0, + "grad_norm": 0.0009489697404205799, + "learning_rate": 5.567901234567901e-06, + "loss": 0.1178, + "step": 1350 + }, + { + "epoch": 15.0, + "eval_accuracy": 0.8485804416403786, + "eval_loss": 1.4265798330307007, + "eval_macro_f1": 0.8314726209463051, + "eval_runtime": 0.6825, + "eval_samples_per_second": 464.443, + "eval_steps_per_second": 29.302, + "eval_weighted_f1": 0.8472253667339172, + "step": 1350 + } + ], + "logging_steps": 25, + "max_steps": 1800, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/special_tokens_map.json b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/tokenizer_config.json b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/trainer_state.json b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..070e42b70a40c0eec1dc157b81318013e3858460 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1620/trainer_state.json @@ -0,0 +1,689 @@ +{ + "best_global_step": 1350, + "best_metric": 0.8314726209463051, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-viclickbait2025-xlmr-source-source-05-s44-20260901T141022Z-6863e5c6/trainer/checkpoint-1350", + "epoch": 18.0, + "eval_steps": 500, + "global_step": 1620, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.2801120448179272, + "grad_norm": 21.286026000976562, + "learning_rate": 2.666666666666667e-06, + "loss": 2.9303, + "step": 25 + }, + { + "epoch": 0.5602240896358543, + "grad_norm": 18.28496742248535, + "learning_rate": 5.444444444444445e-06, + "loss": 2.7658, + "step": 50 + }, + { + "epoch": 0.8403361344537815, + "grad_norm": 23.925424575805664, + "learning_rate": 8.222222222222222e-06, + "loss": 2.427, + "step": 75 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.7665615141955836, + "eval_loss": 0.5131620168685913, + "eval_macro_f1": 0.7311835350201688, + "eval_runtime": 0.7127, + "eval_samples_per_second": 444.81, + "eval_steps_per_second": 28.064, + "eval_weighted_f1": 0.759793552962026, + "step": 90 + }, + { + "epoch": 1.112044817927171, + "grad_norm": 42.43369674682617, + "learning_rate": 1.1000000000000001e-05, + "loss": 2.151, + "step": 100 + }, + { + "epoch": 1.392156862745098, + "grad_norm": 75.90241241455078, + "learning_rate": 1.377777777777778e-05, + "loss": 1.8566, + "step": 125 + }, + { + "epoch": 1.6722689075630253, + "grad_norm": 39.13203048706055, + "learning_rate": 1.6555555555555556e-05, + "loss": 2.0029, + "step": 150 + }, + { + "epoch": 1.9523809523809523, + "grad_norm": 113.73194122314453, + "learning_rate": 1.9333333333333333e-05, + "loss": 1.7688, + "step": 175 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.8107255520504731, + "eval_loss": 0.535222053527832, + "eval_macro_f1": 0.7784456248252726, + "eval_runtime": 0.6496, + "eval_samples_per_second": 488.029, + "eval_steps_per_second": 30.791, + "eval_weighted_f1": 0.803255816824807, + "step": 180 + }, + { + "epoch": 2.224089635854342, + "grad_norm": 58.54833984375, + "learning_rate": 1.9765432098765434e-05, + "loss": 1.5445, + "step": 200 + }, + { + "epoch": 2.504201680672269, + "grad_norm": 121.798095703125, + "learning_rate": 1.9456790123456793e-05, + "loss": 1.6574, + "step": 225 + }, + { + "epoch": 2.784313725490196, + "grad_norm": 34.078330993652344, + "learning_rate": 1.9148148148148148e-05, + "loss": 1.455, + "step": 250 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.8107255520504731, + "eval_loss": 0.5058051943778992, + "eval_macro_f1": 0.803933696190005, + "eval_runtime": 0.6798, + "eval_samples_per_second": 466.328, + "eval_steps_per_second": 29.421, + "eval_weighted_f1": 0.8146395028853194, + "step": 270 + }, + { + "epoch": 3.0560224089635852, + "grad_norm": 53.804298400878906, + "learning_rate": 1.8839506172839507e-05, + "loss": 1.3164, + "step": 275 + }, + { + "epoch": 3.3361344537815127, + "grad_norm": 46.20854949951172, + "learning_rate": 1.8530864197530866e-05, + "loss": 1.3639, + "step": 300 + }, + { + "epoch": 3.61624649859944, + "grad_norm": 45.22013473510742, + "learning_rate": 1.8222222222222224e-05, + "loss": 1.1226, + "step": 325 + }, + { + "epoch": 3.896358543417367, + "grad_norm": 48.710594177246094, + "learning_rate": 1.7913580246913583e-05, + "loss": 1.1293, + "step": 350 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8170347003154574, + "eval_loss": 0.3787948191165924, + "eval_macro_f1": 0.7998040069686412, + "eval_runtime": 0.7137, + "eval_samples_per_second": 444.141, + "eval_steps_per_second": 28.022, + "eval_weighted_f1": 0.8170347003154574, + "step": 360 + }, + { + "epoch": 4.168067226890757, + "grad_norm": 83.72383880615234, + "learning_rate": 1.760493827160494e-05, + "loss": 0.8385, + "step": 375 + }, + { + "epoch": 4.448179271708684, + "grad_norm": 62.41585922241211, + "learning_rate": 1.7296296296296297e-05, + "loss": 0.9109, + "step": 400 + }, + { + "epoch": 4.728291316526611, + "grad_norm": 37.59474563598633, + "learning_rate": 1.6987654320987656e-05, + "loss": 0.9558, + "step": 425 + }, + { + "epoch": 5.0, + "grad_norm": 4.466841697692871, + "learning_rate": 1.6679012345679015e-05, + "loss": 0.8452, + "step": 450 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8138801261829653, + "eval_loss": 0.561353325843811, + "eval_macro_f1": 0.8031532527127867, + "eval_runtime": 0.6452, + "eval_samples_per_second": 491.315, + "eval_steps_per_second": 30.998, + "eval_weighted_f1": 0.8166343234253083, + "step": 450 + }, + { + "epoch": 5.280112044817927, + "grad_norm": 46.44953918457031, + "learning_rate": 1.6370370370370374e-05, + "loss": 0.6156, + "step": 475 + }, + { + "epoch": 5.560224089635854, + "grad_norm": 44.48462677001953, + "learning_rate": 1.606172839506173e-05, + "loss": 0.615, + "step": 500 + }, + { + "epoch": 5.840336134453781, + "grad_norm": 77.2662353515625, + "learning_rate": 1.5753086419753088e-05, + "loss": 0.7461, + "step": 525 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8422712933753943, + "eval_loss": 0.6332255601882935, + "eval_macro_f1": 0.8163979241960894, + "eval_runtime": 0.666, + "eval_samples_per_second": 476.01, + "eval_steps_per_second": 30.032, + "eval_weighted_f1": 0.8366182883446218, + "step": 540 + }, + { + "epoch": 6.1120448179271705, + "grad_norm": 5.867457389831543, + "learning_rate": 1.5444444444444446e-05, + "loss": 0.6319, + "step": 550 + }, + { + "epoch": 6.392156862745098, + "grad_norm": 3.1487138271331787, + "learning_rate": 1.5135802469135803e-05, + "loss": 0.3849, + "step": 575 + }, + { + "epoch": 6.6722689075630255, + "grad_norm": 39.26576232910156, + "learning_rate": 1.4827160493827162e-05, + "loss": 0.5494, + "step": 600 + }, + { + "epoch": 6.9523809523809526, + "grad_norm": 3.9887986183166504, + "learning_rate": 1.4518518518518521e-05, + "loss": 0.6336, + "step": 625 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8391167192429022, + "eval_loss": 0.7031358480453491, + "eval_macro_f1": 0.8197116188819378, + "eval_runtime": 0.6708, + "eval_samples_per_second": 472.594, + "eval_steps_per_second": 29.817, + "eval_weighted_f1": 0.8370642567047232, + "step": 630 + }, + { + "epoch": 7.224089635854342, + "grad_norm": 11.982047080993652, + "learning_rate": 1.4209876543209878e-05, + "loss": 0.3573, + "step": 650 + }, + { + "epoch": 7.504201680672269, + "grad_norm": 40.24326705932617, + "learning_rate": 1.3901234567901237e-05, + "loss": 0.3807, + "step": 675 + }, + { + "epoch": 7.784313725490196, + "grad_norm": 1.3944097757339478, + "learning_rate": 1.3592592592592592e-05, + "loss": 0.4012, + "step": 700 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8170347003154574, + "eval_loss": 1.148803472518921, + "eval_macro_f1": 0.805562605752961, + "eval_runtime": 0.665, + "eval_samples_per_second": 476.714, + "eval_steps_per_second": 30.077, + "eval_weighted_f1": 0.8194185121725995, + "step": 720 + }, + { + "epoch": 8.056022408963585, + "grad_norm": 212.5960235595703, + "learning_rate": 1.3283950617283951e-05, + "loss": 0.4906, + "step": 725 + }, + { + "epoch": 8.336134453781513, + "grad_norm": 318.5392761230469, + "learning_rate": 1.297530864197531e-05, + "loss": 0.3477, + "step": 750 + }, + { + "epoch": 8.61624649859944, + "grad_norm": 46.817447662353516, + "learning_rate": 1.2666666666666667e-05, + "loss": 0.3973, + "step": 775 + }, + { + "epoch": 8.896358543417367, + "grad_norm": 1.5035258531570435, + "learning_rate": 1.2358024691358026e-05, + "loss": 0.2958, + "step": 800 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8454258675078864, + "eval_loss": 0.9726073741912842, + "eval_macro_f1": 0.8290917093029653, + "eval_runtime": 0.6293, + "eval_samples_per_second": 503.774, + "eval_steps_per_second": 31.784, + "eval_weighted_f1": 0.8445924920892679, + "step": 810 + }, + { + "epoch": 9.168067226890756, + "grad_norm": 0.4589764177799225, + "learning_rate": 1.2049382716049384e-05, + "loss": 0.1497, + "step": 825 + }, + { + "epoch": 9.448179271708684, + "grad_norm": 8.860508918762207, + "learning_rate": 1.1740740740740741e-05, + "loss": 0.243, + "step": 850 + }, + { + "epoch": 9.72829131652661, + "grad_norm": 130.12161254882812, + "learning_rate": 1.14320987654321e-05, + "loss": 0.2929, + "step": 875 + }, + { + "epoch": 10.0, + "grad_norm": 0.031857386231422424, + "learning_rate": 1.1123456790123457e-05, + "loss": 0.2966, + "step": 900 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8422712933753943, + "eval_loss": 1.1647114753723145, + "eval_macro_f1": 0.8259771629336847, + "eval_runtime": 0.6312, + "eval_samples_per_second": 502.21, + "eval_steps_per_second": 31.685, + "eval_weighted_f1": 0.8415993704705815, + "step": 900 + }, + { + "epoch": 10.280112044817928, + "grad_norm": 1010.2288208007812, + "learning_rate": 1.0814814814814816e-05, + "loss": 0.1748, + "step": 925 + }, + { + "epoch": 10.560224089635854, + "grad_norm": 0.4934237003326416, + "learning_rate": 1.0506172839506175e-05, + "loss": 0.2865, + "step": 950 + }, + { + "epoch": 10.840336134453782, + "grad_norm": 6.401635646820068, + "learning_rate": 1.0197530864197532e-05, + "loss": 0.2486, + "step": 975 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.8359621451104101, + "eval_loss": 1.4146658182144165, + "eval_macro_f1": 0.8219255034137067, + "eval_runtime": 0.6608, + "eval_samples_per_second": 479.755, + "eval_steps_per_second": 30.268, + "eval_weighted_f1": 0.8365930054113856, + "step": 990 + }, + { + "epoch": 11.11204481792717, + "grad_norm": 438.1229248046875, + "learning_rate": 9.88888888888889e-06, + "loss": 0.1684, + "step": 1000 + }, + { + "epoch": 11.392156862745098, + "grad_norm": 416.82550048828125, + "learning_rate": 9.580246913580248e-06, + "loss": 0.1978, + "step": 1025 + }, + { + "epoch": 11.672268907563025, + "grad_norm": 204.48353576660156, + "learning_rate": 9.271604938271606e-06, + "loss": 0.2504, + "step": 1050 + }, + { + "epoch": 11.952380952380953, + "grad_norm": 3.435577869415283, + "learning_rate": 8.962962962962963e-06, + "loss": 0.2117, + "step": 1075 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8485804416403786, + "eval_loss": 1.2211965322494507, + "eval_macro_f1": 0.8299204149154968, + "eval_runtime": 0.6933, + "eval_samples_per_second": 457.219, + "eval_steps_per_second": 28.847, + "eval_weighted_f1": 0.8464478671575351, + "step": 1080 + }, + { + "epoch": 12.224089635854341, + "grad_norm": 0.07730665057897568, + "learning_rate": 8.654320987654322e-06, + "loss": 0.1066, + "step": 1100 + }, + { + "epoch": 12.504201680672269, + "grad_norm": 0.006811291445046663, + "learning_rate": 8.34567901234568e-06, + "loss": 0.1689, + "step": 1125 + }, + { + "epoch": 12.784313725490197, + "grad_norm": 26.207834243774414, + "learning_rate": 8.037037037037038e-06, + "loss": 0.102, + "step": 1150 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.8454258675078864, + "eval_loss": 1.3052610158920288, + "eval_macro_f1": 0.8290917093029653, + "eval_runtime": 0.6859, + "eval_samples_per_second": 462.174, + "eval_steps_per_second": 29.159, + "eval_weighted_f1": 0.8445924920892679, + "step": 1170 + }, + { + "epoch": 13.056022408963585, + "grad_norm": 161.05384826660156, + "learning_rate": 7.728395061728395e-06, + "loss": 0.199, + "step": 1175 + }, + { + "epoch": 13.336134453781513, + "grad_norm": 158.7040557861328, + "learning_rate": 7.419753086419754e-06, + "loss": 0.1567, + "step": 1200 + }, + { + "epoch": 13.61624649859944, + "grad_norm": 0.009637095965445042, + "learning_rate": 7.111111111111112e-06, + "loss": 0.0224, + "step": 1225 + }, + { + "epoch": 13.896358543417367, + "grad_norm": 4.374759197235107, + "learning_rate": 6.80246913580247e-06, + "loss": 0.1211, + "step": 1250 + }, + { + "epoch": 14.0, + "eval_accuracy": 0.8391167192429022, + "eval_loss": 1.3063347339630127, + "eval_macro_f1": 0.8205361602930565, + "eval_runtime": 0.6592, + "eval_samples_per_second": 480.858, + "eval_steps_per_second": 30.338, + "eval_weighted_f1": 0.8374772581590924, + "step": 1260 + }, + { + "epoch": 14.168067226890756, + "grad_norm": 0.043752290308475494, + "learning_rate": 6.493827160493828e-06, + "loss": 0.1534, + "step": 1275 + }, + { + "epoch": 14.448179271708684, + "grad_norm": 0.2848133146762848, + "learning_rate": 6.1851851851851856e-06, + "loss": 0.1551, + "step": 1300 + }, + { + "epoch": 14.72829131652661, + "grad_norm": 0.20860472321510315, + "learning_rate": 5.876543209876544e-06, + "loss": 0.0696, + "step": 1325 + }, + { + "epoch": 15.0, + "grad_norm": 0.0009489697404205799, + "learning_rate": 5.567901234567901e-06, + "loss": 0.1178, + "step": 1350 + }, + { + "epoch": 15.0, + "eval_accuracy": 0.8485804416403786, + "eval_loss": 1.4265798330307007, + "eval_macro_f1": 0.8314726209463051, + "eval_runtime": 0.6825, + "eval_samples_per_second": 464.443, + "eval_steps_per_second": 29.302, + "eval_weighted_f1": 0.8472253667339172, + "step": 1350 + }, + { + "epoch": 15.280112044817928, + "grad_norm": 0.014971903525292873, + "learning_rate": 5.259259259259259e-06, + "loss": 0.105, + "step": 1375 + }, + { + "epoch": 15.560224089635854, + "grad_norm": 0.008831561543047428, + "learning_rate": 4.950617283950618e-06, + "loss": 0.0525, + "step": 1400 + }, + { + "epoch": 15.840336134453782, + "grad_norm": 0.004474582616239786, + "learning_rate": 4.641975308641975e-06, + "loss": 0.1044, + "step": 1425 + }, + { + "epoch": 16.0, + "eval_accuracy": 0.8485804416403786, + "eval_loss": 1.442765235900879, + "eval_macro_f1": 0.8314726209463051, + "eval_runtime": 0.781, + "eval_samples_per_second": 405.865, + "eval_steps_per_second": 25.607, + "eval_weighted_f1": 0.8472253667339172, + "step": 1440 + }, + { + "epoch": 16.11204481792717, + "grad_norm": 0.01411131490021944, + "learning_rate": 4.333333333333334e-06, + "loss": 0.0241, + "step": 1450 + }, + { + "epoch": 16.392156862745097, + "grad_norm": 0.029822485521435738, + "learning_rate": 4.024691358024692e-06, + "loss": 0.2201, + "step": 1475 + }, + { + "epoch": 16.672268907563026, + "grad_norm": 0.8752092123031616, + "learning_rate": 3.7160493827160497e-06, + "loss": 0.1631, + "step": 1500 + }, + { + "epoch": 16.952380952380953, + "grad_norm": 0.03659430518746376, + "learning_rate": 3.4074074074074077e-06, + "loss": 0.0123, + "step": 1525 + }, + { + "epoch": 17.0, + "eval_accuracy": 0.8454258675078864, + "eval_loss": 1.4357441663742065, + "eval_macro_f1": 0.8283437765916299, + "eval_runtime": 0.7362, + "eval_samples_per_second": 430.61, + "eval_steps_per_second": 27.168, + "eval_weighted_f1": 0.8442301211437485, + "step": 1530 + }, + { + "epoch": 17.22408963585434, + "grad_norm": 0.032733943313360214, + "learning_rate": 3.098765432098766e-06, + "loss": 0.0249, + "step": 1550 + }, + { + "epoch": 17.504201680672267, + "grad_norm": 0.008555321954190731, + "learning_rate": 2.7901234567901235e-06, + "loss": 0.0344, + "step": 1575 + }, + { + "epoch": 17.784313725490197, + "grad_norm": 0.003118259133771062, + "learning_rate": 2.481481481481482e-06, + "loss": 0.0059, + "step": 1600 + }, + { + "epoch": 18.0, + "eval_accuracy": 0.8454258675078864, + "eval_loss": 1.4717018604278564, + "eval_macro_f1": 0.827573957928623, + "eval_runtime": 0.6972, + "eval_samples_per_second": 454.657, + "eval_steps_per_second": 28.685, + "eval_weighted_f1": 0.8438506990155984, + "step": 1620 + } + ], + "logging_steps": 25, + "max_steps": 1800, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/special_tokens_map.json b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/tokenizer_config.json b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/trainer_state.json b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..16881fa661bb70a013cfbd32bd4ba0c90ced4d8c --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516/trainer_state.json @@ -0,0 +1,249 @@ +{ + "best_global_step": 516, + "best_metric": 0.8428641025978643, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516", + "epoch": 6.0, + "eval_steps": 500, + "global_step": 516, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.29069767441860467, + "grad_norm": 23.544675827026367, + "learning_rate": 2.790697674418605e-06, + "loss": 2.9263, + "step": 25 + }, + { + "epoch": 0.5813953488372093, + "grad_norm": 23.54766845703125, + "learning_rate": 5.697674418604652e-06, + "loss": 2.7261, + "step": 50 + }, + { + "epoch": 0.872093023255814, + "grad_norm": 61.286956787109375, + "learning_rate": 8.604651162790698e-06, + "loss": 2.4571, + "step": 75 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.7254901960784313, + "eval_loss": 0.48061713576316833, + "eval_macro_f1": 0.42045454545454547, + "eval_runtime": 0.7673, + "eval_samples_per_second": 398.802, + "eval_steps_per_second": 26.065, + "eval_weighted_f1": 0.6100713012477719, + "step": 86 + }, + { + "epoch": 1.1627906976744187, + "grad_norm": 26.423561096191406, + "learning_rate": 1.1511627906976746e-05, + "loss": 2.3328, + "step": 100 + }, + { + "epoch": 1.4534883720930232, + "grad_norm": 60.45857620239258, + "learning_rate": 1.441860465116279e-05, + "loss": 1.9427, + "step": 125 + }, + { + "epoch": 1.744186046511628, + "grad_norm": 74.30782318115234, + "learning_rate": 1.7325581395348837e-05, + "loss": 1.9428, + "step": 150 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.8562091503267973, + "eval_loss": 0.33015862107276917, + "eval_macro_f1": 0.8039375582479031, + "eval_runtime": 0.663, + "eval_samples_per_second": 461.567, + "eval_steps_per_second": 30.168, + "eval_weighted_f1": 0.8495924931016209, + "step": 172 + }, + { + "epoch": 2.0348837209302326, + "grad_norm": 66.0932846069336, + "learning_rate": 1.9974160206718347e-05, + "loss": 1.6777, + "step": 175 + }, + { + "epoch": 2.3255813953488373, + "grad_norm": 61.038414001464844, + "learning_rate": 1.9651162790697676e-05, + "loss": 1.667, + "step": 200 + }, + { + "epoch": 2.616279069767442, + "grad_norm": 84.13005828857422, + "learning_rate": 1.9328165374677004e-05, + "loss": 1.4979, + "step": 225 + }, + { + "epoch": 2.9069767441860463, + "grad_norm": 40.771522521972656, + "learning_rate": 1.9005167958656333e-05, + "loss": 1.4829, + "step": 250 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7745098039215687, + "eval_loss": 0.4348524510860443, + "eval_macro_f1": 0.7546794938826729, + "eval_runtime": 0.6897, + "eval_samples_per_second": 443.668, + "eval_steps_per_second": 28.998, + "eval_weighted_f1": 0.7861344684271283, + "step": 258 + }, + { + "epoch": 3.197674418604651, + "grad_norm": 63.65130615234375, + "learning_rate": 1.868217054263566e-05, + "loss": 1.2801, + "step": 275 + }, + { + "epoch": 3.488372093023256, + "grad_norm": 68.13922119140625, + "learning_rate": 1.8359173126614988e-05, + "loss": 1.2032, + "step": 300 + }, + { + "epoch": 3.7790697674418605, + "grad_norm": 73.62340545654297, + "learning_rate": 1.8036175710594317e-05, + "loss": 1.3543, + "step": 325 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8006535947712419, + "eval_loss": 0.4494991898536682, + "eval_macro_f1": 0.7740494607255692, + "eval_runtime": 0.6841, + "eval_samples_per_second": 447.328, + "eval_steps_per_second": 29.237, + "eval_weighted_f1": 0.8090148940427389, + "step": 344 + }, + { + "epoch": 4.069767441860465, + "grad_norm": 28.662683486938477, + "learning_rate": 1.7713178294573646e-05, + "loss": 1.135, + "step": 350 + }, + { + "epoch": 4.3604651162790695, + "grad_norm": 68.47048950195312, + "learning_rate": 1.7390180878552975e-05, + "loss": 0.8531, + "step": 375 + }, + { + "epoch": 4.651162790697675, + "grad_norm": 73.94048309326172, + "learning_rate": 1.70671834625323e-05, + "loss": 0.9591, + "step": 400 + }, + { + "epoch": 4.941860465116279, + "grad_norm": 81.35037994384766, + "learning_rate": 1.674418604651163e-05, + "loss": 0.9536, + "step": 425 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8300653594771242, + "eval_loss": 0.46315619349479675, + "eval_macro_f1": 0.8068932038834951, + "eval_runtime": 0.684, + "eval_samples_per_second": 447.359, + "eval_steps_per_second": 29.239, + "eval_weighted_f1": 0.8370607272035028, + "step": 430 + }, + { + "epoch": 5.232558139534884, + "grad_norm": 36.92371368408203, + "learning_rate": 1.6421188630490958e-05, + "loss": 0.719, + "step": 450 + }, + { + "epoch": 5.523255813953488, + "grad_norm": 89.27535247802734, + "learning_rate": 1.6098191214470284e-05, + "loss": 0.7397, + "step": 475 + }, + { + "epoch": 5.813953488372093, + "grad_norm": 102.30037689208984, + "learning_rate": 1.5775193798449613e-05, + "loss": 0.6656, + "step": 500 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8725490196078431, + "eval_loss": 0.4044148027896881, + "eval_macro_f1": 0.8428641025978643, + "eval_runtime": 0.6602, + "eval_samples_per_second": 463.488, + "eval_steps_per_second": 30.293, + "eval_weighted_f1": 0.8736649939315265, + "step": 516 + } + ], + "logging_steps": 25, + "max_steps": 1720, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/special_tokens_map.json b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/tokenizer_config.json b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/trainer_state.json b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..6fb35f4f8fc86b887b2888e01412f33748a0ca46 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-774/trainer_state.json @@ -0,0 +1,352 @@ +{ + "best_global_step": 516, + "best_metric": 0.8428641025978643, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-viclickbait2025-xlmr-source-source-06-s44-20260901T142033Z-6863e5c6/trainer/checkpoint-516", + "epoch": 9.0, + "eval_steps": 500, + "global_step": 774, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.29069767441860467, + "grad_norm": 23.544675827026367, + "learning_rate": 2.790697674418605e-06, + "loss": 2.9263, + "step": 25 + }, + { + "epoch": 0.5813953488372093, + "grad_norm": 23.54766845703125, + "learning_rate": 5.697674418604652e-06, + "loss": 2.7261, + "step": 50 + }, + { + "epoch": 0.872093023255814, + "grad_norm": 61.286956787109375, + "learning_rate": 8.604651162790698e-06, + "loss": 2.4571, + "step": 75 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.7254901960784313, + "eval_loss": 0.48061713576316833, + "eval_macro_f1": 0.42045454545454547, + "eval_runtime": 0.7673, + "eval_samples_per_second": 398.802, + "eval_steps_per_second": 26.065, + "eval_weighted_f1": 0.6100713012477719, + "step": 86 + }, + { + "epoch": 1.1627906976744187, + "grad_norm": 26.423561096191406, + "learning_rate": 1.1511627906976746e-05, + "loss": 2.3328, + "step": 100 + }, + { + "epoch": 1.4534883720930232, + "grad_norm": 60.45857620239258, + "learning_rate": 1.441860465116279e-05, + "loss": 1.9427, + "step": 125 + }, + { + "epoch": 1.744186046511628, + "grad_norm": 74.30782318115234, + "learning_rate": 1.7325581395348837e-05, + "loss": 1.9428, + "step": 150 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.8562091503267973, + "eval_loss": 0.33015862107276917, + "eval_macro_f1": 0.8039375582479031, + "eval_runtime": 0.663, + "eval_samples_per_second": 461.567, + "eval_steps_per_second": 30.168, + "eval_weighted_f1": 0.8495924931016209, + "step": 172 + }, + { + "epoch": 2.0348837209302326, + "grad_norm": 66.0932846069336, + "learning_rate": 1.9974160206718347e-05, + "loss": 1.6777, + "step": 175 + }, + { + "epoch": 2.3255813953488373, + "grad_norm": 61.038414001464844, + "learning_rate": 1.9651162790697676e-05, + "loss": 1.667, + "step": 200 + }, + { + "epoch": 2.616279069767442, + "grad_norm": 84.13005828857422, + "learning_rate": 1.9328165374677004e-05, + "loss": 1.4979, + "step": 225 + }, + { + "epoch": 2.9069767441860463, + "grad_norm": 40.771522521972656, + "learning_rate": 1.9005167958656333e-05, + "loss": 1.4829, + "step": 250 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7745098039215687, + "eval_loss": 0.4348524510860443, + "eval_macro_f1": 0.7546794938826729, + "eval_runtime": 0.6897, + "eval_samples_per_second": 443.668, + "eval_steps_per_second": 28.998, + "eval_weighted_f1": 0.7861344684271283, + "step": 258 + }, + { + "epoch": 3.197674418604651, + "grad_norm": 63.65130615234375, + "learning_rate": 1.868217054263566e-05, + "loss": 1.2801, + "step": 275 + }, + { + "epoch": 3.488372093023256, + "grad_norm": 68.13922119140625, + "learning_rate": 1.8359173126614988e-05, + "loss": 1.2032, + "step": 300 + }, + { + "epoch": 3.7790697674418605, + "grad_norm": 73.62340545654297, + "learning_rate": 1.8036175710594317e-05, + "loss": 1.3543, + "step": 325 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8006535947712419, + "eval_loss": 0.4494991898536682, + "eval_macro_f1": 0.7740494607255692, + "eval_runtime": 0.6841, + "eval_samples_per_second": 447.328, + "eval_steps_per_second": 29.237, + "eval_weighted_f1": 0.8090148940427389, + "step": 344 + }, + { + "epoch": 4.069767441860465, + "grad_norm": 28.662683486938477, + "learning_rate": 1.7713178294573646e-05, + "loss": 1.135, + "step": 350 + }, + { + "epoch": 4.3604651162790695, + "grad_norm": 68.47048950195312, + "learning_rate": 1.7390180878552975e-05, + "loss": 0.8531, + "step": 375 + }, + { + "epoch": 4.651162790697675, + "grad_norm": 73.94048309326172, + "learning_rate": 1.70671834625323e-05, + "loss": 0.9591, + "step": 400 + }, + { + "epoch": 4.941860465116279, + "grad_norm": 81.35037994384766, + "learning_rate": 1.674418604651163e-05, + "loss": 0.9536, + "step": 425 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8300653594771242, + "eval_loss": 0.46315619349479675, + "eval_macro_f1": 0.8068932038834951, + "eval_runtime": 0.684, + "eval_samples_per_second": 447.359, + "eval_steps_per_second": 29.239, + "eval_weighted_f1": 0.8370607272035028, + "step": 430 + }, + { + "epoch": 5.232558139534884, + "grad_norm": 36.92371368408203, + "learning_rate": 1.6421188630490958e-05, + "loss": 0.719, + "step": 450 + }, + { + "epoch": 5.523255813953488, + "grad_norm": 89.27535247802734, + "learning_rate": 1.6098191214470284e-05, + "loss": 0.7397, + "step": 475 + }, + { + "epoch": 5.813953488372093, + "grad_norm": 102.30037689208984, + "learning_rate": 1.5775193798449613e-05, + "loss": 0.6656, + "step": 500 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8725490196078431, + "eval_loss": 0.4044148027896881, + "eval_macro_f1": 0.8428641025978643, + "eval_runtime": 0.6602, + "eval_samples_per_second": 463.488, + "eval_steps_per_second": 30.293, + "eval_weighted_f1": 0.8736649939315265, + "step": 516 + }, + { + "epoch": 6.104651162790698, + "grad_norm": 104.36873626708984, + "learning_rate": 1.545219638242894e-05, + "loss": 0.6109, + "step": 525 + }, + { + "epoch": 6.395348837209302, + "grad_norm": 58.92158889770508, + "learning_rate": 1.5129198966408269e-05, + "loss": 0.3602, + "step": 550 + }, + { + "epoch": 6.686046511627907, + "grad_norm": 34.21396255493164, + "learning_rate": 1.4806201550387598e-05, + "loss": 0.4644, + "step": 575 + }, + { + "epoch": 6.976744186046512, + "grad_norm": 64.25516510009766, + "learning_rate": 1.4483204134366927e-05, + "loss": 0.6714, + "step": 600 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8594771241830066, + "eval_loss": 0.46995070576667786, + "eval_macro_f1": 0.8372883871047523, + "eval_runtime": 0.652, + "eval_samples_per_second": 469.33, + "eval_steps_per_second": 30.675, + "eval_weighted_f1": 0.8643861368109389, + "step": 602 + }, + { + "epoch": 7.267441860465116, + "grad_norm": 77.10468292236328, + "learning_rate": 1.4160206718346254e-05, + "loss": 0.2013, + "step": 625 + }, + { + "epoch": 7.558139534883721, + "grad_norm": 73.71111297607422, + "learning_rate": 1.3837209302325583e-05, + "loss": 0.4848, + "step": 650 + }, + { + "epoch": 7.848837209302325, + "grad_norm": 16.351274490356445, + "learning_rate": 1.3514211886304912e-05, + "loss": 0.367, + "step": 675 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8627450980392157, + "eval_loss": 0.6721513867378235, + "eval_macro_f1": 0.837800999545661, + "eval_runtime": 0.7237, + "eval_samples_per_second": 422.832, + "eval_steps_per_second": 27.636, + "eval_weighted_f1": 0.8664867128132488, + "step": 688 + }, + { + "epoch": 8.13953488372093, + "grad_norm": 162.64071655273438, + "learning_rate": 1.3191214470284239e-05, + "loss": 0.2724, + "step": 700 + }, + { + "epoch": 8.430232558139535, + "grad_norm": 18.183868408203125, + "learning_rate": 1.2868217054263568e-05, + "loss": 0.1065, + "step": 725 + }, + { + "epoch": 8.720930232558139, + "grad_norm": 96.52957916259766, + "learning_rate": 1.2545219638242895e-05, + "loss": 0.6261, + "step": 750 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8398692810457516, + "eval_loss": 0.9877539873123169, + "eval_macro_f1": 0.8184987471402114, + "eval_runtime": 0.7388, + "eval_samples_per_second": 414.212, + "eval_steps_per_second": 27.073, + "eval_weighted_f1": 0.8465857345589215, + "step": 774 + } + ], + "logging_steps": 25, + "max_steps": 1720, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/special_tokens_map.json b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/tokenizer_config.json b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/trainer_state.json b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..6525520829e5bb4946b2f2af1ed7b70adaccafbe --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711/trainer_state.json @@ -0,0 +1,338 @@ +{ + "best_global_step": 711, + "best_metric": 0.8489247311827957, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711", + "epoch": 9.0, + "eval_steps": 500, + "global_step": 711, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.31645569620253167, + "grad_norm": 10.859970092773438, + "learning_rate": 3.037974683544304e-06, + "loss": 2.894, + "step": 25 + }, + { + "epoch": 0.6329113924050633, + "grad_norm": 30.539888381958008, + "learning_rate": 6.20253164556962e-06, + "loss": 2.7578, + "step": 50 + }, + { + "epoch": 0.9493670886075949, + "grad_norm": 17.019699096679688, + "learning_rate": 9.367088607594937e-06, + "loss": 2.5701, + "step": 75 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.6690391459074733, + "eval_loss": 0.5780642032623291, + "eval_macro_f1": 0.40085287846481876, + "eval_runtime": 0.6537, + "eval_samples_per_second": 429.883, + "eval_steps_per_second": 27.537, + "eval_weighted_f1": 0.5363725348853091, + "step": 79 + }, + { + "epoch": 1.2658227848101267, + "grad_norm": 89.87989807128906, + "learning_rate": 1.2531645569620255e-05, + "loss": 2.2146, + "step": 100 + }, + { + "epoch": 1.5822784810126582, + "grad_norm": 57.168434143066406, + "learning_rate": 1.5696202531645572e-05, + "loss": 2.1326, + "step": 125 + }, + { + "epoch": 1.8987341772151898, + "grad_norm": 90.69680786132812, + "learning_rate": 1.886075949367089e-05, + "loss": 1.9353, + "step": 150 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.800711743772242, + "eval_loss": 0.46159303188323975, + "eval_macro_f1": 0.7854259845096542, + "eval_runtime": 0.6264, + "eval_samples_per_second": 448.565, + "eval_steps_per_second": 28.734, + "eval_weighted_f1": 0.8047879462422654, + "step": 158 + }, + { + "epoch": 2.2151898734177213, + "grad_norm": 27.01743507385254, + "learning_rate": 1.977496483825598e-05, + "loss": 1.7965, + "step": 175 + }, + { + "epoch": 2.5316455696202533, + "grad_norm": 53.67705535888672, + "learning_rate": 1.9423347398030942e-05, + "loss": 1.4839, + "step": 200 + }, + { + "epoch": 2.848101265822785, + "grad_norm": 53.62406921386719, + "learning_rate": 1.907172995780591e-05, + "loss": 1.6328, + "step": 225 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.8434163701067615, + "eval_loss": 0.3271748423576355, + "eval_macro_f1": 0.8212261422787739, + "eval_runtime": 0.6805, + "eval_samples_per_second": 412.955, + "eval_steps_per_second": 26.453, + "eval_weighted_f1": 0.8425197952450246, + "step": 237 + }, + { + "epoch": 3.1645569620253164, + "grad_norm": 25.777297973632812, + "learning_rate": 1.872011251758087e-05, + "loss": 1.3556, + "step": 250 + }, + { + "epoch": 3.481012658227848, + "grad_norm": 45.83961868286133, + "learning_rate": 1.8368495077355838e-05, + "loss": 1.2223, + "step": 275 + }, + { + "epoch": 3.7974683544303796, + "grad_norm": 45.81952667236328, + "learning_rate": 1.8016877637130804e-05, + "loss": 1.3262, + "step": 300 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8220640569395018, + "eval_loss": 0.40762826800346375, + "eval_macro_f1": 0.8113384896336878, + "eval_runtime": 0.6108, + "eval_samples_per_second": 460.026, + "eval_steps_per_second": 29.468, + "eval_weighted_f1": 0.8265463835747672, + "step": 316 + }, + { + "epoch": 4.113924050632911, + "grad_norm": 27.679380416870117, + "learning_rate": 1.7665260196905767e-05, + "loss": 1.1138, + "step": 325 + }, + { + "epoch": 4.430379746835443, + "grad_norm": 28.408794403076172, + "learning_rate": 1.7313642756680733e-05, + "loss": 1.0092, + "step": 350 + }, + { + "epoch": 4.746835443037975, + "grad_norm": 65.93917846679688, + "learning_rate": 1.6962025316455696e-05, + "loss": 1.0344, + "step": 375 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8576512455516014, + "eval_loss": 0.407770574092865, + "eval_macro_f1": 0.8433143749303, + "eval_runtime": 0.6449, + "eval_samples_per_second": 435.704, + "eval_steps_per_second": 27.91, + "eval_weighted_f1": 0.859337936212931, + "step": 395 + }, + { + "epoch": 5.063291139240507, + "grad_norm": 81.56014251708984, + "learning_rate": 1.6610407876230663e-05, + "loss": 0.6673, + "step": 400 + }, + { + "epoch": 5.379746835443038, + "grad_norm": 225.88876342773438, + "learning_rate": 1.6258790436005626e-05, + "loss": 0.5635, + "step": 425 + }, + { + "epoch": 5.69620253164557, + "grad_norm": 28.108448028564453, + "learning_rate": 1.5907172995780592e-05, + "loss": 0.7501, + "step": 450 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8612099644128114, + "eval_loss": 0.47222700715065, + "eval_macro_f1": 0.8445465764500617, + "eval_runtime": 0.6649, + "eval_samples_per_second": 422.631, + "eval_steps_per_second": 27.072, + "eval_weighted_f1": 0.8617533357594228, + "step": 474 + }, + { + "epoch": 6.012658227848101, + "grad_norm": 30.494224548339844, + "learning_rate": 1.555555555555556e-05, + "loss": 0.7262, + "step": 475 + }, + { + "epoch": 6.329113924050633, + "grad_norm": 68.9117202758789, + "learning_rate": 1.5203938115330521e-05, + "loss": 0.4915, + "step": 500 + }, + { + "epoch": 6.6455696202531644, + "grad_norm": 126.9439468383789, + "learning_rate": 1.4852320675105488e-05, + "loss": 0.4423, + "step": 525 + }, + { + "epoch": 6.962025316455696, + "grad_norm": 60.305728912353516, + "learning_rate": 1.450070323488045e-05, + "loss": 0.4613, + "step": 550 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8612099644128114, + "eval_loss": 0.5797529220581055, + "eval_macro_f1": 0.8437219251336898, + "eval_runtime": 0.6399, + "eval_samples_per_second": 439.105, + "eval_steps_per_second": 28.128, + "eval_weighted_f1": 0.8613960073838659, + "step": 553 + }, + { + "epoch": 7.2784810126582276, + "grad_norm": 220.0026397705078, + "learning_rate": 1.4149085794655415e-05, + "loss": 0.2764, + "step": 575 + }, + { + "epoch": 7.594936708860759, + "grad_norm": 68.6791763305664, + "learning_rate": 1.3797468354430381e-05, + "loss": 0.5405, + "step": 600 + }, + { + "epoch": 7.911392405063291, + "grad_norm": 381.5016174316406, + "learning_rate": 1.3445850914205344e-05, + "loss": 0.3258, + "step": 625 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8469750889679716, + "eval_loss": 0.8536282777786255, + "eval_macro_f1": 0.8303354536136035, + "eval_runtime": 0.6626, + "eval_samples_per_second": 424.085, + "eval_steps_per_second": 27.166, + "eval_weighted_f1": 0.8482986963257053, + "step": 632 + }, + { + "epoch": 8.227848101265822, + "grad_norm": 131.03399658203125, + "learning_rate": 1.309423347398031e-05, + "loss": 0.2809, + "step": 650 + }, + { + "epoch": 8.544303797468354, + "grad_norm": 273.69573974609375, + "learning_rate": 1.2742616033755275e-05, + "loss": 0.171, + "step": 675 + }, + { + "epoch": 8.860759493670885, + "grad_norm": 287.41156005859375, + "learning_rate": 1.2390998593530238e-05, + "loss": 0.3972, + "step": 700 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8647686832740213, + "eval_loss": 0.957329273223877, + "eval_macro_f1": 0.8489247311827957, + "eval_runtime": 0.6499, + "eval_samples_per_second": 432.37, + "eval_steps_per_second": 27.696, + "eval_weighted_f1": 0.8654651207285807, + "step": 711 + } + ], + "logging_steps": 25, + "max_steps": 1580, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/special_tokens_map.json b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/tokenizer_config.json b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/trainer_state.json b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..1a968c8d2c0e86a141ef00631f687cc4cc59dca3 --- /dev/null +++ b/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-948/trainer_state.json @@ -0,0 +1,434 @@ +{ + "best_global_step": 711, + "best_metric": 0.8489247311827957, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-viclickbait2025-xlmr-source-source-07-s44-20260901T142547Z-6863e5c6/trainer/checkpoint-711", + "epoch": 12.0, + "eval_steps": 500, + "global_step": 948, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.31645569620253167, + "grad_norm": 10.859970092773438, + "learning_rate": 3.037974683544304e-06, + "loss": 2.894, + "step": 25 + }, + { + "epoch": 0.6329113924050633, + "grad_norm": 30.539888381958008, + "learning_rate": 6.20253164556962e-06, + "loss": 2.7578, + "step": 50 + }, + { + "epoch": 0.9493670886075949, + "grad_norm": 17.019699096679688, + "learning_rate": 9.367088607594937e-06, + "loss": 2.5701, + "step": 75 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.6690391459074733, + "eval_loss": 0.5780642032623291, + "eval_macro_f1": 0.40085287846481876, + "eval_runtime": 0.6537, + "eval_samples_per_second": 429.883, + "eval_steps_per_second": 27.537, + "eval_weighted_f1": 0.5363725348853091, + "step": 79 + }, + { + "epoch": 1.2658227848101267, + "grad_norm": 89.87989807128906, + "learning_rate": 1.2531645569620255e-05, + "loss": 2.2146, + "step": 100 + }, + { + "epoch": 1.5822784810126582, + "grad_norm": 57.168434143066406, + "learning_rate": 1.5696202531645572e-05, + "loss": 2.1326, + "step": 125 + }, + { + "epoch": 1.8987341772151898, + "grad_norm": 90.69680786132812, + "learning_rate": 1.886075949367089e-05, + "loss": 1.9353, + "step": 150 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.800711743772242, + "eval_loss": 0.46159303188323975, + "eval_macro_f1": 0.7854259845096542, + "eval_runtime": 0.6264, + "eval_samples_per_second": 448.565, + "eval_steps_per_second": 28.734, + "eval_weighted_f1": 0.8047879462422654, + "step": 158 + }, + { + "epoch": 2.2151898734177213, + "grad_norm": 27.01743507385254, + "learning_rate": 1.977496483825598e-05, + "loss": 1.7965, + "step": 175 + }, + { + "epoch": 2.5316455696202533, + "grad_norm": 53.67705535888672, + "learning_rate": 1.9423347398030942e-05, + "loss": 1.4839, + "step": 200 + }, + { + "epoch": 2.848101265822785, + "grad_norm": 53.62406921386719, + "learning_rate": 1.907172995780591e-05, + "loss": 1.6328, + "step": 225 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.8434163701067615, + "eval_loss": 0.3271748423576355, + "eval_macro_f1": 0.8212261422787739, + "eval_runtime": 0.6805, + "eval_samples_per_second": 412.955, + "eval_steps_per_second": 26.453, + "eval_weighted_f1": 0.8425197952450246, + "step": 237 + }, + { + "epoch": 3.1645569620253164, + "grad_norm": 25.777297973632812, + "learning_rate": 1.872011251758087e-05, + "loss": 1.3556, + "step": 250 + }, + { + "epoch": 3.481012658227848, + "grad_norm": 45.83961868286133, + "learning_rate": 1.8368495077355838e-05, + "loss": 1.2223, + "step": 275 + }, + { + "epoch": 3.7974683544303796, + "grad_norm": 45.81952667236328, + "learning_rate": 1.8016877637130804e-05, + "loss": 1.3262, + "step": 300 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8220640569395018, + "eval_loss": 0.40762826800346375, + "eval_macro_f1": 0.8113384896336878, + "eval_runtime": 0.6108, + "eval_samples_per_second": 460.026, + "eval_steps_per_second": 29.468, + "eval_weighted_f1": 0.8265463835747672, + "step": 316 + }, + { + "epoch": 4.113924050632911, + "grad_norm": 27.679380416870117, + "learning_rate": 1.7665260196905767e-05, + "loss": 1.1138, + "step": 325 + }, + { + "epoch": 4.430379746835443, + "grad_norm": 28.408794403076172, + "learning_rate": 1.7313642756680733e-05, + "loss": 1.0092, + "step": 350 + }, + { + "epoch": 4.746835443037975, + "grad_norm": 65.93917846679688, + "learning_rate": 1.6962025316455696e-05, + "loss": 1.0344, + "step": 375 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8576512455516014, + "eval_loss": 0.407770574092865, + "eval_macro_f1": 0.8433143749303, + "eval_runtime": 0.6449, + "eval_samples_per_second": 435.704, + "eval_steps_per_second": 27.91, + "eval_weighted_f1": 0.859337936212931, + "step": 395 + }, + { + "epoch": 5.063291139240507, + "grad_norm": 81.56014251708984, + "learning_rate": 1.6610407876230663e-05, + "loss": 0.6673, + "step": 400 + }, + { + "epoch": 5.379746835443038, + "grad_norm": 225.88876342773438, + "learning_rate": 1.6258790436005626e-05, + "loss": 0.5635, + "step": 425 + }, + { + "epoch": 5.69620253164557, + "grad_norm": 28.108448028564453, + "learning_rate": 1.5907172995780592e-05, + "loss": 0.7501, + "step": 450 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8612099644128114, + "eval_loss": 0.47222700715065, + "eval_macro_f1": 0.8445465764500617, + "eval_runtime": 0.6649, + "eval_samples_per_second": 422.631, + "eval_steps_per_second": 27.072, + "eval_weighted_f1": 0.8617533357594228, + "step": 474 + }, + { + "epoch": 6.012658227848101, + "grad_norm": 30.494224548339844, + "learning_rate": 1.555555555555556e-05, + "loss": 0.7262, + "step": 475 + }, + { + "epoch": 6.329113924050633, + "grad_norm": 68.9117202758789, + "learning_rate": 1.5203938115330521e-05, + "loss": 0.4915, + "step": 500 + }, + { + "epoch": 6.6455696202531644, + "grad_norm": 126.9439468383789, + "learning_rate": 1.4852320675105488e-05, + "loss": 0.4423, + "step": 525 + }, + { + "epoch": 6.962025316455696, + "grad_norm": 60.305728912353516, + "learning_rate": 1.450070323488045e-05, + "loss": 0.4613, + "step": 550 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8612099644128114, + "eval_loss": 0.5797529220581055, + "eval_macro_f1": 0.8437219251336898, + "eval_runtime": 0.6399, + "eval_samples_per_second": 439.105, + "eval_steps_per_second": 28.128, + "eval_weighted_f1": 0.8613960073838659, + "step": 553 + }, + { + "epoch": 7.2784810126582276, + "grad_norm": 220.0026397705078, + "learning_rate": 1.4149085794655415e-05, + "loss": 0.2764, + "step": 575 + }, + { + "epoch": 7.594936708860759, + "grad_norm": 68.6791763305664, + "learning_rate": 1.3797468354430381e-05, + "loss": 0.5405, + "step": 600 + }, + { + "epoch": 7.911392405063291, + "grad_norm": 381.5016174316406, + "learning_rate": 1.3445850914205344e-05, + "loss": 0.3258, + "step": 625 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8469750889679716, + "eval_loss": 0.8536282777786255, + "eval_macro_f1": 0.8303354536136035, + "eval_runtime": 0.6626, + "eval_samples_per_second": 424.085, + "eval_steps_per_second": 27.166, + "eval_weighted_f1": 0.8482986963257053, + "step": 632 + }, + { + "epoch": 8.227848101265822, + "grad_norm": 131.03399658203125, + "learning_rate": 1.309423347398031e-05, + "loss": 0.2809, + "step": 650 + }, + { + "epoch": 8.544303797468354, + "grad_norm": 273.69573974609375, + "learning_rate": 1.2742616033755275e-05, + "loss": 0.171, + "step": 675 + }, + { + "epoch": 8.860759493670885, + "grad_norm": 287.41156005859375, + "learning_rate": 1.2390998593530238e-05, + "loss": 0.3972, + "step": 700 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8647686832740213, + "eval_loss": 0.957329273223877, + "eval_macro_f1": 0.8489247311827957, + "eval_runtime": 0.6499, + "eval_samples_per_second": 432.37, + "eval_steps_per_second": 27.696, + "eval_weighted_f1": 0.8654651207285807, + "step": 711 + }, + { + "epoch": 9.177215189873417, + "grad_norm": 204.0855255126953, + "learning_rate": 1.2039381153305205e-05, + "loss": 0.4871, + "step": 725 + }, + { + "epoch": 9.49367088607595, + "grad_norm": 0.08458869904279709, + "learning_rate": 1.168776371308017e-05, + "loss": 0.1852, + "step": 750 + }, + { + "epoch": 9.810126582278482, + "grad_norm": 0.16510236263275146, + "learning_rate": 1.1336146272855136e-05, + "loss": 0.2526, + "step": 775 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8612099644128114, + "eval_loss": 1.0146349668502808, + "eval_macro_f1": 0.8360927895185534, + "eval_runtime": 0.6529, + "eval_samples_per_second": 430.404, + "eval_steps_per_second": 27.57, + "eval_weighted_f1": 0.8577848951090489, + "step": 790 + }, + { + "epoch": 10.126582278481013, + "grad_norm": 285.2706298828125, + "learning_rate": 1.0984528832630099e-05, + "loss": 0.3189, + "step": 800 + }, + { + "epoch": 10.443037974683545, + "grad_norm": 74.95194244384766, + "learning_rate": 1.0632911392405063e-05, + "loss": 0.0864, + "step": 825 + }, + { + "epoch": 10.759493670886076, + "grad_norm": 0.11407579481601715, + "learning_rate": 1.028129395218003e-05, + "loss": 0.2649, + "step": 850 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.8469750889679716, + "eval_loss": 1.2006800174713135, + "eval_macro_f1": 0.8294830724940376, + "eval_runtime": 0.6345, + "eval_samples_per_second": 442.838, + "eval_steps_per_second": 28.367, + "eval_weighted_f1": 0.8479468676609678, + "step": 869 + }, + { + "epoch": 11.075949367088608, + "grad_norm": 268.8535461425781, + "learning_rate": 9.929676511954994e-06, + "loss": 0.2889, + "step": 875 + }, + { + "epoch": 11.39240506329114, + "grad_norm": 0.028925402089953423, + "learning_rate": 9.578059071729959e-06, + "loss": 0.2529, + "step": 900 + }, + { + "epoch": 11.708860759493671, + "grad_norm": 0.23740242421627045, + "learning_rate": 9.226441631504923e-06, + "loss": 0.2063, + "step": 925 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8505338078291815, + "eval_loss": 1.2759236097335815, + "eval_macro_f1": 0.8362470862470862, + "eval_runtime": 0.6721, + "eval_samples_per_second": 418.076, + "eval_steps_per_second": 26.781, + "eval_weighted_f1": 0.8525993579374362, + "step": 948 + } + ], + "logging_steps": 25, + "max_steps": 1580, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/trainer_state.json b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..a52204a2e25bc7d21b689527cc89caebfc35bf6a --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548/trainer_state.json @@ -0,0 +1,904 @@ +{ + "best_global_step": 2548, + "best_metric": 0.8084286581699676, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548", + "epoch": 14.0, + "eval_steps": 500, + "global_step": 2548, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13793103448275862, + "grad_norm": 39.19527816772461, + "learning_rate": 1.3186813186813187e-06, + "loss": 4.9093, + "step": 25 + }, + { + "epoch": 0.27586206896551724, + "grad_norm": 52.46269226074219, + "learning_rate": 2.6923076923076923e-06, + "loss": 4.7032, + "step": 50 + }, + { + "epoch": 0.41379310344827586, + "grad_norm": 62.08519744873047, + "learning_rate": 4.065934065934066e-06, + "loss": 4.6141, + "step": 75 + }, + { + "epoch": 0.5517241379310345, + "grad_norm": 52.28927993774414, + "learning_rate": 5.43956043956044e-06, + "loss": 4.4695, + "step": 100 + }, + { + "epoch": 0.6896551724137931, + "grad_norm": 72.97164916992188, + "learning_rate": 6.813186813186814e-06, + "loss": 4.445, + "step": 125 + }, + { + "epoch": 0.8275862068965517, + "grad_norm": 135.36875915527344, + "learning_rate": 8.186813186813188e-06, + "loss": 4.3505, + "step": 150 + }, + { + "epoch": 0.9655172413793104, + "grad_norm": 87.52588653564453, + "learning_rate": 9.560439560439562e-06, + "loss": 4.3182, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5582137161084529, + "eval_loss": 0.9493994116783142, + "eval_macro_f1": 0.5214585756924014, + "eval_runtime": 1.145, + "eval_samples_per_second": 547.614, + "eval_steps_per_second": 34.935, + "eval_weighted_f1": 0.5239640857538724, + "step": 182 + }, + { + "epoch": 1.0993103448275863, + "grad_norm": 18.75910186767578, + "learning_rate": 1.0934065934065933e-05, + "loss": 3.8659, + "step": 200 + }, + { + "epoch": 1.237241379310345, + "grad_norm": 103.00005340576172, + "learning_rate": 1.230769230769231e-05, + "loss": 3.9323, + "step": 225 + }, + { + "epoch": 1.3751724137931034, + "grad_norm": 461.6681823730469, + "learning_rate": 1.3681318681318683e-05, + "loss": 3.9205, + "step": 250 + }, + { + "epoch": 1.513103448275862, + "grad_norm": 86.98558807373047, + "learning_rate": 1.5054945054945056e-05, + "loss": 3.9262, + "step": 275 + }, + { + "epoch": 1.6510344827586207, + "grad_norm": 118.53007507324219, + "learning_rate": 1.642857142857143e-05, + "loss": 3.6482, + "step": 300 + }, + { + "epoch": 1.7889655172413792, + "grad_norm": 61.446815490722656, + "learning_rate": 1.78021978021978e-05, + "loss": 3.3495, + "step": 325 + }, + { + "epoch": 1.926896551724138, + "grad_norm": 36.16039276123047, + "learning_rate": 1.9175824175824178e-05, + "loss": 3.1645, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6778309409888357, + "eval_loss": 0.7728503346443176, + "eval_macro_f1": 0.663754797826036, + "eval_runtime": 1.2012, + "eval_samples_per_second": 521.966, + "eval_steps_per_second": 33.299, + "eval_weighted_f1": 0.664461982274911, + "step": 364 + }, + { + "epoch": 2.0606896551724136, + "grad_norm": 90.64247131347656, + "learning_rate": 1.993894993894994e-05, + "loss": 3.1183, + "step": 375 + }, + { + "epoch": 2.1986206896551725, + "grad_norm": 132.85733032226562, + "learning_rate": 1.9786324786324788e-05, + "loss": 3.028, + "step": 400 + }, + { + "epoch": 2.336551724137931, + "grad_norm": 68.87568664550781, + "learning_rate": 1.9633699633699634e-05, + "loss": 2.7834, + "step": 425 + }, + { + "epoch": 2.47448275862069, + "grad_norm": 213.333251953125, + "learning_rate": 1.948107448107448e-05, + "loss": 2.8182, + "step": 450 + }, + { + "epoch": 2.6124137931034483, + "grad_norm": 68.24656677246094, + "learning_rate": 1.932844932844933e-05, + "loss": 2.7826, + "step": 475 + }, + { + "epoch": 2.750344827586207, + "grad_norm": 57.20626449584961, + "learning_rate": 1.9175824175824178e-05, + "loss": 2.4841, + "step": 500 + }, + { + "epoch": 2.8882758620689657, + "grad_norm": 86.33191680908203, + "learning_rate": 1.9023199023199025e-05, + "loss": 2.3844, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.759170653907496, + "eval_loss": 0.634452223777771, + "eval_macro_f1": 0.7576138458491398, + "eval_runtime": 1.2596, + "eval_samples_per_second": 497.763, + "eval_steps_per_second": 31.755, + "eval_weighted_f1": 0.7578268121471049, + "step": 546 + }, + { + "epoch": 3.0220689655172412, + "grad_norm": 46.42818832397461, + "learning_rate": 1.887057387057387e-05, + "loss": 2.1964, + "step": 550 + }, + { + "epoch": 3.16, + "grad_norm": 85.85521697998047, + "learning_rate": 1.8717948717948718e-05, + "loss": 1.9962, + "step": 575 + }, + { + "epoch": 3.2979310344827586, + "grad_norm": 74.1902847290039, + "learning_rate": 1.8565323565323568e-05, + "loss": 1.9985, + "step": 600 + }, + { + "epoch": 3.435862068965517, + "grad_norm": 55.97420120239258, + "learning_rate": 1.8412698412698415e-05, + "loss": 2.1766, + "step": 625 + }, + { + "epoch": 3.573793103448276, + "grad_norm": 61.05451202392578, + "learning_rate": 1.826007326007326e-05, + "loss": 1.9655, + "step": 650 + }, + { + "epoch": 3.7117241379310344, + "grad_norm": 74.15213012695312, + "learning_rate": 1.8107448107448108e-05, + "loss": 1.8523, + "step": 675 + }, + { + "epoch": 3.849655172413793, + "grad_norm": 123.62406921386719, + "learning_rate": 1.7954822954822958e-05, + "loss": 2.1683, + "step": 700 + }, + { + "epoch": 3.9875862068965517, + "grad_norm": 112.76321411132812, + "learning_rate": 1.78021978021978e-05, + "loss": 2.263, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7623604465709729, + "eval_loss": 0.6158862113952637, + "eval_macro_f1": 0.761909154501918, + "eval_runtime": 1.1601, + "eval_samples_per_second": 540.485, + "eval_steps_per_second": 34.481, + "eval_weighted_f1": 0.7622521658984954, + "step": 728 + }, + { + "epoch": 4.121379310344827, + "grad_norm": 87.3993148803711, + "learning_rate": 1.764957264957265e-05, + "loss": 1.6419, + "step": 750 + }, + { + "epoch": 4.259310344827586, + "grad_norm": 37.41790771484375, + "learning_rate": 1.74969474969475e-05, + "loss": 1.6413, + "step": 775 + }, + { + "epoch": 4.397241379310345, + "grad_norm": 64.60355377197266, + "learning_rate": 1.7344322344322345e-05, + "loss": 1.4674, + "step": 800 + }, + { + "epoch": 4.535172413793103, + "grad_norm": 30.588787078857422, + "learning_rate": 1.7191697191697195e-05, + "loss": 1.7474, + "step": 825 + }, + { + "epoch": 4.673103448275862, + "grad_norm": 92.7600326538086, + "learning_rate": 1.703907203907204e-05, + "loss": 1.4887, + "step": 850 + }, + { + "epoch": 4.811034482758621, + "grad_norm": 119.30988311767578, + "learning_rate": 1.688644688644689e-05, + "loss": 1.7374, + "step": 875 + }, + { + "epoch": 4.94896551724138, + "grad_norm": 58.519447326660156, + "learning_rate": 1.6733821733821735e-05, + "loss": 1.8617, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7783094098883573, + "eval_loss": 0.6327180862426758, + "eval_macro_f1": 0.779588954162009, + "eval_runtime": 1.1419, + "eval_samples_per_second": 549.099, + "eval_steps_per_second": 35.03, + "eval_weighted_f1": 0.779605143168874, + "step": 910 + }, + { + "epoch": 5.082758620689655, + "grad_norm": 42.61728286743164, + "learning_rate": 1.6581196581196585e-05, + "loss": 1.5041, + "step": 925 + }, + { + "epoch": 5.220689655172414, + "grad_norm": 154.97837829589844, + "learning_rate": 1.642857142857143e-05, + "loss": 1.2823, + "step": 950 + }, + { + "epoch": 5.358620689655172, + "grad_norm": 129.80027770996094, + "learning_rate": 1.627594627594628e-05, + "loss": 1.2453, + "step": 975 + }, + { + "epoch": 5.496551724137931, + "grad_norm": 54.04684829711914, + "learning_rate": 1.6123321123321125e-05, + "loss": 1.4213, + "step": 1000 + }, + { + "epoch": 5.63448275862069, + "grad_norm": 80.3543472290039, + "learning_rate": 1.5970695970695972e-05, + "loss": 1.1575, + "step": 1025 + }, + { + "epoch": 5.772413793103448, + "grad_norm": 88.98538208007812, + "learning_rate": 1.581807081807082e-05, + "loss": 1.1128, + "step": 1050 + }, + { + "epoch": 5.910344827586207, + "grad_norm": 85.2730941772461, + "learning_rate": 1.5665445665445665e-05, + "loss": 1.2682, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7910685805422647, + "eval_loss": 0.6248478293418884, + "eval_macro_f1": 0.7915072260253936, + "eval_runtime": 1.1875, + "eval_samples_per_second": 527.995, + "eval_steps_per_second": 33.684, + "eval_weighted_f1": 0.791812298590797, + "step": 1092 + }, + { + "epoch": 6.0441379310344825, + "grad_norm": 53.63780212402344, + "learning_rate": 1.5512820512820516e-05, + "loss": 1.1473, + "step": 1100 + }, + { + "epoch": 6.182068965517241, + "grad_norm": 63.029563903808594, + "learning_rate": 1.5360195360195362e-05, + "loss": 0.8161, + "step": 1125 + }, + { + "epoch": 6.32, + "grad_norm": 28.43917465209961, + "learning_rate": 1.5207570207570207e-05, + "loss": 0.8535, + "step": 1150 + }, + { + "epoch": 6.457931034482758, + "grad_norm": 158.05908203125, + "learning_rate": 1.5054945054945056e-05, + "loss": 1.1149, + "step": 1175 + }, + { + "epoch": 6.595862068965517, + "grad_norm": 70.95556640625, + "learning_rate": 1.4902319902319904e-05, + "loss": 0.8912, + "step": 1200 + }, + { + "epoch": 6.733793103448276, + "grad_norm": 74.47285461425781, + "learning_rate": 1.4749694749694752e-05, + "loss": 0.8943, + "step": 1225 + }, + { + "epoch": 6.871724137931034, + "grad_norm": 163.60885620117188, + "learning_rate": 1.4597069597069597e-05, + "loss": 1.049, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.7990430622009569, + "eval_loss": 0.7322165966033936, + "eval_macro_f1": 0.7991770461017825, + "eval_runtime": 1.304, + "eval_samples_per_second": 480.823, + "eval_steps_per_second": 30.675, + "eval_weighted_f1": 0.7994727057439845, + "step": 1274 + }, + { + "epoch": 7.0055172413793105, + "grad_norm": 54.150089263916016, + "learning_rate": 1.4444444444444446e-05, + "loss": 1.021, + "step": 1275 + }, + { + "epoch": 7.1434482758620685, + "grad_norm": 34.459228515625, + "learning_rate": 1.4291819291819293e-05, + "loss": 0.6191, + "step": 1300 + }, + { + "epoch": 7.281379310344827, + "grad_norm": 67.76248168945312, + "learning_rate": 1.4139194139194141e-05, + "loss": 0.7633, + "step": 1325 + }, + { + "epoch": 7.419310344827586, + "grad_norm": 21.25878143310547, + "learning_rate": 1.3986568986568988e-05, + "loss": 0.6288, + "step": 1350 + }, + { + "epoch": 7.557241379310344, + "grad_norm": 62.710269927978516, + "learning_rate": 1.3833943833943834e-05, + "loss": 0.6954, + "step": 1375 + }, + { + "epoch": 7.695172413793103, + "grad_norm": 194.7527618408203, + "learning_rate": 1.3681318681318683e-05, + "loss": 0.4977, + "step": 1400 + }, + { + "epoch": 7.833103448275862, + "grad_norm": 118.31316375732422, + "learning_rate": 1.3528693528693531e-05, + "loss": 0.9496, + "step": 1425 + }, + { + "epoch": 7.971034482758621, + "grad_norm": 90.14521026611328, + "learning_rate": 1.3376068376068376e-05, + "loss": 0.7248, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8022328548644339, + "eval_loss": 0.7689692378044128, + "eval_macro_f1": 0.8004514623124823, + "eval_runtime": 1.1568, + "eval_samples_per_second": 542.001, + "eval_steps_per_second": 34.577, + "eval_weighted_f1": 0.801057614157989, + "step": 1456 + }, + { + "epoch": 8.104827586206897, + "grad_norm": 301.8183898925781, + "learning_rate": 1.3223443223443224e-05, + "loss": 0.5541, + "step": 1475 + }, + { + "epoch": 8.242758620689655, + "grad_norm": 203.32354736328125, + "learning_rate": 1.3070818070818073e-05, + "loss": 0.5226, + "step": 1500 + }, + { + "epoch": 8.380689655172414, + "grad_norm": 15.354279518127441, + "learning_rate": 1.291819291819292e-05, + "loss": 0.4858, + "step": 1525 + }, + { + "epoch": 8.518620689655172, + "grad_norm": 10.226881980895996, + "learning_rate": 1.2765567765567766e-05, + "loss": 0.5951, + "step": 1550 + }, + { + "epoch": 8.65655172413793, + "grad_norm": 83.22173309326172, + "learning_rate": 1.2612942612942613e-05, + "loss": 0.5951, + "step": 1575 + }, + { + "epoch": 8.79448275862069, + "grad_norm": 12.615023612976074, + "learning_rate": 1.2460317460317461e-05, + "loss": 0.489, + "step": 1600 + }, + { + "epoch": 8.932413793103448, + "grad_norm": 54.15406799316406, + "learning_rate": 1.230769230769231e-05, + "loss": 0.4364, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8022328548644339, + "eval_loss": 0.9064925909042358, + "eval_macro_f1": 0.8025287782234445, + "eval_runtime": 1.1773, + "eval_samples_per_second": 532.59, + "eval_steps_per_second": 33.977, + "eval_weighted_f1": 0.8023587030058814, + "step": 1638 + }, + { + "epoch": 9.066206896551725, + "grad_norm": 87.1186294555664, + "learning_rate": 1.2155067155067155e-05, + "loss": 0.4131, + "step": 1650 + }, + { + "epoch": 9.204137931034483, + "grad_norm": 118.65606689453125, + "learning_rate": 1.2002442002442003e-05, + "loss": 0.3966, + "step": 1675 + }, + { + "epoch": 9.34206896551724, + "grad_norm": 43.27682113647461, + "learning_rate": 1.1849816849816852e-05, + "loss": 0.3994, + "step": 1700 + }, + { + "epoch": 9.48, + "grad_norm": 46.55112075805664, + "learning_rate": 1.1697191697191698e-05, + "loss": 0.3002, + "step": 1725 + }, + { + "epoch": 9.617931034482758, + "grad_norm": 299.67840576171875, + "learning_rate": 1.1544566544566545e-05, + "loss": 0.4456, + "step": 1750 + }, + { + "epoch": 9.755862068965516, + "grad_norm": 94.89724731445312, + "learning_rate": 1.1391941391941392e-05, + "loss": 0.4069, + "step": 1775 + }, + { + "epoch": 9.893793103448276, + "grad_norm": 83.06227111816406, + "learning_rate": 1.123931623931624e-05, + "loss": 0.3748, + "step": 1800 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.773524720893142, + "eval_loss": 1.3610498905181885, + "eval_macro_f1": 0.7716054942384281, + "eval_runtime": 1.1853, + "eval_samples_per_second": 528.97, + "eval_steps_per_second": 33.746, + "eval_weighted_f1": 0.7719216877725383, + "step": 1820 + }, + { + "epoch": 10.027586206896551, + "grad_norm": 171.37094116210938, + "learning_rate": 1.1086691086691088e-05, + "loss": 0.398, + "step": 1825 + }, + { + "epoch": 10.16551724137931, + "grad_norm": 31.03558349609375, + "learning_rate": 1.0934065934065933e-05, + "loss": 0.2433, + "step": 1850 + }, + { + "epoch": 10.303448275862069, + "grad_norm": 5.404947757720947, + "learning_rate": 1.0781440781440782e-05, + "loss": 0.3894, + "step": 1875 + }, + { + "epoch": 10.441379310344828, + "grad_norm": 1.9526923894882202, + "learning_rate": 1.062881562881563e-05, + "loss": 0.2768, + "step": 1900 + }, + { + "epoch": 10.579310344827586, + "grad_norm": 3.5537567138671875, + "learning_rate": 1.0476190476190477e-05, + "loss": 0.3726, + "step": 1925 + }, + { + "epoch": 10.717241379310344, + "grad_norm": 218.355712890625, + "learning_rate": 1.0323565323565325e-05, + "loss": 0.2008, + "step": 1950 + }, + { + "epoch": 10.855172413793104, + "grad_norm": 0.5096381306648254, + "learning_rate": 1.0170940170940172e-05, + "loss": 0.5065, + "step": 1975 + }, + { + "epoch": 10.993103448275862, + "grad_norm": 0.12985798716545105, + "learning_rate": 1.0018315018315019e-05, + "loss": 0.2639, + "step": 2000 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.79585326953748, + "eval_loss": 1.401996374130249, + "eval_macro_f1": 0.794472041204085, + "eval_runtime": 1.1683, + "eval_samples_per_second": 536.672, + "eval_steps_per_second": 34.237, + "eval_weighted_f1": 0.7946472239407022, + "step": 2002 + }, + { + "epoch": 11.126896551724139, + "grad_norm": 244.31393432617188, + "learning_rate": 9.865689865689867e-06, + "loss": 0.4166, + "step": 2025 + }, + { + "epoch": 11.264827586206897, + "grad_norm": 0.7453848123550415, + "learning_rate": 9.713064713064714e-06, + "loss": 0.1639, + "step": 2050 + }, + { + "epoch": 11.402758620689655, + "grad_norm": 1.1450600624084473, + "learning_rate": 9.560439560439562e-06, + "loss": 0.2126, + "step": 2075 + }, + { + "epoch": 11.540689655172415, + "grad_norm": 94.49410247802734, + "learning_rate": 9.407814407814409e-06, + "loss": 0.4276, + "step": 2100 + }, + { + "epoch": 11.678620689655173, + "grad_norm": 53.80741500854492, + "learning_rate": 9.255189255189256e-06, + "loss": 0.4125, + "step": 2125 + }, + { + "epoch": 11.81655172413793, + "grad_norm": 2.2708137035369873, + "learning_rate": 9.102564102564104e-06, + "loss": 0.3553, + "step": 2150 + }, + { + "epoch": 11.95448275862069, + "grad_norm": 16.314722061157227, + "learning_rate": 8.94993894993895e-06, + "loss": 0.213, + "step": 2175 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8086124401913876, + "eval_loss": 1.4752185344696045, + "eval_macro_f1": 0.8073371962771456, + "eval_runtime": 1.1292, + "eval_samples_per_second": 555.264, + "eval_steps_per_second": 35.424, + "eval_weighted_f1": 0.8074475414562201, + "step": 2184 + }, + { + "epoch": 12.088275862068965, + "grad_norm": 4.032958507537842, + "learning_rate": 8.797313797313797e-06, + "loss": 0.1548, + "step": 2200 + }, + { + "epoch": 12.226206896551725, + "grad_norm": 11.657094955444336, + "learning_rate": 8.644688644688646e-06, + "loss": 0.2706, + "step": 2225 + }, + { + "epoch": 12.364137931034483, + "grad_norm": 12.007180213928223, + "learning_rate": 8.492063492063492e-06, + "loss": 0.1725, + "step": 2250 + }, + { + "epoch": 12.50206896551724, + "grad_norm": 3.103738307952881, + "learning_rate": 8.33943833943834e-06, + "loss": 0.1077, + "step": 2275 + }, + { + "epoch": 12.64, + "grad_norm": 0.05534272640943527, + "learning_rate": 8.186813186813188e-06, + "loss": 0.2332, + "step": 2300 + }, + { + "epoch": 12.777931034482759, + "grad_norm": 0.27866512537002563, + "learning_rate": 8.034188034188036e-06, + "loss": 0.0977, + "step": 2325 + }, + { + "epoch": 12.915862068965517, + "grad_norm": 16.027671813964844, + "learning_rate": 7.881562881562883e-06, + "loss": 0.3268, + "step": 2350 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.7974481658692185, + "eval_loss": 1.5549489259719849, + "eval_macro_f1": 0.7969154183788802, + "eval_runtime": 1.1851, + "eval_samples_per_second": 529.085, + "eval_steps_per_second": 33.753, + "eval_weighted_f1": 0.7970637749059234, + "step": 2366 + }, + { + "epoch": 13.049655172413793, + "grad_norm": 167.1862030029297, + "learning_rate": 7.72893772893773e-06, + "loss": 0.1916, + "step": 2375 + }, + { + "epoch": 13.187586206896551, + "grad_norm": 578.3486328125, + "learning_rate": 7.576312576312577e-06, + "loss": 0.2089, + "step": 2400 + }, + { + "epoch": 13.32551724137931, + "grad_norm": 2.9909238815307617, + "learning_rate": 7.423687423687424e-06, + "loss": 0.1702, + "step": 2425 + }, + { + "epoch": 13.463448275862069, + "grad_norm": 12.770936012268066, + "learning_rate": 7.271062271062271e-06, + "loss": 0.1574, + "step": 2450 + }, + { + "epoch": 13.601379310344827, + "grad_norm": 0.8478724360466003, + "learning_rate": 7.1184371184371195e-06, + "loss": 0.1238, + "step": 2475 + }, + { + "epoch": 13.739310344827587, + "grad_norm": 0.18776558339595795, + "learning_rate": 6.965811965811966e-06, + "loss": 0.0629, + "step": 2500 + }, + { + "epoch": 13.877241379310345, + "grad_norm": 10.055283546447754, + "learning_rate": 6.813186813186814e-06, + "loss": 0.3233, + "step": 2525 + }, + { + "epoch": 14.0, + "eval_accuracy": 0.8086124401913876, + "eval_loss": 1.6799039840698242, + "eval_macro_f1": 0.8084286581699676, + "eval_runtime": 1.1687, + "eval_samples_per_second": 536.477, + "eval_steps_per_second": 34.225, + "eval_weighted_f1": 0.8085553790065865, + "step": 2548 + } + ], + "logging_steps": 25, + "max_steps": 3640, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/trainer_state.json b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..b9d680ca24fe3e7f152e0f876cd4d9fdea07174e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-3094/trainer_state.json @@ -0,0 +1,1091 @@ +{ + "best_global_step": 2548, + "best_metric": 0.8084286581699676, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-00-s42-20260901T150602Z-27bda108/trainer/checkpoint-2548", + "epoch": 17.0, + "eval_steps": 500, + "global_step": 3094, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13793103448275862, + "grad_norm": 39.19527816772461, + "learning_rate": 1.3186813186813187e-06, + "loss": 4.9093, + "step": 25 + }, + { + "epoch": 0.27586206896551724, + "grad_norm": 52.46269226074219, + "learning_rate": 2.6923076923076923e-06, + "loss": 4.7032, + "step": 50 + }, + { + "epoch": 0.41379310344827586, + "grad_norm": 62.08519744873047, + "learning_rate": 4.065934065934066e-06, + "loss": 4.6141, + "step": 75 + }, + { + "epoch": 0.5517241379310345, + "grad_norm": 52.28927993774414, + "learning_rate": 5.43956043956044e-06, + "loss": 4.4695, + "step": 100 + }, + { + "epoch": 0.6896551724137931, + "grad_norm": 72.97164916992188, + "learning_rate": 6.813186813186814e-06, + "loss": 4.445, + "step": 125 + }, + { + "epoch": 0.8275862068965517, + "grad_norm": 135.36875915527344, + "learning_rate": 8.186813186813188e-06, + "loss": 4.3505, + "step": 150 + }, + { + "epoch": 0.9655172413793104, + "grad_norm": 87.52588653564453, + "learning_rate": 9.560439560439562e-06, + "loss": 4.3182, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5582137161084529, + "eval_loss": 0.9493994116783142, + "eval_macro_f1": 0.5214585756924014, + "eval_runtime": 1.145, + "eval_samples_per_second": 547.614, + "eval_steps_per_second": 34.935, + "eval_weighted_f1": 0.5239640857538724, + "step": 182 + }, + { + "epoch": 1.0993103448275863, + "grad_norm": 18.75910186767578, + "learning_rate": 1.0934065934065933e-05, + "loss": 3.8659, + "step": 200 + }, + { + "epoch": 1.237241379310345, + "grad_norm": 103.00005340576172, + "learning_rate": 1.230769230769231e-05, + "loss": 3.9323, + "step": 225 + }, + { + "epoch": 1.3751724137931034, + "grad_norm": 461.6681823730469, + "learning_rate": 1.3681318681318683e-05, + "loss": 3.9205, + "step": 250 + }, + { + "epoch": 1.513103448275862, + "grad_norm": 86.98558807373047, + "learning_rate": 1.5054945054945056e-05, + "loss": 3.9262, + "step": 275 + }, + { + "epoch": 1.6510344827586207, + "grad_norm": 118.53007507324219, + "learning_rate": 1.642857142857143e-05, + "loss": 3.6482, + "step": 300 + }, + { + "epoch": 1.7889655172413792, + "grad_norm": 61.446815490722656, + "learning_rate": 1.78021978021978e-05, + "loss": 3.3495, + "step": 325 + }, + { + "epoch": 1.926896551724138, + "grad_norm": 36.16039276123047, + "learning_rate": 1.9175824175824178e-05, + "loss": 3.1645, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6778309409888357, + "eval_loss": 0.7728503346443176, + "eval_macro_f1": 0.663754797826036, + "eval_runtime": 1.2012, + "eval_samples_per_second": 521.966, + "eval_steps_per_second": 33.299, + "eval_weighted_f1": 0.664461982274911, + "step": 364 + }, + { + "epoch": 2.0606896551724136, + "grad_norm": 90.64247131347656, + "learning_rate": 1.993894993894994e-05, + "loss": 3.1183, + "step": 375 + }, + { + "epoch": 2.1986206896551725, + "grad_norm": 132.85733032226562, + "learning_rate": 1.9786324786324788e-05, + "loss": 3.028, + "step": 400 + }, + { + "epoch": 2.336551724137931, + "grad_norm": 68.87568664550781, + "learning_rate": 1.9633699633699634e-05, + "loss": 2.7834, + "step": 425 + }, + { + "epoch": 2.47448275862069, + "grad_norm": 213.333251953125, + "learning_rate": 1.948107448107448e-05, + "loss": 2.8182, + "step": 450 + }, + { + "epoch": 2.6124137931034483, + "grad_norm": 68.24656677246094, + "learning_rate": 1.932844932844933e-05, + "loss": 2.7826, + "step": 475 + }, + { + "epoch": 2.750344827586207, + "grad_norm": 57.20626449584961, + "learning_rate": 1.9175824175824178e-05, + "loss": 2.4841, + "step": 500 + }, + { + "epoch": 2.8882758620689657, + "grad_norm": 86.33191680908203, + "learning_rate": 1.9023199023199025e-05, + "loss": 2.3844, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.759170653907496, + "eval_loss": 0.634452223777771, + "eval_macro_f1": 0.7576138458491398, + "eval_runtime": 1.2596, + "eval_samples_per_second": 497.763, + "eval_steps_per_second": 31.755, + "eval_weighted_f1": 0.7578268121471049, + "step": 546 + }, + { + "epoch": 3.0220689655172412, + "grad_norm": 46.42818832397461, + "learning_rate": 1.887057387057387e-05, + "loss": 2.1964, + "step": 550 + }, + { + "epoch": 3.16, + "grad_norm": 85.85521697998047, + "learning_rate": 1.8717948717948718e-05, + "loss": 1.9962, + "step": 575 + }, + { + "epoch": 3.2979310344827586, + "grad_norm": 74.1902847290039, + "learning_rate": 1.8565323565323568e-05, + "loss": 1.9985, + "step": 600 + }, + { + "epoch": 3.435862068965517, + "grad_norm": 55.97420120239258, + "learning_rate": 1.8412698412698415e-05, + "loss": 2.1766, + "step": 625 + }, + { + "epoch": 3.573793103448276, + "grad_norm": 61.05451202392578, + "learning_rate": 1.826007326007326e-05, + "loss": 1.9655, + "step": 650 + }, + { + "epoch": 3.7117241379310344, + "grad_norm": 74.15213012695312, + "learning_rate": 1.8107448107448108e-05, + "loss": 1.8523, + "step": 675 + }, + { + "epoch": 3.849655172413793, + "grad_norm": 123.62406921386719, + "learning_rate": 1.7954822954822958e-05, + "loss": 2.1683, + "step": 700 + }, + { + "epoch": 3.9875862068965517, + "grad_norm": 112.76321411132812, + "learning_rate": 1.78021978021978e-05, + "loss": 2.263, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7623604465709729, + "eval_loss": 0.6158862113952637, + "eval_macro_f1": 0.761909154501918, + "eval_runtime": 1.1601, + "eval_samples_per_second": 540.485, + "eval_steps_per_second": 34.481, + "eval_weighted_f1": 0.7622521658984954, + "step": 728 + }, + { + "epoch": 4.121379310344827, + "grad_norm": 87.3993148803711, + "learning_rate": 1.764957264957265e-05, + "loss": 1.6419, + "step": 750 + }, + { + "epoch": 4.259310344827586, + "grad_norm": 37.41790771484375, + "learning_rate": 1.74969474969475e-05, + "loss": 1.6413, + "step": 775 + }, + { + "epoch": 4.397241379310345, + "grad_norm": 64.60355377197266, + "learning_rate": 1.7344322344322345e-05, + "loss": 1.4674, + "step": 800 + }, + { + "epoch": 4.535172413793103, + "grad_norm": 30.588787078857422, + "learning_rate": 1.7191697191697195e-05, + "loss": 1.7474, + "step": 825 + }, + { + "epoch": 4.673103448275862, + "grad_norm": 92.7600326538086, + "learning_rate": 1.703907203907204e-05, + "loss": 1.4887, + "step": 850 + }, + { + "epoch": 4.811034482758621, + "grad_norm": 119.30988311767578, + "learning_rate": 1.688644688644689e-05, + "loss": 1.7374, + "step": 875 + }, + { + "epoch": 4.94896551724138, + "grad_norm": 58.519447326660156, + "learning_rate": 1.6733821733821735e-05, + "loss": 1.8617, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7783094098883573, + "eval_loss": 0.6327180862426758, + "eval_macro_f1": 0.779588954162009, + "eval_runtime": 1.1419, + "eval_samples_per_second": 549.099, + "eval_steps_per_second": 35.03, + "eval_weighted_f1": 0.779605143168874, + "step": 910 + }, + { + "epoch": 5.082758620689655, + "grad_norm": 42.61728286743164, + "learning_rate": 1.6581196581196585e-05, + "loss": 1.5041, + "step": 925 + }, + { + "epoch": 5.220689655172414, + "grad_norm": 154.97837829589844, + "learning_rate": 1.642857142857143e-05, + "loss": 1.2823, + "step": 950 + }, + { + "epoch": 5.358620689655172, + "grad_norm": 129.80027770996094, + "learning_rate": 1.627594627594628e-05, + "loss": 1.2453, + "step": 975 + }, + { + "epoch": 5.496551724137931, + "grad_norm": 54.04684829711914, + "learning_rate": 1.6123321123321125e-05, + "loss": 1.4213, + "step": 1000 + }, + { + "epoch": 5.63448275862069, + "grad_norm": 80.3543472290039, + "learning_rate": 1.5970695970695972e-05, + "loss": 1.1575, + "step": 1025 + }, + { + "epoch": 5.772413793103448, + "grad_norm": 88.98538208007812, + "learning_rate": 1.581807081807082e-05, + "loss": 1.1128, + "step": 1050 + }, + { + "epoch": 5.910344827586207, + "grad_norm": 85.2730941772461, + "learning_rate": 1.5665445665445665e-05, + "loss": 1.2682, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7910685805422647, + "eval_loss": 0.6248478293418884, + "eval_macro_f1": 0.7915072260253936, + "eval_runtime": 1.1875, + "eval_samples_per_second": 527.995, + "eval_steps_per_second": 33.684, + "eval_weighted_f1": 0.791812298590797, + "step": 1092 + }, + { + "epoch": 6.0441379310344825, + "grad_norm": 53.63780212402344, + "learning_rate": 1.5512820512820516e-05, + "loss": 1.1473, + "step": 1100 + }, + { + "epoch": 6.182068965517241, + "grad_norm": 63.029563903808594, + "learning_rate": 1.5360195360195362e-05, + "loss": 0.8161, + "step": 1125 + }, + { + "epoch": 6.32, + "grad_norm": 28.43917465209961, + "learning_rate": 1.5207570207570207e-05, + "loss": 0.8535, + "step": 1150 + }, + { + "epoch": 6.457931034482758, + "grad_norm": 158.05908203125, + "learning_rate": 1.5054945054945056e-05, + "loss": 1.1149, + "step": 1175 + }, + { + "epoch": 6.595862068965517, + "grad_norm": 70.95556640625, + "learning_rate": 1.4902319902319904e-05, + "loss": 0.8912, + "step": 1200 + }, + { + "epoch": 6.733793103448276, + "grad_norm": 74.47285461425781, + "learning_rate": 1.4749694749694752e-05, + "loss": 0.8943, + "step": 1225 + }, + { + "epoch": 6.871724137931034, + "grad_norm": 163.60885620117188, + "learning_rate": 1.4597069597069597e-05, + "loss": 1.049, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.7990430622009569, + "eval_loss": 0.7322165966033936, + "eval_macro_f1": 0.7991770461017825, + "eval_runtime": 1.304, + "eval_samples_per_second": 480.823, + "eval_steps_per_second": 30.675, + "eval_weighted_f1": 0.7994727057439845, + "step": 1274 + }, + { + "epoch": 7.0055172413793105, + "grad_norm": 54.150089263916016, + "learning_rate": 1.4444444444444446e-05, + "loss": 1.021, + "step": 1275 + }, + { + "epoch": 7.1434482758620685, + "grad_norm": 34.459228515625, + "learning_rate": 1.4291819291819293e-05, + "loss": 0.6191, + "step": 1300 + }, + { + "epoch": 7.281379310344827, + "grad_norm": 67.76248168945312, + "learning_rate": 1.4139194139194141e-05, + "loss": 0.7633, + "step": 1325 + }, + { + "epoch": 7.419310344827586, + "grad_norm": 21.25878143310547, + "learning_rate": 1.3986568986568988e-05, + "loss": 0.6288, + "step": 1350 + }, + { + "epoch": 7.557241379310344, + "grad_norm": 62.710269927978516, + "learning_rate": 1.3833943833943834e-05, + "loss": 0.6954, + "step": 1375 + }, + { + "epoch": 7.695172413793103, + "grad_norm": 194.7527618408203, + "learning_rate": 1.3681318681318683e-05, + "loss": 0.4977, + "step": 1400 + }, + { + "epoch": 7.833103448275862, + "grad_norm": 118.31316375732422, + "learning_rate": 1.3528693528693531e-05, + "loss": 0.9496, + "step": 1425 + }, + { + "epoch": 7.971034482758621, + "grad_norm": 90.14521026611328, + "learning_rate": 1.3376068376068376e-05, + "loss": 0.7248, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8022328548644339, + "eval_loss": 0.7689692378044128, + "eval_macro_f1": 0.8004514623124823, + "eval_runtime": 1.1568, + "eval_samples_per_second": 542.001, + "eval_steps_per_second": 34.577, + "eval_weighted_f1": 0.801057614157989, + "step": 1456 + }, + { + "epoch": 8.104827586206897, + "grad_norm": 301.8183898925781, + "learning_rate": 1.3223443223443224e-05, + "loss": 0.5541, + "step": 1475 + }, + { + "epoch": 8.242758620689655, + "grad_norm": 203.32354736328125, + "learning_rate": 1.3070818070818073e-05, + "loss": 0.5226, + "step": 1500 + }, + { + "epoch": 8.380689655172414, + "grad_norm": 15.354279518127441, + "learning_rate": 1.291819291819292e-05, + "loss": 0.4858, + "step": 1525 + }, + { + "epoch": 8.518620689655172, + "grad_norm": 10.226881980895996, + "learning_rate": 1.2765567765567766e-05, + "loss": 0.5951, + "step": 1550 + }, + { + "epoch": 8.65655172413793, + "grad_norm": 83.22173309326172, + "learning_rate": 1.2612942612942613e-05, + "loss": 0.5951, + "step": 1575 + }, + { + "epoch": 8.79448275862069, + "grad_norm": 12.615023612976074, + "learning_rate": 1.2460317460317461e-05, + "loss": 0.489, + "step": 1600 + }, + { + "epoch": 8.932413793103448, + "grad_norm": 54.15406799316406, + "learning_rate": 1.230769230769231e-05, + "loss": 0.4364, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8022328548644339, + "eval_loss": 0.9064925909042358, + "eval_macro_f1": 0.8025287782234445, + "eval_runtime": 1.1773, + "eval_samples_per_second": 532.59, + "eval_steps_per_second": 33.977, + "eval_weighted_f1": 0.8023587030058814, + "step": 1638 + }, + { + "epoch": 9.066206896551725, + "grad_norm": 87.1186294555664, + "learning_rate": 1.2155067155067155e-05, + "loss": 0.4131, + "step": 1650 + }, + { + "epoch": 9.204137931034483, + "grad_norm": 118.65606689453125, + "learning_rate": 1.2002442002442003e-05, + "loss": 0.3966, + "step": 1675 + }, + { + "epoch": 9.34206896551724, + "grad_norm": 43.27682113647461, + "learning_rate": 1.1849816849816852e-05, + "loss": 0.3994, + "step": 1700 + }, + { + "epoch": 9.48, + "grad_norm": 46.55112075805664, + "learning_rate": 1.1697191697191698e-05, + "loss": 0.3002, + "step": 1725 + }, + { + "epoch": 9.617931034482758, + "grad_norm": 299.67840576171875, + "learning_rate": 1.1544566544566545e-05, + "loss": 0.4456, + "step": 1750 + }, + { + "epoch": 9.755862068965516, + "grad_norm": 94.89724731445312, + "learning_rate": 1.1391941391941392e-05, + "loss": 0.4069, + "step": 1775 + }, + { + "epoch": 9.893793103448276, + "grad_norm": 83.06227111816406, + "learning_rate": 1.123931623931624e-05, + "loss": 0.3748, + "step": 1800 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.773524720893142, + "eval_loss": 1.3610498905181885, + "eval_macro_f1": 0.7716054942384281, + "eval_runtime": 1.1853, + "eval_samples_per_second": 528.97, + "eval_steps_per_second": 33.746, + "eval_weighted_f1": 0.7719216877725383, + "step": 1820 + }, + { + "epoch": 10.027586206896551, + "grad_norm": 171.37094116210938, + "learning_rate": 1.1086691086691088e-05, + "loss": 0.398, + "step": 1825 + }, + { + "epoch": 10.16551724137931, + "grad_norm": 31.03558349609375, + "learning_rate": 1.0934065934065933e-05, + "loss": 0.2433, + "step": 1850 + }, + { + "epoch": 10.303448275862069, + "grad_norm": 5.404947757720947, + "learning_rate": 1.0781440781440782e-05, + "loss": 0.3894, + "step": 1875 + }, + { + "epoch": 10.441379310344828, + "grad_norm": 1.9526923894882202, + "learning_rate": 1.062881562881563e-05, + "loss": 0.2768, + "step": 1900 + }, + { + "epoch": 10.579310344827586, + "grad_norm": 3.5537567138671875, + "learning_rate": 1.0476190476190477e-05, + "loss": 0.3726, + "step": 1925 + }, + { + "epoch": 10.717241379310344, + "grad_norm": 218.355712890625, + "learning_rate": 1.0323565323565325e-05, + "loss": 0.2008, + "step": 1950 + }, + { + "epoch": 10.855172413793104, + "grad_norm": 0.5096381306648254, + "learning_rate": 1.0170940170940172e-05, + "loss": 0.5065, + "step": 1975 + }, + { + "epoch": 10.993103448275862, + "grad_norm": 0.12985798716545105, + "learning_rate": 1.0018315018315019e-05, + "loss": 0.2639, + "step": 2000 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.79585326953748, + "eval_loss": 1.401996374130249, + "eval_macro_f1": 0.794472041204085, + "eval_runtime": 1.1683, + "eval_samples_per_second": 536.672, + "eval_steps_per_second": 34.237, + "eval_weighted_f1": 0.7946472239407022, + "step": 2002 + }, + { + "epoch": 11.126896551724139, + "grad_norm": 244.31393432617188, + "learning_rate": 9.865689865689867e-06, + "loss": 0.4166, + "step": 2025 + }, + { + "epoch": 11.264827586206897, + "grad_norm": 0.7453848123550415, + "learning_rate": 9.713064713064714e-06, + "loss": 0.1639, + "step": 2050 + }, + { + "epoch": 11.402758620689655, + "grad_norm": 1.1450600624084473, + "learning_rate": 9.560439560439562e-06, + "loss": 0.2126, + "step": 2075 + }, + { + "epoch": 11.540689655172415, + "grad_norm": 94.49410247802734, + "learning_rate": 9.407814407814409e-06, + "loss": 0.4276, + "step": 2100 + }, + { + "epoch": 11.678620689655173, + "grad_norm": 53.80741500854492, + "learning_rate": 9.255189255189256e-06, + "loss": 0.4125, + "step": 2125 + }, + { + "epoch": 11.81655172413793, + "grad_norm": 2.2708137035369873, + "learning_rate": 9.102564102564104e-06, + "loss": 0.3553, + "step": 2150 + }, + { + "epoch": 11.95448275862069, + "grad_norm": 16.314722061157227, + "learning_rate": 8.94993894993895e-06, + "loss": 0.213, + "step": 2175 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8086124401913876, + "eval_loss": 1.4752185344696045, + "eval_macro_f1": 0.8073371962771456, + "eval_runtime": 1.1292, + "eval_samples_per_second": 555.264, + "eval_steps_per_second": 35.424, + "eval_weighted_f1": 0.8074475414562201, + "step": 2184 + }, + { + "epoch": 12.088275862068965, + "grad_norm": 4.032958507537842, + "learning_rate": 8.797313797313797e-06, + "loss": 0.1548, + "step": 2200 + }, + { + "epoch": 12.226206896551725, + "grad_norm": 11.657094955444336, + "learning_rate": 8.644688644688646e-06, + "loss": 0.2706, + "step": 2225 + }, + { + "epoch": 12.364137931034483, + "grad_norm": 12.007180213928223, + "learning_rate": 8.492063492063492e-06, + "loss": 0.1725, + "step": 2250 + }, + { + "epoch": 12.50206896551724, + "grad_norm": 3.103738307952881, + "learning_rate": 8.33943833943834e-06, + "loss": 0.1077, + "step": 2275 + }, + { + "epoch": 12.64, + "grad_norm": 0.05534272640943527, + "learning_rate": 8.186813186813188e-06, + "loss": 0.2332, + "step": 2300 + }, + { + "epoch": 12.777931034482759, + "grad_norm": 0.27866512537002563, + "learning_rate": 8.034188034188036e-06, + "loss": 0.0977, + "step": 2325 + }, + { + "epoch": 12.915862068965517, + "grad_norm": 16.027671813964844, + "learning_rate": 7.881562881562883e-06, + "loss": 0.3268, + "step": 2350 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.7974481658692185, + "eval_loss": 1.5549489259719849, + "eval_macro_f1": 0.7969154183788802, + "eval_runtime": 1.1851, + "eval_samples_per_second": 529.085, + "eval_steps_per_second": 33.753, + "eval_weighted_f1": 0.7970637749059234, + "step": 2366 + }, + { + "epoch": 13.049655172413793, + "grad_norm": 167.1862030029297, + "learning_rate": 7.72893772893773e-06, + "loss": 0.1916, + "step": 2375 + }, + { + "epoch": 13.187586206896551, + "grad_norm": 578.3486328125, + "learning_rate": 7.576312576312577e-06, + "loss": 0.2089, + "step": 2400 + }, + { + "epoch": 13.32551724137931, + "grad_norm": 2.9909238815307617, + "learning_rate": 7.423687423687424e-06, + "loss": 0.1702, + "step": 2425 + }, + { + "epoch": 13.463448275862069, + "grad_norm": 12.770936012268066, + "learning_rate": 7.271062271062271e-06, + "loss": 0.1574, + "step": 2450 + }, + { + "epoch": 13.601379310344827, + "grad_norm": 0.8478724360466003, + "learning_rate": 7.1184371184371195e-06, + "loss": 0.1238, + "step": 2475 + }, + { + "epoch": 13.739310344827587, + "grad_norm": 0.18776558339595795, + "learning_rate": 6.965811965811966e-06, + "loss": 0.0629, + "step": 2500 + }, + { + "epoch": 13.877241379310345, + "grad_norm": 10.055283546447754, + "learning_rate": 6.813186813186814e-06, + "loss": 0.3233, + "step": 2525 + }, + { + "epoch": 14.0, + "eval_accuracy": 0.8086124401913876, + "eval_loss": 1.6799039840698242, + "eval_macro_f1": 0.8084286581699676, + "eval_runtime": 1.1687, + "eval_samples_per_second": 536.477, + "eval_steps_per_second": 34.225, + "eval_weighted_f1": 0.8085553790065865, + "step": 2548 + }, + { + "epoch": 14.011034482758621, + "grad_norm": 276.6418762207031, + "learning_rate": 6.66056166056166e-06, + "loss": 0.2604, + "step": 2550 + }, + { + "epoch": 14.148965517241379, + "grad_norm": 23.674741744995117, + "learning_rate": 6.507936507936509e-06, + "loss": 0.1366, + "step": 2575 + }, + { + "epoch": 14.286896551724137, + "grad_norm": 456.3885803222656, + "learning_rate": 6.3553113553113555e-06, + "loss": 0.2356, + "step": 2600 + }, + { + "epoch": 14.424827586206897, + "grad_norm": 0.32852277159690857, + "learning_rate": 6.202686202686203e-06, + "loss": 0.1828, + "step": 2625 + }, + { + "epoch": 14.562758620689655, + "grad_norm": 0.16003744304180145, + "learning_rate": 6.05006105006105e-06, + "loss": 0.121, + "step": 2650 + }, + { + "epoch": 14.700689655172415, + "grad_norm": 199.4854278564453, + "learning_rate": 5.897435897435898e-06, + "loss": 0.26, + "step": 2675 + }, + { + "epoch": 14.838620689655173, + "grad_norm": 0.23486696183681488, + "learning_rate": 5.744810744810745e-06, + "loss": 0.2724, + "step": 2700 + }, + { + "epoch": 14.97655172413793, + "grad_norm": 552.55029296875, + "learning_rate": 5.592185592185592e-06, + "loss": 0.106, + "step": 2725 + }, + { + "epoch": 15.0, + "eval_accuracy": 0.8006379585326954, + "eval_loss": 1.7526615858078003, + "eval_macro_f1": 0.8013367773515058, + "eval_runtime": 1.1993, + "eval_samples_per_second": 522.785, + "eval_steps_per_second": 33.352, + "eval_weighted_f1": 0.8011896085394442, + "step": 2730 + }, + { + "epoch": 15.110344827586207, + "grad_norm": 0.046297576278448105, + "learning_rate": 5.43956043956044e-06, + "loss": 0.1396, + "step": 2750 + }, + { + "epoch": 15.248275862068965, + "grad_norm": 0.0772584080696106, + "learning_rate": 5.2869352869352875e-06, + "loss": 0.0958, + "step": 2775 + }, + { + "epoch": 15.386206896551725, + "grad_norm": 265.6419372558594, + "learning_rate": 5.134310134310134e-06, + "loss": 0.2009, + "step": 2800 + }, + { + "epoch": 15.524137931034483, + "grad_norm": 19.894752502441406, + "learning_rate": 4.9816849816849826e-06, + "loss": 0.073, + "step": 2825 + }, + { + "epoch": 15.662068965517241, + "grad_norm": 341.0168762207031, + "learning_rate": 4.829059829059829e-06, + "loss": 0.2365, + "step": 2850 + }, + { + "epoch": 15.8, + "grad_norm": 152.31944274902344, + "learning_rate": 4.676434676434677e-06, + "loss": 0.0926, + "step": 2875 + }, + { + "epoch": 15.937931034482759, + "grad_norm": 0.21693311631679535, + "learning_rate": 4.523809523809524e-06, + "loss": 0.159, + "step": 2900 + }, + { + "epoch": 16.0, + "eval_accuracy": 0.7990430622009569, + "eval_loss": 1.797470211982727, + "eval_macro_f1": 0.799145012661444, + "eval_runtime": 1.1154, + "eval_samples_per_second": 562.148, + "eval_steps_per_second": 35.863, + "eval_weighted_f1": 0.7992051011258572, + "step": 2912 + }, + { + "epoch": 16.071724137931035, + "grad_norm": 395.2620849609375, + "learning_rate": 4.371184371184372e-06, + "loss": 0.1296, + "step": 2925 + }, + { + "epoch": 16.209655172413793, + "grad_norm": 0.18170881271362305, + "learning_rate": 4.218559218559219e-06, + "loss": 0.1053, + "step": 2950 + }, + { + "epoch": 16.34758620689655, + "grad_norm": 420.1138610839844, + "learning_rate": 4.065934065934066e-06, + "loss": 0.0738, + "step": 2975 + }, + { + "epoch": 16.48551724137931, + "grad_norm": 0.03433357924222946, + "learning_rate": 3.913308913308914e-06, + "loss": 0.0067, + "step": 3000 + }, + { + "epoch": 16.623448275862067, + "grad_norm": 0.047096770256757736, + "learning_rate": 3.760683760683761e-06, + "loss": 0.1146, + "step": 3025 + }, + { + "epoch": 16.76137931034483, + "grad_norm": 0.02566961944103241, + "learning_rate": 3.6080586080586083e-06, + "loss": 0.0096, + "step": 3050 + }, + { + "epoch": 16.899310344827587, + "grad_norm": 1.223204255104065, + "learning_rate": 3.4554334554334555e-06, + "loss": 0.0274, + "step": 3075 + }, + { + "epoch": 17.0, + "eval_accuracy": 0.8022328548644339, + "eval_loss": 1.8602946996688843, + "eval_macro_f1": 0.8019165313579076, + "eval_runtime": 1.18, + "eval_samples_per_second": 531.372, + "eval_steps_per_second": 33.899, + "eval_weighted_f1": 0.802071879926345, + "step": 3094 + } + ], + "logging_steps": 25, + "max_steps": 3640, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/trainer_state.json b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..f87096143a824e9e56d5af87f1fd8788a720c8c4 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548/trainer_state.json @@ -0,0 +1,904 @@ +{ + "best_global_step": 2548, + "best_metric": 0.8244676436475006, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548", + "epoch": 14.0, + "eval_steps": 500, + "global_step": 2548, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13793103448275862, + "grad_norm": 20.084362030029297, + "learning_rate": 1.3186813186813187e-06, + "loss": 4.7288, + "step": 25 + }, + { + "epoch": 0.27586206896551724, + "grad_norm": 5.729361534118652, + "learning_rate": 2.6923076923076923e-06, + "loss": 4.684, + "step": 50 + }, + { + "epoch": 0.41379310344827586, + "grad_norm": 7.694893836975098, + "learning_rate": 4.065934065934066e-06, + "loss": 4.5741, + "step": 75 + }, + { + "epoch": 0.5517241379310345, + "grad_norm": 8.551424980163574, + "learning_rate": 5.43956043956044e-06, + "loss": 4.4816, + "step": 100 + }, + { + "epoch": 0.6896551724137931, + "grad_norm": 35.368587493896484, + "learning_rate": 6.813186813186814e-06, + "loss": 4.5308, + "step": 125 + }, + { + "epoch": 0.8275862068965517, + "grad_norm": 44.43610763549805, + "learning_rate": 8.186813186813188e-06, + "loss": 4.444, + "step": 150 + }, + { + "epoch": 0.9655172413793104, + "grad_norm": 24.932857513427734, + "learning_rate": 9.560439560439562e-06, + "loss": 4.3658, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.4258373205741627, + "eval_loss": 1.0057348012924194, + "eval_macro_f1": 0.3397112691230338, + "eval_runtime": 1.1171, + "eval_samples_per_second": 561.294, + "eval_steps_per_second": 35.808, + "eval_weighted_f1": 0.3524679654710614, + "step": 182 + }, + { + "epoch": 1.0993103448275863, + "grad_norm": 61.700660705566406, + "learning_rate": 1.0934065934065933e-05, + "loss": 4.1429, + "step": 200 + }, + { + "epoch": 1.237241379310345, + "grad_norm": 46.13859558105469, + "learning_rate": 1.230769230769231e-05, + "loss": 3.9467, + "step": 225 + }, + { + "epoch": 1.3751724137931034, + "grad_norm": 46.452125549316406, + "learning_rate": 1.3681318681318683e-05, + "loss": 3.6127, + "step": 250 + }, + { + "epoch": 1.513103448275862, + "grad_norm": 59.73109436035156, + "learning_rate": 1.5054945054945056e-05, + "loss": 3.5063, + "step": 275 + }, + { + "epoch": 1.6510344827586207, + "grad_norm": 145.90650939941406, + "learning_rate": 1.642857142857143e-05, + "loss": 3.3779, + "step": 300 + }, + { + "epoch": 1.7889655172413792, + "grad_norm": 54.86250305175781, + "learning_rate": 1.78021978021978e-05, + "loss": 3.0618, + "step": 325 + }, + { + "epoch": 1.926896551724138, + "grad_norm": 52.89915084838867, + "learning_rate": 1.9175824175824178e-05, + "loss": 3.1781, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.7113237639553429, + "eval_loss": 0.7212908864021301, + "eval_macro_f1": 0.7123417832526394, + "eval_runtime": 1.0768, + "eval_samples_per_second": 582.292, + "eval_steps_per_second": 37.148, + "eval_weighted_f1": 0.712662576565873, + "step": 364 + }, + { + "epoch": 2.0606896551724136, + "grad_norm": 84.83079528808594, + "learning_rate": 1.993894993894994e-05, + "loss": 2.6491, + "step": 375 + }, + { + "epoch": 2.1986206896551725, + "grad_norm": 108.39063262939453, + "learning_rate": 1.9786324786324788e-05, + "loss": 2.7958, + "step": 400 + }, + { + "epoch": 2.336551724137931, + "grad_norm": 96.23841094970703, + "learning_rate": 1.9633699633699634e-05, + "loss": 2.4406, + "step": 425 + }, + { + "epoch": 2.47448275862069, + "grad_norm": 85.79228973388672, + "learning_rate": 1.948107448107448e-05, + "loss": 2.6577, + "step": 450 + }, + { + "epoch": 2.6124137931034483, + "grad_norm": 42.99919509887695, + "learning_rate": 1.932844932844933e-05, + "loss": 2.5698, + "step": 475 + }, + { + "epoch": 2.750344827586207, + "grad_norm": 82.11553192138672, + "learning_rate": 1.9175824175824178e-05, + "loss": 2.465, + "step": 500 + }, + { + "epoch": 2.8882758620689657, + "grad_norm": 64.50354766845703, + "learning_rate": 1.9023199023199025e-05, + "loss": 2.6201, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7543859649122807, + "eval_loss": 0.5967164039611816, + "eval_macro_f1": 0.7556428199642172, + "eval_runtime": 1.0678, + "eval_samples_per_second": 587.187, + "eval_steps_per_second": 37.46, + "eval_weighted_f1": 0.7552112588758818, + "step": 546 + }, + { + "epoch": 3.0220689655172412, + "grad_norm": 57.78512954711914, + "learning_rate": 1.887057387057387e-05, + "loss": 2.2413, + "step": 550 + }, + { + "epoch": 3.16, + "grad_norm": 49.69960403442383, + "learning_rate": 1.8717948717948718e-05, + "loss": 2.0975, + "step": 575 + }, + { + "epoch": 3.2979310344827586, + "grad_norm": 70.23579406738281, + "learning_rate": 1.8565323565323568e-05, + "loss": 1.8645, + "step": 600 + }, + { + "epoch": 3.435862068965517, + "grad_norm": 139.99871826171875, + "learning_rate": 1.8412698412698415e-05, + "loss": 1.9666, + "step": 625 + }, + { + "epoch": 3.573793103448276, + "grad_norm": 41.8527717590332, + "learning_rate": 1.826007326007326e-05, + "loss": 1.8205, + "step": 650 + }, + { + "epoch": 3.7117241379310344, + "grad_norm": 109.82010650634766, + "learning_rate": 1.8107448107448108e-05, + "loss": 1.6387, + "step": 675 + }, + { + "epoch": 3.849655172413793, + "grad_norm": 107.30518341064453, + "learning_rate": 1.7954822954822958e-05, + "loss": 2.1496, + "step": 700 + }, + { + "epoch": 3.9875862068965517, + "grad_norm": 83.41844177246094, + "learning_rate": 1.78021978021978e-05, + "loss": 1.9329, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7767145135566188, + "eval_loss": 0.5591686964035034, + "eval_macro_f1": 0.7763996474814833, + "eval_runtime": 1.1489, + "eval_samples_per_second": 545.719, + "eval_steps_per_second": 34.815, + "eval_weighted_f1": 0.7766707740817129, + "step": 728 + }, + { + "epoch": 4.121379310344827, + "grad_norm": 38.20743942260742, + "learning_rate": 1.764957264957265e-05, + "loss": 1.501, + "step": 750 + }, + { + "epoch": 4.259310344827586, + "grad_norm": 101.95227813720703, + "learning_rate": 1.74969474969475e-05, + "loss": 1.5836, + "step": 775 + }, + { + "epoch": 4.397241379310345, + "grad_norm": 48.2017936706543, + "learning_rate": 1.7344322344322345e-05, + "loss": 1.3153, + "step": 800 + }, + { + "epoch": 4.535172413793103, + "grad_norm": 56.81407928466797, + "learning_rate": 1.7191697191697195e-05, + "loss": 1.5429, + "step": 825 + }, + { + "epoch": 4.673103448275862, + "grad_norm": 101.91703796386719, + "learning_rate": 1.703907203907204e-05, + "loss": 1.3026, + "step": 850 + }, + { + "epoch": 4.811034482758621, + "grad_norm": 75.37382507324219, + "learning_rate": 1.688644688644689e-05, + "loss": 1.2796, + "step": 875 + }, + { + "epoch": 4.94896551724138, + "grad_norm": 123.11703491210938, + "learning_rate": 1.6733821733821735e-05, + "loss": 1.4378, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7862838915470495, + "eval_loss": 0.7190022468566895, + "eval_macro_f1": 0.7867678171448537, + "eval_runtime": 1.1303, + "eval_samples_per_second": 554.703, + "eval_steps_per_second": 35.388, + "eval_weighted_f1": 0.7868702555186253, + "step": 910 + }, + { + "epoch": 5.082758620689655, + "grad_norm": 111.80149841308594, + "learning_rate": 1.6581196581196585e-05, + "loss": 1.0587, + "step": 925 + }, + { + "epoch": 5.220689655172414, + "grad_norm": 101.65257263183594, + "learning_rate": 1.642857142857143e-05, + "loss": 1.0306, + "step": 950 + }, + { + "epoch": 5.358620689655172, + "grad_norm": 58.97084045410156, + "learning_rate": 1.627594627594628e-05, + "loss": 1.1596, + "step": 975 + }, + { + "epoch": 5.496551724137931, + "grad_norm": 63.96665573120117, + "learning_rate": 1.6123321123321125e-05, + "loss": 1.2873, + "step": 1000 + }, + { + "epoch": 5.63448275862069, + "grad_norm": 104.3017349243164, + "learning_rate": 1.5970695970695972e-05, + "loss": 1.0208, + "step": 1025 + }, + { + "epoch": 5.772413793103448, + "grad_norm": 108.5999526977539, + "learning_rate": 1.581807081807082e-05, + "loss": 1.2335, + "step": 1050 + }, + { + "epoch": 5.910344827586207, + "grad_norm": 94.29696655273438, + "learning_rate": 1.5665445665445665e-05, + "loss": 1.2207, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7878787878787878, + "eval_loss": 0.6836919784545898, + "eval_macro_f1": 0.7879398797194143, + "eval_runtime": 1.1922, + "eval_samples_per_second": 525.905, + "eval_steps_per_second": 33.551, + "eval_weighted_f1": 0.7879111724300937, + "step": 1092 + }, + { + "epoch": 6.0441379310344825, + "grad_norm": 103.21680450439453, + "learning_rate": 1.5512820512820516e-05, + "loss": 0.9723, + "step": 1100 + }, + { + "epoch": 6.182068965517241, + "grad_norm": 104.1303482055664, + "learning_rate": 1.5360195360195362e-05, + "loss": 0.7962, + "step": 1125 + }, + { + "epoch": 6.32, + "grad_norm": 56.65522003173828, + "learning_rate": 1.5207570207570207e-05, + "loss": 0.7656, + "step": 1150 + }, + { + "epoch": 6.457931034482758, + "grad_norm": 136.6097869873047, + "learning_rate": 1.5054945054945056e-05, + "loss": 0.6219, + "step": 1175 + }, + { + "epoch": 6.595862068965517, + "grad_norm": 116.55268096923828, + "learning_rate": 1.4902319902319904e-05, + "loss": 0.7777, + "step": 1200 + }, + { + "epoch": 6.733793103448276, + "grad_norm": 61.935054779052734, + "learning_rate": 1.4749694749694752e-05, + "loss": 0.8592, + "step": 1225 + }, + { + "epoch": 6.871724137931034, + "grad_norm": 34.48295593261719, + "learning_rate": 1.4597069597069597e-05, + "loss": 0.9455, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.7830940988835726, + "eval_loss": 0.7971011996269226, + "eval_macro_f1": 0.7841797417521805, + "eval_runtime": 1.1578, + "eval_samples_per_second": 541.551, + "eval_steps_per_second": 34.549, + "eval_weighted_f1": 0.7845438631046493, + "step": 1274 + }, + { + "epoch": 7.0055172413793105, + "grad_norm": 178.21438598632812, + "learning_rate": 1.4444444444444446e-05, + "loss": 0.7349, + "step": 1275 + }, + { + "epoch": 7.1434482758620685, + "grad_norm": 101.17195892333984, + "learning_rate": 1.4291819291819293e-05, + "loss": 0.6712, + "step": 1300 + }, + { + "epoch": 7.281379310344827, + "grad_norm": 60.77378845214844, + "learning_rate": 1.4139194139194141e-05, + "loss": 0.6681, + "step": 1325 + }, + { + "epoch": 7.419310344827586, + "grad_norm": 227.9224090576172, + "learning_rate": 1.3986568986568988e-05, + "loss": 0.5815, + "step": 1350 + }, + { + "epoch": 7.557241379310344, + "grad_norm": 93.1969223022461, + "learning_rate": 1.3833943833943834e-05, + "loss": 0.7168, + "step": 1375 + }, + { + "epoch": 7.695172413793103, + "grad_norm": 53.236114501953125, + "learning_rate": 1.3681318681318683e-05, + "loss": 0.8488, + "step": 1400 + }, + { + "epoch": 7.833103448275862, + "grad_norm": 84.6240005493164, + "learning_rate": 1.3528693528693531e-05, + "loss": 0.5962, + "step": 1425 + }, + { + "epoch": 7.971034482758621, + "grad_norm": 70.12305450439453, + "learning_rate": 1.3376068376068376e-05, + "loss": 0.6278, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.7910685805422647, + "eval_loss": 0.9096579551696777, + "eval_macro_f1": 0.7890004378104845, + "eval_runtime": 1.0983, + "eval_samples_per_second": 570.88, + "eval_steps_per_second": 36.42, + "eval_weighted_f1": 0.7894828129259822, + "step": 1456 + }, + { + "epoch": 8.104827586206897, + "grad_norm": 181.92086791992188, + "learning_rate": 1.3223443223443224e-05, + "loss": 0.5441, + "step": 1475 + }, + { + "epoch": 8.242758620689655, + "grad_norm": 91.28218841552734, + "learning_rate": 1.3070818070818073e-05, + "loss": 0.6527, + "step": 1500 + }, + { + "epoch": 8.380689655172414, + "grad_norm": 211.7768096923828, + "learning_rate": 1.291819291819292e-05, + "loss": 0.4916, + "step": 1525 + }, + { + "epoch": 8.518620689655172, + "grad_norm": 32.13935089111328, + "learning_rate": 1.2765567765567766e-05, + "loss": 0.4391, + "step": 1550 + }, + { + "epoch": 8.65655172413793, + "grad_norm": 295.61663818359375, + "learning_rate": 1.2612942612942613e-05, + "loss": 0.5198, + "step": 1575 + }, + { + "epoch": 8.79448275862069, + "grad_norm": 32.63080596923828, + "learning_rate": 1.2460317460317461e-05, + "loss": 0.4442, + "step": 1600 + }, + { + "epoch": 8.932413793103448, + "grad_norm": 83.26895904541016, + "learning_rate": 1.230769230769231e-05, + "loss": 0.6965, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.7974481658692185, + "eval_loss": 1.1046565771102905, + "eval_macro_f1": 0.7953705175062238, + "eval_runtime": 1.1067, + "eval_samples_per_second": 566.552, + "eval_steps_per_second": 36.144, + "eval_weighted_f1": 0.7957420601762267, + "step": 1638 + }, + { + "epoch": 9.066206896551725, + "grad_norm": 0.9721019268035889, + "learning_rate": 1.2155067155067155e-05, + "loss": 0.3916, + "step": 1650 + }, + { + "epoch": 9.204137931034483, + "grad_norm": 246.64122009277344, + "learning_rate": 1.2002442002442003e-05, + "loss": 0.5128, + "step": 1675 + }, + { + "epoch": 9.34206896551724, + "grad_norm": 66.1498794555664, + "learning_rate": 1.1849816849816852e-05, + "loss": 0.2748, + "step": 1700 + }, + { + "epoch": 9.48, + "grad_norm": 18.061092376708984, + "learning_rate": 1.1697191697191698e-05, + "loss": 0.421, + "step": 1725 + }, + { + "epoch": 9.617931034482758, + "grad_norm": 7.748656272888184, + "learning_rate": 1.1544566544566545e-05, + "loss": 0.3691, + "step": 1750 + }, + { + "epoch": 9.755862068965516, + "grad_norm": 139.55853271484375, + "learning_rate": 1.1391941391941392e-05, + "loss": 0.3909, + "step": 1775 + }, + { + "epoch": 9.893793103448276, + "grad_norm": 165.58914184570312, + "learning_rate": 1.123931623931624e-05, + "loss": 0.3689, + "step": 1800 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.7926634768740032, + "eval_loss": 1.3370226621627808, + "eval_macro_f1": 0.7928866612894071, + "eval_runtime": 1.0674, + "eval_samples_per_second": 587.409, + "eval_steps_per_second": 37.474, + "eval_weighted_f1": 0.7930633611359907, + "step": 1820 + }, + { + "epoch": 10.027586206896551, + "grad_norm": 2.90893292427063, + "learning_rate": 1.1086691086691088e-05, + "loss": 0.2536, + "step": 1825 + }, + { + "epoch": 10.16551724137931, + "grad_norm": 69.7308120727539, + "learning_rate": 1.0934065934065933e-05, + "loss": 0.4142, + "step": 1850 + }, + { + "epoch": 10.303448275862069, + "grad_norm": 0.6135218739509583, + "learning_rate": 1.0781440781440782e-05, + "loss": 0.2497, + "step": 1875 + }, + { + "epoch": 10.441379310344828, + "grad_norm": 1.3770861625671387, + "learning_rate": 1.062881562881563e-05, + "loss": 0.2115, + "step": 1900 + }, + { + "epoch": 10.579310344827586, + "grad_norm": 284.72198486328125, + "learning_rate": 1.0476190476190477e-05, + "loss": 0.596, + "step": 1925 + }, + { + "epoch": 10.717241379310344, + "grad_norm": 2.9492316246032715, + "learning_rate": 1.0323565323565325e-05, + "loss": 0.4342, + "step": 1950 + }, + { + "epoch": 10.855172413793104, + "grad_norm": 68.05549621582031, + "learning_rate": 1.0170940170940172e-05, + "loss": 0.2668, + "step": 1975 + }, + { + "epoch": 10.993103448275862, + "grad_norm": 106.35601043701172, + "learning_rate": 1.0018315018315019e-05, + "loss": 0.4008, + "step": 2000 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.773524720893142, + "eval_loss": 1.292651653289795, + "eval_macro_f1": 0.7737703583664469, + "eval_runtime": 1.2845, + "eval_samples_per_second": 488.141, + "eval_steps_per_second": 31.141, + "eval_weighted_f1": 0.7740535430140322, + "step": 2002 + }, + { + "epoch": 11.126896551724139, + "grad_norm": 12.272939682006836, + "learning_rate": 9.865689865689867e-06, + "loss": 0.1852, + "step": 2025 + }, + { + "epoch": 11.264827586206897, + "grad_norm": 0.17907734215259552, + "learning_rate": 9.713064713064714e-06, + "loss": 0.2774, + "step": 2050 + }, + { + "epoch": 11.402758620689655, + "grad_norm": 143.93467712402344, + "learning_rate": 9.560439560439562e-06, + "loss": 0.2405, + "step": 2075 + }, + { + "epoch": 11.540689655172415, + "grad_norm": 139.8834991455078, + "learning_rate": 9.407814407814409e-06, + "loss": 0.2925, + "step": 2100 + }, + { + "epoch": 11.678620689655173, + "grad_norm": 12.48167610168457, + "learning_rate": 9.255189255189256e-06, + "loss": 0.3291, + "step": 2125 + }, + { + "epoch": 11.81655172413793, + "grad_norm": 213.54396057128906, + "learning_rate": 9.102564102564104e-06, + "loss": 0.3811, + "step": 2150 + }, + { + "epoch": 11.95448275862069, + "grad_norm": 3.9226012229919434, + "learning_rate": 8.94993894993895e-06, + "loss": 0.272, + "step": 2175 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8070175438596491, + "eval_loss": 1.3421677350997925, + "eval_macro_f1": 0.8049096987990539, + "eval_runtime": 1.1904, + "eval_samples_per_second": 526.708, + "eval_steps_per_second": 33.602, + "eval_weighted_f1": 0.8049808148241474, + "step": 2184 + }, + { + "epoch": 12.088275862068965, + "grad_norm": 164.14903259277344, + "learning_rate": 8.797313797313797e-06, + "loss": 0.2303, + "step": 2200 + }, + { + "epoch": 12.226206896551725, + "grad_norm": 1.2080793380737305, + "learning_rate": 8.644688644688646e-06, + "loss": 0.2818, + "step": 2225 + }, + { + "epoch": 12.364137931034483, + "grad_norm": 0.2107601761817932, + "learning_rate": 8.492063492063492e-06, + "loss": 0.166, + "step": 2250 + }, + { + "epoch": 12.50206896551724, + "grad_norm": 597.9608154296875, + "learning_rate": 8.33943833943834e-06, + "loss": 0.2335, + "step": 2275 + }, + { + "epoch": 12.64, + "grad_norm": 0.06262794137001038, + "learning_rate": 8.186813186813188e-06, + "loss": 0.3141, + "step": 2300 + }, + { + "epoch": 12.777931034482759, + "grad_norm": 183.90464782714844, + "learning_rate": 8.034188034188036e-06, + "loss": 0.1393, + "step": 2325 + }, + { + "epoch": 12.915862068965517, + "grad_norm": 37.281959533691406, + "learning_rate": 7.881562881562883e-06, + "loss": 0.1174, + "step": 2350 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.79585326953748, + "eval_loss": 1.623354434967041, + "eval_macro_f1": 0.7948055254552955, + "eval_runtime": 1.2964, + "eval_samples_per_second": 483.644, + "eval_steps_per_second": 30.854, + "eval_weighted_f1": 0.7949443345209599, + "step": 2366 + }, + { + "epoch": 13.049655172413793, + "grad_norm": 3.0382513999938965, + "learning_rate": 7.72893772893773e-06, + "loss": 0.1771, + "step": 2375 + }, + { + "epoch": 13.187586206896551, + "grad_norm": 0.5465341210365295, + "learning_rate": 7.576312576312577e-06, + "loss": 0.2488, + "step": 2400 + }, + { + "epoch": 13.32551724137931, + "grad_norm": 44.515506744384766, + "learning_rate": 7.423687423687424e-06, + "loss": 0.2077, + "step": 2425 + }, + { + "epoch": 13.463448275862069, + "grad_norm": 27.11043357849121, + "learning_rate": 7.271062271062271e-06, + "loss": 0.1455, + "step": 2450 + }, + { + "epoch": 13.601379310344827, + "grad_norm": 313.0473937988281, + "learning_rate": 7.1184371184371195e-06, + "loss": 0.1181, + "step": 2475 + }, + { + "epoch": 13.739310344827587, + "grad_norm": 0.022376621142029762, + "learning_rate": 6.965811965811966e-06, + "loss": 0.1992, + "step": 2500 + }, + { + "epoch": 13.877241379310345, + "grad_norm": 0.028782488778233528, + "learning_rate": 6.813186813186814e-06, + "loss": 0.0495, + "step": 2525 + }, + { + "epoch": 14.0, + "eval_accuracy": 0.8261562998405104, + "eval_loss": 1.483603596687317, + "eval_macro_f1": 0.8244676436475006, + "eval_runtime": 1.1393, + "eval_samples_per_second": 550.344, + "eval_steps_per_second": 35.11, + "eval_weighted_f1": 0.82470379163135, + "step": 2548 + } + ], + "logging_steps": 25, + "max_steps": 3640, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/trainer_state.json b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..7c8c4eb00ed78e2e5b4e7e205f4dee9000dfa9ff --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-3094/trainer_state.json @@ -0,0 +1,1091 @@ +{ + "best_global_step": 2548, + "best_metric": 0.8244676436475006, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-00-s43-20260901T164831Z-fbd63439/trainer/checkpoint-2548", + "epoch": 17.0, + "eval_steps": 500, + "global_step": 3094, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13793103448275862, + "grad_norm": 20.084362030029297, + "learning_rate": 1.3186813186813187e-06, + "loss": 4.7288, + "step": 25 + }, + { + "epoch": 0.27586206896551724, + "grad_norm": 5.729361534118652, + "learning_rate": 2.6923076923076923e-06, + "loss": 4.684, + "step": 50 + }, + { + "epoch": 0.41379310344827586, + "grad_norm": 7.694893836975098, + "learning_rate": 4.065934065934066e-06, + "loss": 4.5741, + "step": 75 + }, + { + "epoch": 0.5517241379310345, + "grad_norm": 8.551424980163574, + "learning_rate": 5.43956043956044e-06, + "loss": 4.4816, + "step": 100 + }, + { + "epoch": 0.6896551724137931, + "grad_norm": 35.368587493896484, + "learning_rate": 6.813186813186814e-06, + "loss": 4.5308, + "step": 125 + }, + { + "epoch": 0.8275862068965517, + "grad_norm": 44.43610763549805, + "learning_rate": 8.186813186813188e-06, + "loss": 4.444, + "step": 150 + }, + { + "epoch": 0.9655172413793104, + "grad_norm": 24.932857513427734, + "learning_rate": 9.560439560439562e-06, + "loss": 4.3658, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.4258373205741627, + "eval_loss": 1.0057348012924194, + "eval_macro_f1": 0.3397112691230338, + "eval_runtime": 1.1171, + "eval_samples_per_second": 561.294, + "eval_steps_per_second": 35.808, + "eval_weighted_f1": 0.3524679654710614, + "step": 182 + }, + { + "epoch": 1.0993103448275863, + "grad_norm": 61.700660705566406, + "learning_rate": 1.0934065934065933e-05, + "loss": 4.1429, + "step": 200 + }, + { + "epoch": 1.237241379310345, + "grad_norm": 46.13859558105469, + "learning_rate": 1.230769230769231e-05, + "loss": 3.9467, + "step": 225 + }, + { + "epoch": 1.3751724137931034, + "grad_norm": 46.452125549316406, + "learning_rate": 1.3681318681318683e-05, + "loss": 3.6127, + "step": 250 + }, + { + "epoch": 1.513103448275862, + "grad_norm": 59.73109436035156, + "learning_rate": 1.5054945054945056e-05, + "loss": 3.5063, + "step": 275 + }, + { + "epoch": 1.6510344827586207, + "grad_norm": 145.90650939941406, + "learning_rate": 1.642857142857143e-05, + "loss": 3.3779, + "step": 300 + }, + { + "epoch": 1.7889655172413792, + "grad_norm": 54.86250305175781, + "learning_rate": 1.78021978021978e-05, + "loss": 3.0618, + "step": 325 + }, + { + "epoch": 1.926896551724138, + "grad_norm": 52.89915084838867, + "learning_rate": 1.9175824175824178e-05, + "loss": 3.1781, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.7113237639553429, + "eval_loss": 0.7212908864021301, + "eval_macro_f1": 0.7123417832526394, + "eval_runtime": 1.0768, + "eval_samples_per_second": 582.292, + "eval_steps_per_second": 37.148, + "eval_weighted_f1": 0.712662576565873, + "step": 364 + }, + { + "epoch": 2.0606896551724136, + "grad_norm": 84.83079528808594, + "learning_rate": 1.993894993894994e-05, + "loss": 2.6491, + "step": 375 + }, + { + "epoch": 2.1986206896551725, + "grad_norm": 108.39063262939453, + "learning_rate": 1.9786324786324788e-05, + "loss": 2.7958, + "step": 400 + }, + { + "epoch": 2.336551724137931, + "grad_norm": 96.23841094970703, + "learning_rate": 1.9633699633699634e-05, + "loss": 2.4406, + "step": 425 + }, + { + "epoch": 2.47448275862069, + "grad_norm": 85.79228973388672, + "learning_rate": 1.948107448107448e-05, + "loss": 2.6577, + "step": 450 + }, + { + "epoch": 2.6124137931034483, + "grad_norm": 42.99919509887695, + "learning_rate": 1.932844932844933e-05, + "loss": 2.5698, + "step": 475 + }, + { + "epoch": 2.750344827586207, + "grad_norm": 82.11553192138672, + "learning_rate": 1.9175824175824178e-05, + "loss": 2.465, + "step": 500 + }, + { + "epoch": 2.8882758620689657, + "grad_norm": 64.50354766845703, + "learning_rate": 1.9023199023199025e-05, + "loss": 2.6201, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7543859649122807, + "eval_loss": 0.5967164039611816, + "eval_macro_f1": 0.7556428199642172, + "eval_runtime": 1.0678, + "eval_samples_per_second": 587.187, + "eval_steps_per_second": 37.46, + "eval_weighted_f1": 0.7552112588758818, + "step": 546 + }, + { + "epoch": 3.0220689655172412, + "grad_norm": 57.78512954711914, + "learning_rate": 1.887057387057387e-05, + "loss": 2.2413, + "step": 550 + }, + { + "epoch": 3.16, + "grad_norm": 49.69960403442383, + "learning_rate": 1.8717948717948718e-05, + "loss": 2.0975, + "step": 575 + }, + { + "epoch": 3.2979310344827586, + "grad_norm": 70.23579406738281, + "learning_rate": 1.8565323565323568e-05, + "loss": 1.8645, + "step": 600 + }, + { + "epoch": 3.435862068965517, + "grad_norm": 139.99871826171875, + "learning_rate": 1.8412698412698415e-05, + "loss": 1.9666, + "step": 625 + }, + { + "epoch": 3.573793103448276, + "grad_norm": 41.8527717590332, + "learning_rate": 1.826007326007326e-05, + "loss": 1.8205, + "step": 650 + }, + { + "epoch": 3.7117241379310344, + "grad_norm": 109.82010650634766, + "learning_rate": 1.8107448107448108e-05, + "loss": 1.6387, + "step": 675 + }, + { + "epoch": 3.849655172413793, + "grad_norm": 107.30518341064453, + "learning_rate": 1.7954822954822958e-05, + "loss": 2.1496, + "step": 700 + }, + { + "epoch": 3.9875862068965517, + "grad_norm": 83.41844177246094, + "learning_rate": 1.78021978021978e-05, + "loss": 1.9329, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7767145135566188, + "eval_loss": 0.5591686964035034, + "eval_macro_f1": 0.7763996474814833, + "eval_runtime": 1.1489, + "eval_samples_per_second": 545.719, + "eval_steps_per_second": 34.815, + "eval_weighted_f1": 0.7766707740817129, + "step": 728 + }, + { + "epoch": 4.121379310344827, + "grad_norm": 38.20743942260742, + "learning_rate": 1.764957264957265e-05, + "loss": 1.501, + "step": 750 + }, + { + "epoch": 4.259310344827586, + "grad_norm": 101.95227813720703, + "learning_rate": 1.74969474969475e-05, + "loss": 1.5836, + "step": 775 + }, + { + "epoch": 4.397241379310345, + "grad_norm": 48.2017936706543, + "learning_rate": 1.7344322344322345e-05, + "loss": 1.3153, + "step": 800 + }, + { + "epoch": 4.535172413793103, + "grad_norm": 56.81407928466797, + "learning_rate": 1.7191697191697195e-05, + "loss": 1.5429, + "step": 825 + }, + { + "epoch": 4.673103448275862, + "grad_norm": 101.91703796386719, + "learning_rate": 1.703907203907204e-05, + "loss": 1.3026, + "step": 850 + }, + { + "epoch": 4.811034482758621, + "grad_norm": 75.37382507324219, + "learning_rate": 1.688644688644689e-05, + "loss": 1.2796, + "step": 875 + }, + { + "epoch": 4.94896551724138, + "grad_norm": 123.11703491210938, + "learning_rate": 1.6733821733821735e-05, + "loss": 1.4378, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7862838915470495, + "eval_loss": 0.7190022468566895, + "eval_macro_f1": 0.7867678171448537, + "eval_runtime": 1.1303, + "eval_samples_per_second": 554.703, + "eval_steps_per_second": 35.388, + "eval_weighted_f1": 0.7868702555186253, + "step": 910 + }, + { + "epoch": 5.082758620689655, + "grad_norm": 111.80149841308594, + "learning_rate": 1.6581196581196585e-05, + "loss": 1.0587, + "step": 925 + }, + { + "epoch": 5.220689655172414, + "grad_norm": 101.65257263183594, + "learning_rate": 1.642857142857143e-05, + "loss": 1.0306, + "step": 950 + }, + { + "epoch": 5.358620689655172, + "grad_norm": 58.97084045410156, + "learning_rate": 1.627594627594628e-05, + "loss": 1.1596, + "step": 975 + }, + { + "epoch": 5.496551724137931, + "grad_norm": 63.96665573120117, + "learning_rate": 1.6123321123321125e-05, + "loss": 1.2873, + "step": 1000 + }, + { + "epoch": 5.63448275862069, + "grad_norm": 104.3017349243164, + "learning_rate": 1.5970695970695972e-05, + "loss": 1.0208, + "step": 1025 + }, + { + "epoch": 5.772413793103448, + "grad_norm": 108.5999526977539, + "learning_rate": 1.581807081807082e-05, + "loss": 1.2335, + "step": 1050 + }, + { + "epoch": 5.910344827586207, + "grad_norm": 94.29696655273438, + "learning_rate": 1.5665445665445665e-05, + "loss": 1.2207, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7878787878787878, + "eval_loss": 0.6836919784545898, + "eval_macro_f1": 0.7879398797194143, + "eval_runtime": 1.1922, + "eval_samples_per_second": 525.905, + "eval_steps_per_second": 33.551, + "eval_weighted_f1": 0.7879111724300937, + "step": 1092 + }, + { + "epoch": 6.0441379310344825, + "grad_norm": 103.21680450439453, + "learning_rate": 1.5512820512820516e-05, + "loss": 0.9723, + "step": 1100 + }, + { + "epoch": 6.182068965517241, + "grad_norm": 104.1303482055664, + "learning_rate": 1.5360195360195362e-05, + "loss": 0.7962, + "step": 1125 + }, + { + "epoch": 6.32, + "grad_norm": 56.65522003173828, + "learning_rate": 1.5207570207570207e-05, + "loss": 0.7656, + "step": 1150 + }, + { + "epoch": 6.457931034482758, + "grad_norm": 136.6097869873047, + "learning_rate": 1.5054945054945056e-05, + "loss": 0.6219, + "step": 1175 + }, + { + "epoch": 6.595862068965517, + "grad_norm": 116.55268096923828, + "learning_rate": 1.4902319902319904e-05, + "loss": 0.7777, + "step": 1200 + }, + { + "epoch": 6.733793103448276, + "grad_norm": 61.935054779052734, + "learning_rate": 1.4749694749694752e-05, + "loss": 0.8592, + "step": 1225 + }, + { + "epoch": 6.871724137931034, + "grad_norm": 34.48295593261719, + "learning_rate": 1.4597069597069597e-05, + "loss": 0.9455, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.7830940988835726, + "eval_loss": 0.7971011996269226, + "eval_macro_f1": 0.7841797417521805, + "eval_runtime": 1.1578, + "eval_samples_per_second": 541.551, + "eval_steps_per_second": 34.549, + "eval_weighted_f1": 0.7845438631046493, + "step": 1274 + }, + { + "epoch": 7.0055172413793105, + "grad_norm": 178.21438598632812, + "learning_rate": 1.4444444444444446e-05, + "loss": 0.7349, + "step": 1275 + }, + { + "epoch": 7.1434482758620685, + "grad_norm": 101.17195892333984, + "learning_rate": 1.4291819291819293e-05, + "loss": 0.6712, + "step": 1300 + }, + { + "epoch": 7.281379310344827, + "grad_norm": 60.77378845214844, + "learning_rate": 1.4139194139194141e-05, + "loss": 0.6681, + "step": 1325 + }, + { + "epoch": 7.419310344827586, + "grad_norm": 227.9224090576172, + "learning_rate": 1.3986568986568988e-05, + "loss": 0.5815, + "step": 1350 + }, + { + "epoch": 7.557241379310344, + "grad_norm": 93.1969223022461, + "learning_rate": 1.3833943833943834e-05, + "loss": 0.7168, + "step": 1375 + }, + { + "epoch": 7.695172413793103, + "grad_norm": 53.236114501953125, + "learning_rate": 1.3681318681318683e-05, + "loss": 0.8488, + "step": 1400 + }, + { + "epoch": 7.833103448275862, + "grad_norm": 84.6240005493164, + "learning_rate": 1.3528693528693531e-05, + "loss": 0.5962, + "step": 1425 + }, + { + "epoch": 7.971034482758621, + "grad_norm": 70.12305450439453, + "learning_rate": 1.3376068376068376e-05, + "loss": 0.6278, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.7910685805422647, + "eval_loss": 0.9096579551696777, + "eval_macro_f1": 0.7890004378104845, + "eval_runtime": 1.0983, + "eval_samples_per_second": 570.88, + "eval_steps_per_second": 36.42, + "eval_weighted_f1": 0.7894828129259822, + "step": 1456 + }, + { + "epoch": 8.104827586206897, + "grad_norm": 181.92086791992188, + "learning_rate": 1.3223443223443224e-05, + "loss": 0.5441, + "step": 1475 + }, + { + "epoch": 8.242758620689655, + "grad_norm": 91.28218841552734, + "learning_rate": 1.3070818070818073e-05, + "loss": 0.6527, + "step": 1500 + }, + { + "epoch": 8.380689655172414, + "grad_norm": 211.7768096923828, + "learning_rate": 1.291819291819292e-05, + "loss": 0.4916, + "step": 1525 + }, + { + "epoch": 8.518620689655172, + "grad_norm": 32.13935089111328, + "learning_rate": 1.2765567765567766e-05, + "loss": 0.4391, + "step": 1550 + }, + { + "epoch": 8.65655172413793, + "grad_norm": 295.61663818359375, + "learning_rate": 1.2612942612942613e-05, + "loss": 0.5198, + "step": 1575 + }, + { + "epoch": 8.79448275862069, + "grad_norm": 32.63080596923828, + "learning_rate": 1.2460317460317461e-05, + "loss": 0.4442, + "step": 1600 + }, + { + "epoch": 8.932413793103448, + "grad_norm": 83.26895904541016, + "learning_rate": 1.230769230769231e-05, + "loss": 0.6965, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.7974481658692185, + "eval_loss": 1.1046565771102905, + "eval_macro_f1": 0.7953705175062238, + "eval_runtime": 1.1067, + "eval_samples_per_second": 566.552, + "eval_steps_per_second": 36.144, + "eval_weighted_f1": 0.7957420601762267, + "step": 1638 + }, + { + "epoch": 9.066206896551725, + "grad_norm": 0.9721019268035889, + "learning_rate": 1.2155067155067155e-05, + "loss": 0.3916, + "step": 1650 + }, + { + "epoch": 9.204137931034483, + "grad_norm": 246.64122009277344, + "learning_rate": 1.2002442002442003e-05, + "loss": 0.5128, + "step": 1675 + }, + { + "epoch": 9.34206896551724, + "grad_norm": 66.1498794555664, + "learning_rate": 1.1849816849816852e-05, + "loss": 0.2748, + "step": 1700 + }, + { + "epoch": 9.48, + "grad_norm": 18.061092376708984, + "learning_rate": 1.1697191697191698e-05, + "loss": 0.421, + "step": 1725 + }, + { + "epoch": 9.617931034482758, + "grad_norm": 7.748656272888184, + "learning_rate": 1.1544566544566545e-05, + "loss": 0.3691, + "step": 1750 + }, + { + "epoch": 9.755862068965516, + "grad_norm": 139.55853271484375, + "learning_rate": 1.1391941391941392e-05, + "loss": 0.3909, + "step": 1775 + }, + { + "epoch": 9.893793103448276, + "grad_norm": 165.58914184570312, + "learning_rate": 1.123931623931624e-05, + "loss": 0.3689, + "step": 1800 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.7926634768740032, + "eval_loss": 1.3370226621627808, + "eval_macro_f1": 0.7928866612894071, + "eval_runtime": 1.0674, + "eval_samples_per_second": 587.409, + "eval_steps_per_second": 37.474, + "eval_weighted_f1": 0.7930633611359907, + "step": 1820 + }, + { + "epoch": 10.027586206896551, + "grad_norm": 2.90893292427063, + "learning_rate": 1.1086691086691088e-05, + "loss": 0.2536, + "step": 1825 + }, + { + "epoch": 10.16551724137931, + "grad_norm": 69.7308120727539, + "learning_rate": 1.0934065934065933e-05, + "loss": 0.4142, + "step": 1850 + }, + { + "epoch": 10.303448275862069, + "grad_norm": 0.6135218739509583, + "learning_rate": 1.0781440781440782e-05, + "loss": 0.2497, + "step": 1875 + }, + { + "epoch": 10.441379310344828, + "grad_norm": 1.3770861625671387, + "learning_rate": 1.062881562881563e-05, + "loss": 0.2115, + "step": 1900 + }, + { + "epoch": 10.579310344827586, + "grad_norm": 284.72198486328125, + "learning_rate": 1.0476190476190477e-05, + "loss": 0.596, + "step": 1925 + }, + { + "epoch": 10.717241379310344, + "grad_norm": 2.9492316246032715, + "learning_rate": 1.0323565323565325e-05, + "loss": 0.4342, + "step": 1950 + }, + { + "epoch": 10.855172413793104, + "grad_norm": 68.05549621582031, + "learning_rate": 1.0170940170940172e-05, + "loss": 0.2668, + "step": 1975 + }, + { + "epoch": 10.993103448275862, + "grad_norm": 106.35601043701172, + "learning_rate": 1.0018315018315019e-05, + "loss": 0.4008, + "step": 2000 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.773524720893142, + "eval_loss": 1.292651653289795, + "eval_macro_f1": 0.7737703583664469, + "eval_runtime": 1.2845, + "eval_samples_per_second": 488.141, + "eval_steps_per_second": 31.141, + "eval_weighted_f1": 0.7740535430140322, + "step": 2002 + }, + { + "epoch": 11.126896551724139, + "grad_norm": 12.272939682006836, + "learning_rate": 9.865689865689867e-06, + "loss": 0.1852, + "step": 2025 + }, + { + "epoch": 11.264827586206897, + "grad_norm": 0.17907734215259552, + "learning_rate": 9.713064713064714e-06, + "loss": 0.2774, + "step": 2050 + }, + { + "epoch": 11.402758620689655, + "grad_norm": 143.93467712402344, + "learning_rate": 9.560439560439562e-06, + "loss": 0.2405, + "step": 2075 + }, + { + "epoch": 11.540689655172415, + "grad_norm": 139.8834991455078, + "learning_rate": 9.407814407814409e-06, + "loss": 0.2925, + "step": 2100 + }, + { + "epoch": 11.678620689655173, + "grad_norm": 12.48167610168457, + "learning_rate": 9.255189255189256e-06, + "loss": 0.3291, + "step": 2125 + }, + { + "epoch": 11.81655172413793, + "grad_norm": 213.54396057128906, + "learning_rate": 9.102564102564104e-06, + "loss": 0.3811, + "step": 2150 + }, + { + "epoch": 11.95448275862069, + "grad_norm": 3.9226012229919434, + "learning_rate": 8.94993894993895e-06, + "loss": 0.272, + "step": 2175 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8070175438596491, + "eval_loss": 1.3421677350997925, + "eval_macro_f1": 0.8049096987990539, + "eval_runtime": 1.1904, + "eval_samples_per_second": 526.708, + "eval_steps_per_second": 33.602, + "eval_weighted_f1": 0.8049808148241474, + "step": 2184 + }, + { + "epoch": 12.088275862068965, + "grad_norm": 164.14903259277344, + "learning_rate": 8.797313797313797e-06, + "loss": 0.2303, + "step": 2200 + }, + { + "epoch": 12.226206896551725, + "grad_norm": 1.2080793380737305, + "learning_rate": 8.644688644688646e-06, + "loss": 0.2818, + "step": 2225 + }, + { + "epoch": 12.364137931034483, + "grad_norm": 0.2107601761817932, + "learning_rate": 8.492063492063492e-06, + "loss": 0.166, + "step": 2250 + }, + { + "epoch": 12.50206896551724, + "grad_norm": 597.9608154296875, + "learning_rate": 8.33943833943834e-06, + "loss": 0.2335, + "step": 2275 + }, + { + "epoch": 12.64, + "grad_norm": 0.06262794137001038, + "learning_rate": 8.186813186813188e-06, + "loss": 0.3141, + "step": 2300 + }, + { + "epoch": 12.777931034482759, + "grad_norm": 183.90464782714844, + "learning_rate": 8.034188034188036e-06, + "loss": 0.1393, + "step": 2325 + }, + { + "epoch": 12.915862068965517, + "grad_norm": 37.281959533691406, + "learning_rate": 7.881562881562883e-06, + "loss": 0.1174, + "step": 2350 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.79585326953748, + "eval_loss": 1.623354434967041, + "eval_macro_f1": 0.7948055254552955, + "eval_runtime": 1.2964, + "eval_samples_per_second": 483.644, + "eval_steps_per_second": 30.854, + "eval_weighted_f1": 0.7949443345209599, + "step": 2366 + }, + { + "epoch": 13.049655172413793, + "grad_norm": 3.0382513999938965, + "learning_rate": 7.72893772893773e-06, + "loss": 0.1771, + "step": 2375 + }, + { + "epoch": 13.187586206896551, + "grad_norm": 0.5465341210365295, + "learning_rate": 7.576312576312577e-06, + "loss": 0.2488, + "step": 2400 + }, + { + "epoch": 13.32551724137931, + "grad_norm": 44.515506744384766, + "learning_rate": 7.423687423687424e-06, + "loss": 0.2077, + "step": 2425 + }, + { + "epoch": 13.463448275862069, + "grad_norm": 27.11043357849121, + "learning_rate": 7.271062271062271e-06, + "loss": 0.1455, + "step": 2450 + }, + { + "epoch": 13.601379310344827, + "grad_norm": 313.0473937988281, + "learning_rate": 7.1184371184371195e-06, + "loss": 0.1181, + "step": 2475 + }, + { + "epoch": 13.739310344827587, + "grad_norm": 0.022376621142029762, + "learning_rate": 6.965811965811966e-06, + "loss": 0.1992, + "step": 2500 + }, + { + "epoch": 13.877241379310345, + "grad_norm": 0.028782488778233528, + "learning_rate": 6.813186813186814e-06, + "loss": 0.0495, + "step": 2525 + }, + { + "epoch": 14.0, + "eval_accuracy": 0.8261562998405104, + "eval_loss": 1.483603596687317, + "eval_macro_f1": 0.8244676436475006, + "eval_runtime": 1.1393, + "eval_samples_per_second": 550.344, + "eval_steps_per_second": 35.11, + "eval_weighted_f1": 0.82470379163135, + "step": 2548 + }, + { + "epoch": 14.011034482758621, + "grad_norm": 167.15940856933594, + "learning_rate": 6.66056166056166e-06, + "loss": 0.1243, + "step": 2550 + }, + { + "epoch": 14.148965517241379, + "grad_norm": 0.06580586731433868, + "learning_rate": 6.507936507936509e-06, + "loss": 0.1277, + "step": 2575 + }, + { + "epoch": 14.286896551724137, + "grad_norm": 0.0872478038072586, + "learning_rate": 6.3553113553113555e-06, + "loss": 0.0469, + "step": 2600 + }, + { + "epoch": 14.424827586206897, + "grad_norm": 0.4317820966243744, + "learning_rate": 6.202686202686203e-06, + "loss": 0.274, + "step": 2625 + }, + { + "epoch": 14.562758620689655, + "grad_norm": 0.05667780339717865, + "learning_rate": 6.05006105006105e-06, + "loss": 0.2174, + "step": 2650 + }, + { + "epoch": 14.700689655172415, + "grad_norm": 30.911951065063477, + "learning_rate": 5.897435897435898e-06, + "loss": 0.1159, + "step": 2675 + }, + { + "epoch": 14.838620689655173, + "grad_norm": 0.06982552260160446, + "learning_rate": 5.744810744810745e-06, + "loss": 0.0219, + "step": 2700 + }, + { + "epoch": 14.97655172413793, + "grad_norm": 0.06446123123168945, + "learning_rate": 5.592185592185592e-06, + "loss": 0.2657, + "step": 2725 + }, + { + "epoch": 15.0, + "eval_accuracy": 0.8070175438596491, + "eval_loss": 1.6130914688110352, + "eval_macro_f1": 0.8066404339513583, + "eval_runtime": 1.21, + "eval_samples_per_second": 518.198, + "eval_steps_per_second": 33.059, + "eval_weighted_f1": 0.8065686880812932, + "step": 2730 + }, + { + "epoch": 15.110344827586207, + "grad_norm": 0.00978496577590704, + "learning_rate": 5.43956043956044e-06, + "loss": 0.0828, + "step": 2750 + }, + { + "epoch": 15.248275862068965, + "grad_norm": 0.05871890112757683, + "learning_rate": 5.2869352869352875e-06, + "loss": 0.1256, + "step": 2775 + }, + { + "epoch": 15.386206896551725, + "grad_norm": 0.015405900776386261, + "learning_rate": 5.134310134310134e-06, + "loss": 0.063, + "step": 2800 + }, + { + "epoch": 15.524137931034483, + "grad_norm": 417.1138916015625, + "learning_rate": 4.9816849816849826e-06, + "loss": 0.1314, + "step": 2825 + }, + { + "epoch": 15.662068965517241, + "grad_norm": 2.9514923095703125, + "learning_rate": 4.829059829059829e-06, + "loss": 0.0709, + "step": 2850 + }, + { + "epoch": 15.8, + "grad_norm": 106.02448272705078, + "learning_rate": 4.676434676434677e-06, + "loss": 0.1232, + "step": 2875 + }, + { + "epoch": 15.937931034482759, + "grad_norm": 102.89264678955078, + "learning_rate": 4.523809523809524e-06, + "loss": 0.1052, + "step": 2900 + }, + { + "epoch": 16.0, + "eval_accuracy": 0.8022328548644339, + "eval_loss": 1.6951606273651123, + "eval_macro_f1": 0.8008641234238222, + "eval_runtime": 1.2038, + "eval_samples_per_second": 520.843, + "eval_steps_per_second": 33.228, + "eval_weighted_f1": 0.8008428378737971, + "step": 2912 + }, + { + "epoch": 16.071724137931035, + "grad_norm": 0.15309341251850128, + "learning_rate": 4.371184371184372e-06, + "loss": 0.1384, + "step": 2925 + }, + { + "epoch": 16.209655172413793, + "grad_norm": 1.8501631021499634, + "learning_rate": 4.218559218559219e-06, + "loss": 0.2515, + "step": 2950 + }, + { + "epoch": 16.34758620689655, + "grad_norm": 0.022675300016999245, + "learning_rate": 4.065934065934066e-06, + "loss": 0.0442, + "step": 2975 + }, + { + "epoch": 16.48551724137931, + "grad_norm": 0.04241953790187836, + "learning_rate": 3.913308913308914e-06, + "loss": 0.0218, + "step": 3000 + }, + { + "epoch": 16.623448275862067, + "grad_norm": 0.03193862363696098, + "learning_rate": 3.760683760683761e-06, + "loss": 0.0313, + "step": 3025 + }, + { + "epoch": 16.76137931034483, + "grad_norm": 1.043158769607544, + "learning_rate": 3.6080586080586083e-06, + "loss": 0.101, + "step": 3050 + }, + { + "epoch": 16.899310344827587, + "grad_norm": 0.015490840189158916, + "learning_rate": 3.4554334554334555e-06, + "loss": 0.1246, + "step": 3075 + }, + { + "epoch": 17.0, + "eval_accuracy": 0.8038277511961722, + "eval_loss": 1.728859305381775, + "eval_macro_f1": 0.8024189355890387, + "eval_runtime": 1.1977, + "eval_samples_per_second": 523.508, + "eval_steps_per_second": 33.398, + "eval_weighted_f1": 0.802411980314856, + "step": 3094 + } + ], + "logging_steps": 25, + "max_steps": 3640, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/trainer_state.json b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..80f37260c4835176137cac3e7f620110be8eab54 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197/trainer_state.json @@ -0,0 +1,449 @@ +{ + "best_global_step": 1197, + "best_metric": 0.8295240383812666, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197", + "epoch": 7.0, + "eval_steps": 500, + "global_step": 1197, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.14684287812041116, + "grad_norm": 8.276679992675781, + "learning_rate": 1.4035087719298246e-06, + "loss": 4.7457, + "step": 25 + }, + { + "epoch": 0.2936857562408223, + "grad_norm": 32.80095672607422, + "learning_rate": 2.865497076023392e-06, + "loss": 4.8125, + "step": 50 + }, + { + "epoch": 0.44052863436123346, + "grad_norm": 23.083709716796875, + "learning_rate": 4.3274853801169596e-06, + "loss": 4.6282, + "step": 75 + }, + { + "epoch": 0.5873715124816447, + "grad_norm": 25.51261329650879, + "learning_rate": 5.789473684210527e-06, + "loss": 4.6667, + "step": 100 + }, + { + "epoch": 0.7342143906020558, + "grad_norm": 41.84646224975586, + "learning_rate": 7.251461988304094e-06, + "loss": 4.593, + "step": 125 + }, + { + "epoch": 0.8810572687224669, + "grad_norm": 29.750471115112305, + "learning_rate": 8.713450292397661e-06, + "loss": 4.447, + "step": 150 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.3736654804270463, + "eval_loss": 1.0819200277328491, + "eval_macro_f1": 0.29437285769412447, + "eval_runtime": 1.0765, + "eval_samples_per_second": 522.066, + "eval_steps_per_second": 33.442, + "eval_weighted_f1": 0.2922100579819657, + "step": 171 + }, + { + "epoch": 1.0234948604992657, + "grad_norm": 33.48249816894531, + "learning_rate": 1.017543859649123e-05, + "loss": 4.2602, + "step": 175 + }, + { + "epoch": 1.170337738619677, + "grad_norm": 30.328693389892578, + "learning_rate": 1.1637426900584797e-05, + "loss": 4.2628, + "step": 200 + }, + { + "epoch": 1.3171806167400881, + "grad_norm": 53.562705993652344, + "learning_rate": 1.3099415204678362e-05, + "loss": 3.9631, + "step": 225 + }, + { + "epoch": 1.4640234948604993, + "grad_norm": 64.7986831665039, + "learning_rate": 1.4561403508771931e-05, + "loss": 4.2405, + "step": 250 + }, + { + "epoch": 1.6108663729809103, + "grad_norm": 63.56324005126953, + "learning_rate": 1.60233918128655e-05, + "loss": 3.8987, + "step": 275 + }, + { + "epoch": 1.7577092511013217, + "grad_norm": 60.7152214050293, + "learning_rate": 1.7485380116959064e-05, + "loss": 3.6177, + "step": 300 + }, + { + "epoch": 1.9045521292217327, + "grad_norm": 60.10673141479492, + "learning_rate": 1.894736842105263e-05, + "loss": 3.4457, + "step": 325 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.701067615658363, + "eval_loss": 0.6930170655250549, + "eval_macro_f1": 0.6986001077282614, + "eval_runtime": 1.1239, + "eval_samples_per_second": 500.04, + "eval_steps_per_second": 32.031, + "eval_weighted_f1": 0.6982019631794543, + "step": 342 + }, + { + "epoch": 2.0469897209985315, + "grad_norm": 32.58366012573242, + "learning_rate": 1.99545159194282e-05, + "loss": 3.0666, + "step": 350 + }, + { + "epoch": 2.193832599118943, + "grad_norm": 94.89712524414062, + "learning_rate": 1.9792072774528916e-05, + "loss": 3.0562, + "step": 375 + }, + { + "epoch": 2.340675477239354, + "grad_norm": 50.320068359375, + "learning_rate": 1.962962962962963e-05, + "loss": 2.7734, + "step": 400 + }, + { + "epoch": 2.4875183553597653, + "grad_norm": 93.8223648071289, + "learning_rate": 1.9467186484730346e-05, + "loss": 2.6324, + "step": 425 + }, + { + "epoch": 2.6343612334801763, + "grad_norm": 115.90371704101562, + "learning_rate": 1.930474333983106e-05, + "loss": 2.5846, + "step": 450 + }, + { + "epoch": 2.7812041116005872, + "grad_norm": 57.7183952331543, + "learning_rate": 1.9142300194931776e-05, + "loss": 2.7807, + "step": 475 + }, + { + "epoch": 2.9280469897209986, + "grad_norm": 52.94435501098633, + "learning_rate": 1.897985705003249e-05, + "loss": 2.4913, + "step": 500 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7740213523131673, + "eval_loss": 0.5719616413116455, + "eval_macro_f1": 0.7688477830945377, + "eval_runtime": 1.2147, + "eval_samples_per_second": 462.655, + "eval_steps_per_second": 29.636, + "eval_weighted_f1": 0.7688406479623778, + "step": 513 + }, + { + "epoch": 3.0704845814977975, + "grad_norm": 93.01853942871094, + "learning_rate": 1.8817413905133203e-05, + "loss": 2.4572, + "step": 525 + }, + { + "epoch": 3.2173274596182084, + "grad_norm": 63.457862854003906, + "learning_rate": 1.8654970760233918e-05, + "loss": 2.3328, + "step": 550 + }, + { + "epoch": 3.36417033773862, + "grad_norm": 54.87974548339844, + "learning_rate": 1.8492527615334633e-05, + "loss": 2.0633, + "step": 575 + }, + { + "epoch": 3.511013215859031, + "grad_norm": 84.1146469116211, + "learning_rate": 1.833008447043535e-05, + "loss": 2.1194, + "step": 600 + }, + { + "epoch": 3.657856093979442, + "grad_norm": 111.54753112792969, + "learning_rate": 1.8167641325536064e-05, + "loss": 2.2266, + "step": 625 + }, + { + "epoch": 3.804698972099853, + "grad_norm": 34.23565673828125, + "learning_rate": 1.800519818063678e-05, + "loss": 2.006, + "step": 650 + }, + { + "epoch": 3.951541850220264, + "grad_norm": 96.39556121826172, + "learning_rate": 1.7842755035737494e-05, + "loss": 2.0311, + "step": 675 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7900355871886121, + "eval_loss": 0.4983833432197571, + "eval_macro_f1": 0.7929140773256975, + "eval_runtime": 1.0853, + "eval_samples_per_second": 517.822, + "eval_steps_per_second": 33.17, + "eval_weighted_f1": 0.7928571005707712, + "step": 684 + }, + { + "epoch": 4.093979441997063, + "grad_norm": 71.5910415649414, + "learning_rate": 1.768031189083821e-05, + "loss": 2.0338, + "step": 700 + }, + { + "epoch": 4.240822320117474, + "grad_norm": 110.11136627197266, + "learning_rate": 1.751786874593892e-05, + "loss": 1.7973, + "step": 725 + }, + { + "epoch": 4.387665198237886, + "grad_norm": 70.80055236816406, + "learning_rate": 1.7355425601039636e-05, + "loss": 1.6429, + "step": 750 + }, + { + "epoch": 4.534508076358296, + "grad_norm": 84.0125961303711, + "learning_rate": 1.719298245614035e-05, + "loss": 1.5697, + "step": 775 + }, + { + "epoch": 4.681350954478708, + "grad_norm": 61.205753326416016, + "learning_rate": 1.7030539311241066e-05, + "loss": 1.503, + "step": 800 + }, + { + "epoch": 4.828193832599119, + "grad_norm": 90.11736297607422, + "learning_rate": 1.686809616634178e-05, + "loss": 1.5126, + "step": 825 + }, + { + "epoch": 4.975036710719531, + "grad_norm": 114.29306030273438, + "learning_rate": 1.6705653021442497e-05, + "loss": 1.6841, + "step": 850 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8185053380782918, + "eval_loss": 0.5628421306610107, + "eval_macro_f1": 0.8165680404186829, + "eval_runtime": 1.2562, + "eval_samples_per_second": 447.37, + "eval_steps_per_second": 28.657, + "eval_weighted_f1": 0.8158045472592881, + "step": 855 + }, + { + "epoch": 5.117474302496329, + "grad_norm": 73.93684387207031, + "learning_rate": 1.654320987654321e-05, + "loss": 1.3527, + "step": 875 + }, + { + "epoch": 5.26431718061674, + "grad_norm": 65.49266815185547, + "learning_rate": 1.6380766731643927e-05, + "loss": 1.4592, + "step": 900 + }, + { + "epoch": 5.411160058737151, + "grad_norm": 89.7956771850586, + "learning_rate": 1.621832358674464e-05, + "loss": 1.4047, + "step": 925 + }, + { + "epoch": 5.558002936857562, + "grad_norm": 201.89585876464844, + "learning_rate": 1.6055880441845354e-05, + "loss": 1.2312, + "step": 950 + }, + { + "epoch": 5.704845814977974, + "grad_norm": 67.38060760498047, + "learning_rate": 1.589343729694607e-05, + "loss": 1.3537, + "step": 975 + }, + { + "epoch": 5.851688693098385, + "grad_norm": 129.84413146972656, + "learning_rate": 1.5730994152046787e-05, + "loss": 1.4086, + "step": 1000 + }, + { + "epoch": 5.998531571218796, + "grad_norm": 103.0914535522461, + "learning_rate": 1.55685510071475e-05, + "loss": 1.3403, + "step": 1025 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8256227758007118, + "eval_loss": 0.566539466381073, + "eval_macro_f1": 0.8219632408216548, + "eval_runtime": 1.0559, + "eval_samples_per_second": 532.269, + "eval_steps_per_second": 34.096, + "eval_weighted_f1": 0.8219667601620793, + "step": 1026 + }, + { + "epoch": 6.140969162995595, + "grad_norm": 162.61659240722656, + "learning_rate": 1.5406107862248214e-05, + "loss": 1.1035, + "step": 1050 + }, + { + "epoch": 6.2878120411160054, + "grad_norm": 38.425254821777344, + "learning_rate": 1.524366471734893e-05, + "loss": 1.0021, + "step": 1075 + }, + { + "epoch": 6.434654919236417, + "grad_norm": 116.96482849121094, + "learning_rate": 1.5081221572449645e-05, + "loss": 0.9721, + "step": 1100 + }, + { + "epoch": 6.581497797356828, + "grad_norm": 20.544830322265625, + "learning_rate": 1.4918778427550358e-05, + "loss": 0.75, + "step": 1125 + }, + { + "epoch": 6.72834067547724, + "grad_norm": 50.387577056884766, + "learning_rate": 1.4756335282651073e-05, + "loss": 1.0023, + "step": 1150 + }, + { + "epoch": 6.87518355359765, + "grad_norm": 42.758583068847656, + "learning_rate": 1.4593892137751788e-05, + "loss": 0.9204, + "step": 1175 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8309608540925267, + "eval_loss": 0.5813337564468384, + "eval_macro_f1": 0.8295240383812666, + "eval_runtime": 1.02, + "eval_samples_per_second": 550.964, + "eval_steps_per_second": 35.293, + "eval_weighted_f1": 0.8292286443388538, + "step": 1197 + } + ], + "logging_steps": 25, + "max_steps": 3420, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/trainer_state.json b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..574572378aa3c31eddd430364a0fb227f075f5b2 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1710/trainer_state.json @@ -0,0 +1,629 @@ +{ + "best_global_step": 1197, + "best_metric": 0.8295240383812666, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-01-s42-20260901T152349Z-27bda108/trainer/checkpoint-1197", + "epoch": 10.0, + "eval_steps": 500, + "global_step": 1710, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.14684287812041116, + "grad_norm": 8.276679992675781, + "learning_rate": 1.4035087719298246e-06, + "loss": 4.7457, + "step": 25 + }, + { + "epoch": 0.2936857562408223, + "grad_norm": 32.80095672607422, + "learning_rate": 2.865497076023392e-06, + "loss": 4.8125, + "step": 50 + }, + { + "epoch": 0.44052863436123346, + "grad_norm": 23.083709716796875, + "learning_rate": 4.3274853801169596e-06, + "loss": 4.6282, + "step": 75 + }, + { + "epoch": 0.5873715124816447, + "grad_norm": 25.51261329650879, + "learning_rate": 5.789473684210527e-06, + "loss": 4.6667, + "step": 100 + }, + { + "epoch": 0.7342143906020558, + "grad_norm": 41.84646224975586, + "learning_rate": 7.251461988304094e-06, + "loss": 4.593, + "step": 125 + }, + { + "epoch": 0.8810572687224669, + "grad_norm": 29.750471115112305, + "learning_rate": 8.713450292397661e-06, + "loss": 4.447, + "step": 150 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.3736654804270463, + "eval_loss": 1.0819200277328491, + "eval_macro_f1": 0.29437285769412447, + "eval_runtime": 1.0765, + "eval_samples_per_second": 522.066, + "eval_steps_per_second": 33.442, + "eval_weighted_f1": 0.2922100579819657, + "step": 171 + }, + { + "epoch": 1.0234948604992657, + "grad_norm": 33.48249816894531, + "learning_rate": 1.017543859649123e-05, + "loss": 4.2602, + "step": 175 + }, + { + "epoch": 1.170337738619677, + "grad_norm": 30.328693389892578, + "learning_rate": 1.1637426900584797e-05, + "loss": 4.2628, + "step": 200 + }, + { + "epoch": 1.3171806167400881, + "grad_norm": 53.562705993652344, + "learning_rate": 1.3099415204678362e-05, + "loss": 3.9631, + "step": 225 + }, + { + "epoch": 1.4640234948604993, + "grad_norm": 64.7986831665039, + "learning_rate": 1.4561403508771931e-05, + "loss": 4.2405, + "step": 250 + }, + { + "epoch": 1.6108663729809103, + "grad_norm": 63.56324005126953, + "learning_rate": 1.60233918128655e-05, + "loss": 3.8987, + "step": 275 + }, + { + "epoch": 1.7577092511013217, + "grad_norm": 60.7152214050293, + "learning_rate": 1.7485380116959064e-05, + "loss": 3.6177, + "step": 300 + }, + { + "epoch": 1.9045521292217327, + "grad_norm": 60.10673141479492, + "learning_rate": 1.894736842105263e-05, + "loss": 3.4457, + "step": 325 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.701067615658363, + "eval_loss": 0.6930170655250549, + "eval_macro_f1": 0.6986001077282614, + "eval_runtime": 1.1239, + "eval_samples_per_second": 500.04, + "eval_steps_per_second": 32.031, + "eval_weighted_f1": 0.6982019631794543, + "step": 342 + }, + { + "epoch": 2.0469897209985315, + "grad_norm": 32.58366012573242, + "learning_rate": 1.99545159194282e-05, + "loss": 3.0666, + "step": 350 + }, + { + "epoch": 2.193832599118943, + "grad_norm": 94.89712524414062, + "learning_rate": 1.9792072774528916e-05, + "loss": 3.0562, + "step": 375 + }, + { + "epoch": 2.340675477239354, + "grad_norm": 50.320068359375, + "learning_rate": 1.962962962962963e-05, + "loss": 2.7734, + "step": 400 + }, + { + "epoch": 2.4875183553597653, + "grad_norm": 93.8223648071289, + "learning_rate": 1.9467186484730346e-05, + "loss": 2.6324, + "step": 425 + }, + { + "epoch": 2.6343612334801763, + "grad_norm": 115.90371704101562, + "learning_rate": 1.930474333983106e-05, + "loss": 2.5846, + "step": 450 + }, + { + "epoch": 2.7812041116005872, + "grad_norm": 57.7183952331543, + "learning_rate": 1.9142300194931776e-05, + "loss": 2.7807, + "step": 475 + }, + { + "epoch": 2.9280469897209986, + "grad_norm": 52.94435501098633, + "learning_rate": 1.897985705003249e-05, + "loss": 2.4913, + "step": 500 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7740213523131673, + "eval_loss": 0.5719616413116455, + "eval_macro_f1": 0.7688477830945377, + "eval_runtime": 1.2147, + "eval_samples_per_second": 462.655, + "eval_steps_per_second": 29.636, + "eval_weighted_f1": 0.7688406479623778, + "step": 513 + }, + { + "epoch": 3.0704845814977975, + "grad_norm": 93.01853942871094, + "learning_rate": 1.8817413905133203e-05, + "loss": 2.4572, + "step": 525 + }, + { + "epoch": 3.2173274596182084, + "grad_norm": 63.457862854003906, + "learning_rate": 1.8654970760233918e-05, + "loss": 2.3328, + "step": 550 + }, + { + "epoch": 3.36417033773862, + "grad_norm": 54.87974548339844, + "learning_rate": 1.8492527615334633e-05, + "loss": 2.0633, + "step": 575 + }, + { + "epoch": 3.511013215859031, + "grad_norm": 84.1146469116211, + "learning_rate": 1.833008447043535e-05, + "loss": 2.1194, + "step": 600 + }, + { + "epoch": 3.657856093979442, + "grad_norm": 111.54753112792969, + "learning_rate": 1.8167641325536064e-05, + "loss": 2.2266, + "step": 625 + }, + { + "epoch": 3.804698972099853, + "grad_norm": 34.23565673828125, + "learning_rate": 1.800519818063678e-05, + "loss": 2.006, + "step": 650 + }, + { + "epoch": 3.951541850220264, + "grad_norm": 96.39556121826172, + "learning_rate": 1.7842755035737494e-05, + "loss": 2.0311, + "step": 675 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7900355871886121, + "eval_loss": 0.4983833432197571, + "eval_macro_f1": 0.7929140773256975, + "eval_runtime": 1.0853, + "eval_samples_per_second": 517.822, + "eval_steps_per_second": 33.17, + "eval_weighted_f1": 0.7928571005707712, + "step": 684 + }, + { + "epoch": 4.093979441997063, + "grad_norm": 71.5910415649414, + "learning_rate": 1.768031189083821e-05, + "loss": 2.0338, + "step": 700 + }, + { + "epoch": 4.240822320117474, + "grad_norm": 110.11136627197266, + "learning_rate": 1.751786874593892e-05, + "loss": 1.7973, + "step": 725 + }, + { + "epoch": 4.387665198237886, + "grad_norm": 70.80055236816406, + "learning_rate": 1.7355425601039636e-05, + "loss": 1.6429, + "step": 750 + }, + { + "epoch": 4.534508076358296, + "grad_norm": 84.0125961303711, + "learning_rate": 1.719298245614035e-05, + "loss": 1.5697, + "step": 775 + }, + { + "epoch": 4.681350954478708, + "grad_norm": 61.205753326416016, + "learning_rate": 1.7030539311241066e-05, + "loss": 1.503, + "step": 800 + }, + { + "epoch": 4.828193832599119, + "grad_norm": 90.11736297607422, + "learning_rate": 1.686809616634178e-05, + "loss": 1.5126, + "step": 825 + }, + { + "epoch": 4.975036710719531, + "grad_norm": 114.29306030273438, + "learning_rate": 1.6705653021442497e-05, + "loss": 1.6841, + "step": 850 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8185053380782918, + "eval_loss": 0.5628421306610107, + "eval_macro_f1": 0.8165680404186829, + "eval_runtime": 1.2562, + "eval_samples_per_second": 447.37, + "eval_steps_per_second": 28.657, + "eval_weighted_f1": 0.8158045472592881, + "step": 855 + }, + { + "epoch": 5.117474302496329, + "grad_norm": 73.93684387207031, + "learning_rate": 1.654320987654321e-05, + "loss": 1.3527, + "step": 875 + }, + { + "epoch": 5.26431718061674, + "grad_norm": 65.49266815185547, + "learning_rate": 1.6380766731643927e-05, + "loss": 1.4592, + "step": 900 + }, + { + "epoch": 5.411160058737151, + "grad_norm": 89.7956771850586, + "learning_rate": 1.621832358674464e-05, + "loss": 1.4047, + "step": 925 + }, + { + "epoch": 5.558002936857562, + "grad_norm": 201.89585876464844, + "learning_rate": 1.6055880441845354e-05, + "loss": 1.2312, + "step": 950 + }, + { + "epoch": 5.704845814977974, + "grad_norm": 67.38060760498047, + "learning_rate": 1.589343729694607e-05, + "loss": 1.3537, + "step": 975 + }, + { + "epoch": 5.851688693098385, + "grad_norm": 129.84413146972656, + "learning_rate": 1.5730994152046787e-05, + "loss": 1.4086, + "step": 1000 + }, + { + "epoch": 5.998531571218796, + "grad_norm": 103.0914535522461, + "learning_rate": 1.55685510071475e-05, + "loss": 1.3403, + "step": 1025 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8256227758007118, + "eval_loss": 0.566539466381073, + "eval_macro_f1": 0.8219632408216548, + "eval_runtime": 1.0559, + "eval_samples_per_second": 532.269, + "eval_steps_per_second": 34.096, + "eval_weighted_f1": 0.8219667601620793, + "step": 1026 + }, + { + "epoch": 6.140969162995595, + "grad_norm": 162.61659240722656, + "learning_rate": 1.5406107862248214e-05, + "loss": 1.1035, + "step": 1050 + }, + { + "epoch": 6.2878120411160054, + "grad_norm": 38.425254821777344, + "learning_rate": 1.524366471734893e-05, + "loss": 1.0021, + "step": 1075 + }, + { + "epoch": 6.434654919236417, + "grad_norm": 116.96482849121094, + "learning_rate": 1.5081221572449645e-05, + "loss": 0.9721, + "step": 1100 + }, + { + "epoch": 6.581497797356828, + "grad_norm": 20.544830322265625, + "learning_rate": 1.4918778427550358e-05, + "loss": 0.75, + "step": 1125 + }, + { + "epoch": 6.72834067547724, + "grad_norm": 50.387577056884766, + "learning_rate": 1.4756335282651073e-05, + "loss": 1.0023, + "step": 1150 + }, + { + "epoch": 6.87518355359765, + "grad_norm": 42.758583068847656, + "learning_rate": 1.4593892137751788e-05, + "loss": 0.9204, + "step": 1175 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8309608540925267, + "eval_loss": 0.5813337564468384, + "eval_macro_f1": 0.8295240383812666, + "eval_runtime": 1.02, + "eval_samples_per_second": 550.964, + "eval_steps_per_second": 35.293, + "eval_weighted_f1": 0.8292286443388538, + "step": 1197 + }, + { + "epoch": 7.0176211453744495, + "grad_norm": 169.1999053955078, + "learning_rate": 1.4431448992852503e-05, + "loss": 1.1155, + "step": 1200 + }, + { + "epoch": 7.164464023494861, + "grad_norm": 72.44261932373047, + "learning_rate": 1.4269005847953217e-05, + "loss": 0.6562, + "step": 1225 + }, + { + "epoch": 7.311306901615271, + "grad_norm": 70.4012451171875, + "learning_rate": 1.4106562703053932e-05, + "loss": 0.7291, + "step": 1250 + }, + { + "epoch": 7.458149779735683, + "grad_norm": 46.89229965209961, + "learning_rate": 1.3944119558154647e-05, + "loss": 0.6769, + "step": 1275 + }, + { + "epoch": 7.604992657856094, + "grad_norm": 80.89586639404297, + "learning_rate": 1.3781676413255362e-05, + "loss": 0.7291, + "step": 1300 + }, + { + "epoch": 7.751835535976505, + "grad_norm": 144.99386596679688, + "learning_rate": 1.3619233268356076e-05, + "loss": 0.7643, + "step": 1325 + }, + { + "epoch": 7.898678414096916, + "grad_norm": 59.30048370361328, + "learning_rate": 1.345679012345679e-05, + "loss": 0.6184, + "step": 1350 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8202846975088968, + "eval_loss": 0.6676938533782959, + "eval_macro_f1": 0.8193423338206208, + "eval_runtime": 1.2476, + "eval_samples_per_second": 450.467, + "eval_steps_per_second": 28.856, + "eval_weighted_f1": 0.81896413525029, + "step": 1368 + }, + { + "epoch": 8.041116005873715, + "grad_norm": 107.24308776855469, + "learning_rate": 1.3294346978557506e-05, + "loss": 0.7401, + "step": 1375 + }, + { + "epoch": 8.187958883994126, + "grad_norm": 8.43368148803711, + "learning_rate": 1.3131903833658221e-05, + "loss": 0.5523, + "step": 1400 + }, + { + "epoch": 8.334801762114537, + "grad_norm": 91.36034393310547, + "learning_rate": 1.2969460688758935e-05, + "loss": 0.5694, + "step": 1425 + }, + { + "epoch": 8.481644640234949, + "grad_norm": 76.99760437011719, + "learning_rate": 1.280701754385965e-05, + "loss": 0.5402, + "step": 1450 + }, + { + "epoch": 8.62848751835536, + "grad_norm": 19.6176815032959, + "learning_rate": 1.2644574398960365e-05, + "loss": 0.454, + "step": 1475 + }, + { + "epoch": 8.775330396475772, + "grad_norm": 89.50540924072266, + "learning_rate": 1.248213125406108e-05, + "loss": 0.7104, + "step": 1500 + }, + { + "epoch": 8.922173274596181, + "grad_norm": 54.60981369018555, + "learning_rate": 1.2319688109161793e-05, + "loss": 0.5546, + "step": 1525 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8185053380782918, + "eval_loss": 0.7808212041854858, + "eval_macro_f1": 0.8177603560755428, + "eval_runtime": 1.2467, + "eval_samples_per_second": 450.784, + "eval_steps_per_second": 28.876, + "eval_weighted_f1": 0.8177147662693375, + "step": 1539 + }, + { + "epoch": 9.06461086637298, + "grad_norm": 49.732154846191406, + "learning_rate": 1.2157244964262509e-05, + "loss": 0.6509, + "step": 1550 + }, + { + "epoch": 9.211453744493392, + "grad_norm": 120.88252258300781, + "learning_rate": 1.1994801819363224e-05, + "loss": 0.4404, + "step": 1575 + }, + { + "epoch": 9.358296622613803, + "grad_norm": 16.407711029052734, + "learning_rate": 1.1832358674463939e-05, + "loss": 0.2951, + "step": 1600 + }, + { + "epoch": 9.505139500734215, + "grad_norm": 194.71717834472656, + "learning_rate": 1.1669915529564652e-05, + "loss": 0.4514, + "step": 1625 + }, + { + "epoch": 9.651982378854626, + "grad_norm": 97.11231231689453, + "learning_rate": 1.1507472384665367e-05, + "loss": 0.3223, + "step": 1650 + }, + { + "epoch": 9.798825256975036, + "grad_norm": 35.31584167480469, + "learning_rate": 1.1345029239766083e-05, + "loss": 0.6108, + "step": 1675 + }, + { + "epoch": 9.945668135095447, + "grad_norm": 122.03335571289062, + "learning_rate": 1.11825860948668e-05, + "loss": 0.492, + "step": 1700 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8274021352313167, + "eval_loss": 0.8296598792076111, + "eval_macro_f1": 0.8281855148698446, + "eval_runtime": 1.2851, + "eval_samples_per_second": 437.305, + "eval_steps_per_second": 28.012, + "eval_weighted_f1": 0.8277867192540188, + "step": 1710 + } + ], + "logging_steps": 25, + "max_steps": 3420, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/trainer_state.json b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..df9001ea7010abf384e31be7b463ace0d5fc1b2a --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197/trainer_state.json @@ -0,0 +1,449 @@ +{ + "best_global_step": 1197, + "best_metric": 0.8120160172108976, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197", + "epoch": 7.0, + "eval_steps": 500, + "global_step": 1197, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.14684287812041116, + "grad_norm": 19.898252487182617, + "learning_rate": 1.4035087719298246e-06, + "loss": 4.8259, + "step": 25 + }, + { + "epoch": 0.2936857562408223, + "grad_norm": 28.20109748840332, + "learning_rate": 2.865497076023392e-06, + "loss": 4.6946, + "step": 50 + }, + { + "epoch": 0.44052863436123346, + "grad_norm": 19.910593032836914, + "learning_rate": 4.3274853801169596e-06, + "loss": 4.5263, + "step": 75 + }, + { + "epoch": 0.5873715124816447, + "grad_norm": 23.893070220947266, + "learning_rate": 5.789473684210527e-06, + "loss": 4.4799, + "step": 100 + }, + { + "epoch": 0.7342143906020558, + "grad_norm": 21.93558120727539, + "learning_rate": 7.251461988304094e-06, + "loss": 4.5204, + "step": 125 + }, + { + "epoch": 0.8810572687224669, + "grad_norm": 20.283721923828125, + "learning_rate": 8.713450292397661e-06, + "loss": 4.3808, + "step": 150 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.3274021352313167, + "eval_loss": 1.1216636896133423, + "eval_macro_f1": 0.1680264699233441, + "eval_runtime": 1.1499, + "eval_samples_per_second": 488.741, + "eval_steps_per_second": 31.307, + "eval_weighted_f1": 0.16437457594901667, + "step": 171 + }, + { + "epoch": 1.0234948604992657, + "grad_norm": 18.82137680053711, + "learning_rate": 1.017543859649123e-05, + "loss": 4.3132, + "step": 175 + }, + { + "epoch": 1.170337738619677, + "grad_norm": 39.67106628417969, + "learning_rate": 1.1637426900584797e-05, + "loss": 4.3691, + "step": 200 + }, + { + "epoch": 1.3171806167400881, + "grad_norm": 25.474512100219727, + "learning_rate": 1.3099415204678362e-05, + "loss": 4.3744, + "step": 225 + }, + { + "epoch": 1.4640234948604993, + "grad_norm": 29.966018676757812, + "learning_rate": 1.4561403508771931e-05, + "loss": 4.1576, + "step": 250 + }, + { + "epoch": 1.6108663729809103, + "grad_norm": 38.67800521850586, + "learning_rate": 1.60233918128655e-05, + "loss": 3.9165, + "step": 275 + }, + { + "epoch": 1.7577092511013217, + "grad_norm": 120.15225219726562, + "learning_rate": 1.7485380116959064e-05, + "loss": 3.5492, + "step": 300 + }, + { + "epoch": 1.9045521292217327, + "grad_norm": 145.66622924804688, + "learning_rate": 1.894736842105263e-05, + "loss": 3.2797, + "step": 325 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6814946619217082, + "eval_loss": 0.8266801238059998, + "eval_macro_f1": 0.6695599664073812, + "eval_runtime": 1.0467, + "eval_samples_per_second": 536.94, + "eval_steps_per_second": 34.395, + "eval_weighted_f1": 0.6707705283670244, + "step": 342 + }, + { + "epoch": 2.0469897209985315, + "grad_norm": 61.46903610229492, + "learning_rate": 1.99545159194282e-05, + "loss": 3.1801, + "step": 350 + }, + { + "epoch": 2.193832599118943, + "grad_norm": 94.04988861083984, + "learning_rate": 1.9792072774528916e-05, + "loss": 2.9369, + "step": 375 + }, + { + "epoch": 2.340675477239354, + "grad_norm": 131.471923828125, + "learning_rate": 1.962962962962963e-05, + "loss": 2.8236, + "step": 400 + }, + { + "epoch": 2.4875183553597653, + "grad_norm": 176.6563262939453, + "learning_rate": 1.9467186484730346e-05, + "loss": 2.9867, + "step": 425 + }, + { + "epoch": 2.6343612334801763, + "grad_norm": 59.89003372192383, + "learning_rate": 1.930474333983106e-05, + "loss": 2.9233, + "step": 450 + }, + { + "epoch": 2.7812041116005872, + "grad_norm": 68.92916870117188, + "learning_rate": 1.9142300194931776e-05, + "loss": 2.8271, + "step": 475 + }, + { + "epoch": 2.9280469897209986, + "grad_norm": 103.18282318115234, + "learning_rate": 1.897985705003249e-05, + "loss": 2.7263, + "step": 500 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7686832740213523, + "eval_loss": 0.6141501069068909, + "eval_macro_f1": 0.7704908189767045, + "eval_runtime": 1.1127, + "eval_samples_per_second": 505.085, + "eval_steps_per_second": 32.354, + "eval_weighted_f1": 0.7697335190679598, + "step": 513 + }, + { + "epoch": 3.0704845814977975, + "grad_norm": 106.83403778076172, + "learning_rate": 1.8817413905133203e-05, + "loss": 2.5354, + "step": 525 + }, + { + "epoch": 3.2173274596182084, + "grad_norm": 103.26848602294922, + "learning_rate": 1.8654970760233918e-05, + "loss": 2.4149, + "step": 550 + }, + { + "epoch": 3.36417033773862, + "grad_norm": 84.16307067871094, + "learning_rate": 1.8492527615334633e-05, + "loss": 2.1616, + "step": 575 + }, + { + "epoch": 3.511013215859031, + "grad_norm": 51.95928955078125, + "learning_rate": 1.833008447043535e-05, + "loss": 2.2616, + "step": 600 + }, + { + "epoch": 3.657856093979442, + "grad_norm": 48.234153747558594, + "learning_rate": 1.8167641325536064e-05, + "loss": 1.916, + "step": 625 + }, + { + "epoch": 3.804698972099853, + "grad_norm": 97.79098510742188, + "learning_rate": 1.800519818063678e-05, + "loss": 2.1566, + "step": 650 + }, + { + "epoch": 3.951541850220264, + "grad_norm": 83.2856674194336, + "learning_rate": 1.7842755035737494e-05, + "loss": 2.0654, + "step": 675 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.791814946619217, + "eval_loss": 0.5504909753799438, + "eval_macro_f1": 0.7894521890156985, + "eval_runtime": 1.1111, + "eval_samples_per_second": 505.812, + "eval_steps_per_second": 32.401, + "eval_weighted_f1": 0.7894752915277294, + "step": 684 + }, + { + "epoch": 4.093979441997063, + "grad_norm": 51.651371002197266, + "learning_rate": 1.768031189083821e-05, + "loss": 1.7594, + "step": 700 + }, + { + "epoch": 4.240822320117474, + "grad_norm": 96.34539794921875, + "learning_rate": 1.751786874593892e-05, + "loss": 1.7873, + "step": 725 + }, + { + "epoch": 4.387665198237886, + "grad_norm": 113.27198791503906, + "learning_rate": 1.7355425601039636e-05, + "loss": 1.6555, + "step": 750 + }, + { + "epoch": 4.534508076358296, + "grad_norm": 52.46745681762695, + "learning_rate": 1.719298245614035e-05, + "loss": 1.5451, + "step": 775 + }, + { + "epoch": 4.681350954478708, + "grad_norm": 81.5442886352539, + "learning_rate": 1.7030539311241066e-05, + "loss": 1.5964, + "step": 800 + }, + { + "epoch": 4.828193832599119, + "grad_norm": 155.65516662597656, + "learning_rate": 1.686809616634178e-05, + "loss": 1.7224, + "step": 825 + }, + { + "epoch": 4.975036710719531, + "grad_norm": 48.24635314941406, + "learning_rate": 1.6705653021442497e-05, + "loss": 1.7126, + "step": 850 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8078291814946619, + "eval_loss": 0.5076857805252075, + "eval_macro_f1": 0.8081435252166961, + "eval_runtime": 1.0205, + "eval_samples_per_second": 550.698, + "eval_steps_per_second": 35.276, + "eval_weighted_f1": 0.8077148227916391, + "step": 855 + }, + { + "epoch": 5.117474302496329, + "grad_norm": 90.79500579833984, + "learning_rate": 1.654320987654321e-05, + "loss": 1.4157, + "step": 875 + }, + { + "epoch": 5.26431718061674, + "grad_norm": 74.45726013183594, + "learning_rate": 1.6380766731643927e-05, + "loss": 1.3348, + "step": 900 + }, + { + "epoch": 5.411160058737151, + "grad_norm": 72.0616455078125, + "learning_rate": 1.621832358674464e-05, + "loss": 1.1797, + "step": 925 + }, + { + "epoch": 5.558002936857562, + "grad_norm": 83.30821228027344, + "learning_rate": 1.6055880441845354e-05, + "loss": 1.0697, + "step": 950 + }, + { + "epoch": 5.704845814977974, + "grad_norm": 72.64716339111328, + "learning_rate": 1.589343729694607e-05, + "loss": 1.3893, + "step": 975 + }, + { + "epoch": 5.851688693098385, + "grad_norm": 99.64823150634766, + "learning_rate": 1.5730994152046787e-05, + "loss": 1.3584, + "step": 1000 + }, + { + "epoch": 5.998531571218796, + "grad_norm": 97.09939575195312, + "learning_rate": 1.55685510071475e-05, + "loss": 1.5412, + "step": 1025 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8078291814946619, + "eval_loss": 0.5981741547584534, + "eval_macro_f1": 0.8021303737721648, + "eval_runtime": 1.1109, + "eval_samples_per_second": 505.911, + "eval_steps_per_second": 32.407, + "eval_weighted_f1": 0.802081094663556, + "step": 1026 + }, + { + "epoch": 6.140969162995595, + "grad_norm": 95.87456512451172, + "learning_rate": 1.5406107862248214e-05, + "loss": 1.0143, + "step": 1050 + }, + { + "epoch": 6.2878120411160054, + "grad_norm": 68.2560806274414, + "learning_rate": 1.524366471734893e-05, + "loss": 0.8491, + "step": 1075 + }, + { + "epoch": 6.434654919236417, + "grad_norm": 77.93789672851562, + "learning_rate": 1.5081221572449645e-05, + "loss": 1.0485, + "step": 1100 + }, + { + "epoch": 6.581497797356828, + "grad_norm": 16.324548721313477, + "learning_rate": 1.4918778427550358e-05, + "loss": 0.935, + "step": 1125 + }, + { + "epoch": 6.72834067547724, + "grad_norm": 65.43302917480469, + "learning_rate": 1.4756335282651073e-05, + "loss": 1.0821, + "step": 1150 + }, + { + "epoch": 6.87518355359765, + "grad_norm": 124.80599975585938, + "learning_rate": 1.4593892137751788e-05, + "loss": 0.9082, + "step": 1175 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8113879003558719, + "eval_loss": 0.6220943331718445, + "eval_macro_f1": 0.8120160172108976, + "eval_runtime": 1.0895, + "eval_samples_per_second": 515.83, + "eval_steps_per_second": 33.042, + "eval_weighted_f1": 0.8116930414813697, + "step": 1197 + } + ], + "logging_steps": 25, + "max_steps": 3420, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/trainer_state.json b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..1e493138507cca6d4bec6cd9e6fbe782a693b459 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1710/trainer_state.json @@ -0,0 +1,629 @@ +{ + "best_global_step": 1197, + "best_metric": 0.8120160172108976, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-01-s43-20260901T170609Z-fbd63439/trainer/checkpoint-1197", + "epoch": 10.0, + "eval_steps": 500, + "global_step": 1710, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.14684287812041116, + "grad_norm": 19.898252487182617, + "learning_rate": 1.4035087719298246e-06, + "loss": 4.8259, + "step": 25 + }, + { + "epoch": 0.2936857562408223, + "grad_norm": 28.20109748840332, + "learning_rate": 2.865497076023392e-06, + "loss": 4.6946, + "step": 50 + }, + { + "epoch": 0.44052863436123346, + "grad_norm": 19.910593032836914, + "learning_rate": 4.3274853801169596e-06, + "loss": 4.5263, + "step": 75 + }, + { + "epoch": 0.5873715124816447, + "grad_norm": 23.893070220947266, + "learning_rate": 5.789473684210527e-06, + "loss": 4.4799, + "step": 100 + }, + { + "epoch": 0.7342143906020558, + "grad_norm": 21.93558120727539, + "learning_rate": 7.251461988304094e-06, + "loss": 4.5204, + "step": 125 + }, + { + "epoch": 0.8810572687224669, + "grad_norm": 20.283721923828125, + "learning_rate": 8.713450292397661e-06, + "loss": 4.3808, + "step": 150 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.3274021352313167, + "eval_loss": 1.1216636896133423, + "eval_macro_f1": 0.1680264699233441, + "eval_runtime": 1.1499, + "eval_samples_per_second": 488.741, + "eval_steps_per_second": 31.307, + "eval_weighted_f1": 0.16437457594901667, + "step": 171 + }, + { + "epoch": 1.0234948604992657, + "grad_norm": 18.82137680053711, + "learning_rate": 1.017543859649123e-05, + "loss": 4.3132, + "step": 175 + }, + { + "epoch": 1.170337738619677, + "grad_norm": 39.67106628417969, + "learning_rate": 1.1637426900584797e-05, + "loss": 4.3691, + "step": 200 + }, + { + "epoch": 1.3171806167400881, + "grad_norm": 25.474512100219727, + "learning_rate": 1.3099415204678362e-05, + "loss": 4.3744, + "step": 225 + }, + { + "epoch": 1.4640234948604993, + "grad_norm": 29.966018676757812, + "learning_rate": 1.4561403508771931e-05, + "loss": 4.1576, + "step": 250 + }, + { + "epoch": 1.6108663729809103, + "grad_norm": 38.67800521850586, + "learning_rate": 1.60233918128655e-05, + "loss": 3.9165, + "step": 275 + }, + { + "epoch": 1.7577092511013217, + "grad_norm": 120.15225219726562, + "learning_rate": 1.7485380116959064e-05, + "loss": 3.5492, + "step": 300 + }, + { + "epoch": 1.9045521292217327, + "grad_norm": 145.66622924804688, + "learning_rate": 1.894736842105263e-05, + "loss": 3.2797, + "step": 325 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6814946619217082, + "eval_loss": 0.8266801238059998, + "eval_macro_f1": 0.6695599664073812, + "eval_runtime": 1.0467, + "eval_samples_per_second": 536.94, + "eval_steps_per_second": 34.395, + "eval_weighted_f1": 0.6707705283670244, + "step": 342 + }, + { + "epoch": 2.0469897209985315, + "grad_norm": 61.46903610229492, + "learning_rate": 1.99545159194282e-05, + "loss": 3.1801, + "step": 350 + }, + { + "epoch": 2.193832599118943, + "grad_norm": 94.04988861083984, + "learning_rate": 1.9792072774528916e-05, + "loss": 2.9369, + "step": 375 + }, + { + "epoch": 2.340675477239354, + "grad_norm": 131.471923828125, + "learning_rate": 1.962962962962963e-05, + "loss": 2.8236, + "step": 400 + }, + { + "epoch": 2.4875183553597653, + "grad_norm": 176.6563262939453, + "learning_rate": 1.9467186484730346e-05, + "loss": 2.9867, + "step": 425 + }, + { + "epoch": 2.6343612334801763, + "grad_norm": 59.89003372192383, + "learning_rate": 1.930474333983106e-05, + "loss": 2.9233, + "step": 450 + }, + { + "epoch": 2.7812041116005872, + "grad_norm": 68.92916870117188, + "learning_rate": 1.9142300194931776e-05, + "loss": 2.8271, + "step": 475 + }, + { + "epoch": 2.9280469897209986, + "grad_norm": 103.18282318115234, + "learning_rate": 1.897985705003249e-05, + "loss": 2.7263, + "step": 500 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7686832740213523, + "eval_loss": 0.6141501069068909, + "eval_macro_f1": 0.7704908189767045, + "eval_runtime": 1.1127, + "eval_samples_per_second": 505.085, + "eval_steps_per_second": 32.354, + "eval_weighted_f1": 0.7697335190679598, + "step": 513 + }, + { + "epoch": 3.0704845814977975, + "grad_norm": 106.83403778076172, + "learning_rate": 1.8817413905133203e-05, + "loss": 2.5354, + "step": 525 + }, + { + "epoch": 3.2173274596182084, + "grad_norm": 103.26848602294922, + "learning_rate": 1.8654970760233918e-05, + "loss": 2.4149, + "step": 550 + }, + { + "epoch": 3.36417033773862, + "grad_norm": 84.16307067871094, + "learning_rate": 1.8492527615334633e-05, + "loss": 2.1616, + "step": 575 + }, + { + "epoch": 3.511013215859031, + "grad_norm": 51.95928955078125, + "learning_rate": 1.833008447043535e-05, + "loss": 2.2616, + "step": 600 + }, + { + "epoch": 3.657856093979442, + "grad_norm": 48.234153747558594, + "learning_rate": 1.8167641325536064e-05, + "loss": 1.916, + "step": 625 + }, + { + "epoch": 3.804698972099853, + "grad_norm": 97.79098510742188, + "learning_rate": 1.800519818063678e-05, + "loss": 2.1566, + "step": 650 + }, + { + "epoch": 3.951541850220264, + "grad_norm": 83.2856674194336, + "learning_rate": 1.7842755035737494e-05, + "loss": 2.0654, + "step": 675 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.791814946619217, + "eval_loss": 0.5504909753799438, + "eval_macro_f1": 0.7894521890156985, + "eval_runtime": 1.1111, + "eval_samples_per_second": 505.812, + "eval_steps_per_second": 32.401, + "eval_weighted_f1": 0.7894752915277294, + "step": 684 + }, + { + "epoch": 4.093979441997063, + "grad_norm": 51.651371002197266, + "learning_rate": 1.768031189083821e-05, + "loss": 1.7594, + "step": 700 + }, + { + "epoch": 4.240822320117474, + "grad_norm": 96.34539794921875, + "learning_rate": 1.751786874593892e-05, + "loss": 1.7873, + "step": 725 + }, + { + "epoch": 4.387665198237886, + "grad_norm": 113.27198791503906, + "learning_rate": 1.7355425601039636e-05, + "loss": 1.6555, + "step": 750 + }, + { + "epoch": 4.534508076358296, + "grad_norm": 52.46745681762695, + "learning_rate": 1.719298245614035e-05, + "loss": 1.5451, + "step": 775 + }, + { + "epoch": 4.681350954478708, + "grad_norm": 81.5442886352539, + "learning_rate": 1.7030539311241066e-05, + "loss": 1.5964, + "step": 800 + }, + { + "epoch": 4.828193832599119, + "grad_norm": 155.65516662597656, + "learning_rate": 1.686809616634178e-05, + "loss": 1.7224, + "step": 825 + }, + { + "epoch": 4.975036710719531, + "grad_norm": 48.24635314941406, + "learning_rate": 1.6705653021442497e-05, + "loss": 1.7126, + "step": 850 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8078291814946619, + "eval_loss": 0.5076857805252075, + "eval_macro_f1": 0.8081435252166961, + "eval_runtime": 1.0205, + "eval_samples_per_second": 550.698, + "eval_steps_per_second": 35.276, + "eval_weighted_f1": 0.8077148227916391, + "step": 855 + }, + { + "epoch": 5.117474302496329, + "grad_norm": 90.79500579833984, + "learning_rate": 1.654320987654321e-05, + "loss": 1.4157, + "step": 875 + }, + { + "epoch": 5.26431718061674, + "grad_norm": 74.45726013183594, + "learning_rate": 1.6380766731643927e-05, + "loss": 1.3348, + "step": 900 + }, + { + "epoch": 5.411160058737151, + "grad_norm": 72.0616455078125, + "learning_rate": 1.621832358674464e-05, + "loss": 1.1797, + "step": 925 + }, + { + "epoch": 5.558002936857562, + "grad_norm": 83.30821228027344, + "learning_rate": 1.6055880441845354e-05, + "loss": 1.0697, + "step": 950 + }, + { + "epoch": 5.704845814977974, + "grad_norm": 72.64716339111328, + "learning_rate": 1.589343729694607e-05, + "loss": 1.3893, + "step": 975 + }, + { + "epoch": 5.851688693098385, + "grad_norm": 99.64823150634766, + "learning_rate": 1.5730994152046787e-05, + "loss": 1.3584, + "step": 1000 + }, + { + "epoch": 5.998531571218796, + "grad_norm": 97.09939575195312, + "learning_rate": 1.55685510071475e-05, + "loss": 1.5412, + "step": 1025 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8078291814946619, + "eval_loss": 0.5981741547584534, + "eval_macro_f1": 0.8021303737721648, + "eval_runtime": 1.1109, + "eval_samples_per_second": 505.911, + "eval_steps_per_second": 32.407, + "eval_weighted_f1": 0.802081094663556, + "step": 1026 + }, + { + "epoch": 6.140969162995595, + "grad_norm": 95.87456512451172, + "learning_rate": 1.5406107862248214e-05, + "loss": 1.0143, + "step": 1050 + }, + { + "epoch": 6.2878120411160054, + "grad_norm": 68.2560806274414, + "learning_rate": 1.524366471734893e-05, + "loss": 0.8491, + "step": 1075 + }, + { + "epoch": 6.434654919236417, + "grad_norm": 77.93789672851562, + "learning_rate": 1.5081221572449645e-05, + "loss": 1.0485, + "step": 1100 + }, + { + "epoch": 6.581497797356828, + "grad_norm": 16.324548721313477, + "learning_rate": 1.4918778427550358e-05, + "loss": 0.935, + "step": 1125 + }, + { + "epoch": 6.72834067547724, + "grad_norm": 65.43302917480469, + "learning_rate": 1.4756335282651073e-05, + "loss": 1.0821, + "step": 1150 + }, + { + "epoch": 6.87518355359765, + "grad_norm": 124.80599975585938, + "learning_rate": 1.4593892137751788e-05, + "loss": 0.9082, + "step": 1175 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8113879003558719, + "eval_loss": 0.6220943331718445, + "eval_macro_f1": 0.8120160172108976, + "eval_runtime": 1.0895, + "eval_samples_per_second": 515.83, + "eval_steps_per_second": 33.042, + "eval_weighted_f1": 0.8116930414813697, + "step": 1197 + }, + { + "epoch": 7.0176211453744495, + "grad_norm": 71.75617980957031, + "learning_rate": 1.4431448992852503e-05, + "loss": 1.1683, + "step": 1200 + }, + { + "epoch": 7.164464023494861, + "grad_norm": 99.44718170166016, + "learning_rate": 1.4269005847953217e-05, + "loss": 0.7722, + "step": 1225 + }, + { + "epoch": 7.311306901615271, + "grad_norm": 85.54470825195312, + "learning_rate": 1.4106562703053932e-05, + "loss": 0.7766, + "step": 1250 + }, + { + "epoch": 7.458149779735683, + "grad_norm": 21.999500274658203, + "learning_rate": 1.3944119558154647e-05, + "loss": 0.6444, + "step": 1275 + }, + { + "epoch": 7.604992657856094, + "grad_norm": 89.1247787475586, + "learning_rate": 1.3781676413255362e-05, + "loss": 0.6769, + "step": 1300 + }, + { + "epoch": 7.751835535976505, + "grad_norm": 89.86659240722656, + "learning_rate": 1.3619233268356076e-05, + "loss": 1.0367, + "step": 1325 + }, + { + "epoch": 7.898678414096916, + "grad_norm": 135.19757080078125, + "learning_rate": 1.345679012345679e-05, + "loss": 0.5959, + "step": 1350 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.7882562277580071, + "eval_loss": 0.7940157055854797, + "eval_macro_f1": 0.7912761397548417, + "eval_runtime": 1.0699, + "eval_samples_per_second": 525.305, + "eval_steps_per_second": 33.649, + "eval_weighted_f1": 0.79101162234856, + "step": 1368 + }, + { + "epoch": 8.041116005873715, + "grad_norm": 123.22237396240234, + "learning_rate": 1.3294346978557506e-05, + "loss": 0.8356, + "step": 1375 + }, + { + "epoch": 8.187958883994126, + "grad_norm": 68.84619903564453, + "learning_rate": 1.3131903833658221e-05, + "loss": 0.4695, + "step": 1400 + }, + { + "epoch": 8.334801762114537, + "grad_norm": 154.56805419921875, + "learning_rate": 1.2969460688758935e-05, + "loss": 0.6034, + "step": 1425 + }, + { + "epoch": 8.481644640234949, + "grad_norm": 86.3011245727539, + "learning_rate": 1.280701754385965e-05, + "loss": 0.5087, + "step": 1450 + }, + { + "epoch": 8.62848751835536, + "grad_norm": 24.14583969116211, + "learning_rate": 1.2644574398960365e-05, + "loss": 0.422, + "step": 1475 + }, + { + "epoch": 8.775330396475772, + "grad_norm": 310.94207763671875, + "learning_rate": 1.248213125406108e-05, + "loss": 0.6256, + "step": 1500 + }, + { + "epoch": 8.922173274596181, + "grad_norm": 139.9342041015625, + "learning_rate": 1.2319688109161793e-05, + "loss": 0.6883, + "step": 1525 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.802491103202847, + "eval_loss": 0.9116710424423218, + "eval_macro_f1": 0.8018396686159844, + "eval_runtime": 1.0243, + "eval_samples_per_second": 548.666, + "eval_steps_per_second": 35.146, + "eval_weighted_f1": 0.8012830768350294, + "step": 1539 + }, + { + "epoch": 9.06461086637298, + "grad_norm": 46.106658935546875, + "learning_rate": 1.2157244964262509e-05, + "loss": 0.5617, + "step": 1550 + }, + { + "epoch": 9.211453744493392, + "grad_norm": 26.815168380737305, + "learning_rate": 1.1994801819363224e-05, + "loss": 0.4145, + "step": 1575 + }, + { + "epoch": 9.358296622613803, + "grad_norm": 113.18115997314453, + "learning_rate": 1.1832358674463939e-05, + "loss": 0.6228, + "step": 1600 + }, + { + "epoch": 9.505139500734215, + "grad_norm": 15.208633422851562, + "learning_rate": 1.1669915529564652e-05, + "loss": 0.4904, + "step": 1625 + }, + { + "epoch": 9.651982378854626, + "grad_norm": 120.86727142333984, + "learning_rate": 1.1507472384665367e-05, + "loss": 0.3946, + "step": 1650 + }, + { + "epoch": 9.798825256975036, + "grad_norm": 340.96685791015625, + "learning_rate": 1.1345029239766083e-05, + "loss": 0.6081, + "step": 1675 + }, + { + "epoch": 9.945668135095447, + "grad_norm": 213.39102172851562, + "learning_rate": 1.11825860948668e-05, + "loss": 0.3996, + "step": 1700 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8096085409252669, + "eval_loss": 1.0028072595596313, + "eval_macro_f1": 0.8115669703691483, + "eval_runtime": 1.128, + "eval_samples_per_second": 498.24, + "eval_steps_per_second": 31.916, + "eval_weighted_f1": 0.8112526763459069, + "step": 1710 + } + ], + "logging_steps": 25, + "max_steps": 3420, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/trainer_state.json b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..631518a20e9a2df3b8993761dfbc3bd7d56a55d9 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-1448/trainer_state.json @@ -0,0 +1,530 @@ +{ + "best_global_step": 905, + "best_metric": 0.8079147423599644, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905", + "epoch": 8.0, + "eval_steps": 500, + "global_step": 1448, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13812154696132597, + "grad_norm": 12.728955268859863, + "learning_rate": 1.3259668508287293e-06, + "loss": 4.8048, + "step": 25 + }, + { + "epoch": 0.27624309392265195, + "grad_norm": 56.0223274230957, + "learning_rate": 2.707182320441989e-06, + "loss": 4.711, + "step": 50 + }, + { + "epoch": 0.4143646408839779, + "grad_norm": 39.932518005371094, + "learning_rate": 4.088397790055249e-06, + "loss": 4.5371, + "step": 75 + }, + { + "epoch": 0.5524861878453039, + "grad_norm": 42.93894577026367, + "learning_rate": 5.469613259668509e-06, + "loss": 4.678, + "step": 100 + }, + { + "epoch": 0.6906077348066298, + "grad_norm": 37.85771179199219, + "learning_rate": 6.850828729281769e-06, + "loss": 4.662, + "step": 125 + }, + { + "epoch": 0.8287292817679558, + "grad_norm": 48.18323516845703, + "learning_rate": 8.232044198895029e-06, + "loss": 4.4032, + "step": 150 + }, + { + "epoch": 0.9668508287292817, + "grad_norm": 99.214599609375, + "learning_rate": 9.613259668508288e-06, + "loss": 4.2769, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5567010309278351, + "eval_loss": 0.9564104080200195, + "eval_macro_f1": 0.531338415603882, + "eval_runtime": 1.078, + "eval_samples_per_second": 539.897, + "eval_steps_per_second": 34.323, + "eval_weighted_f1": 0.5352146874056312, + "step": 181 + }, + { + "epoch": 1.1049723756906078, + "grad_norm": 68.16238403320312, + "learning_rate": 1.0994475138121547e-05, + "loss": 4.0766, + "step": 200 + }, + { + "epoch": 1.2430939226519337, + "grad_norm": 36.474945068359375, + "learning_rate": 1.2375690607734806e-05, + "loss": 4.0206, + "step": 225 + }, + { + "epoch": 1.3812154696132597, + "grad_norm": 41.88302230834961, + "learning_rate": 1.3756906077348067e-05, + "loss": 4.0765, + "step": 250 + }, + { + "epoch": 1.5193370165745856, + "grad_norm": 73.10688781738281, + "learning_rate": 1.5138121546961326e-05, + "loss": 3.8742, + "step": 275 + }, + { + "epoch": 1.6574585635359116, + "grad_norm": 57.27144241333008, + "learning_rate": 1.6519337016574586e-05, + "loss": 3.7189, + "step": 300 + }, + { + "epoch": 1.7955801104972375, + "grad_norm": 80.90764617919922, + "learning_rate": 1.7900552486187846e-05, + "loss": 3.3838, + "step": 325 + }, + { + "epoch": 1.9337016574585635, + "grad_norm": 100.98841094970703, + "learning_rate": 1.9281767955801107e-05, + "loss": 3.2471, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.7096219931271478, + "eval_loss": 0.7144930362701416, + "eval_macro_f1": 0.708526703261887, + "eval_runtime": 1.087, + "eval_samples_per_second": 535.442, + "eval_steps_per_second": 34.04, + "eval_weighted_f1": 0.7085504083724671, + "step": 362 + }, + { + "epoch": 2.0718232044198897, + "grad_norm": 61.62918472290039, + "learning_rate": 1.9926335174953962e-05, + "loss": 2.9424, + "step": 375 + }, + { + "epoch": 2.2099447513812156, + "grad_norm": 66.43571472167969, + "learning_rate": 1.9772866789441377e-05, + "loss": 2.7574, + "step": 400 + }, + { + "epoch": 2.3480662983425415, + "grad_norm": 124.33316802978516, + "learning_rate": 1.9619398403928794e-05, + "loss": 2.8673, + "step": 425 + }, + { + "epoch": 2.4861878453038675, + "grad_norm": 52.74990463256836, + "learning_rate": 1.946593001841621e-05, + "loss": 2.7332, + "step": 450 + }, + { + "epoch": 2.6243093922651934, + "grad_norm": 120.99431610107422, + "learning_rate": 1.9312461632903623e-05, + "loss": 2.5593, + "step": 475 + }, + { + "epoch": 2.7624309392265194, + "grad_norm": 85.82913970947266, + "learning_rate": 1.915899324739104e-05, + "loss": 2.4126, + "step": 500 + }, + { + "epoch": 2.9005524861878453, + "grad_norm": 105.4771499633789, + "learning_rate": 1.9005524861878455e-05, + "loss": 2.4222, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7663230240549829, + "eval_loss": 0.5739344358444214, + "eval_macro_f1": 0.7566652570915214, + "eval_runtime": 1.0416, + "eval_samples_per_second": 558.739, + "eval_steps_per_second": 35.521, + "eval_weighted_f1": 0.758295393308871, + "step": 543 + }, + { + "epoch": 3.0386740331491713, + "grad_norm": 104.53805541992188, + "learning_rate": 1.885205647636587e-05, + "loss": 2.2542, + "step": 550 + }, + { + "epoch": 3.1767955801104972, + "grad_norm": 119.23262786865234, + "learning_rate": 1.8698588090853287e-05, + "loss": 2.2797, + "step": 575 + }, + { + "epoch": 3.314917127071823, + "grad_norm": 67.51237487792969, + "learning_rate": 1.85451197053407e-05, + "loss": 1.8071, + "step": 600 + }, + { + "epoch": 3.453038674033149, + "grad_norm": 48.214908599853516, + "learning_rate": 1.8391651319828116e-05, + "loss": 1.9983, + "step": 625 + }, + { + "epoch": 3.591160220994475, + "grad_norm": 67.250732421875, + "learning_rate": 1.823818293431553e-05, + "loss": 1.8777, + "step": 650 + }, + { + "epoch": 3.729281767955801, + "grad_norm": 115.72201538085938, + "learning_rate": 1.8084714548802948e-05, + "loss": 1.9912, + "step": 675 + }, + { + "epoch": 3.867403314917127, + "grad_norm": 47.928062438964844, + "learning_rate": 1.7931246163290362e-05, + "loss": 1.9777, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.788659793814433, + "eval_loss": 0.5242534875869751, + "eval_macro_f1": 0.787816514118535, + "eval_runtime": 1.1404, + "eval_samples_per_second": 510.334, + "eval_steps_per_second": 32.444, + "eval_weighted_f1": 0.7883303300363687, + "step": 724 + }, + { + "epoch": 4.005524861878453, + "grad_norm": 58.05070495605469, + "learning_rate": 1.7777777777777777e-05, + "loss": 1.9077, + "step": 725 + }, + { + "epoch": 4.143646408839779, + "grad_norm": 108.5123062133789, + "learning_rate": 1.7624309392265194e-05, + "loss": 1.3115, + "step": 750 + }, + { + "epoch": 4.281767955801105, + "grad_norm": 139.55747985839844, + "learning_rate": 1.747084100675261e-05, + "loss": 1.5837, + "step": 775 + }, + { + "epoch": 4.419889502762431, + "grad_norm": 84.59967803955078, + "learning_rate": 1.7317372621240026e-05, + "loss": 1.5453, + "step": 800 + }, + { + "epoch": 4.558011049723757, + "grad_norm": 60.05119323730469, + "learning_rate": 1.716390423572744e-05, + "loss": 1.7393, + "step": 825 + }, + { + "epoch": 4.696132596685083, + "grad_norm": 85.76810455322266, + "learning_rate": 1.701043585021486e-05, + "loss": 1.5069, + "step": 850 + }, + { + "epoch": 4.834254143646409, + "grad_norm": 140.66427612304688, + "learning_rate": 1.6856967464702273e-05, + "loss": 1.6284, + "step": 875 + }, + { + "epoch": 4.972375690607735, + "grad_norm": 116.09174346923828, + "learning_rate": 1.670349907918969e-05, + "loss": 1.4866, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8092783505154639, + "eval_loss": 0.5569475889205933, + "eval_macro_f1": 0.8079147423599644, + "eval_runtime": 1.1105, + "eval_samples_per_second": 524.085, + "eval_steps_per_second": 33.318, + "eval_weighted_f1": 0.8088169111415616, + "step": 905 + }, + { + "epoch": 5.110497237569061, + "grad_norm": 75.29962921142578, + "learning_rate": 1.6550030693677105e-05, + "loss": 1.174, + "step": 925 + }, + { + "epoch": 5.248618784530387, + "grad_norm": 86.1469497680664, + "learning_rate": 1.639656230816452e-05, + "loss": 1.1958, + "step": 950 + }, + { + "epoch": 5.386740331491713, + "grad_norm": 151.677978515625, + "learning_rate": 1.6243093922651937e-05, + "loss": 1.0569, + "step": 975 + }, + { + "epoch": 5.524861878453039, + "grad_norm": 50.731685638427734, + "learning_rate": 1.608962553713935e-05, + "loss": 1.1108, + "step": 1000 + }, + { + "epoch": 5.662983425414365, + "grad_norm": 33.633480072021484, + "learning_rate": 1.5936157151626766e-05, + "loss": 1.013, + "step": 1025 + }, + { + "epoch": 5.801104972375691, + "grad_norm": 105.0028076171875, + "learning_rate": 1.5782688766114183e-05, + "loss": 1.2053, + "step": 1050 + }, + { + "epoch": 5.939226519337017, + "grad_norm": 95.74874114990234, + "learning_rate": 1.5629220380601598e-05, + "loss": 1.1072, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.802405498281787, + "eval_loss": 0.6753490567207336, + "eval_macro_f1": 0.8016931014910001, + "eval_runtime": 1.1063, + "eval_samples_per_second": 526.071, + "eval_steps_per_second": 33.444, + "eval_weighted_f1": 0.8022709706811875, + "step": 1086 + }, + { + "epoch": 6.077348066298343, + "grad_norm": 141.87071228027344, + "learning_rate": 1.5475751995089012e-05, + "loss": 1.1294, + "step": 1100 + }, + { + "epoch": 6.2154696132596685, + "grad_norm": 55.73676300048828, + "learning_rate": 1.532228360957643e-05, + "loss": 0.9051, + "step": 1125 + }, + { + "epoch": 6.3535911602209945, + "grad_norm": 53.82463836669922, + "learning_rate": 1.5168815224063844e-05, + "loss": 0.9204, + "step": 1150 + }, + { + "epoch": 6.49171270718232, + "grad_norm": 113.02780151367188, + "learning_rate": 1.5015346838551258e-05, + "loss": 0.8417, + "step": 1175 + }, + { + "epoch": 6.629834254143646, + "grad_norm": 96.89856719970703, + "learning_rate": 1.4861878453038676e-05, + "loss": 0.7389, + "step": 1200 + }, + { + "epoch": 6.767955801104972, + "grad_norm": 66.34088134765625, + "learning_rate": 1.470841006752609e-05, + "loss": 0.8814, + "step": 1225 + }, + { + "epoch": 6.906077348066298, + "grad_norm": 86.4268569946289, + "learning_rate": 1.4554941682013506e-05, + "loss": 0.7623, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.7938144329896907, + "eval_loss": 0.7672864198684692, + "eval_macro_f1": 0.791636724497827, + "eval_runtime": 1.2192, + "eval_samples_per_second": 477.353, + "eval_steps_per_second": 30.347, + "eval_weighted_f1": 0.7921651883437386, + "step": 1267 + }, + { + "epoch": 7.044198895027624, + "grad_norm": 10.062307357788086, + "learning_rate": 1.4401473296500922e-05, + "loss": 1.2103, + "step": 1275 + }, + { + "epoch": 7.18232044198895, + "grad_norm": 176.4008331298828, + "learning_rate": 1.4248004910988338e-05, + "loss": 0.7116, + "step": 1300 + }, + { + "epoch": 7.320441988950276, + "grad_norm": 62.07518005371094, + "learning_rate": 1.4094536525475753e-05, + "loss": 0.6019, + "step": 1325 + }, + { + "epoch": 7.458563535911602, + "grad_norm": 127.64053344726562, + "learning_rate": 1.394106813996317e-05, + "loss": 0.5221, + "step": 1350 + }, + { + "epoch": 7.596685082872928, + "grad_norm": 73.96366882324219, + "learning_rate": 1.3787599754450585e-05, + "loss": 0.6994, + "step": 1375 + }, + { + "epoch": 7.734806629834254, + "grad_norm": 109.97692108154297, + "learning_rate": 1.3634131368937999e-05, + "loss": 0.5523, + "step": 1400 + }, + { + "epoch": 7.87292817679558, + "grad_norm": 101.4031753540039, + "learning_rate": 1.3480662983425417e-05, + "loss": 0.5238, + "step": 1425 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.7989690721649485, + "eval_loss": 0.8612092137336731, + "eval_macro_f1": 0.7970887738341478, + "eval_runtime": 1.0604, + "eval_samples_per_second": 548.846, + "eval_steps_per_second": 34.892, + "eval_weighted_f1": 0.7979701647109763, + "step": 1448 + } + ], + "logging_steps": 25, + "max_steps": 3620, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/trainer_state.json b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..82c5cc547c220828de8a5a8547a6c972928a1c3c --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905/trainer_state.json @@ -0,0 +1,350 @@ +{ + "best_global_step": 905, + "best_metric": 0.8079147423599644, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-02-s42-20260901T153352Z-27bda108/trainer/checkpoint-905", + "epoch": 5.0, + "eval_steps": 500, + "global_step": 905, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13812154696132597, + "grad_norm": 12.728955268859863, + "learning_rate": 1.3259668508287293e-06, + "loss": 4.8048, + "step": 25 + }, + { + "epoch": 0.27624309392265195, + "grad_norm": 56.0223274230957, + "learning_rate": 2.707182320441989e-06, + "loss": 4.711, + "step": 50 + }, + { + "epoch": 0.4143646408839779, + "grad_norm": 39.932518005371094, + "learning_rate": 4.088397790055249e-06, + "loss": 4.5371, + "step": 75 + }, + { + "epoch": 0.5524861878453039, + "grad_norm": 42.93894577026367, + "learning_rate": 5.469613259668509e-06, + "loss": 4.678, + "step": 100 + }, + { + "epoch": 0.6906077348066298, + "grad_norm": 37.85771179199219, + "learning_rate": 6.850828729281769e-06, + "loss": 4.662, + "step": 125 + }, + { + "epoch": 0.8287292817679558, + "grad_norm": 48.18323516845703, + "learning_rate": 8.232044198895029e-06, + "loss": 4.4032, + "step": 150 + }, + { + "epoch": 0.9668508287292817, + "grad_norm": 99.214599609375, + "learning_rate": 9.613259668508288e-06, + "loss": 4.2769, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5567010309278351, + "eval_loss": 0.9564104080200195, + "eval_macro_f1": 0.531338415603882, + "eval_runtime": 1.078, + "eval_samples_per_second": 539.897, + "eval_steps_per_second": 34.323, + "eval_weighted_f1": 0.5352146874056312, + "step": 181 + }, + { + "epoch": 1.1049723756906078, + "grad_norm": 68.16238403320312, + "learning_rate": 1.0994475138121547e-05, + "loss": 4.0766, + "step": 200 + }, + { + "epoch": 1.2430939226519337, + "grad_norm": 36.474945068359375, + "learning_rate": 1.2375690607734806e-05, + "loss": 4.0206, + "step": 225 + }, + { + "epoch": 1.3812154696132597, + "grad_norm": 41.88302230834961, + "learning_rate": 1.3756906077348067e-05, + "loss": 4.0765, + "step": 250 + }, + { + "epoch": 1.5193370165745856, + "grad_norm": 73.10688781738281, + "learning_rate": 1.5138121546961326e-05, + "loss": 3.8742, + "step": 275 + }, + { + "epoch": 1.6574585635359116, + "grad_norm": 57.27144241333008, + "learning_rate": 1.6519337016574586e-05, + "loss": 3.7189, + "step": 300 + }, + { + "epoch": 1.7955801104972375, + "grad_norm": 80.90764617919922, + "learning_rate": 1.7900552486187846e-05, + "loss": 3.3838, + "step": 325 + }, + { + "epoch": 1.9337016574585635, + "grad_norm": 100.98841094970703, + "learning_rate": 1.9281767955801107e-05, + "loss": 3.2471, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.7096219931271478, + "eval_loss": 0.7144930362701416, + "eval_macro_f1": 0.708526703261887, + "eval_runtime": 1.087, + "eval_samples_per_second": 535.442, + "eval_steps_per_second": 34.04, + "eval_weighted_f1": 0.7085504083724671, + "step": 362 + }, + { + "epoch": 2.0718232044198897, + "grad_norm": 61.62918472290039, + "learning_rate": 1.9926335174953962e-05, + "loss": 2.9424, + "step": 375 + }, + { + "epoch": 2.2099447513812156, + "grad_norm": 66.43571472167969, + "learning_rate": 1.9772866789441377e-05, + "loss": 2.7574, + "step": 400 + }, + { + "epoch": 2.3480662983425415, + "grad_norm": 124.33316802978516, + "learning_rate": 1.9619398403928794e-05, + "loss": 2.8673, + "step": 425 + }, + { + "epoch": 2.4861878453038675, + "grad_norm": 52.74990463256836, + "learning_rate": 1.946593001841621e-05, + "loss": 2.7332, + "step": 450 + }, + { + "epoch": 2.6243093922651934, + "grad_norm": 120.99431610107422, + "learning_rate": 1.9312461632903623e-05, + "loss": 2.5593, + "step": 475 + }, + { + "epoch": 2.7624309392265194, + "grad_norm": 85.82913970947266, + "learning_rate": 1.915899324739104e-05, + "loss": 2.4126, + "step": 500 + }, + { + "epoch": 2.9005524861878453, + "grad_norm": 105.4771499633789, + "learning_rate": 1.9005524861878455e-05, + "loss": 2.4222, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7663230240549829, + "eval_loss": 0.5739344358444214, + "eval_macro_f1": 0.7566652570915214, + "eval_runtime": 1.0416, + "eval_samples_per_second": 558.739, + "eval_steps_per_second": 35.521, + "eval_weighted_f1": 0.758295393308871, + "step": 543 + }, + { + "epoch": 3.0386740331491713, + "grad_norm": 104.53805541992188, + "learning_rate": 1.885205647636587e-05, + "loss": 2.2542, + "step": 550 + }, + { + "epoch": 3.1767955801104972, + "grad_norm": 119.23262786865234, + "learning_rate": 1.8698588090853287e-05, + "loss": 2.2797, + "step": 575 + }, + { + "epoch": 3.314917127071823, + "grad_norm": 67.51237487792969, + "learning_rate": 1.85451197053407e-05, + "loss": 1.8071, + "step": 600 + }, + { + "epoch": 3.453038674033149, + "grad_norm": 48.214908599853516, + "learning_rate": 1.8391651319828116e-05, + "loss": 1.9983, + "step": 625 + }, + { + "epoch": 3.591160220994475, + "grad_norm": 67.250732421875, + "learning_rate": 1.823818293431553e-05, + "loss": 1.8777, + "step": 650 + }, + { + "epoch": 3.729281767955801, + "grad_norm": 115.72201538085938, + "learning_rate": 1.8084714548802948e-05, + "loss": 1.9912, + "step": 675 + }, + { + "epoch": 3.867403314917127, + "grad_norm": 47.928062438964844, + "learning_rate": 1.7931246163290362e-05, + "loss": 1.9777, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.788659793814433, + "eval_loss": 0.5242534875869751, + "eval_macro_f1": 0.787816514118535, + "eval_runtime": 1.1404, + "eval_samples_per_second": 510.334, + "eval_steps_per_second": 32.444, + "eval_weighted_f1": 0.7883303300363687, + "step": 724 + }, + { + "epoch": 4.005524861878453, + "grad_norm": 58.05070495605469, + "learning_rate": 1.7777777777777777e-05, + "loss": 1.9077, + "step": 725 + }, + { + "epoch": 4.143646408839779, + "grad_norm": 108.5123062133789, + "learning_rate": 1.7624309392265194e-05, + "loss": 1.3115, + "step": 750 + }, + { + "epoch": 4.281767955801105, + "grad_norm": 139.55747985839844, + "learning_rate": 1.747084100675261e-05, + "loss": 1.5837, + "step": 775 + }, + { + "epoch": 4.419889502762431, + "grad_norm": 84.59967803955078, + "learning_rate": 1.7317372621240026e-05, + "loss": 1.5453, + "step": 800 + }, + { + "epoch": 4.558011049723757, + "grad_norm": 60.05119323730469, + "learning_rate": 1.716390423572744e-05, + "loss": 1.7393, + "step": 825 + }, + { + "epoch": 4.696132596685083, + "grad_norm": 85.76810455322266, + "learning_rate": 1.701043585021486e-05, + "loss": 1.5069, + "step": 850 + }, + { + "epoch": 4.834254143646409, + "grad_norm": 140.66427612304688, + "learning_rate": 1.6856967464702273e-05, + "loss": 1.6284, + "step": 875 + }, + { + "epoch": 4.972375690607735, + "grad_norm": 116.09174346923828, + "learning_rate": 1.670349907918969e-05, + "loss": 1.4866, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8092783505154639, + "eval_loss": 0.5569475889205933, + "eval_macro_f1": 0.8079147423599644, + "eval_runtime": 1.1105, + "eval_samples_per_second": 524.085, + "eval_steps_per_second": 33.318, + "eval_weighted_f1": 0.8088169111415616, + "step": 905 + } + ], + "logging_steps": 25, + "max_steps": 3620, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/trainer_state.json b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..687d0174ec7bba6a033a218d3d6664e41bdc2291 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267/trainer_state.json @@ -0,0 +1,470 @@ +{ + "best_global_step": 1267, + "best_metric": 0.7954733120448395, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267", + "epoch": 7.0, + "eval_steps": 500, + "global_step": 1267, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13812154696132597, + "grad_norm": 20.876405715942383, + "learning_rate": 1.3259668508287293e-06, + "loss": 4.7917, + "step": 25 + }, + { + "epoch": 0.27624309392265195, + "grad_norm": 17.433170318603516, + "learning_rate": 2.707182320441989e-06, + "loss": 4.6316, + "step": 50 + }, + { + "epoch": 0.4143646408839779, + "grad_norm": 13.11660099029541, + "learning_rate": 4.088397790055249e-06, + "loss": 4.5535, + "step": 75 + }, + { + "epoch": 0.5524861878453039, + "grad_norm": 24.04007339477539, + "learning_rate": 5.469613259668509e-06, + "loss": 4.5854, + "step": 100 + }, + { + "epoch": 0.6906077348066298, + "grad_norm": 14.225695610046387, + "learning_rate": 6.850828729281769e-06, + "loss": 4.4616, + "step": 125 + }, + { + "epoch": 0.8287292817679558, + "grad_norm": 58.386573791503906, + "learning_rate": 8.232044198895029e-06, + "loss": 4.4235, + "step": 150 + }, + { + "epoch": 0.9668508287292817, + "grad_norm": 75.67534637451172, + "learning_rate": 9.613259668508288e-06, + "loss": 4.4501, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.4209621993127148, + "eval_loss": 1.0621777772903442, + "eval_macro_f1": 0.3264749162286163, + "eval_runtime": 1.0829, + "eval_samples_per_second": 537.44, + "eval_steps_per_second": 34.167, + "eval_weighted_f1": 0.3351642565705843, + "step": 181 + }, + { + "epoch": 1.1049723756906078, + "grad_norm": 55.63917541503906, + "learning_rate": 1.0994475138121547e-05, + "loss": 4.205, + "step": 200 + }, + { + "epoch": 1.2430939226519337, + "grad_norm": 42.331546783447266, + "learning_rate": 1.2375690607734806e-05, + "loss": 3.7812, + "step": 225 + }, + { + "epoch": 1.3812154696132597, + "grad_norm": 107.68368530273438, + "learning_rate": 1.3756906077348067e-05, + "loss": 3.559, + "step": 250 + }, + { + "epoch": 1.5193370165745856, + "grad_norm": 88.61949157714844, + "learning_rate": 1.5138121546961326e-05, + "loss": 3.449, + "step": 275 + }, + { + "epoch": 1.6574585635359116, + "grad_norm": 82.10724639892578, + "learning_rate": 1.6519337016574586e-05, + "loss": 3.2662, + "step": 300 + }, + { + "epoch": 1.7955801104972375, + "grad_norm": 50.293724060058594, + "learning_rate": 1.7900552486187846e-05, + "loss": 3.222, + "step": 325 + }, + { + "epoch": 1.9337016574585635, + "grad_norm": 155.03953552246094, + "learning_rate": 1.9281767955801107e-05, + "loss": 3.463, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6872852233676976, + "eval_loss": 0.698262095451355, + "eval_macro_f1": 0.6712137376689805, + "eval_runtime": 1.0621, + "eval_samples_per_second": 547.989, + "eval_steps_per_second": 34.838, + "eval_weighted_f1": 0.6732908155642355, + "step": 362 + }, + { + "epoch": 2.0718232044198897, + "grad_norm": 81.68114471435547, + "learning_rate": 1.9926335174953962e-05, + "loss": 3.2279, + "step": 375 + }, + { + "epoch": 2.2099447513812156, + "grad_norm": 68.81178283691406, + "learning_rate": 1.9772866789441377e-05, + "loss": 3.0296, + "step": 400 + }, + { + "epoch": 2.3480662983425415, + "grad_norm": 82.61100006103516, + "learning_rate": 1.9619398403928794e-05, + "loss": 2.7194, + "step": 425 + }, + { + "epoch": 2.4861878453038675, + "grad_norm": 175.10655212402344, + "learning_rate": 1.946593001841621e-05, + "loss": 2.8714, + "step": 450 + }, + { + "epoch": 2.6243093922651934, + "grad_norm": 60.468971252441406, + "learning_rate": 1.9312461632903623e-05, + "loss": 2.702, + "step": 475 + }, + { + "epoch": 2.7624309392265194, + "grad_norm": 71.27595520019531, + "learning_rate": 1.915899324739104e-05, + "loss": 2.6361, + "step": 500 + }, + { + "epoch": 2.9005524861878453, + "grad_norm": 63.07265853881836, + "learning_rate": 1.9005524861878455e-05, + "loss": 2.4736, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7353951890034365, + "eval_loss": 0.6232973337173462, + "eval_macro_f1": 0.7353321251064063, + "eval_runtime": 1.1101, + "eval_samples_per_second": 524.3, + "eval_steps_per_second": 33.332, + "eval_weighted_f1": 0.735417741346018, + "step": 543 + }, + { + "epoch": 3.0386740331491713, + "grad_norm": 44.54671859741211, + "learning_rate": 1.885205647636587e-05, + "loss": 2.3959, + "step": 550 + }, + { + "epoch": 3.1767955801104972, + "grad_norm": 50.67565155029297, + "learning_rate": 1.8698588090853287e-05, + "loss": 2.1091, + "step": 575 + }, + { + "epoch": 3.314917127071823, + "grad_norm": 37.43223571777344, + "learning_rate": 1.85451197053407e-05, + "loss": 2.2882, + "step": 600 + }, + { + "epoch": 3.453038674033149, + "grad_norm": 143.8233184814453, + "learning_rate": 1.8391651319828116e-05, + "loss": 2.187, + "step": 625 + }, + { + "epoch": 3.591160220994475, + "grad_norm": 87.6942367553711, + "learning_rate": 1.823818293431553e-05, + "loss": 2.1955, + "step": 650 + }, + { + "epoch": 3.729281767955801, + "grad_norm": 153.69772338867188, + "learning_rate": 1.8084714548802948e-05, + "loss": 2.0048, + "step": 675 + }, + { + "epoch": 3.867403314917127, + "grad_norm": 87.95222473144531, + "learning_rate": 1.7931246163290362e-05, + "loss": 2.0396, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7697594501718213, + "eval_loss": 0.6113913655281067, + "eval_macro_f1": 0.7717941663342183, + "eval_runtime": 1.1763, + "eval_samples_per_second": 494.768, + "eval_steps_per_second": 31.454, + "eval_weighted_f1": 0.7727020067399202, + "step": 724 + }, + { + "epoch": 4.005524861878453, + "grad_norm": 98.54376220703125, + "learning_rate": 1.7777777777777777e-05, + "loss": 2.0418, + "step": 725 + }, + { + "epoch": 4.143646408839779, + "grad_norm": 96.78535461425781, + "learning_rate": 1.7624309392265194e-05, + "loss": 1.7481, + "step": 750 + }, + { + "epoch": 4.281767955801105, + "grad_norm": 100.56416320800781, + "learning_rate": 1.747084100675261e-05, + "loss": 1.6409, + "step": 775 + }, + { + "epoch": 4.419889502762431, + "grad_norm": 84.92568969726562, + "learning_rate": 1.7317372621240026e-05, + "loss": 1.4778, + "step": 800 + }, + { + "epoch": 4.558011049723757, + "grad_norm": 67.47079467773438, + "learning_rate": 1.716390423572744e-05, + "loss": 1.6417, + "step": 825 + }, + { + "epoch": 4.696132596685083, + "grad_norm": 63.104026794433594, + "learning_rate": 1.701043585021486e-05, + "loss": 1.5646, + "step": 850 + }, + { + "epoch": 4.834254143646409, + "grad_norm": 32.19059753417969, + "learning_rate": 1.6856967464702273e-05, + "loss": 1.6532, + "step": 875 + }, + { + "epoch": 4.972375690607735, + "grad_norm": 91.98654174804688, + "learning_rate": 1.670349907918969e-05, + "loss": 1.5656, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7405498281786942, + "eval_loss": 0.686448872089386, + "eval_macro_f1": 0.7401846883095243, + "eval_runtime": 1.1008, + "eval_samples_per_second": 528.715, + "eval_steps_per_second": 33.612, + "eval_weighted_f1": 0.7395879576258151, + "step": 905 + }, + { + "epoch": 5.110497237569061, + "grad_norm": 53.78845977783203, + "learning_rate": 1.6550030693677105e-05, + "loss": 1.5646, + "step": 925 + }, + { + "epoch": 5.248618784530387, + "grad_norm": 97.8853988647461, + "learning_rate": 1.639656230816452e-05, + "loss": 1.2122, + "step": 950 + }, + { + "epoch": 5.386740331491713, + "grad_norm": 141.8799591064453, + "learning_rate": 1.6243093922651937e-05, + "loss": 1.3174, + "step": 975 + }, + { + "epoch": 5.524861878453039, + "grad_norm": 141.72279357910156, + "learning_rate": 1.608962553713935e-05, + "loss": 1.3151, + "step": 1000 + }, + { + "epoch": 5.662983425414365, + "grad_norm": 42.930763244628906, + "learning_rate": 1.5936157151626766e-05, + "loss": 1.3529, + "step": 1025 + }, + { + "epoch": 5.801104972375691, + "grad_norm": 95.8829574584961, + "learning_rate": 1.5782688766114183e-05, + "loss": 1.2065, + "step": 1050 + }, + { + "epoch": 5.939226519337017, + "grad_norm": 94.2139892578125, + "learning_rate": 1.5629220380601598e-05, + "loss": 1.2515, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7869415807560137, + "eval_loss": 0.7194661498069763, + "eval_macro_f1": 0.7836261253161675, + "eval_runtime": 1.0939, + "eval_samples_per_second": 532.046, + "eval_steps_per_second": 33.824, + "eval_weighted_f1": 0.7845392600592425, + "step": 1086 + }, + { + "epoch": 6.077348066298343, + "grad_norm": 69.97132873535156, + "learning_rate": 1.5475751995089012e-05, + "loss": 1.0752, + "step": 1100 + }, + { + "epoch": 6.2154696132596685, + "grad_norm": 243.30223083496094, + "learning_rate": 1.532228360957643e-05, + "loss": 0.8988, + "step": 1125 + }, + { + "epoch": 6.3535911602209945, + "grad_norm": 107.98970031738281, + "learning_rate": 1.5168815224063844e-05, + "loss": 0.9617, + "step": 1150 + }, + { + "epoch": 6.49171270718232, + "grad_norm": 108.35636901855469, + "learning_rate": 1.5015346838551258e-05, + "loss": 0.7218, + "step": 1175 + }, + { + "epoch": 6.629834254143646, + "grad_norm": 134.66993713378906, + "learning_rate": 1.4861878453038676e-05, + "loss": 1.0372, + "step": 1200 + }, + { + "epoch": 6.767955801104972, + "grad_norm": 94.36707305908203, + "learning_rate": 1.470841006752609e-05, + "loss": 1.0096, + "step": 1225 + }, + { + "epoch": 6.906077348066298, + "grad_norm": 43.02191925048828, + "learning_rate": 1.4554941682013506e-05, + "loss": 1.062, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.7955326460481099, + "eval_loss": 0.7213481068611145, + "eval_macro_f1": 0.7954733120448395, + "eval_runtime": 1.2019, + "eval_samples_per_second": 484.219, + "eval_steps_per_second": 30.784, + "eval_weighted_f1": 0.7954606340769154, + "step": 1267 + } + ], + "logging_steps": 25, + "max_steps": 3620, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/trainer_state.json b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4d94a54000c45c463fb624700a1501f53ebd38ba --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1810/trainer_state.json @@ -0,0 +1,657 @@ +{ + "best_global_step": 1267, + "best_metric": 0.7954733120448395, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-02-s43-20260901T171619Z-fbd63439/trainer/checkpoint-1267", + "epoch": 10.0, + "eval_steps": 500, + "global_step": 1810, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13812154696132597, + "grad_norm": 20.876405715942383, + "learning_rate": 1.3259668508287293e-06, + "loss": 4.7917, + "step": 25 + }, + { + "epoch": 0.27624309392265195, + "grad_norm": 17.433170318603516, + "learning_rate": 2.707182320441989e-06, + "loss": 4.6316, + "step": 50 + }, + { + "epoch": 0.4143646408839779, + "grad_norm": 13.11660099029541, + "learning_rate": 4.088397790055249e-06, + "loss": 4.5535, + "step": 75 + }, + { + "epoch": 0.5524861878453039, + "grad_norm": 24.04007339477539, + "learning_rate": 5.469613259668509e-06, + "loss": 4.5854, + "step": 100 + }, + { + "epoch": 0.6906077348066298, + "grad_norm": 14.225695610046387, + "learning_rate": 6.850828729281769e-06, + "loss": 4.4616, + "step": 125 + }, + { + "epoch": 0.8287292817679558, + "grad_norm": 58.386573791503906, + "learning_rate": 8.232044198895029e-06, + "loss": 4.4235, + "step": 150 + }, + { + "epoch": 0.9668508287292817, + "grad_norm": 75.67534637451172, + "learning_rate": 9.613259668508288e-06, + "loss": 4.4501, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.4209621993127148, + "eval_loss": 1.0621777772903442, + "eval_macro_f1": 0.3264749162286163, + "eval_runtime": 1.0829, + "eval_samples_per_second": 537.44, + "eval_steps_per_second": 34.167, + "eval_weighted_f1": 0.3351642565705843, + "step": 181 + }, + { + "epoch": 1.1049723756906078, + "grad_norm": 55.63917541503906, + "learning_rate": 1.0994475138121547e-05, + "loss": 4.205, + "step": 200 + }, + { + "epoch": 1.2430939226519337, + "grad_norm": 42.331546783447266, + "learning_rate": 1.2375690607734806e-05, + "loss": 3.7812, + "step": 225 + }, + { + "epoch": 1.3812154696132597, + "grad_norm": 107.68368530273438, + "learning_rate": 1.3756906077348067e-05, + "loss": 3.559, + "step": 250 + }, + { + "epoch": 1.5193370165745856, + "grad_norm": 88.61949157714844, + "learning_rate": 1.5138121546961326e-05, + "loss": 3.449, + "step": 275 + }, + { + "epoch": 1.6574585635359116, + "grad_norm": 82.10724639892578, + "learning_rate": 1.6519337016574586e-05, + "loss": 3.2662, + "step": 300 + }, + { + "epoch": 1.7955801104972375, + "grad_norm": 50.293724060058594, + "learning_rate": 1.7900552486187846e-05, + "loss": 3.222, + "step": 325 + }, + { + "epoch": 1.9337016574585635, + "grad_norm": 155.03953552246094, + "learning_rate": 1.9281767955801107e-05, + "loss": 3.463, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6872852233676976, + "eval_loss": 0.698262095451355, + "eval_macro_f1": 0.6712137376689805, + "eval_runtime": 1.0621, + "eval_samples_per_second": 547.989, + "eval_steps_per_second": 34.838, + "eval_weighted_f1": 0.6732908155642355, + "step": 362 + }, + { + "epoch": 2.0718232044198897, + "grad_norm": 81.68114471435547, + "learning_rate": 1.9926335174953962e-05, + "loss": 3.2279, + "step": 375 + }, + { + "epoch": 2.2099447513812156, + "grad_norm": 68.81178283691406, + "learning_rate": 1.9772866789441377e-05, + "loss": 3.0296, + "step": 400 + }, + { + "epoch": 2.3480662983425415, + "grad_norm": 82.61100006103516, + "learning_rate": 1.9619398403928794e-05, + "loss": 2.7194, + "step": 425 + }, + { + "epoch": 2.4861878453038675, + "grad_norm": 175.10655212402344, + "learning_rate": 1.946593001841621e-05, + "loss": 2.8714, + "step": 450 + }, + { + "epoch": 2.6243093922651934, + "grad_norm": 60.468971252441406, + "learning_rate": 1.9312461632903623e-05, + "loss": 2.702, + "step": 475 + }, + { + "epoch": 2.7624309392265194, + "grad_norm": 71.27595520019531, + "learning_rate": 1.915899324739104e-05, + "loss": 2.6361, + "step": 500 + }, + { + "epoch": 2.9005524861878453, + "grad_norm": 63.07265853881836, + "learning_rate": 1.9005524861878455e-05, + "loss": 2.4736, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7353951890034365, + "eval_loss": 0.6232973337173462, + "eval_macro_f1": 0.7353321251064063, + "eval_runtime": 1.1101, + "eval_samples_per_second": 524.3, + "eval_steps_per_second": 33.332, + "eval_weighted_f1": 0.735417741346018, + "step": 543 + }, + { + "epoch": 3.0386740331491713, + "grad_norm": 44.54671859741211, + "learning_rate": 1.885205647636587e-05, + "loss": 2.3959, + "step": 550 + }, + { + "epoch": 3.1767955801104972, + "grad_norm": 50.67565155029297, + "learning_rate": 1.8698588090853287e-05, + "loss": 2.1091, + "step": 575 + }, + { + "epoch": 3.314917127071823, + "grad_norm": 37.43223571777344, + "learning_rate": 1.85451197053407e-05, + "loss": 2.2882, + "step": 600 + }, + { + "epoch": 3.453038674033149, + "grad_norm": 143.8233184814453, + "learning_rate": 1.8391651319828116e-05, + "loss": 2.187, + "step": 625 + }, + { + "epoch": 3.591160220994475, + "grad_norm": 87.6942367553711, + "learning_rate": 1.823818293431553e-05, + "loss": 2.1955, + "step": 650 + }, + { + "epoch": 3.729281767955801, + "grad_norm": 153.69772338867188, + "learning_rate": 1.8084714548802948e-05, + "loss": 2.0048, + "step": 675 + }, + { + "epoch": 3.867403314917127, + "grad_norm": 87.95222473144531, + "learning_rate": 1.7931246163290362e-05, + "loss": 2.0396, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7697594501718213, + "eval_loss": 0.6113913655281067, + "eval_macro_f1": 0.7717941663342183, + "eval_runtime": 1.1763, + "eval_samples_per_second": 494.768, + "eval_steps_per_second": 31.454, + "eval_weighted_f1": 0.7727020067399202, + "step": 724 + }, + { + "epoch": 4.005524861878453, + "grad_norm": 98.54376220703125, + "learning_rate": 1.7777777777777777e-05, + "loss": 2.0418, + "step": 725 + }, + { + "epoch": 4.143646408839779, + "grad_norm": 96.78535461425781, + "learning_rate": 1.7624309392265194e-05, + "loss": 1.7481, + "step": 750 + }, + { + "epoch": 4.281767955801105, + "grad_norm": 100.56416320800781, + "learning_rate": 1.747084100675261e-05, + "loss": 1.6409, + "step": 775 + }, + { + "epoch": 4.419889502762431, + "grad_norm": 84.92568969726562, + "learning_rate": 1.7317372621240026e-05, + "loss": 1.4778, + "step": 800 + }, + { + "epoch": 4.558011049723757, + "grad_norm": 67.47079467773438, + "learning_rate": 1.716390423572744e-05, + "loss": 1.6417, + "step": 825 + }, + { + "epoch": 4.696132596685083, + "grad_norm": 63.104026794433594, + "learning_rate": 1.701043585021486e-05, + "loss": 1.5646, + "step": 850 + }, + { + "epoch": 4.834254143646409, + "grad_norm": 32.19059753417969, + "learning_rate": 1.6856967464702273e-05, + "loss": 1.6532, + "step": 875 + }, + { + "epoch": 4.972375690607735, + "grad_norm": 91.98654174804688, + "learning_rate": 1.670349907918969e-05, + "loss": 1.5656, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7405498281786942, + "eval_loss": 0.686448872089386, + "eval_macro_f1": 0.7401846883095243, + "eval_runtime": 1.1008, + "eval_samples_per_second": 528.715, + "eval_steps_per_second": 33.612, + "eval_weighted_f1": 0.7395879576258151, + "step": 905 + }, + { + "epoch": 5.110497237569061, + "grad_norm": 53.78845977783203, + "learning_rate": 1.6550030693677105e-05, + "loss": 1.5646, + "step": 925 + }, + { + "epoch": 5.248618784530387, + "grad_norm": 97.8853988647461, + "learning_rate": 1.639656230816452e-05, + "loss": 1.2122, + "step": 950 + }, + { + "epoch": 5.386740331491713, + "grad_norm": 141.8799591064453, + "learning_rate": 1.6243093922651937e-05, + "loss": 1.3174, + "step": 975 + }, + { + "epoch": 5.524861878453039, + "grad_norm": 141.72279357910156, + "learning_rate": 1.608962553713935e-05, + "loss": 1.3151, + "step": 1000 + }, + { + "epoch": 5.662983425414365, + "grad_norm": 42.930763244628906, + "learning_rate": 1.5936157151626766e-05, + "loss": 1.3529, + "step": 1025 + }, + { + "epoch": 5.801104972375691, + "grad_norm": 95.8829574584961, + "learning_rate": 1.5782688766114183e-05, + "loss": 1.2065, + "step": 1050 + }, + { + "epoch": 5.939226519337017, + "grad_norm": 94.2139892578125, + "learning_rate": 1.5629220380601598e-05, + "loss": 1.2515, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7869415807560137, + "eval_loss": 0.7194661498069763, + "eval_macro_f1": 0.7836261253161675, + "eval_runtime": 1.0939, + "eval_samples_per_second": 532.046, + "eval_steps_per_second": 33.824, + "eval_weighted_f1": 0.7845392600592425, + "step": 1086 + }, + { + "epoch": 6.077348066298343, + "grad_norm": 69.97132873535156, + "learning_rate": 1.5475751995089012e-05, + "loss": 1.0752, + "step": 1100 + }, + { + "epoch": 6.2154696132596685, + "grad_norm": 243.30223083496094, + "learning_rate": 1.532228360957643e-05, + "loss": 0.8988, + "step": 1125 + }, + { + "epoch": 6.3535911602209945, + "grad_norm": 107.98970031738281, + "learning_rate": 1.5168815224063844e-05, + "loss": 0.9617, + "step": 1150 + }, + { + "epoch": 6.49171270718232, + "grad_norm": 108.35636901855469, + "learning_rate": 1.5015346838551258e-05, + "loss": 0.7218, + "step": 1175 + }, + { + "epoch": 6.629834254143646, + "grad_norm": 134.66993713378906, + "learning_rate": 1.4861878453038676e-05, + "loss": 1.0372, + "step": 1200 + }, + { + "epoch": 6.767955801104972, + "grad_norm": 94.36707305908203, + "learning_rate": 1.470841006752609e-05, + "loss": 1.0096, + "step": 1225 + }, + { + "epoch": 6.906077348066298, + "grad_norm": 43.02191925048828, + "learning_rate": 1.4554941682013506e-05, + "loss": 1.062, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.7955326460481099, + "eval_loss": 0.7213481068611145, + "eval_macro_f1": 0.7954733120448395, + "eval_runtime": 1.2019, + "eval_samples_per_second": 484.219, + "eval_steps_per_second": 30.784, + "eval_weighted_f1": 0.7954606340769154, + "step": 1267 + }, + { + "epoch": 7.044198895027624, + "grad_norm": 92.72286987304688, + "learning_rate": 1.4401473296500922e-05, + "loss": 0.6996, + "step": 1275 + }, + { + "epoch": 7.18232044198895, + "grad_norm": 107.59213256835938, + "learning_rate": 1.4248004910988338e-05, + "loss": 0.6418, + "step": 1300 + }, + { + "epoch": 7.320441988950276, + "grad_norm": 121.2758560180664, + "learning_rate": 1.4094536525475753e-05, + "loss": 0.7162, + "step": 1325 + }, + { + "epoch": 7.458563535911602, + "grad_norm": 109.18578338623047, + "learning_rate": 1.394106813996317e-05, + "loss": 0.5579, + "step": 1350 + }, + { + "epoch": 7.596685082872928, + "grad_norm": 151.84164428710938, + "learning_rate": 1.3787599754450585e-05, + "loss": 0.9806, + "step": 1375 + }, + { + "epoch": 7.734806629834254, + "grad_norm": 51.866580963134766, + "learning_rate": 1.3634131368937999e-05, + "loss": 0.6476, + "step": 1400 + }, + { + "epoch": 7.87292817679558, + "grad_norm": 42.80246353149414, + "learning_rate": 1.3480662983425417e-05, + "loss": 0.6789, + "step": 1425 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.7835051546391752, + "eval_loss": 0.8676192760467529, + "eval_macro_f1": 0.7831063996889714, + "eval_runtime": 1.2809, + "eval_samples_per_second": 454.35, + "eval_steps_per_second": 28.885, + "eval_weighted_f1": 0.783752317724552, + "step": 1448 + }, + { + "epoch": 8.011049723756907, + "grad_norm": 50.81716537475586, + "learning_rate": 1.3327194597912831e-05, + "loss": 0.8686, + "step": 1450 + }, + { + "epoch": 8.149171270718233, + "grad_norm": 149.4502716064453, + "learning_rate": 1.3173726212400245e-05, + "loss": 0.4912, + "step": 1475 + }, + { + "epoch": 8.287292817679559, + "grad_norm": 93.06773376464844, + "learning_rate": 1.3020257826887662e-05, + "loss": 0.6658, + "step": 1500 + }, + { + "epoch": 8.425414364640885, + "grad_norm": 93.24884033203125, + "learning_rate": 1.2866789441375078e-05, + "loss": 0.6, + "step": 1525 + }, + { + "epoch": 8.56353591160221, + "grad_norm": 72.1173324584961, + "learning_rate": 1.2713321055862494e-05, + "loss": 0.4449, + "step": 1550 + }, + { + "epoch": 8.701657458563536, + "grad_norm": 108.58943176269531, + "learning_rate": 1.2559852670349908e-05, + "loss": 0.5033, + "step": 1575 + }, + { + "epoch": 8.839779005524862, + "grad_norm": 244.68243408203125, + "learning_rate": 1.2406384284837326e-05, + "loss": 0.7072, + "step": 1600 + }, + { + "epoch": 8.977900552486188, + "grad_norm": 92.26280975341797, + "learning_rate": 1.225291589932474e-05, + "loss": 0.6048, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.7920962199312714, + "eval_loss": 0.8034908175468445, + "eval_macro_f1": 0.7899251549604602, + "eval_runtime": 1.2928, + "eval_samples_per_second": 450.202, + "eval_steps_per_second": 28.621, + "eval_weighted_f1": 0.7905565780174157, + "step": 1629 + }, + { + "epoch": 9.116022099447514, + "grad_norm": 132.67178344726562, + "learning_rate": 1.2099447513812154e-05, + "loss": 0.4016, + "step": 1650 + }, + { + "epoch": 9.25414364640884, + "grad_norm": 265.9836120605469, + "learning_rate": 1.1945979128299572e-05, + "loss": 0.3719, + "step": 1675 + }, + { + "epoch": 9.392265193370166, + "grad_norm": 106.52467346191406, + "learning_rate": 1.1792510742786986e-05, + "loss": 0.4098, + "step": 1700 + }, + { + "epoch": 9.530386740331492, + "grad_norm": 5.532192707061768, + "learning_rate": 1.1639042357274402e-05, + "loss": 0.3524, + "step": 1725 + }, + { + "epoch": 9.668508287292818, + "grad_norm": 48.053131103515625, + "learning_rate": 1.1485573971761818e-05, + "loss": 0.4597, + "step": 1750 + }, + { + "epoch": 9.806629834254144, + "grad_norm": 165.1774444580078, + "learning_rate": 1.1332105586249233e-05, + "loss": 0.3277, + "step": 1775 + }, + { + "epoch": 9.94475138121547, + "grad_norm": 123.62344360351562, + "learning_rate": 1.1178637200736649e-05, + "loss": 0.437, + "step": 1800 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.781786941580756, + "eval_loss": 1.2036364078521729, + "eval_macro_f1": 0.779737549421618, + "eval_runtime": 1.5055, + "eval_samples_per_second": 386.57, + "eval_steps_per_second": 24.576, + "eval_weighted_f1": 0.7805753164794587, + "step": 1810 + } + ], + "logging_steps": 25, + "max_steps": 3620, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/trainer_state.json b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..f506cbf374cb1cac7429baa6420c4321739d7679 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379/trainer_state.json @@ -0,0 +1,851 @@ +{ + "best_global_step": 2379, + "best_metric": 0.8268937732765768, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379", + "epoch": 13.0, + "eval_steps": 500, + "global_step": 2379, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13717421124828533, + "grad_norm": 21.41527557373047, + "learning_rate": 1.3114754098360657e-06, + "loss": 4.873, + "step": 25 + }, + { + "epoch": 0.27434842249657065, + "grad_norm": 27.969419479370117, + "learning_rate": 2.677595628415301e-06, + "loss": 4.7529, + "step": 50 + }, + { + "epoch": 0.411522633744856, + "grad_norm": 47.10126495361328, + "learning_rate": 4.043715846994536e-06, + "loss": 4.6215, + "step": 75 + }, + { + "epoch": 0.5486968449931413, + "grad_norm": 40.9450569152832, + "learning_rate": 5.409836065573772e-06, + "loss": 4.5727, + "step": 100 + }, + { + "epoch": 0.6858710562414266, + "grad_norm": 44.236454010009766, + "learning_rate": 6.775956284153006e-06, + "loss": 4.4553, + "step": 125 + }, + { + "epoch": 0.823045267489712, + "grad_norm": 24.01091194152832, + "learning_rate": 8.14207650273224e-06, + "loss": 4.3509, + "step": 150 + }, + { + "epoch": 0.9602194787379973, + "grad_norm": 45.515621185302734, + "learning_rate": 9.508196721311476e-06, + "loss": 4.3353, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5374376039933444, + "eval_loss": 0.9826720952987671, + "eval_macro_f1": 0.5282305995955081, + "eval_runtime": 1.1721, + "eval_samples_per_second": 512.75, + "eval_steps_per_second": 32.42, + "eval_weighted_f1": 0.5290473590640427, + "step": 183 + }, + { + "epoch": 1.093278463648834, + "grad_norm": 65.18627166748047, + "learning_rate": 1.0874316939890712e-05, + "loss": 3.8781, + "step": 200 + }, + { + "epoch": 1.2304526748971194, + "grad_norm": 41.4657096862793, + "learning_rate": 1.2240437158469946e-05, + "loss": 3.9997, + "step": 225 + }, + { + "epoch": 1.3676268861454046, + "grad_norm": 55.039772033691406, + "learning_rate": 1.3606557377049181e-05, + "loss": 3.8176, + "step": 250 + }, + { + "epoch": 1.50480109739369, + "grad_norm": 76.37615966796875, + "learning_rate": 1.4972677595628417e-05, + "loss": 3.6584, + "step": 275 + }, + { + "epoch": 1.6419753086419753, + "grad_norm": 61.83112716674805, + "learning_rate": 1.633879781420765e-05, + "loss": 3.6136, + "step": 300 + }, + { + "epoch": 1.7791495198902605, + "grad_norm": 46.20899963378906, + "learning_rate": 1.7704918032786887e-05, + "loss": 3.5587, + "step": 325 + }, + { + "epoch": 1.916323731138546, + "grad_norm": 294.0778503417969, + "learning_rate": 1.9071038251366122e-05, + "loss": 3.2291, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6306156405990017, + "eval_loss": 0.8852547407150269, + "eval_macro_f1": 0.6110913322659918, + "eval_runtime": 1.1233, + "eval_samples_per_second": 535.039, + "eval_steps_per_second": 33.829, + "eval_weighted_f1": 0.6135340457256488, + "step": 366 + }, + { + "epoch": 2.049382716049383, + "grad_norm": 117.1589584350586, + "learning_rate": 1.995142683667274e-05, + "loss": 2.9501, + "step": 375 + }, + { + "epoch": 2.186556927297668, + "grad_norm": 90.38821411132812, + "learning_rate": 1.9799635701275045e-05, + "loss": 2.7705, + "step": 400 + }, + { + "epoch": 2.3237311385459534, + "grad_norm": 74.47439575195312, + "learning_rate": 1.9647844565877354e-05, + "loss": 2.807, + "step": 425 + }, + { + "epoch": 2.460905349794239, + "grad_norm": 63.590675354003906, + "learning_rate": 1.9496053430479663e-05, + "loss": 2.6937, + "step": 450 + }, + { + "epoch": 2.598079561042524, + "grad_norm": 166.7218475341797, + "learning_rate": 1.934426229508197e-05, + "loss": 2.7012, + "step": 475 + }, + { + "epoch": 2.7352537722908092, + "grad_norm": 53.819976806640625, + "learning_rate": 1.9192471159684274e-05, + "loss": 2.76, + "step": 500 + }, + { + "epoch": 2.8724279835390947, + "grad_norm": 102.4853515625, + "learning_rate": 1.9040680024286583e-05, + "loss": 2.5279, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7770382695507487, + "eval_loss": 0.5550937056541443, + "eval_macro_f1": 0.776457130792325, + "eval_runtime": 1.106, + "eval_samples_per_second": 543.406, + "eval_steps_per_second": 34.358, + "eval_weighted_f1": 0.7770852058772176, + "step": 549 + }, + { + "epoch": 3.0054869684499312, + "grad_norm": 71.87889862060547, + "learning_rate": 1.888888888888889e-05, + "loss": 2.5189, + "step": 550 + }, + { + "epoch": 3.1426611796982167, + "grad_norm": 111.80260467529297, + "learning_rate": 1.8737097753491198e-05, + "loss": 2.2251, + "step": 575 + }, + { + "epoch": 3.279835390946502, + "grad_norm": 45.8128776550293, + "learning_rate": 1.8585306618093504e-05, + "loss": 2.0583, + "step": 600 + }, + { + "epoch": 3.4170096021947876, + "grad_norm": 130.3729705810547, + "learning_rate": 1.8433515482695813e-05, + "loss": 2.1355, + "step": 625 + }, + { + "epoch": 3.5541838134430725, + "grad_norm": 97.28055572509766, + "learning_rate": 1.828172434729812e-05, + "loss": 2.127, + "step": 650 + }, + { + "epoch": 3.691358024691358, + "grad_norm": 130.78652954101562, + "learning_rate": 1.8129933211900424e-05, + "loss": 2.0981, + "step": 675 + }, + { + "epoch": 3.8285322359396434, + "grad_norm": 39.792720794677734, + "learning_rate": 1.7978142076502733e-05, + "loss": 2.0602, + "step": 700 + }, + { + "epoch": 3.9657064471879284, + "grad_norm": 65.91027069091797, + "learning_rate": 1.7826350941105042e-05, + "loss": 2.1078, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7870216306156406, + "eval_loss": 0.5595181584358215, + "eval_macro_f1": 0.7858719460975324, + "eval_runtime": 1.0677, + "eval_samples_per_second": 562.899, + "eval_steps_per_second": 35.591, + "eval_weighted_f1": 0.7866899055646851, + "step": 732 + }, + { + "epoch": 4.098765432098766, + "grad_norm": 123.3680419921875, + "learning_rate": 1.7674559805707348e-05, + "loss": 1.7121, + "step": 750 + }, + { + "epoch": 4.23593964334705, + "grad_norm": 96.52764129638672, + "learning_rate": 1.7522768670309657e-05, + "loss": 1.7204, + "step": 775 + }, + { + "epoch": 4.373113854595336, + "grad_norm": 110.54822540283203, + "learning_rate": 1.7370977534911962e-05, + "loss": 1.6999, + "step": 800 + }, + { + "epoch": 4.510288065843621, + "grad_norm": 68.0560531616211, + "learning_rate": 1.7219186399514268e-05, + "loss": 1.7622, + "step": 825 + }, + { + "epoch": 4.647462277091907, + "grad_norm": 37.093326568603516, + "learning_rate": 1.7067395264116577e-05, + "loss": 1.5891, + "step": 850 + }, + { + "epoch": 4.784636488340192, + "grad_norm": 62.84049987792969, + "learning_rate": 1.6915604128718883e-05, + "loss": 1.6429, + "step": 875 + }, + { + "epoch": 4.921810699588478, + "grad_norm": 109.30193328857422, + "learning_rate": 1.6763812993321192e-05, + "loss": 1.3749, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8119800332778702, + "eval_loss": 0.5119051933288574, + "eval_macro_f1": 0.8097023513901606, + "eval_runtime": 1.1117, + "eval_samples_per_second": 540.626, + "eval_steps_per_second": 34.183, + "eval_weighted_f1": 0.8102721630526095, + "step": 915 + }, + { + "epoch": 5.054869684499314, + "grad_norm": 87.6075668334961, + "learning_rate": 1.66120218579235e-05, + "loss": 1.4281, + "step": 925 + }, + { + "epoch": 5.1920438957476, + "grad_norm": 188.70724487304688, + "learning_rate": 1.6460230722525807e-05, + "loss": 1.3955, + "step": 950 + }, + { + "epoch": 5.329218106995885, + "grad_norm": 133.2298126220703, + "learning_rate": 1.6308439587128112e-05, + "loss": 1.1853, + "step": 975 + }, + { + "epoch": 5.4663923182441705, + "grad_norm": 38.23005294799805, + "learning_rate": 1.615664845173042e-05, + "loss": 1.4112, + "step": 1000 + }, + { + "epoch": 5.603566529492455, + "grad_norm": 95.95108032226562, + "learning_rate": 1.6004857316332727e-05, + "loss": 1.1255, + "step": 1025 + }, + { + "epoch": 5.7407407407407405, + "grad_norm": 61.02497482299805, + "learning_rate": 1.5853066180935036e-05, + "loss": 1.3015, + "step": 1050 + }, + { + "epoch": 5.877914951989026, + "grad_norm": 130.4191436767578, + "learning_rate": 1.570127504553734e-05, + "loss": 1.3481, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8053244592346089, + "eval_loss": 0.5796843767166138, + "eval_macro_f1": 0.8030593982695837, + "eval_runtime": 1.1006, + "eval_samples_per_second": 546.052, + "eval_steps_per_second": 34.526, + "eval_weighted_f1": 0.8032471405385242, + "step": 1098 + }, + { + "epoch": 6.0109739368998625, + "grad_norm": 83.11539459228516, + "learning_rate": 1.5549483910139647e-05, + "loss": 1.2312, + "step": 1100 + }, + { + "epoch": 6.148148148148148, + "grad_norm": 85.27437591552734, + "learning_rate": 1.5397692774741956e-05, + "loss": 1.0447, + "step": 1125 + }, + { + "epoch": 6.285322359396433, + "grad_norm": 113.0291519165039, + "learning_rate": 1.5245901639344264e-05, + "loss": 1.0277, + "step": 1150 + }, + { + "epoch": 6.422496570644719, + "grad_norm": 151.12319946289062, + "learning_rate": 1.5094110503946571e-05, + "loss": 1.053, + "step": 1175 + }, + { + "epoch": 6.559670781893004, + "grad_norm": 27.48976707458496, + "learning_rate": 1.4942319368548878e-05, + "loss": 0.9557, + "step": 1200 + }, + { + "epoch": 6.69684499314129, + "grad_norm": 75.97730255126953, + "learning_rate": 1.4790528233151184e-05, + "loss": 0.854, + "step": 1225 + }, + { + "epoch": 6.834019204389575, + "grad_norm": 45.34609603881836, + "learning_rate": 1.4638737097753493e-05, + "loss": 1.0633, + "step": 1250 + }, + { + "epoch": 6.97119341563786, + "grad_norm": 49.1536750793457, + "learning_rate": 1.44869459623558e-05, + "loss": 0.9677, + "step": 1275 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8202995008319468, + "eval_loss": 0.6126930117607117, + "eval_macro_f1": 0.8186216714076312, + "eval_runtime": 1.0826, + "eval_samples_per_second": 555.138, + "eval_steps_per_second": 35.1, + "eval_weighted_f1": 0.818687691247028, + "step": 1281 + }, + { + "epoch": 7.104252400548697, + "grad_norm": 209.2027587890625, + "learning_rate": 1.4335154826958106e-05, + "loss": 0.7649, + "step": 1300 + }, + { + "epoch": 7.2414266117969825, + "grad_norm": 73.82598876953125, + "learning_rate": 1.4183363691560415e-05, + "loss": 0.6279, + "step": 1325 + }, + { + "epoch": 7.378600823045267, + "grad_norm": 241.08839416503906, + "learning_rate": 1.403157255616272e-05, + "loss": 0.7884, + "step": 1350 + }, + { + "epoch": 7.5157750342935525, + "grad_norm": 166.5994873046875, + "learning_rate": 1.3879781420765028e-05, + "loss": 0.7833, + "step": 1375 + }, + { + "epoch": 7.652949245541838, + "grad_norm": 92.84734344482422, + "learning_rate": 1.3727990285367337e-05, + "loss": 0.7483, + "step": 1400 + }, + { + "epoch": 7.790123456790123, + "grad_norm": 101.81146240234375, + "learning_rate": 1.3576199149969643e-05, + "loss": 0.8631, + "step": 1425 + }, + { + "epoch": 7.927297668038409, + "grad_norm": 143.8651123046875, + "learning_rate": 1.342440801457195e-05, + "loss": 0.9062, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8236272878535774, + "eval_loss": 0.6300873756408691, + "eval_macro_f1": 0.8229516649507022, + "eval_runtime": 1.0782, + "eval_samples_per_second": 557.396, + "eval_steps_per_second": 35.243, + "eval_weighted_f1": 0.8226756348885598, + "step": 1464 + }, + { + "epoch": 8.060356652949245, + "grad_norm": 185.40240478515625, + "learning_rate": 1.3272616879174259e-05, + "loss": 0.7586, + "step": 1475 + }, + { + "epoch": 8.197530864197532, + "grad_norm": 111.78955078125, + "learning_rate": 1.3120825743776565e-05, + "loss": 0.5159, + "step": 1500 + }, + { + "epoch": 8.334705075445816, + "grad_norm": 120.39631652832031, + "learning_rate": 1.2969034608378872e-05, + "loss": 0.6553, + "step": 1525 + }, + { + "epoch": 8.4718792866941, + "grad_norm": 40.821014404296875, + "learning_rate": 1.2817243472981178e-05, + "loss": 0.4416, + "step": 1550 + }, + { + "epoch": 8.609053497942387, + "grad_norm": 45.94298553466797, + "learning_rate": 1.2665452337583487e-05, + "loss": 0.6887, + "step": 1575 + }, + { + "epoch": 8.746227709190672, + "grad_norm": 47.62234115600586, + "learning_rate": 1.2513661202185794e-05, + "loss": 0.5083, + "step": 1600 + }, + { + "epoch": 8.883401920438958, + "grad_norm": 140.30810546875, + "learning_rate": 1.23618700667881e-05, + "loss": 0.6989, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8153078202995009, + "eval_loss": 0.8748893141746521, + "eval_macro_f1": 0.8140017520902462, + "eval_runtime": 1.0952, + "eval_samples_per_second": 548.759, + "eval_steps_per_second": 34.697, + "eval_weighted_f1": 0.8138648294262724, + "step": 1647 + }, + { + "epoch": 9.016460905349794, + "grad_norm": 75.1722412109375, + "learning_rate": 1.2210078931390409e-05, + "loss": 0.3881, + "step": 1650 + }, + { + "epoch": 9.15363511659808, + "grad_norm": 74.72351837158203, + "learning_rate": 1.2058287795992716e-05, + "loss": 0.3156, + "step": 1675 + }, + { + "epoch": 9.290809327846365, + "grad_norm": 165.40896606445312, + "learning_rate": 1.1906496660595022e-05, + "loss": 0.5666, + "step": 1700 + }, + { + "epoch": 9.427983539094651, + "grad_norm": 9.87984848022461, + "learning_rate": 1.1754705525197329e-05, + "loss": 0.5101, + "step": 1725 + }, + { + "epoch": 9.565157750342935, + "grad_norm": 115.18270874023438, + "learning_rate": 1.1602914389799638e-05, + "loss": 0.4209, + "step": 1750 + }, + { + "epoch": 9.70233196159122, + "grad_norm": 221.89083862304688, + "learning_rate": 1.1451123254401944e-05, + "loss": 0.4536, + "step": 1775 + }, + { + "epoch": 9.839506172839506, + "grad_norm": 225.4305877685547, + "learning_rate": 1.1299332119004251e-05, + "loss": 0.3954, + "step": 1800 + }, + { + "epoch": 9.976680384087791, + "grad_norm": 222.6260986328125, + "learning_rate": 1.1147540983606557e-05, + "loss": 0.5151, + "step": 1825 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8219633943427621, + "eval_loss": 0.934634804725647, + "eval_macro_f1": 0.8217934616964605, + "eval_runtime": 1.1187, + "eval_samples_per_second": 537.254, + "eval_steps_per_second": 33.969, + "eval_weighted_f1": 0.8217971218520704, + "step": 1830 + }, + { + "epoch": 10.109739368998628, + "grad_norm": 16.582561492919922, + "learning_rate": 1.0995749848208866e-05, + "loss": 0.322, + "step": 1850 + }, + { + "epoch": 10.246913580246913, + "grad_norm": 61.523067474365234, + "learning_rate": 1.0843958712811173e-05, + "loss": 0.291, + "step": 1875 + }, + { + "epoch": 10.3840877914952, + "grad_norm": 161.82211303710938, + "learning_rate": 1.0692167577413479e-05, + "loss": 0.3396, + "step": 1900 + }, + { + "epoch": 10.521262002743484, + "grad_norm": 100.72151947021484, + "learning_rate": 1.0540376442015788e-05, + "loss": 0.4572, + "step": 1925 + }, + { + "epoch": 10.65843621399177, + "grad_norm": 99.69158172607422, + "learning_rate": 1.0388585306618095e-05, + "loss": 0.4253, + "step": 1950 + }, + { + "epoch": 10.795610425240055, + "grad_norm": 81.34037017822266, + "learning_rate": 1.02367941712204e-05, + "loss": 0.3459, + "step": 1975 + }, + { + "epoch": 10.932784636488341, + "grad_norm": 125.5887451171875, + "learning_rate": 1.008500303582271e-05, + "loss": 0.3968, + "step": 2000 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.826955074875208, + "eval_loss": 1.1937121152877808, + "eval_macro_f1": 0.825301093057823, + "eval_runtime": 1.1086, + "eval_samples_per_second": 542.148, + "eval_steps_per_second": 34.279, + "eval_weighted_f1": 0.8254728469689576, + "step": 2013 + }, + { + "epoch": 11.065843621399177, + "grad_norm": 3.588291883468628, + "learning_rate": 9.933211900425015e-06, + "loss": 0.3265, + "step": 2025 + }, + { + "epoch": 11.203017832647463, + "grad_norm": 283.16986083984375, + "learning_rate": 9.781420765027323e-06, + "loss": 0.387, + "step": 2050 + }, + { + "epoch": 11.340192043895748, + "grad_norm": 12.006860733032227, + "learning_rate": 9.62962962962963e-06, + "loss": 0.127, + "step": 2075 + }, + { + "epoch": 11.477366255144032, + "grad_norm": 1.6399710178375244, + "learning_rate": 9.477838494231937e-06, + "loss": 0.128, + "step": 2100 + }, + { + "epoch": 11.614540466392318, + "grad_norm": 47.6485710144043, + "learning_rate": 9.326047358834245e-06, + "loss": 0.3986, + "step": 2125 + }, + { + "epoch": 11.751714677640603, + "grad_norm": 7.119675159454346, + "learning_rate": 9.174256223436552e-06, + "loss": 0.3773, + "step": 2150 + }, + { + "epoch": 11.88888888888889, + "grad_norm": 1.0404638051986694, + "learning_rate": 9.02246508803886e-06, + "loss": 0.3928, + "step": 2175 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.826955074875208, + "eval_loss": 1.356019139289856, + "eval_macro_f1": 0.8265982383629442, + "eval_runtime": 1.1093, + "eval_samples_per_second": 541.775, + "eval_steps_per_second": 34.255, + "eval_weighted_f1": 0.8263752820863519, + "step": 2196 + }, + { + "epoch": 12.021947873799725, + "grad_norm": 0.7447333335876465, + "learning_rate": 8.870673952641167e-06, + "loss": 0.2426, + "step": 2200 + }, + { + "epoch": 12.159122085048011, + "grad_norm": 213.7488250732422, + "learning_rate": 8.718882817243474e-06, + "loss": 0.2819, + "step": 2225 + }, + { + "epoch": 12.296296296296296, + "grad_norm": 0.06055834889411926, + "learning_rate": 8.567091681845781e-06, + "loss": 0.175, + "step": 2250 + }, + { + "epoch": 12.433470507544582, + "grad_norm": 316.73077392578125, + "learning_rate": 8.415300546448089e-06, + "loss": 0.1941, + "step": 2275 + }, + { + "epoch": 12.570644718792867, + "grad_norm": 389.9907531738281, + "learning_rate": 8.263509411050396e-06, + "loss": 0.212, + "step": 2300 + }, + { + "epoch": 12.707818930041153, + "grad_norm": 0.12231023609638214, + "learning_rate": 8.111718275652702e-06, + "loss": 0.1736, + "step": 2325 + }, + { + "epoch": 12.844993141289438, + "grad_norm": 55.67293930053711, + "learning_rate": 7.959927140255009e-06, + "loss": 0.2743, + "step": 2350 + }, + { + "epoch": 12.982167352537722, + "grad_norm": 0.07137472182512283, + "learning_rate": 7.808136004857318e-06, + "loss": 0.1396, + "step": 2375 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.826955074875208, + "eval_loss": 1.5114561319351196, + "eval_macro_f1": 0.8268937732765768, + "eval_runtime": 1.1418, + "eval_samples_per_second": 526.375, + "eval_steps_per_second": 33.282, + "eval_weighted_f1": 0.826584598306716, + "step": 2379 + } + ], + "logging_steps": 25, + "max_steps": 3660, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/trainer_state.json b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..b05308e8b3d972c4ac6b22d8d0af4a889cd986f5 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2928/trainer_state.json @@ -0,0 +1,1038 @@ +{ + "best_global_step": 2379, + "best_metric": 0.8268937732765768, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-03-s42-20260901T154226Z-27bda108/trainer/checkpoint-2379", + "epoch": 16.0, + "eval_steps": 500, + "global_step": 2928, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13717421124828533, + "grad_norm": 21.41527557373047, + "learning_rate": 1.3114754098360657e-06, + "loss": 4.873, + "step": 25 + }, + { + "epoch": 0.27434842249657065, + "grad_norm": 27.969419479370117, + "learning_rate": 2.677595628415301e-06, + "loss": 4.7529, + "step": 50 + }, + { + "epoch": 0.411522633744856, + "grad_norm": 47.10126495361328, + "learning_rate": 4.043715846994536e-06, + "loss": 4.6215, + "step": 75 + }, + { + "epoch": 0.5486968449931413, + "grad_norm": 40.9450569152832, + "learning_rate": 5.409836065573772e-06, + "loss": 4.5727, + "step": 100 + }, + { + "epoch": 0.6858710562414266, + "grad_norm": 44.236454010009766, + "learning_rate": 6.775956284153006e-06, + "loss": 4.4553, + "step": 125 + }, + { + "epoch": 0.823045267489712, + "grad_norm": 24.01091194152832, + "learning_rate": 8.14207650273224e-06, + "loss": 4.3509, + "step": 150 + }, + { + "epoch": 0.9602194787379973, + "grad_norm": 45.515621185302734, + "learning_rate": 9.508196721311476e-06, + "loss": 4.3353, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5374376039933444, + "eval_loss": 0.9826720952987671, + "eval_macro_f1": 0.5282305995955081, + "eval_runtime": 1.1721, + "eval_samples_per_second": 512.75, + "eval_steps_per_second": 32.42, + "eval_weighted_f1": 0.5290473590640427, + "step": 183 + }, + { + "epoch": 1.093278463648834, + "grad_norm": 65.18627166748047, + "learning_rate": 1.0874316939890712e-05, + "loss": 3.8781, + "step": 200 + }, + { + "epoch": 1.2304526748971194, + "grad_norm": 41.4657096862793, + "learning_rate": 1.2240437158469946e-05, + "loss": 3.9997, + "step": 225 + }, + { + "epoch": 1.3676268861454046, + "grad_norm": 55.039772033691406, + "learning_rate": 1.3606557377049181e-05, + "loss": 3.8176, + "step": 250 + }, + { + "epoch": 1.50480109739369, + "grad_norm": 76.37615966796875, + "learning_rate": 1.4972677595628417e-05, + "loss": 3.6584, + "step": 275 + }, + { + "epoch": 1.6419753086419753, + "grad_norm": 61.83112716674805, + "learning_rate": 1.633879781420765e-05, + "loss": 3.6136, + "step": 300 + }, + { + "epoch": 1.7791495198902605, + "grad_norm": 46.20899963378906, + "learning_rate": 1.7704918032786887e-05, + "loss": 3.5587, + "step": 325 + }, + { + "epoch": 1.916323731138546, + "grad_norm": 294.0778503417969, + "learning_rate": 1.9071038251366122e-05, + "loss": 3.2291, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6306156405990017, + "eval_loss": 0.8852547407150269, + "eval_macro_f1": 0.6110913322659918, + "eval_runtime": 1.1233, + "eval_samples_per_second": 535.039, + "eval_steps_per_second": 33.829, + "eval_weighted_f1": 0.6135340457256488, + "step": 366 + }, + { + "epoch": 2.049382716049383, + "grad_norm": 117.1589584350586, + "learning_rate": 1.995142683667274e-05, + "loss": 2.9501, + "step": 375 + }, + { + "epoch": 2.186556927297668, + "grad_norm": 90.38821411132812, + "learning_rate": 1.9799635701275045e-05, + "loss": 2.7705, + "step": 400 + }, + { + "epoch": 2.3237311385459534, + "grad_norm": 74.47439575195312, + "learning_rate": 1.9647844565877354e-05, + "loss": 2.807, + "step": 425 + }, + { + "epoch": 2.460905349794239, + "grad_norm": 63.590675354003906, + "learning_rate": 1.9496053430479663e-05, + "loss": 2.6937, + "step": 450 + }, + { + "epoch": 2.598079561042524, + "grad_norm": 166.7218475341797, + "learning_rate": 1.934426229508197e-05, + "loss": 2.7012, + "step": 475 + }, + { + "epoch": 2.7352537722908092, + "grad_norm": 53.819976806640625, + "learning_rate": 1.9192471159684274e-05, + "loss": 2.76, + "step": 500 + }, + { + "epoch": 2.8724279835390947, + "grad_norm": 102.4853515625, + "learning_rate": 1.9040680024286583e-05, + "loss": 2.5279, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7770382695507487, + "eval_loss": 0.5550937056541443, + "eval_macro_f1": 0.776457130792325, + "eval_runtime": 1.106, + "eval_samples_per_second": 543.406, + "eval_steps_per_second": 34.358, + "eval_weighted_f1": 0.7770852058772176, + "step": 549 + }, + { + "epoch": 3.0054869684499312, + "grad_norm": 71.87889862060547, + "learning_rate": 1.888888888888889e-05, + "loss": 2.5189, + "step": 550 + }, + { + "epoch": 3.1426611796982167, + "grad_norm": 111.80260467529297, + "learning_rate": 1.8737097753491198e-05, + "loss": 2.2251, + "step": 575 + }, + { + "epoch": 3.279835390946502, + "grad_norm": 45.8128776550293, + "learning_rate": 1.8585306618093504e-05, + "loss": 2.0583, + "step": 600 + }, + { + "epoch": 3.4170096021947876, + "grad_norm": 130.3729705810547, + "learning_rate": 1.8433515482695813e-05, + "loss": 2.1355, + "step": 625 + }, + { + "epoch": 3.5541838134430725, + "grad_norm": 97.28055572509766, + "learning_rate": 1.828172434729812e-05, + "loss": 2.127, + "step": 650 + }, + { + "epoch": 3.691358024691358, + "grad_norm": 130.78652954101562, + "learning_rate": 1.8129933211900424e-05, + "loss": 2.0981, + "step": 675 + }, + { + "epoch": 3.8285322359396434, + "grad_norm": 39.792720794677734, + "learning_rate": 1.7978142076502733e-05, + "loss": 2.0602, + "step": 700 + }, + { + "epoch": 3.9657064471879284, + "grad_norm": 65.91027069091797, + "learning_rate": 1.7826350941105042e-05, + "loss": 2.1078, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7870216306156406, + "eval_loss": 0.5595181584358215, + "eval_macro_f1": 0.7858719460975324, + "eval_runtime": 1.0677, + "eval_samples_per_second": 562.899, + "eval_steps_per_second": 35.591, + "eval_weighted_f1": 0.7866899055646851, + "step": 732 + }, + { + "epoch": 4.098765432098766, + "grad_norm": 123.3680419921875, + "learning_rate": 1.7674559805707348e-05, + "loss": 1.7121, + "step": 750 + }, + { + "epoch": 4.23593964334705, + "grad_norm": 96.52764129638672, + "learning_rate": 1.7522768670309657e-05, + "loss": 1.7204, + "step": 775 + }, + { + "epoch": 4.373113854595336, + "grad_norm": 110.54822540283203, + "learning_rate": 1.7370977534911962e-05, + "loss": 1.6999, + "step": 800 + }, + { + "epoch": 4.510288065843621, + "grad_norm": 68.0560531616211, + "learning_rate": 1.7219186399514268e-05, + "loss": 1.7622, + "step": 825 + }, + { + "epoch": 4.647462277091907, + "grad_norm": 37.093326568603516, + "learning_rate": 1.7067395264116577e-05, + "loss": 1.5891, + "step": 850 + }, + { + "epoch": 4.784636488340192, + "grad_norm": 62.84049987792969, + "learning_rate": 1.6915604128718883e-05, + "loss": 1.6429, + "step": 875 + }, + { + "epoch": 4.921810699588478, + "grad_norm": 109.30193328857422, + "learning_rate": 1.6763812993321192e-05, + "loss": 1.3749, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8119800332778702, + "eval_loss": 0.5119051933288574, + "eval_macro_f1": 0.8097023513901606, + "eval_runtime": 1.1117, + "eval_samples_per_second": 540.626, + "eval_steps_per_second": 34.183, + "eval_weighted_f1": 0.8102721630526095, + "step": 915 + }, + { + "epoch": 5.054869684499314, + "grad_norm": 87.6075668334961, + "learning_rate": 1.66120218579235e-05, + "loss": 1.4281, + "step": 925 + }, + { + "epoch": 5.1920438957476, + "grad_norm": 188.70724487304688, + "learning_rate": 1.6460230722525807e-05, + "loss": 1.3955, + "step": 950 + }, + { + "epoch": 5.329218106995885, + "grad_norm": 133.2298126220703, + "learning_rate": 1.6308439587128112e-05, + "loss": 1.1853, + "step": 975 + }, + { + "epoch": 5.4663923182441705, + "grad_norm": 38.23005294799805, + "learning_rate": 1.615664845173042e-05, + "loss": 1.4112, + "step": 1000 + }, + { + "epoch": 5.603566529492455, + "grad_norm": 95.95108032226562, + "learning_rate": 1.6004857316332727e-05, + "loss": 1.1255, + "step": 1025 + }, + { + "epoch": 5.7407407407407405, + "grad_norm": 61.02497482299805, + "learning_rate": 1.5853066180935036e-05, + "loss": 1.3015, + "step": 1050 + }, + { + "epoch": 5.877914951989026, + "grad_norm": 130.4191436767578, + "learning_rate": 1.570127504553734e-05, + "loss": 1.3481, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8053244592346089, + "eval_loss": 0.5796843767166138, + "eval_macro_f1": 0.8030593982695837, + "eval_runtime": 1.1006, + "eval_samples_per_second": 546.052, + "eval_steps_per_second": 34.526, + "eval_weighted_f1": 0.8032471405385242, + "step": 1098 + }, + { + "epoch": 6.0109739368998625, + "grad_norm": 83.11539459228516, + "learning_rate": 1.5549483910139647e-05, + "loss": 1.2312, + "step": 1100 + }, + { + "epoch": 6.148148148148148, + "grad_norm": 85.27437591552734, + "learning_rate": 1.5397692774741956e-05, + "loss": 1.0447, + "step": 1125 + }, + { + "epoch": 6.285322359396433, + "grad_norm": 113.0291519165039, + "learning_rate": 1.5245901639344264e-05, + "loss": 1.0277, + "step": 1150 + }, + { + "epoch": 6.422496570644719, + "grad_norm": 151.12319946289062, + "learning_rate": 1.5094110503946571e-05, + "loss": 1.053, + "step": 1175 + }, + { + "epoch": 6.559670781893004, + "grad_norm": 27.48976707458496, + "learning_rate": 1.4942319368548878e-05, + "loss": 0.9557, + "step": 1200 + }, + { + "epoch": 6.69684499314129, + "grad_norm": 75.97730255126953, + "learning_rate": 1.4790528233151184e-05, + "loss": 0.854, + "step": 1225 + }, + { + "epoch": 6.834019204389575, + "grad_norm": 45.34609603881836, + "learning_rate": 1.4638737097753493e-05, + "loss": 1.0633, + "step": 1250 + }, + { + "epoch": 6.97119341563786, + "grad_norm": 49.1536750793457, + "learning_rate": 1.44869459623558e-05, + "loss": 0.9677, + "step": 1275 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8202995008319468, + "eval_loss": 0.6126930117607117, + "eval_macro_f1": 0.8186216714076312, + "eval_runtime": 1.0826, + "eval_samples_per_second": 555.138, + "eval_steps_per_second": 35.1, + "eval_weighted_f1": 0.818687691247028, + "step": 1281 + }, + { + "epoch": 7.104252400548697, + "grad_norm": 209.2027587890625, + "learning_rate": 1.4335154826958106e-05, + "loss": 0.7649, + "step": 1300 + }, + { + "epoch": 7.2414266117969825, + "grad_norm": 73.82598876953125, + "learning_rate": 1.4183363691560415e-05, + "loss": 0.6279, + "step": 1325 + }, + { + "epoch": 7.378600823045267, + "grad_norm": 241.08839416503906, + "learning_rate": 1.403157255616272e-05, + "loss": 0.7884, + "step": 1350 + }, + { + "epoch": 7.5157750342935525, + "grad_norm": 166.5994873046875, + "learning_rate": 1.3879781420765028e-05, + "loss": 0.7833, + "step": 1375 + }, + { + "epoch": 7.652949245541838, + "grad_norm": 92.84734344482422, + "learning_rate": 1.3727990285367337e-05, + "loss": 0.7483, + "step": 1400 + }, + { + "epoch": 7.790123456790123, + "grad_norm": 101.81146240234375, + "learning_rate": 1.3576199149969643e-05, + "loss": 0.8631, + "step": 1425 + }, + { + "epoch": 7.927297668038409, + "grad_norm": 143.8651123046875, + "learning_rate": 1.342440801457195e-05, + "loss": 0.9062, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8236272878535774, + "eval_loss": 0.6300873756408691, + "eval_macro_f1": 0.8229516649507022, + "eval_runtime": 1.0782, + "eval_samples_per_second": 557.396, + "eval_steps_per_second": 35.243, + "eval_weighted_f1": 0.8226756348885598, + "step": 1464 + }, + { + "epoch": 8.060356652949245, + "grad_norm": 185.40240478515625, + "learning_rate": 1.3272616879174259e-05, + "loss": 0.7586, + "step": 1475 + }, + { + "epoch": 8.197530864197532, + "grad_norm": 111.78955078125, + "learning_rate": 1.3120825743776565e-05, + "loss": 0.5159, + "step": 1500 + }, + { + "epoch": 8.334705075445816, + "grad_norm": 120.39631652832031, + "learning_rate": 1.2969034608378872e-05, + "loss": 0.6553, + "step": 1525 + }, + { + "epoch": 8.4718792866941, + "grad_norm": 40.821014404296875, + "learning_rate": 1.2817243472981178e-05, + "loss": 0.4416, + "step": 1550 + }, + { + "epoch": 8.609053497942387, + "grad_norm": 45.94298553466797, + "learning_rate": 1.2665452337583487e-05, + "loss": 0.6887, + "step": 1575 + }, + { + "epoch": 8.746227709190672, + "grad_norm": 47.62234115600586, + "learning_rate": 1.2513661202185794e-05, + "loss": 0.5083, + "step": 1600 + }, + { + "epoch": 8.883401920438958, + "grad_norm": 140.30810546875, + "learning_rate": 1.23618700667881e-05, + "loss": 0.6989, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8153078202995009, + "eval_loss": 0.8748893141746521, + "eval_macro_f1": 0.8140017520902462, + "eval_runtime": 1.0952, + "eval_samples_per_second": 548.759, + "eval_steps_per_second": 34.697, + "eval_weighted_f1": 0.8138648294262724, + "step": 1647 + }, + { + "epoch": 9.016460905349794, + "grad_norm": 75.1722412109375, + "learning_rate": 1.2210078931390409e-05, + "loss": 0.3881, + "step": 1650 + }, + { + "epoch": 9.15363511659808, + "grad_norm": 74.72351837158203, + "learning_rate": 1.2058287795992716e-05, + "loss": 0.3156, + "step": 1675 + }, + { + "epoch": 9.290809327846365, + "grad_norm": 165.40896606445312, + "learning_rate": 1.1906496660595022e-05, + "loss": 0.5666, + "step": 1700 + }, + { + "epoch": 9.427983539094651, + "grad_norm": 9.87984848022461, + "learning_rate": 1.1754705525197329e-05, + "loss": 0.5101, + "step": 1725 + }, + { + "epoch": 9.565157750342935, + "grad_norm": 115.18270874023438, + "learning_rate": 1.1602914389799638e-05, + "loss": 0.4209, + "step": 1750 + }, + { + "epoch": 9.70233196159122, + "grad_norm": 221.89083862304688, + "learning_rate": 1.1451123254401944e-05, + "loss": 0.4536, + "step": 1775 + }, + { + "epoch": 9.839506172839506, + "grad_norm": 225.4305877685547, + "learning_rate": 1.1299332119004251e-05, + "loss": 0.3954, + "step": 1800 + }, + { + "epoch": 9.976680384087791, + "grad_norm": 222.6260986328125, + "learning_rate": 1.1147540983606557e-05, + "loss": 0.5151, + "step": 1825 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8219633943427621, + "eval_loss": 0.934634804725647, + "eval_macro_f1": 0.8217934616964605, + "eval_runtime": 1.1187, + "eval_samples_per_second": 537.254, + "eval_steps_per_second": 33.969, + "eval_weighted_f1": 0.8217971218520704, + "step": 1830 + }, + { + "epoch": 10.109739368998628, + "grad_norm": 16.582561492919922, + "learning_rate": 1.0995749848208866e-05, + "loss": 0.322, + "step": 1850 + }, + { + "epoch": 10.246913580246913, + "grad_norm": 61.523067474365234, + "learning_rate": 1.0843958712811173e-05, + "loss": 0.291, + "step": 1875 + }, + { + "epoch": 10.3840877914952, + "grad_norm": 161.82211303710938, + "learning_rate": 1.0692167577413479e-05, + "loss": 0.3396, + "step": 1900 + }, + { + "epoch": 10.521262002743484, + "grad_norm": 100.72151947021484, + "learning_rate": 1.0540376442015788e-05, + "loss": 0.4572, + "step": 1925 + }, + { + "epoch": 10.65843621399177, + "grad_norm": 99.69158172607422, + "learning_rate": 1.0388585306618095e-05, + "loss": 0.4253, + "step": 1950 + }, + { + "epoch": 10.795610425240055, + "grad_norm": 81.34037017822266, + "learning_rate": 1.02367941712204e-05, + "loss": 0.3459, + "step": 1975 + }, + { + "epoch": 10.932784636488341, + "grad_norm": 125.5887451171875, + "learning_rate": 1.008500303582271e-05, + "loss": 0.3968, + "step": 2000 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.826955074875208, + "eval_loss": 1.1937121152877808, + "eval_macro_f1": 0.825301093057823, + "eval_runtime": 1.1086, + "eval_samples_per_second": 542.148, + "eval_steps_per_second": 34.279, + "eval_weighted_f1": 0.8254728469689576, + "step": 2013 + }, + { + "epoch": 11.065843621399177, + "grad_norm": 3.588291883468628, + "learning_rate": 9.933211900425015e-06, + "loss": 0.3265, + "step": 2025 + }, + { + "epoch": 11.203017832647463, + "grad_norm": 283.16986083984375, + "learning_rate": 9.781420765027323e-06, + "loss": 0.387, + "step": 2050 + }, + { + "epoch": 11.340192043895748, + "grad_norm": 12.006860733032227, + "learning_rate": 9.62962962962963e-06, + "loss": 0.127, + "step": 2075 + }, + { + "epoch": 11.477366255144032, + "grad_norm": 1.6399710178375244, + "learning_rate": 9.477838494231937e-06, + "loss": 0.128, + "step": 2100 + }, + { + "epoch": 11.614540466392318, + "grad_norm": 47.6485710144043, + "learning_rate": 9.326047358834245e-06, + "loss": 0.3986, + "step": 2125 + }, + { + "epoch": 11.751714677640603, + "grad_norm": 7.119675159454346, + "learning_rate": 9.174256223436552e-06, + "loss": 0.3773, + "step": 2150 + }, + { + "epoch": 11.88888888888889, + "grad_norm": 1.0404638051986694, + "learning_rate": 9.02246508803886e-06, + "loss": 0.3928, + "step": 2175 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.826955074875208, + "eval_loss": 1.356019139289856, + "eval_macro_f1": 0.8265982383629442, + "eval_runtime": 1.1093, + "eval_samples_per_second": 541.775, + "eval_steps_per_second": 34.255, + "eval_weighted_f1": 0.8263752820863519, + "step": 2196 + }, + { + "epoch": 12.021947873799725, + "grad_norm": 0.7447333335876465, + "learning_rate": 8.870673952641167e-06, + "loss": 0.2426, + "step": 2200 + }, + { + "epoch": 12.159122085048011, + "grad_norm": 213.7488250732422, + "learning_rate": 8.718882817243474e-06, + "loss": 0.2819, + "step": 2225 + }, + { + "epoch": 12.296296296296296, + "grad_norm": 0.06055834889411926, + "learning_rate": 8.567091681845781e-06, + "loss": 0.175, + "step": 2250 + }, + { + "epoch": 12.433470507544582, + "grad_norm": 316.73077392578125, + "learning_rate": 8.415300546448089e-06, + "loss": 0.1941, + "step": 2275 + }, + { + "epoch": 12.570644718792867, + "grad_norm": 389.9907531738281, + "learning_rate": 8.263509411050396e-06, + "loss": 0.212, + "step": 2300 + }, + { + "epoch": 12.707818930041153, + "grad_norm": 0.12231023609638214, + "learning_rate": 8.111718275652702e-06, + "loss": 0.1736, + "step": 2325 + }, + { + "epoch": 12.844993141289438, + "grad_norm": 55.67293930053711, + "learning_rate": 7.959927140255009e-06, + "loss": 0.2743, + "step": 2350 + }, + { + "epoch": 12.982167352537722, + "grad_norm": 0.07137472182512283, + "learning_rate": 7.808136004857318e-06, + "loss": 0.1396, + "step": 2375 + }, + { + "epoch": 13.0, + "eval_accuracy": 0.826955074875208, + "eval_loss": 1.5114561319351196, + "eval_macro_f1": 0.8268937732765768, + "eval_runtime": 1.1418, + "eval_samples_per_second": 526.375, + "eval_steps_per_second": 33.282, + "eval_weighted_f1": 0.826584598306716, + "step": 2379 + }, + { + "epoch": 13.11522633744856, + "grad_norm": 0.06866336613893509, + "learning_rate": 7.656344869459624e-06, + "loss": 0.3038, + "step": 2400 + }, + { + "epoch": 13.252400548696844, + "grad_norm": 1.099043607711792, + "learning_rate": 7.504553734061931e-06, + "loss": 0.2297, + "step": 2425 + }, + { + "epoch": 13.38957475994513, + "grad_norm": 0.23049405217170715, + "learning_rate": 7.352762598664238e-06, + "loss": 0.2211, + "step": 2450 + }, + { + "epoch": 13.526748971193415, + "grad_norm": 355.7548522949219, + "learning_rate": 7.2009714632665465e-06, + "loss": 0.2186, + "step": 2475 + }, + { + "epoch": 13.663923182441701, + "grad_norm": 26.790739059448242, + "learning_rate": 7.049180327868853e-06, + "loss": 0.1129, + "step": 2500 + }, + { + "epoch": 13.801097393689986, + "grad_norm": 4.913447856903076, + "learning_rate": 6.89738919247116e-06, + "loss": 0.2847, + "step": 2525 + }, + { + "epoch": 13.938271604938272, + "grad_norm": 0.08487499505281448, + "learning_rate": 6.745598057073467e-06, + "loss": 0.1848, + "step": 2550 + }, + { + "epoch": 14.0, + "eval_accuracy": 0.8202995008319468, + "eval_loss": 1.6732088327407837, + "eval_macro_f1": 0.8194117824128891, + "eval_runtime": 1.1426, + "eval_samples_per_second": 525.977, + "eval_steps_per_second": 33.256, + "eval_weighted_f1": 0.8193587313610873, + "step": 2562 + }, + { + "epoch": 14.071330589849108, + "grad_norm": 27.14590072631836, + "learning_rate": 6.593806921675775e-06, + "loss": 0.1607, + "step": 2575 + }, + { + "epoch": 14.208504801097394, + "grad_norm": 0.42849478125572205, + "learning_rate": 6.4420157862780815e-06, + "loss": 0.2167, + "step": 2600 + }, + { + "epoch": 14.345679012345679, + "grad_norm": 0.6335909366607666, + "learning_rate": 6.290224650880389e-06, + "loss": 0.149, + "step": 2625 + }, + { + "epoch": 14.482853223593965, + "grad_norm": 36.388126373291016, + "learning_rate": 6.138433515482697e-06, + "loss": 0.2526, + "step": 2650 + }, + { + "epoch": 14.62002743484225, + "grad_norm": 0.6127989888191223, + "learning_rate": 5.9866423800850036e-06, + "loss": 0.1268, + "step": 2675 + }, + { + "epoch": 14.757201646090534, + "grad_norm": 0.47106754779815674, + "learning_rate": 5.834851244687311e-06, + "loss": 0.3489, + "step": 2700 + }, + { + "epoch": 14.89437585733882, + "grad_norm": 446.8983459472656, + "learning_rate": 5.683060109289617e-06, + "loss": 0.1664, + "step": 2725 + }, + { + "epoch": 15.0, + "eval_accuracy": 0.8186356073211315, + "eval_loss": 1.6188260316848755, + "eval_macro_f1": 0.8188601288127355, + "eval_runtime": 1.159, + "eval_samples_per_second": 518.569, + "eval_steps_per_second": 32.788, + "eval_weighted_f1": 0.8185505625270314, + "step": 2745 + }, + { + "epoch": 15.027434842249656, + "grad_norm": 154.9385223388672, + "learning_rate": 5.5312689738919256e-06, + "loss": 0.1307, + "step": 2750 + }, + { + "epoch": 15.164609053497943, + "grad_norm": 4.390649795532227, + "learning_rate": 5.379477838494233e-06, + "loss": 0.1564, + "step": 2775 + }, + { + "epoch": 15.301783264746227, + "grad_norm": 0.0765179693698883, + "learning_rate": 5.227686703096539e-06, + "loss": 0.1977, + "step": 2800 + }, + { + "epoch": 15.438957475994513, + "grad_norm": 88.90672302246094, + "learning_rate": 5.075895567698847e-06, + "loss": 0.156, + "step": 2825 + }, + { + "epoch": 15.576131687242798, + "grad_norm": 0.23931804299354553, + "learning_rate": 4.924104432301154e-06, + "loss": 0.1153, + "step": 2850 + }, + { + "epoch": 15.713305898491084, + "grad_norm": 0.9773799180984497, + "learning_rate": 4.772313296903461e-06, + "loss": 0.0571, + "step": 2875 + }, + { + "epoch": 15.850480109739369, + "grad_norm": 1.4206039905548096, + "learning_rate": 4.620522161505768e-06, + "loss": 0.0586, + "step": 2900 + }, + { + "epoch": 15.987654320987655, + "grad_norm": 172.25982666015625, + "learning_rate": 4.468731026108076e-06, + "loss": 0.2666, + "step": 2925 + }, + { + "epoch": 16.0, + "eval_accuracy": 0.8169717138103162, + "eval_loss": 1.6360419988632202, + "eval_macro_f1": 0.8165766575416269, + "eval_runtime": 1.0593, + "eval_samples_per_second": 567.363, + "eval_steps_per_second": 35.873, + "eval_weighted_f1": 0.8164301215866757, + "step": 2928 + } + ], + "logging_steps": 25, + "max_steps": 3660, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/trainer_state.json b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..894206bc093e2c1f03d8809e4971e71b16d6f12f --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647/trainer_state.json @@ -0,0 +1,597 @@ +{ + "best_global_step": 1647, + "best_metric": 0.8294399520814615, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647", + "epoch": 9.0, + "eval_steps": 500, + "global_step": 1647, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13717421124828533, + "grad_norm": 31.024362564086914, + "learning_rate": 1.3114754098360657e-06, + "loss": 4.8233, + "step": 25 + }, + { + "epoch": 0.27434842249657065, + "grad_norm": 25.978130340576172, + "learning_rate": 2.677595628415301e-06, + "loss": 4.7401, + "step": 50 + }, + { + "epoch": 0.411522633744856, + "grad_norm": 5.003451347351074, + "learning_rate": 4.043715846994536e-06, + "loss": 4.5149, + "step": 75 + }, + { + "epoch": 0.5486968449931413, + "grad_norm": 27.491491317749023, + "learning_rate": 5.409836065573772e-06, + "loss": 4.4615, + "step": 100 + }, + { + "epoch": 0.6858710562414266, + "grad_norm": 24.514368057250977, + "learning_rate": 6.775956284153006e-06, + "loss": 4.4217, + "step": 125 + }, + { + "epoch": 0.823045267489712, + "grad_norm": 86.98007202148438, + "learning_rate": 8.14207650273224e-06, + "loss": 4.434, + "step": 150 + }, + { + "epoch": 0.9602194787379973, + "grad_norm": 28.279203414916992, + "learning_rate": 9.508196721311476e-06, + "loss": 4.4505, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.33111480865224624, + "eval_loss": 1.1600522994995117, + "eval_macro_f1": 0.18939391425252924, + "eval_runtime": 1.3755, + "eval_samples_per_second": 436.943, + "eval_steps_per_second": 27.627, + "eval_weighted_f1": 0.18587730833323093, + "step": 183 + }, + { + "epoch": 1.093278463648834, + "grad_norm": 30.445919036865234, + "learning_rate": 1.0874316939890712e-05, + "loss": 4.2721, + "step": 200 + }, + { + "epoch": 1.2304526748971194, + "grad_norm": 56.098751068115234, + "learning_rate": 1.2240437158469946e-05, + "loss": 4.1866, + "step": 225 + }, + { + "epoch": 1.3676268861454046, + "grad_norm": 31.6022891998291, + "learning_rate": 1.3606557377049181e-05, + "loss": 4.2595, + "step": 250 + }, + { + "epoch": 1.50480109739369, + "grad_norm": 29.9930477142334, + "learning_rate": 1.4972677595628417e-05, + "loss": 4.2462, + "step": 275 + }, + { + "epoch": 1.6419753086419753, + "grad_norm": 53.22201919555664, + "learning_rate": 1.633879781420765e-05, + "loss": 3.9987, + "step": 300 + }, + { + "epoch": 1.7791495198902605, + "grad_norm": 78.16007995605469, + "learning_rate": 1.7704918032786887e-05, + "loss": 3.7552, + "step": 325 + }, + { + "epoch": 1.916323731138546, + "grad_norm": 69.35794830322266, + "learning_rate": 1.9071038251366122e-05, + "loss": 3.4302, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6405990016638935, + "eval_loss": 0.8634985685348511, + "eval_macro_f1": 0.6242585826667983, + "eval_runtime": 1.226, + "eval_samples_per_second": 490.205, + "eval_steps_per_second": 30.995, + "eval_weighted_f1": 0.625333916498407, + "step": 366 + }, + { + "epoch": 2.049382716049383, + "grad_norm": 70.52806091308594, + "learning_rate": 1.995142683667274e-05, + "loss": 3.1297, + "step": 375 + }, + { + "epoch": 2.186556927297668, + "grad_norm": 497.9524230957031, + "learning_rate": 1.9799635701275045e-05, + "loss": 2.9898, + "step": 400 + }, + { + "epoch": 2.3237311385459534, + "grad_norm": 55.19970703125, + "learning_rate": 1.9647844565877354e-05, + "loss": 2.8493, + "step": 425 + }, + { + "epoch": 2.460905349794239, + "grad_norm": 84.73726654052734, + "learning_rate": 1.9496053430479663e-05, + "loss": 2.5896, + "step": 450 + }, + { + "epoch": 2.598079561042524, + "grad_norm": 91.0534896850586, + "learning_rate": 1.934426229508197e-05, + "loss": 2.6912, + "step": 475 + }, + { + "epoch": 2.7352537722908092, + "grad_norm": 108.56490325927734, + "learning_rate": 1.9192471159684274e-05, + "loss": 2.543, + "step": 500 + }, + { + "epoch": 2.8724279835390947, + "grad_norm": 73.8829345703125, + "learning_rate": 1.9040680024286583e-05, + "loss": 2.5349, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7853577371048253, + "eval_loss": 0.5557334423065186, + "eval_macro_f1": 0.7823013176454139, + "eval_runtime": 1.2521, + "eval_samples_per_second": 480.01, + "eval_steps_per_second": 30.35, + "eval_weighted_f1": 0.7827840548821874, + "step": 549 + }, + { + "epoch": 3.0054869684499312, + "grad_norm": 49.06962203979492, + "learning_rate": 1.888888888888889e-05, + "loss": 2.4599, + "step": 550 + }, + { + "epoch": 3.1426611796982167, + "grad_norm": 90.70523071289062, + "learning_rate": 1.8737097753491198e-05, + "loss": 2.0827, + "step": 575 + }, + { + "epoch": 3.279835390946502, + "grad_norm": 81.72300720214844, + "learning_rate": 1.8585306618093504e-05, + "loss": 1.9883, + "step": 600 + }, + { + "epoch": 3.4170096021947876, + "grad_norm": 84.00972747802734, + "learning_rate": 1.8433515482695813e-05, + "loss": 2.171, + "step": 625 + }, + { + "epoch": 3.5541838134430725, + "grad_norm": 93.31878662109375, + "learning_rate": 1.828172434729812e-05, + "loss": 2.0457, + "step": 650 + }, + { + "epoch": 3.691358024691358, + "grad_norm": 48.66339111328125, + "learning_rate": 1.8129933211900424e-05, + "loss": 1.9985, + "step": 675 + }, + { + "epoch": 3.8285322359396434, + "grad_norm": 99.85594940185547, + "learning_rate": 1.7978142076502733e-05, + "loss": 1.9116, + "step": 700 + }, + { + "epoch": 3.9657064471879284, + "grad_norm": 70.8326644897461, + "learning_rate": 1.7826350941105042e-05, + "loss": 1.8741, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7986688851913477, + "eval_loss": 0.5442688465118408, + "eval_macro_f1": 0.7968380577563866, + "eval_runtime": 1.2575, + "eval_samples_per_second": 477.949, + "eval_steps_per_second": 30.22, + "eval_weighted_f1": 0.7970038415048132, + "step": 732 + }, + { + "epoch": 4.098765432098766, + "grad_norm": 64.25154113769531, + "learning_rate": 1.7674559805707348e-05, + "loss": 1.7298, + "step": 750 + }, + { + "epoch": 4.23593964334705, + "grad_norm": 38.59468078613281, + "learning_rate": 1.7522768670309657e-05, + "loss": 1.6404, + "step": 775 + }, + { + "epoch": 4.373113854595336, + "grad_norm": 60.223514556884766, + "learning_rate": 1.7370977534911962e-05, + "loss": 1.6301, + "step": 800 + }, + { + "epoch": 4.510288065843621, + "grad_norm": 321.6264343261719, + "learning_rate": 1.7219186399514268e-05, + "loss": 1.6313, + "step": 825 + }, + { + "epoch": 4.647462277091907, + "grad_norm": 66.87923431396484, + "learning_rate": 1.7067395264116577e-05, + "loss": 1.6531, + "step": 850 + }, + { + "epoch": 4.784636488340192, + "grad_norm": 80.2781982421875, + "learning_rate": 1.6915604128718883e-05, + "loss": 1.6012, + "step": 875 + }, + { + "epoch": 4.921810699588478, + "grad_norm": 60.99646759033203, + "learning_rate": 1.6763812993321192e-05, + "loss": 1.5348, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8069883527454242, + "eval_loss": 0.5002604722976685, + "eval_macro_f1": 0.8065063701645238, + "eval_runtime": 1.2607, + "eval_samples_per_second": 476.723, + "eval_steps_per_second": 30.142, + "eval_weighted_f1": 0.8060523145094759, + "step": 915 + }, + { + "epoch": 5.054869684499314, + "grad_norm": 38.93902587890625, + "learning_rate": 1.66120218579235e-05, + "loss": 1.3239, + "step": 925 + }, + { + "epoch": 5.1920438957476, + "grad_norm": 60.717933654785156, + "learning_rate": 1.6460230722525807e-05, + "loss": 1.2898, + "step": 950 + }, + { + "epoch": 5.329218106995885, + "grad_norm": 80.94599151611328, + "learning_rate": 1.6308439587128112e-05, + "loss": 1.1648, + "step": 975 + }, + { + "epoch": 5.4663923182441705, + "grad_norm": 169.59779357910156, + "learning_rate": 1.615664845173042e-05, + "loss": 1.2149, + "step": 1000 + }, + { + "epoch": 5.603566529492455, + "grad_norm": 76.24026489257812, + "learning_rate": 1.6004857316332727e-05, + "loss": 1.0994, + "step": 1025 + }, + { + "epoch": 5.7407407407407405, + "grad_norm": 92.1307144165039, + "learning_rate": 1.5853066180935036e-05, + "loss": 1.0024, + "step": 1050 + }, + { + "epoch": 5.877914951989026, + "grad_norm": 66.15457916259766, + "learning_rate": 1.570127504553734e-05, + "loss": 1.1377, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8086522462562395, + "eval_loss": 0.5345150232315063, + "eval_macro_f1": 0.8082046779197087, + "eval_runtime": 1.2125, + "eval_samples_per_second": 495.676, + "eval_steps_per_second": 31.341, + "eval_weighted_f1": 0.8078122640413898, + "step": 1098 + }, + { + "epoch": 6.0109739368998625, + "grad_norm": 39.52352523803711, + "learning_rate": 1.5549483910139647e-05, + "loss": 1.187, + "step": 1100 + }, + { + "epoch": 6.148148148148148, + "grad_norm": 120.31623077392578, + "learning_rate": 1.5397692774741956e-05, + "loss": 0.7544, + "step": 1125 + }, + { + "epoch": 6.285322359396433, + "grad_norm": 210.23890686035156, + "learning_rate": 1.5245901639344264e-05, + "loss": 0.7701, + "step": 1150 + }, + { + "epoch": 6.422496570644719, + "grad_norm": 84.77848052978516, + "learning_rate": 1.5094110503946571e-05, + "loss": 1.0074, + "step": 1175 + }, + { + "epoch": 6.559670781893004, + "grad_norm": 72.13880157470703, + "learning_rate": 1.4942319368548878e-05, + "loss": 1.0031, + "step": 1200 + }, + { + "epoch": 6.69684499314129, + "grad_norm": 34.91520309448242, + "learning_rate": 1.4790528233151184e-05, + "loss": 0.7269, + "step": 1225 + }, + { + "epoch": 6.834019204389575, + "grad_norm": 77.64958190917969, + "learning_rate": 1.4638737097753493e-05, + "loss": 0.9424, + "step": 1250 + }, + { + "epoch": 6.97119341563786, + "grad_norm": 137.78585815429688, + "learning_rate": 1.44869459623558e-05, + "loss": 0.9632, + "step": 1275 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8119800332778702, + "eval_loss": 0.6703364253044128, + "eval_macro_f1": 0.8120019085553066, + "eval_runtime": 1.3144, + "eval_samples_per_second": 457.231, + "eval_steps_per_second": 28.91, + "eval_weighted_f1": 0.8113007612820222, + "step": 1281 + }, + { + "epoch": 7.104252400548697, + "grad_norm": 95.81270599365234, + "learning_rate": 1.4335154826958106e-05, + "loss": 0.5562, + "step": 1300 + }, + { + "epoch": 7.2414266117969825, + "grad_norm": 18.943357467651367, + "learning_rate": 1.4183363691560415e-05, + "loss": 0.8241, + "step": 1325 + }, + { + "epoch": 7.378600823045267, + "grad_norm": 99.89557647705078, + "learning_rate": 1.403157255616272e-05, + "loss": 0.5968, + "step": 1350 + }, + { + "epoch": 7.5157750342935525, + "grad_norm": 47.54081726074219, + "learning_rate": 1.3879781420765028e-05, + "loss": 0.6925, + "step": 1375 + }, + { + "epoch": 7.652949245541838, + "grad_norm": 170.8661346435547, + "learning_rate": 1.3727990285367337e-05, + "loss": 0.695, + "step": 1400 + }, + { + "epoch": 7.790123456790123, + "grad_norm": 85.30807495117188, + "learning_rate": 1.3576199149969643e-05, + "loss": 0.7502, + "step": 1425 + }, + { + "epoch": 7.927297668038409, + "grad_norm": 60.53986358642578, + "learning_rate": 1.342440801457195e-05, + "loss": 0.654, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.7970049916805324, + "eval_loss": 0.9338895082473755, + "eval_macro_f1": 0.7955020666066308, + "eval_runtime": 1.2642, + "eval_samples_per_second": 475.391, + "eval_steps_per_second": 30.058, + "eval_weighted_f1": 0.7952025911410133, + "step": 1464 + }, + { + "epoch": 8.060356652949245, + "grad_norm": 103.39309692382812, + "learning_rate": 1.3272616879174259e-05, + "loss": 0.5181, + "step": 1475 + }, + { + "epoch": 8.197530864197532, + "grad_norm": 62.19187927246094, + "learning_rate": 1.3120825743776565e-05, + "loss": 0.3961, + "step": 1500 + }, + { + "epoch": 8.334705075445816, + "grad_norm": 135.94586181640625, + "learning_rate": 1.2969034608378872e-05, + "loss": 0.6136, + "step": 1525 + }, + { + "epoch": 8.4718792866941, + "grad_norm": 120.57879638671875, + "learning_rate": 1.2817243472981178e-05, + "loss": 0.4972, + "step": 1550 + }, + { + "epoch": 8.609053497942387, + "grad_norm": 162.00804138183594, + "learning_rate": 1.2665452337583487e-05, + "loss": 0.5196, + "step": 1575 + }, + { + "epoch": 8.746227709190672, + "grad_norm": 190.614013671875, + "learning_rate": 1.2513661202185794e-05, + "loss": 0.5183, + "step": 1600 + }, + { + "epoch": 8.883401920438958, + "grad_norm": 104.42936706542969, + "learning_rate": 1.23618700667881e-05, + "loss": 0.4854, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8302828618968386, + "eval_loss": 0.7689645886421204, + "eval_macro_f1": 0.8294399520814615, + "eval_runtime": 1.3555, + "eval_samples_per_second": 443.367, + "eval_steps_per_second": 28.033, + "eval_weighted_f1": 0.8295039863181012, + "step": 1647 + } + ], + "logging_steps": 25, + "max_steps": 3660, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/trainer_state.json b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..32769541b783bd290cb796db4d621e90dea4df5a --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-2196/trainer_state.json @@ -0,0 +1,784 @@ +{ + "best_global_step": 1647, + "best_metric": 0.8294399520814615, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-03-s43-20260901T172732Z-fbd63439/trainer/checkpoint-1647", + "epoch": 12.0, + "eval_steps": 500, + "global_step": 2196, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13717421124828533, + "grad_norm": 31.024362564086914, + "learning_rate": 1.3114754098360657e-06, + "loss": 4.8233, + "step": 25 + }, + { + "epoch": 0.27434842249657065, + "grad_norm": 25.978130340576172, + "learning_rate": 2.677595628415301e-06, + "loss": 4.7401, + "step": 50 + }, + { + "epoch": 0.411522633744856, + "grad_norm": 5.003451347351074, + "learning_rate": 4.043715846994536e-06, + "loss": 4.5149, + "step": 75 + }, + { + "epoch": 0.5486968449931413, + "grad_norm": 27.491491317749023, + "learning_rate": 5.409836065573772e-06, + "loss": 4.4615, + "step": 100 + }, + { + "epoch": 0.6858710562414266, + "grad_norm": 24.514368057250977, + "learning_rate": 6.775956284153006e-06, + "loss": 4.4217, + "step": 125 + }, + { + "epoch": 0.823045267489712, + "grad_norm": 86.98007202148438, + "learning_rate": 8.14207650273224e-06, + "loss": 4.434, + "step": 150 + }, + { + "epoch": 0.9602194787379973, + "grad_norm": 28.279203414916992, + "learning_rate": 9.508196721311476e-06, + "loss": 4.4505, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.33111480865224624, + "eval_loss": 1.1600522994995117, + "eval_macro_f1": 0.18939391425252924, + "eval_runtime": 1.3755, + "eval_samples_per_second": 436.943, + "eval_steps_per_second": 27.627, + "eval_weighted_f1": 0.18587730833323093, + "step": 183 + }, + { + "epoch": 1.093278463648834, + "grad_norm": 30.445919036865234, + "learning_rate": 1.0874316939890712e-05, + "loss": 4.2721, + "step": 200 + }, + { + "epoch": 1.2304526748971194, + "grad_norm": 56.098751068115234, + "learning_rate": 1.2240437158469946e-05, + "loss": 4.1866, + "step": 225 + }, + { + "epoch": 1.3676268861454046, + "grad_norm": 31.6022891998291, + "learning_rate": 1.3606557377049181e-05, + "loss": 4.2595, + "step": 250 + }, + { + "epoch": 1.50480109739369, + "grad_norm": 29.9930477142334, + "learning_rate": 1.4972677595628417e-05, + "loss": 4.2462, + "step": 275 + }, + { + "epoch": 1.6419753086419753, + "grad_norm": 53.22201919555664, + "learning_rate": 1.633879781420765e-05, + "loss": 3.9987, + "step": 300 + }, + { + "epoch": 1.7791495198902605, + "grad_norm": 78.16007995605469, + "learning_rate": 1.7704918032786887e-05, + "loss": 3.7552, + "step": 325 + }, + { + "epoch": 1.916323731138546, + "grad_norm": 69.35794830322266, + "learning_rate": 1.9071038251366122e-05, + "loss": 3.4302, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6405990016638935, + "eval_loss": 0.8634985685348511, + "eval_macro_f1": 0.6242585826667983, + "eval_runtime": 1.226, + "eval_samples_per_second": 490.205, + "eval_steps_per_second": 30.995, + "eval_weighted_f1": 0.625333916498407, + "step": 366 + }, + { + "epoch": 2.049382716049383, + "grad_norm": 70.52806091308594, + "learning_rate": 1.995142683667274e-05, + "loss": 3.1297, + "step": 375 + }, + { + "epoch": 2.186556927297668, + "grad_norm": 497.9524230957031, + "learning_rate": 1.9799635701275045e-05, + "loss": 2.9898, + "step": 400 + }, + { + "epoch": 2.3237311385459534, + "grad_norm": 55.19970703125, + "learning_rate": 1.9647844565877354e-05, + "loss": 2.8493, + "step": 425 + }, + { + "epoch": 2.460905349794239, + "grad_norm": 84.73726654052734, + "learning_rate": 1.9496053430479663e-05, + "loss": 2.5896, + "step": 450 + }, + { + "epoch": 2.598079561042524, + "grad_norm": 91.0534896850586, + "learning_rate": 1.934426229508197e-05, + "loss": 2.6912, + "step": 475 + }, + { + "epoch": 2.7352537722908092, + "grad_norm": 108.56490325927734, + "learning_rate": 1.9192471159684274e-05, + "loss": 2.543, + "step": 500 + }, + { + "epoch": 2.8724279835390947, + "grad_norm": 73.8829345703125, + "learning_rate": 1.9040680024286583e-05, + "loss": 2.5349, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7853577371048253, + "eval_loss": 0.5557334423065186, + "eval_macro_f1": 0.7823013176454139, + "eval_runtime": 1.2521, + "eval_samples_per_second": 480.01, + "eval_steps_per_second": 30.35, + "eval_weighted_f1": 0.7827840548821874, + "step": 549 + }, + { + "epoch": 3.0054869684499312, + "grad_norm": 49.06962203979492, + "learning_rate": 1.888888888888889e-05, + "loss": 2.4599, + "step": 550 + }, + { + "epoch": 3.1426611796982167, + "grad_norm": 90.70523071289062, + "learning_rate": 1.8737097753491198e-05, + "loss": 2.0827, + "step": 575 + }, + { + "epoch": 3.279835390946502, + "grad_norm": 81.72300720214844, + "learning_rate": 1.8585306618093504e-05, + "loss": 1.9883, + "step": 600 + }, + { + "epoch": 3.4170096021947876, + "grad_norm": 84.00972747802734, + "learning_rate": 1.8433515482695813e-05, + "loss": 2.171, + "step": 625 + }, + { + "epoch": 3.5541838134430725, + "grad_norm": 93.31878662109375, + "learning_rate": 1.828172434729812e-05, + "loss": 2.0457, + "step": 650 + }, + { + "epoch": 3.691358024691358, + "grad_norm": 48.66339111328125, + "learning_rate": 1.8129933211900424e-05, + "loss": 1.9985, + "step": 675 + }, + { + "epoch": 3.8285322359396434, + "grad_norm": 99.85594940185547, + "learning_rate": 1.7978142076502733e-05, + "loss": 1.9116, + "step": 700 + }, + { + "epoch": 3.9657064471879284, + "grad_norm": 70.8326644897461, + "learning_rate": 1.7826350941105042e-05, + "loss": 1.8741, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7986688851913477, + "eval_loss": 0.5442688465118408, + "eval_macro_f1": 0.7968380577563866, + "eval_runtime": 1.2575, + "eval_samples_per_second": 477.949, + "eval_steps_per_second": 30.22, + "eval_weighted_f1": 0.7970038415048132, + "step": 732 + }, + { + "epoch": 4.098765432098766, + "grad_norm": 64.25154113769531, + "learning_rate": 1.7674559805707348e-05, + "loss": 1.7298, + "step": 750 + }, + { + "epoch": 4.23593964334705, + "grad_norm": 38.59468078613281, + "learning_rate": 1.7522768670309657e-05, + "loss": 1.6404, + "step": 775 + }, + { + "epoch": 4.373113854595336, + "grad_norm": 60.223514556884766, + "learning_rate": 1.7370977534911962e-05, + "loss": 1.6301, + "step": 800 + }, + { + "epoch": 4.510288065843621, + "grad_norm": 321.6264343261719, + "learning_rate": 1.7219186399514268e-05, + "loss": 1.6313, + "step": 825 + }, + { + "epoch": 4.647462277091907, + "grad_norm": 66.87923431396484, + "learning_rate": 1.7067395264116577e-05, + "loss": 1.6531, + "step": 850 + }, + { + "epoch": 4.784636488340192, + "grad_norm": 80.2781982421875, + "learning_rate": 1.6915604128718883e-05, + "loss": 1.6012, + "step": 875 + }, + { + "epoch": 4.921810699588478, + "grad_norm": 60.99646759033203, + "learning_rate": 1.6763812993321192e-05, + "loss": 1.5348, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8069883527454242, + "eval_loss": 0.5002604722976685, + "eval_macro_f1": 0.8065063701645238, + "eval_runtime": 1.2607, + "eval_samples_per_second": 476.723, + "eval_steps_per_second": 30.142, + "eval_weighted_f1": 0.8060523145094759, + "step": 915 + }, + { + "epoch": 5.054869684499314, + "grad_norm": 38.93902587890625, + "learning_rate": 1.66120218579235e-05, + "loss": 1.3239, + "step": 925 + }, + { + "epoch": 5.1920438957476, + "grad_norm": 60.717933654785156, + "learning_rate": 1.6460230722525807e-05, + "loss": 1.2898, + "step": 950 + }, + { + "epoch": 5.329218106995885, + "grad_norm": 80.94599151611328, + "learning_rate": 1.6308439587128112e-05, + "loss": 1.1648, + "step": 975 + }, + { + "epoch": 5.4663923182441705, + "grad_norm": 169.59779357910156, + "learning_rate": 1.615664845173042e-05, + "loss": 1.2149, + "step": 1000 + }, + { + "epoch": 5.603566529492455, + "grad_norm": 76.24026489257812, + "learning_rate": 1.6004857316332727e-05, + "loss": 1.0994, + "step": 1025 + }, + { + "epoch": 5.7407407407407405, + "grad_norm": 92.1307144165039, + "learning_rate": 1.5853066180935036e-05, + "loss": 1.0024, + "step": 1050 + }, + { + "epoch": 5.877914951989026, + "grad_norm": 66.15457916259766, + "learning_rate": 1.570127504553734e-05, + "loss": 1.1377, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8086522462562395, + "eval_loss": 0.5345150232315063, + "eval_macro_f1": 0.8082046779197087, + "eval_runtime": 1.2125, + "eval_samples_per_second": 495.676, + "eval_steps_per_second": 31.341, + "eval_weighted_f1": 0.8078122640413898, + "step": 1098 + }, + { + "epoch": 6.0109739368998625, + "grad_norm": 39.52352523803711, + "learning_rate": 1.5549483910139647e-05, + "loss": 1.187, + "step": 1100 + }, + { + "epoch": 6.148148148148148, + "grad_norm": 120.31623077392578, + "learning_rate": 1.5397692774741956e-05, + "loss": 0.7544, + "step": 1125 + }, + { + "epoch": 6.285322359396433, + "grad_norm": 210.23890686035156, + "learning_rate": 1.5245901639344264e-05, + "loss": 0.7701, + "step": 1150 + }, + { + "epoch": 6.422496570644719, + "grad_norm": 84.77848052978516, + "learning_rate": 1.5094110503946571e-05, + "loss": 1.0074, + "step": 1175 + }, + { + "epoch": 6.559670781893004, + "grad_norm": 72.13880157470703, + "learning_rate": 1.4942319368548878e-05, + "loss": 1.0031, + "step": 1200 + }, + { + "epoch": 6.69684499314129, + "grad_norm": 34.91520309448242, + "learning_rate": 1.4790528233151184e-05, + "loss": 0.7269, + "step": 1225 + }, + { + "epoch": 6.834019204389575, + "grad_norm": 77.64958190917969, + "learning_rate": 1.4638737097753493e-05, + "loss": 0.9424, + "step": 1250 + }, + { + "epoch": 6.97119341563786, + "grad_norm": 137.78585815429688, + "learning_rate": 1.44869459623558e-05, + "loss": 0.9632, + "step": 1275 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8119800332778702, + "eval_loss": 0.6703364253044128, + "eval_macro_f1": 0.8120019085553066, + "eval_runtime": 1.3144, + "eval_samples_per_second": 457.231, + "eval_steps_per_second": 28.91, + "eval_weighted_f1": 0.8113007612820222, + "step": 1281 + }, + { + "epoch": 7.104252400548697, + "grad_norm": 95.81270599365234, + "learning_rate": 1.4335154826958106e-05, + "loss": 0.5562, + "step": 1300 + }, + { + "epoch": 7.2414266117969825, + "grad_norm": 18.943357467651367, + "learning_rate": 1.4183363691560415e-05, + "loss": 0.8241, + "step": 1325 + }, + { + "epoch": 7.378600823045267, + "grad_norm": 99.89557647705078, + "learning_rate": 1.403157255616272e-05, + "loss": 0.5968, + "step": 1350 + }, + { + "epoch": 7.5157750342935525, + "grad_norm": 47.54081726074219, + "learning_rate": 1.3879781420765028e-05, + "loss": 0.6925, + "step": 1375 + }, + { + "epoch": 7.652949245541838, + "grad_norm": 170.8661346435547, + "learning_rate": 1.3727990285367337e-05, + "loss": 0.695, + "step": 1400 + }, + { + "epoch": 7.790123456790123, + "grad_norm": 85.30807495117188, + "learning_rate": 1.3576199149969643e-05, + "loss": 0.7502, + "step": 1425 + }, + { + "epoch": 7.927297668038409, + "grad_norm": 60.53986358642578, + "learning_rate": 1.342440801457195e-05, + "loss": 0.654, + "step": 1450 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.7970049916805324, + "eval_loss": 0.9338895082473755, + "eval_macro_f1": 0.7955020666066308, + "eval_runtime": 1.2642, + "eval_samples_per_second": 475.391, + "eval_steps_per_second": 30.058, + "eval_weighted_f1": 0.7952025911410133, + "step": 1464 + }, + { + "epoch": 8.060356652949245, + "grad_norm": 103.39309692382812, + "learning_rate": 1.3272616879174259e-05, + "loss": 0.5181, + "step": 1475 + }, + { + "epoch": 8.197530864197532, + "grad_norm": 62.19187927246094, + "learning_rate": 1.3120825743776565e-05, + "loss": 0.3961, + "step": 1500 + }, + { + "epoch": 8.334705075445816, + "grad_norm": 135.94586181640625, + "learning_rate": 1.2969034608378872e-05, + "loss": 0.6136, + "step": 1525 + }, + { + "epoch": 8.4718792866941, + "grad_norm": 120.57879638671875, + "learning_rate": 1.2817243472981178e-05, + "loss": 0.4972, + "step": 1550 + }, + { + "epoch": 8.609053497942387, + "grad_norm": 162.00804138183594, + "learning_rate": 1.2665452337583487e-05, + "loss": 0.5196, + "step": 1575 + }, + { + "epoch": 8.746227709190672, + "grad_norm": 190.614013671875, + "learning_rate": 1.2513661202185794e-05, + "loss": 0.5183, + "step": 1600 + }, + { + "epoch": 8.883401920438958, + "grad_norm": 104.42936706542969, + "learning_rate": 1.23618700667881e-05, + "loss": 0.4854, + "step": 1625 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8302828618968386, + "eval_loss": 0.7689645886421204, + "eval_macro_f1": 0.8294399520814615, + "eval_runtime": 1.3555, + "eval_samples_per_second": 443.367, + "eval_steps_per_second": 28.033, + "eval_weighted_f1": 0.8295039863181012, + "step": 1647 + }, + { + "epoch": 9.016460905349794, + "grad_norm": 98.02400970458984, + "learning_rate": 1.2210078931390409e-05, + "loss": 0.6494, + "step": 1650 + }, + { + "epoch": 9.15363511659808, + "grad_norm": 50.38234329223633, + "learning_rate": 1.2058287795992716e-05, + "loss": 0.4291, + "step": 1675 + }, + { + "epoch": 9.290809327846365, + "grad_norm": 1.6103564500808716, + "learning_rate": 1.1906496660595022e-05, + "loss": 0.4249, + "step": 1700 + }, + { + "epoch": 9.427983539094651, + "grad_norm": 245.33624267578125, + "learning_rate": 1.1754705525197329e-05, + "loss": 0.3066, + "step": 1725 + }, + { + "epoch": 9.565157750342935, + "grad_norm": 34.69443893432617, + "learning_rate": 1.1602914389799638e-05, + "loss": 0.3616, + "step": 1750 + }, + { + "epoch": 9.70233196159122, + "grad_norm": 65.777587890625, + "learning_rate": 1.1451123254401944e-05, + "loss": 0.3139, + "step": 1775 + }, + { + "epoch": 9.839506172839506, + "grad_norm": 451.5834045410156, + "learning_rate": 1.1299332119004251e-05, + "loss": 0.4528, + "step": 1800 + }, + { + "epoch": 9.976680384087791, + "grad_norm": 74.06531524658203, + "learning_rate": 1.1147540983606557e-05, + "loss": 0.5241, + "step": 1825 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.826955074875208, + "eval_loss": 1.0812312364578247, + "eval_macro_f1": 0.8259042719563596, + "eval_runtime": 1.3059, + "eval_samples_per_second": 460.236, + "eval_steps_per_second": 29.1, + "eval_weighted_f1": 0.8256400818088346, + "step": 1830 + }, + { + "epoch": 10.109739368998628, + "grad_norm": 75.8895263671875, + "learning_rate": 1.0995749848208866e-05, + "loss": 0.2742, + "step": 1850 + }, + { + "epoch": 10.246913580246913, + "grad_norm": 166.20535278320312, + "learning_rate": 1.0843958712811173e-05, + "loss": 0.3337, + "step": 1875 + }, + { + "epoch": 10.3840877914952, + "grad_norm": 116.20963287353516, + "learning_rate": 1.0692167577413479e-05, + "loss": 0.415, + "step": 1900 + }, + { + "epoch": 10.521262002743484, + "grad_norm": 1.40334153175354, + "learning_rate": 1.0540376442015788e-05, + "loss": 0.2412, + "step": 1925 + }, + { + "epoch": 10.65843621399177, + "grad_norm": 31.528650283813477, + "learning_rate": 1.0388585306618095e-05, + "loss": 0.3993, + "step": 1950 + }, + { + "epoch": 10.795610425240055, + "grad_norm": 127.42987823486328, + "learning_rate": 1.02367941712204e-05, + "loss": 0.3733, + "step": 1975 + }, + { + "epoch": 10.932784636488341, + "grad_norm": 8.481429100036621, + "learning_rate": 1.008500303582271e-05, + "loss": 0.2571, + "step": 2000 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.8236272878535774, + "eval_loss": 1.1412168741226196, + "eval_macro_f1": 0.8232142857142857, + "eval_runtime": 1.2533, + "eval_samples_per_second": 479.545, + "eval_steps_per_second": 30.321, + "eval_weighted_f1": 0.8228547658664132, + "step": 2013 + }, + { + "epoch": 11.065843621399177, + "grad_norm": 0.4736940562725067, + "learning_rate": 9.933211900425015e-06, + "loss": 0.3189, + "step": 2025 + }, + { + "epoch": 11.203017832647463, + "grad_norm": 0.3013084828853607, + "learning_rate": 9.781420765027323e-06, + "loss": 0.434, + "step": 2050 + }, + { + "epoch": 11.340192043895748, + "grad_norm": 244.99014282226562, + "learning_rate": 9.62962962962963e-06, + "loss": 0.2345, + "step": 2075 + }, + { + "epoch": 11.477366255144032, + "grad_norm": 250.29115295410156, + "learning_rate": 9.477838494231937e-06, + "loss": 0.2848, + "step": 2100 + }, + { + "epoch": 11.614540466392318, + "grad_norm": 52.71842956542969, + "learning_rate": 9.326047358834245e-06, + "loss": 0.2421, + "step": 2125 + }, + { + "epoch": 11.751714677640603, + "grad_norm": 187.28933715820312, + "learning_rate": 9.174256223436552e-06, + "loss": 0.3685, + "step": 2150 + }, + { + "epoch": 11.88888888888889, + "grad_norm": 216.94105529785156, + "learning_rate": 9.02246508803886e-06, + "loss": 0.3261, + "step": 2175 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8103161397670549, + "eval_loss": 1.3759733438491821, + "eval_macro_f1": 0.8094502425217273, + "eval_runtime": 1.272, + "eval_samples_per_second": 472.477, + "eval_steps_per_second": 29.874, + "eval_weighted_f1": 0.8090147637950267, + "step": 2196 + } + ], + "logging_steps": 25, + "max_steps": 3660, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/trainer_state.json b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d9901714eba9040a04d3883175892845cb3d7613 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232/trainer_state.json @@ -0,0 +1,463 @@ +{ + "best_global_step": 1232, + "best_metric": 0.8220538813074404, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232", + "epoch": 7.0, + "eval_steps": 500, + "global_step": 1232, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.14245014245014245, + "grad_norm": 27.217100143432617, + "learning_rate": 1.3636363636363636e-06, + "loss": 4.8518, + "step": 25 + }, + { + "epoch": 0.2849002849002849, + "grad_norm": 39.04031753540039, + "learning_rate": 2.784090909090909e-06, + "loss": 4.7352, + "step": 50 + }, + { + "epoch": 0.42735042735042733, + "grad_norm": 36.9318962097168, + "learning_rate": 4.204545454545455e-06, + "loss": 4.5955, + "step": 75 + }, + { + "epoch": 0.5698005698005698, + "grad_norm": 41.124755859375, + "learning_rate": 5.625e-06, + "loss": 4.5589, + "step": 100 + }, + { + "epoch": 0.7122507122507122, + "grad_norm": 42.625362396240234, + "learning_rate": 7.045454545454546e-06, + "loss": 4.4615, + "step": 125 + }, + { + "epoch": 0.8547008547008547, + "grad_norm": 66.03889465332031, + "learning_rate": 8.465909090909091e-06, + "loss": 4.3691, + "step": 150 + }, + { + "epoch": 0.9971509971509972, + "grad_norm": 89.74577331542969, + "learning_rate": 9.886363636363637e-06, + "loss": 4.1944, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5674891146589259, + "eval_loss": 0.9591799974441528, + "eval_macro_f1": 0.5742638366234435, + "eval_runtime": 1.4945, + "eval_samples_per_second": 461.029, + "eval_steps_per_second": 29.442, + "eval_weighted_f1": 0.5732643363993044, + "step": 176 + }, + { + "epoch": 1.1367521367521367, + "grad_norm": 63.877506256103516, + "learning_rate": 1.1306818181818183e-05, + "loss": 4.1315, + "step": 200 + }, + { + "epoch": 1.2792022792022792, + "grad_norm": 165.04998779296875, + "learning_rate": 1.2727272727272728e-05, + "loss": 4.0353, + "step": 225 + }, + { + "epoch": 1.4216524216524218, + "grad_norm": 144.60411071777344, + "learning_rate": 1.4147727272727274e-05, + "loss": 3.9202, + "step": 250 + }, + { + "epoch": 1.564102564102564, + "grad_norm": 134.73329162597656, + "learning_rate": 1.5568181818181822e-05, + "loss": 3.7649, + "step": 275 + }, + { + "epoch": 1.7065527065527064, + "grad_norm": 63.84787368774414, + "learning_rate": 1.6988636363636368e-05, + "loss": 3.7085, + "step": 300 + }, + { + "epoch": 1.849002849002849, + "grad_norm": 51.40576934814453, + "learning_rate": 1.840909090909091e-05, + "loss": 3.4848, + "step": 325 + }, + { + "epoch": 1.9914529914529915, + "grad_norm": 59.681373596191406, + "learning_rate": 1.9829545454545456e-05, + "loss": 3.2577, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6632801161103048, + "eval_loss": 0.7946692109107971, + "eval_macro_f1": 0.6213893000642483, + "eval_runtime": 1.3864, + "eval_samples_per_second": 496.969, + "eval_steps_per_second": 31.737, + "eval_weighted_f1": 0.6237192518110943, + "step": 352 + }, + { + "epoch": 2.131054131054131, + "grad_norm": 108.29476165771484, + "learning_rate": 1.9861111111111114e-05, + "loss": 3.1482, + "step": 375 + }, + { + "epoch": 2.2735042735042734, + "grad_norm": 74.50999450683594, + "learning_rate": 1.9703282828282832e-05, + "loss": 3.1278, + "step": 400 + }, + { + "epoch": 2.415954415954416, + "grad_norm": 101.41131591796875, + "learning_rate": 1.9545454545454546e-05, + "loss": 2.8185, + "step": 425 + }, + { + "epoch": 2.5584045584045585, + "grad_norm": 102.26775360107422, + "learning_rate": 1.9387626262626263e-05, + "loss": 2.8325, + "step": 450 + }, + { + "epoch": 2.700854700854701, + "grad_norm": 81.17192077636719, + "learning_rate": 1.922979797979798e-05, + "loss": 2.6993, + "step": 475 + }, + { + "epoch": 2.8433048433048436, + "grad_norm": 80.05029296875, + "learning_rate": 1.90719696969697e-05, + "loss": 2.7077, + "step": 500 + }, + { + "epoch": 2.9857549857549857, + "grad_norm": 52.80065155029297, + "learning_rate": 1.8914141414141416e-05, + "loss": 2.6758, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.760522496371553, + "eval_loss": 0.5608870387077332, + "eval_macro_f1": 0.752402755129104, + "eval_runtime": 1.2668, + "eval_samples_per_second": 543.873, + "eval_steps_per_second": 34.732, + "eval_weighted_f1": 0.7532129950855451, + "step": 528 + }, + { + "epoch": 3.1253561253561255, + "grad_norm": 98.53296661376953, + "learning_rate": 1.8756313131313133e-05, + "loss": 2.475, + "step": 550 + }, + { + "epoch": 3.267806267806268, + "grad_norm": 50.69923782348633, + "learning_rate": 1.859848484848485e-05, + "loss": 2.1013, + "step": 575 + }, + { + "epoch": 3.41025641025641, + "grad_norm": 81.4959945678711, + "learning_rate": 1.844065656565657e-05, + "loss": 2.1695, + "step": 600 + }, + { + "epoch": 3.5527065527065527, + "grad_norm": 113.66731262207031, + "learning_rate": 1.8282828282828286e-05, + "loss": 2.2256, + "step": 625 + }, + { + "epoch": 3.695156695156695, + "grad_norm": 53.84329605102539, + "learning_rate": 1.8125e-05, + "loss": 2.0442, + "step": 650 + }, + { + "epoch": 3.8376068376068377, + "grad_norm": 50.35622024536133, + "learning_rate": 1.7967171717171718e-05, + "loss": 2.1707, + "step": 675 + }, + { + "epoch": 3.98005698005698, + "grad_norm": 75.28170776367188, + "learning_rate": 1.7809343434343435e-05, + "loss": 2.2009, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8040638606676342, + "eval_loss": 0.49773257970809937, + "eval_macro_f1": 0.8030923193858746, + "eval_runtime": 1.4558, + "eval_samples_per_second": 473.264, + "eval_steps_per_second": 30.223, + "eval_weighted_f1": 0.8034722165090973, + "step": 704 + }, + { + "epoch": 4.119658119658119, + "grad_norm": 47.5344352722168, + "learning_rate": 1.7651515151515153e-05, + "loss": 1.6864, + "step": 725 + }, + { + "epoch": 4.262108262108262, + "grad_norm": 61.259498596191406, + "learning_rate": 1.749368686868687e-05, + "loss": 1.6896, + "step": 750 + }, + { + "epoch": 4.404558404558404, + "grad_norm": 89.69329833984375, + "learning_rate": 1.7335858585858588e-05, + "loss": 1.7854, + "step": 775 + }, + { + "epoch": 4.547008547008547, + "grad_norm": 97.24896240234375, + "learning_rate": 1.7178030303030305e-05, + "loss": 1.6547, + "step": 800 + }, + { + "epoch": 4.689458689458689, + "grad_norm": 82.40167236328125, + "learning_rate": 1.7020202020202023e-05, + "loss": 1.5816, + "step": 825 + }, + { + "epoch": 4.831908831908832, + "grad_norm": 58.9371337890625, + "learning_rate": 1.6862373737373737e-05, + "loss": 1.4979, + "step": 850 + }, + { + "epoch": 4.9743589743589745, + "grad_norm": 78.72428894042969, + "learning_rate": 1.6704545454545454e-05, + "loss": 1.7305, + "step": 875 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.818577648766328, + "eval_loss": 0.5488840341567993, + "eval_macro_f1": 0.8215258814369704, + "eval_runtime": 1.3142, + "eval_samples_per_second": 524.254, + "eval_steps_per_second": 33.479, + "eval_weighted_f1": 0.8222285630302564, + "step": 880 + }, + { + "epoch": 5.113960113960114, + "grad_norm": 34.06904220581055, + "learning_rate": 1.6546717171717172e-05, + "loss": 1.4071, + "step": 900 + }, + { + "epoch": 5.256410256410256, + "grad_norm": 101.27243041992188, + "learning_rate": 1.638888888888889e-05, + "loss": 1.3481, + "step": 925 + }, + { + "epoch": 5.398860398860399, + "grad_norm": 328.5030517578125, + "learning_rate": 1.6231060606060607e-05, + "loss": 1.1791, + "step": 950 + }, + { + "epoch": 5.5413105413105415, + "grad_norm": 64.46382141113281, + "learning_rate": 1.6073232323232324e-05, + "loss": 1.3329, + "step": 975 + }, + { + "epoch": 5.683760683760684, + "grad_norm": 133.52960205078125, + "learning_rate": 1.5915404040404042e-05, + "loss": 1.4345, + "step": 1000 + }, + { + "epoch": 5.8262108262108265, + "grad_norm": 114.03165435791016, + "learning_rate": 1.575757575757576e-05, + "loss": 1.19, + "step": 1025 + }, + { + "epoch": 5.968660968660968, + "grad_norm": 93.84526062011719, + "learning_rate": 1.5599747474747477e-05, + "loss": 1.1826, + "step": 1050 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8040638606676342, + "eval_loss": 0.6089228391647339, + "eval_macro_f1": 0.7990900704026928, + "eval_runtime": 1.4643, + "eval_samples_per_second": 470.541, + "eval_steps_per_second": 30.049, + "eval_weighted_f1": 0.8000976998394738, + "step": 1056 + }, + { + "epoch": 6.1082621082621085, + "grad_norm": 181.9176025390625, + "learning_rate": 1.544191919191919e-05, + "loss": 1.1669, + "step": 1075 + }, + { + "epoch": 6.250712250712251, + "grad_norm": 40.773719787597656, + "learning_rate": 1.528409090909091e-05, + "loss": 0.9327, + "step": 1100 + }, + { + "epoch": 6.3931623931623935, + "grad_norm": 74.51333618164062, + "learning_rate": 1.5126262626262626e-05, + "loss": 0.8394, + "step": 1125 + }, + { + "epoch": 6.535612535612536, + "grad_norm": 110.24671173095703, + "learning_rate": 1.4968434343434344e-05, + "loss": 1.0755, + "step": 1150 + }, + { + "epoch": 6.678062678062678, + "grad_norm": 63.552757263183594, + "learning_rate": 1.4810606060606061e-05, + "loss": 1.0349, + "step": 1175 + }, + { + "epoch": 6.82051282051282, + "grad_norm": 41.187042236328125, + "learning_rate": 1.4652777777777779e-05, + "loss": 1.2311, + "step": 1200 + }, + { + "epoch": 6.962962962962963, + "grad_norm": 68.12137603759766, + "learning_rate": 1.4494949494949494e-05, + "loss": 0.9405, + "step": 1225 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8243831640058055, + "eval_loss": 0.634580135345459, + "eval_macro_f1": 0.8220538813074404, + "eval_runtime": 1.4188, + "eval_samples_per_second": 485.62, + "eval_steps_per_second": 31.012, + "eval_weighted_f1": 0.8226227831663885, + "step": 1232 + } + ], + "logging_steps": 25, + "max_steps": 3520, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/trainer_state.json b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..82d4c07d8fd56c57cdcfba84b576401aa30ea2b3 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1760/trainer_state.json @@ -0,0 +1,643 @@ +{ + "best_global_step": 1232, + "best_metric": 0.8220538813074404, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-04-s42-20260901T155905Z-27bda108/trainer/checkpoint-1232", + "epoch": 10.0, + "eval_steps": 500, + "global_step": 1760, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.14245014245014245, + "grad_norm": 27.217100143432617, + "learning_rate": 1.3636363636363636e-06, + "loss": 4.8518, + "step": 25 + }, + { + "epoch": 0.2849002849002849, + "grad_norm": 39.04031753540039, + "learning_rate": 2.784090909090909e-06, + "loss": 4.7352, + "step": 50 + }, + { + "epoch": 0.42735042735042733, + "grad_norm": 36.9318962097168, + "learning_rate": 4.204545454545455e-06, + "loss": 4.5955, + "step": 75 + }, + { + "epoch": 0.5698005698005698, + "grad_norm": 41.124755859375, + "learning_rate": 5.625e-06, + "loss": 4.5589, + "step": 100 + }, + { + "epoch": 0.7122507122507122, + "grad_norm": 42.625362396240234, + "learning_rate": 7.045454545454546e-06, + "loss": 4.4615, + "step": 125 + }, + { + "epoch": 0.8547008547008547, + "grad_norm": 66.03889465332031, + "learning_rate": 8.465909090909091e-06, + "loss": 4.3691, + "step": 150 + }, + { + "epoch": 0.9971509971509972, + "grad_norm": 89.74577331542969, + "learning_rate": 9.886363636363637e-06, + "loss": 4.1944, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5674891146589259, + "eval_loss": 0.9591799974441528, + "eval_macro_f1": 0.5742638366234435, + "eval_runtime": 1.4945, + "eval_samples_per_second": 461.029, + "eval_steps_per_second": 29.442, + "eval_weighted_f1": 0.5732643363993044, + "step": 176 + }, + { + "epoch": 1.1367521367521367, + "grad_norm": 63.877506256103516, + "learning_rate": 1.1306818181818183e-05, + "loss": 4.1315, + "step": 200 + }, + { + "epoch": 1.2792022792022792, + "grad_norm": 165.04998779296875, + "learning_rate": 1.2727272727272728e-05, + "loss": 4.0353, + "step": 225 + }, + { + "epoch": 1.4216524216524218, + "grad_norm": 144.60411071777344, + "learning_rate": 1.4147727272727274e-05, + "loss": 3.9202, + "step": 250 + }, + { + "epoch": 1.564102564102564, + "grad_norm": 134.73329162597656, + "learning_rate": 1.5568181818181822e-05, + "loss": 3.7649, + "step": 275 + }, + { + "epoch": 1.7065527065527064, + "grad_norm": 63.84787368774414, + "learning_rate": 1.6988636363636368e-05, + "loss": 3.7085, + "step": 300 + }, + { + "epoch": 1.849002849002849, + "grad_norm": 51.40576934814453, + "learning_rate": 1.840909090909091e-05, + "loss": 3.4848, + "step": 325 + }, + { + "epoch": 1.9914529914529915, + "grad_norm": 59.681373596191406, + "learning_rate": 1.9829545454545456e-05, + "loss": 3.2577, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6632801161103048, + "eval_loss": 0.7946692109107971, + "eval_macro_f1": 0.6213893000642483, + "eval_runtime": 1.3864, + "eval_samples_per_second": 496.969, + "eval_steps_per_second": 31.737, + "eval_weighted_f1": 0.6237192518110943, + "step": 352 + }, + { + "epoch": 2.131054131054131, + "grad_norm": 108.29476165771484, + "learning_rate": 1.9861111111111114e-05, + "loss": 3.1482, + "step": 375 + }, + { + "epoch": 2.2735042735042734, + "grad_norm": 74.50999450683594, + "learning_rate": 1.9703282828282832e-05, + "loss": 3.1278, + "step": 400 + }, + { + "epoch": 2.415954415954416, + "grad_norm": 101.41131591796875, + "learning_rate": 1.9545454545454546e-05, + "loss": 2.8185, + "step": 425 + }, + { + "epoch": 2.5584045584045585, + "grad_norm": 102.26775360107422, + "learning_rate": 1.9387626262626263e-05, + "loss": 2.8325, + "step": 450 + }, + { + "epoch": 2.700854700854701, + "grad_norm": 81.17192077636719, + "learning_rate": 1.922979797979798e-05, + "loss": 2.6993, + "step": 475 + }, + { + "epoch": 2.8433048433048436, + "grad_norm": 80.05029296875, + "learning_rate": 1.90719696969697e-05, + "loss": 2.7077, + "step": 500 + }, + { + "epoch": 2.9857549857549857, + "grad_norm": 52.80065155029297, + "learning_rate": 1.8914141414141416e-05, + "loss": 2.6758, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.760522496371553, + "eval_loss": 0.5608870387077332, + "eval_macro_f1": 0.752402755129104, + "eval_runtime": 1.2668, + "eval_samples_per_second": 543.873, + "eval_steps_per_second": 34.732, + "eval_weighted_f1": 0.7532129950855451, + "step": 528 + }, + { + "epoch": 3.1253561253561255, + "grad_norm": 98.53296661376953, + "learning_rate": 1.8756313131313133e-05, + "loss": 2.475, + "step": 550 + }, + { + "epoch": 3.267806267806268, + "grad_norm": 50.69923782348633, + "learning_rate": 1.859848484848485e-05, + "loss": 2.1013, + "step": 575 + }, + { + "epoch": 3.41025641025641, + "grad_norm": 81.4959945678711, + "learning_rate": 1.844065656565657e-05, + "loss": 2.1695, + "step": 600 + }, + { + "epoch": 3.5527065527065527, + "grad_norm": 113.66731262207031, + "learning_rate": 1.8282828282828286e-05, + "loss": 2.2256, + "step": 625 + }, + { + "epoch": 3.695156695156695, + "grad_norm": 53.84329605102539, + "learning_rate": 1.8125e-05, + "loss": 2.0442, + "step": 650 + }, + { + "epoch": 3.8376068376068377, + "grad_norm": 50.35622024536133, + "learning_rate": 1.7967171717171718e-05, + "loss": 2.1707, + "step": 675 + }, + { + "epoch": 3.98005698005698, + "grad_norm": 75.28170776367188, + "learning_rate": 1.7809343434343435e-05, + "loss": 2.2009, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8040638606676342, + "eval_loss": 0.49773257970809937, + "eval_macro_f1": 0.8030923193858746, + "eval_runtime": 1.4558, + "eval_samples_per_second": 473.264, + "eval_steps_per_second": 30.223, + "eval_weighted_f1": 0.8034722165090973, + "step": 704 + }, + { + "epoch": 4.119658119658119, + "grad_norm": 47.5344352722168, + "learning_rate": 1.7651515151515153e-05, + "loss": 1.6864, + "step": 725 + }, + { + "epoch": 4.262108262108262, + "grad_norm": 61.259498596191406, + "learning_rate": 1.749368686868687e-05, + "loss": 1.6896, + "step": 750 + }, + { + "epoch": 4.404558404558404, + "grad_norm": 89.69329833984375, + "learning_rate": 1.7335858585858588e-05, + "loss": 1.7854, + "step": 775 + }, + { + "epoch": 4.547008547008547, + "grad_norm": 97.24896240234375, + "learning_rate": 1.7178030303030305e-05, + "loss": 1.6547, + "step": 800 + }, + { + "epoch": 4.689458689458689, + "grad_norm": 82.40167236328125, + "learning_rate": 1.7020202020202023e-05, + "loss": 1.5816, + "step": 825 + }, + { + "epoch": 4.831908831908832, + "grad_norm": 58.9371337890625, + "learning_rate": 1.6862373737373737e-05, + "loss": 1.4979, + "step": 850 + }, + { + "epoch": 4.9743589743589745, + "grad_norm": 78.72428894042969, + "learning_rate": 1.6704545454545454e-05, + "loss": 1.7305, + "step": 875 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.818577648766328, + "eval_loss": 0.5488840341567993, + "eval_macro_f1": 0.8215258814369704, + "eval_runtime": 1.3142, + "eval_samples_per_second": 524.254, + "eval_steps_per_second": 33.479, + "eval_weighted_f1": 0.8222285630302564, + "step": 880 + }, + { + "epoch": 5.113960113960114, + "grad_norm": 34.06904220581055, + "learning_rate": 1.6546717171717172e-05, + "loss": 1.4071, + "step": 900 + }, + { + "epoch": 5.256410256410256, + "grad_norm": 101.27243041992188, + "learning_rate": 1.638888888888889e-05, + "loss": 1.3481, + "step": 925 + }, + { + "epoch": 5.398860398860399, + "grad_norm": 328.5030517578125, + "learning_rate": 1.6231060606060607e-05, + "loss": 1.1791, + "step": 950 + }, + { + "epoch": 5.5413105413105415, + "grad_norm": 64.46382141113281, + "learning_rate": 1.6073232323232324e-05, + "loss": 1.3329, + "step": 975 + }, + { + "epoch": 5.683760683760684, + "grad_norm": 133.52960205078125, + "learning_rate": 1.5915404040404042e-05, + "loss": 1.4345, + "step": 1000 + }, + { + "epoch": 5.8262108262108265, + "grad_norm": 114.03165435791016, + "learning_rate": 1.575757575757576e-05, + "loss": 1.19, + "step": 1025 + }, + { + "epoch": 5.968660968660968, + "grad_norm": 93.84526062011719, + "learning_rate": 1.5599747474747477e-05, + "loss": 1.1826, + "step": 1050 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8040638606676342, + "eval_loss": 0.6089228391647339, + "eval_macro_f1": 0.7990900704026928, + "eval_runtime": 1.4643, + "eval_samples_per_second": 470.541, + "eval_steps_per_second": 30.049, + "eval_weighted_f1": 0.8000976998394738, + "step": 1056 + }, + { + "epoch": 6.1082621082621085, + "grad_norm": 181.9176025390625, + "learning_rate": 1.544191919191919e-05, + "loss": 1.1669, + "step": 1075 + }, + { + "epoch": 6.250712250712251, + "grad_norm": 40.773719787597656, + "learning_rate": 1.528409090909091e-05, + "loss": 0.9327, + "step": 1100 + }, + { + "epoch": 6.3931623931623935, + "grad_norm": 74.51333618164062, + "learning_rate": 1.5126262626262626e-05, + "loss": 0.8394, + "step": 1125 + }, + { + "epoch": 6.535612535612536, + "grad_norm": 110.24671173095703, + "learning_rate": 1.4968434343434344e-05, + "loss": 1.0755, + "step": 1150 + }, + { + "epoch": 6.678062678062678, + "grad_norm": 63.552757263183594, + "learning_rate": 1.4810606060606061e-05, + "loss": 1.0349, + "step": 1175 + }, + { + "epoch": 6.82051282051282, + "grad_norm": 41.187042236328125, + "learning_rate": 1.4652777777777779e-05, + "loss": 1.2311, + "step": 1200 + }, + { + "epoch": 6.962962962962963, + "grad_norm": 68.12137603759766, + "learning_rate": 1.4494949494949494e-05, + "loss": 0.9405, + "step": 1225 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8243831640058055, + "eval_loss": 0.634580135345459, + "eval_macro_f1": 0.8220538813074404, + "eval_runtime": 1.4188, + "eval_samples_per_second": 485.62, + "eval_steps_per_second": 31.012, + "eval_weighted_f1": 0.8226227831663885, + "step": 1232 + }, + { + "epoch": 7.102564102564102, + "grad_norm": 56.86425018310547, + "learning_rate": 1.4337121212121212e-05, + "loss": 0.7002, + "step": 1250 + }, + { + "epoch": 7.245014245014245, + "grad_norm": 122.55549621582031, + "learning_rate": 1.4179292929292931e-05, + "loss": 0.6749, + "step": 1275 + }, + { + "epoch": 7.387464387464387, + "grad_norm": 111.5029525756836, + "learning_rate": 1.4021464646464649e-05, + "loss": 0.685, + "step": 1300 + }, + { + "epoch": 7.52991452991453, + "grad_norm": 89.68938446044922, + "learning_rate": 1.3863636363636366e-05, + "loss": 0.753, + "step": 1325 + }, + { + "epoch": 7.672364672364672, + "grad_norm": 147.90426635742188, + "learning_rate": 1.3705808080808084e-05, + "loss": 0.8433, + "step": 1350 + }, + { + "epoch": 7.814814814814815, + "grad_norm": 110.82835388183594, + "learning_rate": 1.35479797979798e-05, + "loss": 0.7559, + "step": 1375 + }, + { + "epoch": 7.957264957264957, + "grad_norm": 44.29365158081055, + "learning_rate": 1.3390151515151517e-05, + "loss": 0.793, + "step": 1400 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8113207547169812, + "eval_loss": 0.6649388670921326, + "eval_macro_f1": 0.8097692636027644, + "eval_runtime": 1.1659, + "eval_samples_per_second": 590.984, + "eval_steps_per_second": 37.741, + "eval_weighted_f1": 0.8102066166282731, + "step": 1408 + }, + { + "epoch": 8.096866096866098, + "grad_norm": 130.97434997558594, + "learning_rate": 1.3232323232323234e-05, + "loss": 0.5839, + "step": 1425 + }, + { + "epoch": 8.239316239316238, + "grad_norm": 110.22476196289062, + "learning_rate": 1.3074494949494952e-05, + "loss": 0.5381, + "step": 1450 + }, + { + "epoch": 8.381766381766381, + "grad_norm": 28.667129516601562, + "learning_rate": 1.2916666666666668e-05, + "loss": 0.5855, + "step": 1475 + }, + { + "epoch": 8.524216524216524, + "grad_norm": 82.33951568603516, + "learning_rate": 1.2758838383838385e-05, + "loss": 0.4591, + "step": 1500 + }, + { + "epoch": 8.666666666666666, + "grad_norm": 16.435546875, + "learning_rate": 1.2601010101010103e-05, + "loss": 0.5493, + "step": 1525 + }, + { + "epoch": 8.809116809116809, + "grad_norm": 11.185385704040527, + "learning_rate": 1.244318181818182e-05, + "loss": 0.6177, + "step": 1550 + }, + { + "epoch": 8.951566951566951, + "grad_norm": 24.901023864746094, + "learning_rate": 1.2285353535353536e-05, + "loss": 0.6191, + "step": 1575 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8113207547169812, + "eval_loss": 0.8255329728126526, + "eval_macro_f1": 0.8085767648579907, + "eval_runtime": 1.2381, + "eval_samples_per_second": 556.492, + "eval_steps_per_second": 35.538, + "eval_weighted_f1": 0.8090666810756689, + "step": 1584 + }, + { + "epoch": 9.091168091168091, + "grad_norm": 226.4795684814453, + "learning_rate": 1.2127525252525254e-05, + "loss": 0.4356, + "step": 1600 + }, + { + "epoch": 9.233618233618234, + "grad_norm": 18.89518928527832, + "learning_rate": 1.1969696969696971e-05, + "loss": 0.3263, + "step": 1625 + }, + { + "epoch": 9.376068376068377, + "grad_norm": 201.86769104003906, + "learning_rate": 1.1811868686868689e-05, + "loss": 0.4137, + "step": 1650 + }, + { + "epoch": 9.518518518518519, + "grad_norm": 9.775382995605469, + "learning_rate": 1.1654040404040406e-05, + "loss": 0.2994, + "step": 1675 + }, + { + "epoch": 9.660968660968662, + "grad_norm": 26.552865982055664, + "learning_rate": 1.1496212121212122e-05, + "loss": 0.3385, + "step": 1700 + }, + { + "epoch": 9.803418803418804, + "grad_norm": 56.45022201538086, + "learning_rate": 1.133838383838384e-05, + "loss": 0.4971, + "step": 1725 + }, + { + "epoch": 9.945868945868947, + "grad_norm": 204.9696807861328, + "learning_rate": 1.1180555555555557e-05, + "loss": 0.5579, + "step": 1750 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.806966618287373, + "eval_loss": 0.9372882843017578, + "eval_macro_f1": 0.8078326937497242, + "eval_runtime": 1.1751, + "eval_samples_per_second": 586.311, + "eval_steps_per_second": 37.442, + "eval_weighted_f1": 0.8079208945732799, + "step": 1760 + } + ], + "logging_steps": 25, + "max_steps": 3520, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/trainer_state.json b/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..a07fbc5df6a20051a12c0062550198977aa36b4a --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936/trainer_state.json @@ -0,0 +1,703 @@ +{ + "best_global_step": 1936, + "best_metric": 0.8344651818959735, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-04-s43-20260901T174119Z-fbd63439/trainer/checkpoint-1936", + "epoch": 11.0, + "eval_steps": 500, + "global_step": 1936, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.14245014245014245, + "grad_norm": 19.109556198120117, + "learning_rate": 1.3636363636363636e-06, + "loss": 4.6934, + "step": 25 + }, + { + "epoch": 0.2849002849002849, + "grad_norm": 13.637514114379883, + "learning_rate": 2.784090909090909e-06, + "loss": 4.6004, + "step": 50 + }, + { + "epoch": 0.42735042735042733, + "grad_norm": 8.681859970092773, + "learning_rate": 4.204545454545455e-06, + "loss": 4.6307, + "step": 75 + }, + { + "epoch": 0.5698005698005698, + "grad_norm": 21.15650177001953, + "learning_rate": 5.625e-06, + "loss": 4.5261, + "step": 100 + }, + { + "epoch": 0.7122507122507122, + "grad_norm": 50.862762451171875, + "learning_rate": 7.045454545454546e-06, + "loss": 4.5219, + "step": 125 + }, + { + "epoch": 0.8547008547008547, + "grad_norm": 22.722986221313477, + "learning_rate": 8.465909090909091e-06, + "loss": 4.5212, + "step": 150 + }, + { + "epoch": 0.9971509971509972, + "grad_norm": 208.30953979492188, + "learning_rate": 9.886363636363637e-06, + "loss": 4.3795, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.41654571843251087, + "eval_loss": 1.0518839359283447, + "eval_macro_f1": 0.33222330225246305, + "eval_runtime": 1.583, + "eval_samples_per_second": 435.256, + "eval_steps_per_second": 27.796, + "eval_weighted_f1": 0.3385708791137879, + "step": 176 + }, + { + "epoch": 1.1367521367521367, + "grad_norm": 67.8524398803711, + "learning_rate": 1.1306818181818183e-05, + "loss": 4.1814, + "step": 200 + }, + { + "epoch": 1.2792022792022792, + "grad_norm": 63.24119567871094, + "learning_rate": 1.2727272727272728e-05, + "loss": 3.9363, + "step": 225 + }, + { + "epoch": 1.4216524216524218, + "grad_norm": 44.609622955322266, + "learning_rate": 1.4147727272727274e-05, + "loss": 3.4515, + "step": 250 + }, + { + "epoch": 1.564102564102564, + "grad_norm": 78.82655334472656, + "learning_rate": 1.5568181818181822e-05, + "loss": 3.4225, + "step": 275 + }, + { + "epoch": 1.7065527065527064, + "grad_norm": 60.25815963745117, + "learning_rate": 1.6988636363636368e-05, + "loss": 3.2167, + "step": 300 + }, + { + "epoch": 1.849002849002849, + "grad_norm": 140.99090576171875, + "learning_rate": 1.840909090909091e-05, + "loss": 3.2183, + "step": 325 + }, + { + "epoch": 1.9914529914529915, + "grad_norm": 37.75986099243164, + "learning_rate": 1.9829545454545456e-05, + "loss": 2.9674, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.706821480406386, + "eval_loss": 0.6712866425514221, + "eval_macro_f1": 0.7003515314496495, + "eval_runtime": 1.2594, + "eval_samples_per_second": 547.068, + "eval_steps_per_second": 34.936, + "eval_weighted_f1": 0.7010788530128889, + "step": 352 + }, + { + "epoch": 2.131054131054131, + "grad_norm": 91.7500991821289, + "learning_rate": 1.9861111111111114e-05, + "loss": 2.7854, + "step": 375 + }, + { + "epoch": 2.2735042735042734, + "grad_norm": 98.97259521484375, + "learning_rate": 1.9703282828282832e-05, + "loss": 2.6409, + "step": 400 + }, + { + "epoch": 2.415954415954416, + "grad_norm": 105.08370208740234, + "learning_rate": 1.9545454545454546e-05, + "loss": 2.626, + "step": 425 + }, + { + "epoch": 2.5584045584045585, + "grad_norm": 74.43781280517578, + "learning_rate": 1.9387626262626263e-05, + "loss": 2.5168, + "step": 450 + }, + { + "epoch": 2.700854700854701, + "grad_norm": 78.37004089355469, + "learning_rate": 1.922979797979798e-05, + "loss": 2.35, + "step": 475 + }, + { + "epoch": 2.8433048433048436, + "grad_norm": 80.54888153076172, + "learning_rate": 1.90719696969697e-05, + "loss": 2.4516, + "step": 500 + }, + { + "epoch": 2.9857549857549857, + "grad_norm": 48.9096794128418, + "learning_rate": 1.8914141414141416e-05, + "loss": 2.4067, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7706821480406386, + "eval_loss": 0.5558003187179565, + "eval_macro_f1": 0.7662371441561094, + "eval_runtime": 1.3657, + "eval_samples_per_second": 504.488, + "eval_steps_per_second": 32.217, + "eval_weighted_f1": 0.7670256905268935, + "step": 528 + }, + { + "epoch": 3.1253561253561255, + "grad_norm": 67.07841491699219, + "learning_rate": 1.8756313131313133e-05, + "loss": 1.9461, + "step": 550 + }, + { + "epoch": 3.267806267806268, + "grad_norm": 46.26266860961914, + "learning_rate": 1.859848484848485e-05, + "loss": 2.0284, + "step": 575 + }, + { + "epoch": 3.41025641025641, + "grad_norm": 73.0999526977539, + "learning_rate": 1.844065656565657e-05, + "loss": 2.0565, + "step": 600 + }, + { + "epoch": 3.5527065527065527, + "grad_norm": 44.275291442871094, + "learning_rate": 1.8282828282828286e-05, + "loss": 2.1275, + "step": 625 + }, + { + "epoch": 3.695156695156695, + "grad_norm": 83.3443832397461, + "learning_rate": 1.8125e-05, + "loss": 1.9817, + "step": 650 + }, + { + "epoch": 3.8376068376068377, + "grad_norm": 131.52674865722656, + "learning_rate": 1.7967171717171718e-05, + "loss": 1.7871, + "step": 675 + }, + { + "epoch": 3.98005698005698, + "grad_norm": 51.05097961425781, + "learning_rate": 1.7809343434343435e-05, + "loss": 1.8753, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8026124818577649, + "eval_loss": 0.5317463278770447, + "eval_macro_f1": 0.8047833784827647, + "eval_runtime": 1.3653, + "eval_samples_per_second": 504.657, + "eval_steps_per_second": 32.228, + "eval_weighted_f1": 0.8048986598769678, + "step": 704 + }, + { + "epoch": 4.119658119658119, + "grad_norm": 65.76126098632812, + "learning_rate": 1.7651515151515153e-05, + "loss": 1.5264, + "step": 725 + }, + { + "epoch": 4.262108262108262, + "grad_norm": 97.57429504394531, + "learning_rate": 1.749368686868687e-05, + "loss": 1.6532, + "step": 750 + }, + { + "epoch": 4.404558404558404, + "grad_norm": 88.40670013427734, + "learning_rate": 1.7335858585858588e-05, + "loss": 1.3995, + "step": 775 + }, + { + "epoch": 4.547008547008547, + "grad_norm": 55.178443908691406, + "learning_rate": 1.7178030303030305e-05, + "loss": 1.4966, + "step": 800 + }, + { + "epoch": 4.689458689458689, + "grad_norm": 106.28096771240234, + "learning_rate": 1.7020202020202023e-05, + "loss": 1.3918, + "step": 825 + }, + { + "epoch": 4.831908831908832, + "grad_norm": 81.64993286132812, + "learning_rate": 1.6862373737373737e-05, + "loss": 1.4582, + "step": 850 + }, + { + "epoch": 4.9743589743589745, + "grad_norm": 161.23573303222656, + "learning_rate": 1.6704545454545454e-05, + "loss": 1.6138, + "step": 875 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8084179970972424, + "eval_loss": 0.5031147003173828, + "eval_macro_f1": 0.8054171079479321, + "eval_runtime": 1.5896, + "eval_samples_per_second": 433.447, + "eval_steps_per_second": 27.68, + "eval_weighted_f1": 0.8058400315131427, + "step": 880 + }, + { + "epoch": 5.113960113960114, + "grad_norm": 203.06906127929688, + "learning_rate": 1.6546717171717172e-05, + "loss": 1.4729, + "step": 900 + }, + { + "epoch": 5.256410256410256, + "grad_norm": 72.82058715820312, + "learning_rate": 1.638888888888889e-05, + "loss": 1.1817, + "step": 925 + }, + { + "epoch": 5.398860398860399, + "grad_norm": 111.98212432861328, + "learning_rate": 1.6231060606060607e-05, + "loss": 1.2548, + "step": 950 + }, + { + "epoch": 5.5413105413105415, + "grad_norm": 80.76667022705078, + "learning_rate": 1.6073232323232324e-05, + "loss": 1.1891, + "step": 975 + }, + { + "epoch": 5.683760683760684, + "grad_norm": 77.15080261230469, + "learning_rate": 1.5915404040404042e-05, + "loss": 1.1713, + "step": 1000 + }, + { + "epoch": 5.8262108262108265, + "grad_norm": 133.8670654296875, + "learning_rate": 1.575757575757576e-05, + "loss": 1.2551, + "step": 1025 + }, + { + "epoch": 5.968660968660968, + "grad_norm": 162.1934051513672, + "learning_rate": 1.5599747474747477e-05, + "loss": 1.0077, + "step": 1050 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8258345428156749, + "eval_loss": 0.5771624445915222, + "eval_macro_f1": 0.8246685551531355, + "eval_runtime": 1.2927, + "eval_samples_per_second": 532.994, + "eval_steps_per_second": 34.037, + "eval_weighted_f1": 0.8252586117065356, + "step": 1056 + }, + { + "epoch": 6.1082621082621085, + "grad_norm": 78.01618194580078, + "learning_rate": 1.544191919191919e-05, + "loss": 0.7831, + "step": 1075 + }, + { + "epoch": 6.250712250712251, + "grad_norm": 92.61405944824219, + "learning_rate": 1.528409090909091e-05, + "loss": 0.7623, + "step": 1100 + }, + { + "epoch": 6.3931623931623935, + "grad_norm": 88.09264373779297, + "learning_rate": 1.5126262626262626e-05, + "loss": 0.8616, + "step": 1125 + }, + { + "epoch": 6.535612535612536, + "grad_norm": 92.74222564697266, + "learning_rate": 1.4968434343434344e-05, + "loss": 1.0544, + "step": 1150 + }, + { + "epoch": 6.678062678062678, + "grad_norm": 80.24437713623047, + "learning_rate": 1.4810606060606061e-05, + "loss": 0.9701, + "step": 1175 + }, + { + "epoch": 6.82051282051282, + "grad_norm": 83.69192504882812, + "learning_rate": 1.4652777777777779e-05, + "loss": 1.0175, + "step": 1200 + }, + { + "epoch": 6.962962962962963, + "grad_norm": 47.23429489135742, + "learning_rate": 1.4494949494949494e-05, + "loss": 0.9753, + "step": 1225 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8171262699564587, + "eval_loss": 0.5948315262794495, + "eval_macro_f1": 0.8156146342511511, + "eval_runtime": 1.3962, + "eval_samples_per_second": 493.479, + "eval_steps_per_second": 31.514, + "eval_weighted_f1": 0.8160222903408281, + "step": 1232 + }, + { + "epoch": 7.102564102564102, + "grad_norm": 136.78518676757812, + "learning_rate": 1.4337121212121212e-05, + "loss": 0.6318, + "step": 1250 + }, + { + "epoch": 7.245014245014245, + "grad_norm": 85.39682006835938, + "learning_rate": 1.4179292929292931e-05, + "loss": 0.6985, + "step": 1275 + }, + { + "epoch": 7.387464387464387, + "grad_norm": 101.58167266845703, + "learning_rate": 1.4021464646464649e-05, + "loss": 0.8447, + "step": 1300 + }, + { + "epoch": 7.52991452991453, + "grad_norm": 18.505325317382812, + "learning_rate": 1.3863636363636366e-05, + "loss": 0.5677, + "step": 1325 + }, + { + "epoch": 7.672364672364672, + "grad_norm": 118.62303924560547, + "learning_rate": 1.3705808080808084e-05, + "loss": 0.5851, + "step": 1350 + }, + { + "epoch": 7.814814814814815, + "grad_norm": 223.8522491455078, + "learning_rate": 1.35479797979798e-05, + "loss": 0.6691, + "step": 1375 + }, + { + "epoch": 7.957264957264957, + "grad_norm": 169.23779296875, + "learning_rate": 1.3390151515151517e-05, + "loss": 0.5999, + "step": 1400 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8214804063860668, + "eval_loss": 0.7902704477310181, + "eval_macro_f1": 0.8174286257213589, + "eval_runtime": 1.2923, + "eval_samples_per_second": 533.172, + "eval_steps_per_second": 34.049, + "eval_weighted_f1": 0.8179624489815055, + "step": 1408 + }, + { + "epoch": 8.096866096866098, + "grad_norm": 102.27836608886719, + "learning_rate": 1.3232323232323234e-05, + "loss": 0.4776, + "step": 1425 + }, + { + "epoch": 8.239316239316238, + "grad_norm": 154.9219207763672, + "learning_rate": 1.3074494949494952e-05, + "loss": 0.4854, + "step": 1450 + }, + { + "epoch": 8.381766381766381, + "grad_norm": 15.329522132873535, + "learning_rate": 1.2916666666666668e-05, + "loss": 0.5001, + "step": 1475 + }, + { + "epoch": 8.524216524216524, + "grad_norm": 49.64401626586914, + "learning_rate": 1.2758838383838385e-05, + "loss": 0.5187, + "step": 1500 + }, + { + "epoch": 8.666666666666666, + "grad_norm": 71.68310546875, + "learning_rate": 1.2601010101010103e-05, + "loss": 0.6416, + "step": 1525 + }, + { + "epoch": 8.809116809116809, + "grad_norm": 36.46543884277344, + "learning_rate": 1.244318181818182e-05, + "loss": 0.5966, + "step": 1550 + }, + { + "epoch": 8.951566951566951, + "grad_norm": 34.157325744628906, + "learning_rate": 1.2285353535353536e-05, + "loss": 0.7163, + "step": 1575 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8272859216255443, + "eval_loss": 0.6640575528144836, + "eval_macro_f1": 0.8271706167585559, + "eval_runtime": 1.2998, + "eval_samples_per_second": 530.076, + "eval_steps_per_second": 33.851, + "eval_weighted_f1": 0.8273207229076359, + "step": 1584 + }, + { + "epoch": 9.091168091168091, + "grad_norm": 113.74295806884766, + "learning_rate": 1.2127525252525254e-05, + "loss": 0.4985, + "step": 1600 + }, + { + "epoch": 9.233618233618234, + "grad_norm": 98.48761749267578, + "learning_rate": 1.1969696969696971e-05, + "loss": 0.4517, + "step": 1625 + }, + { + "epoch": 9.376068376068377, + "grad_norm": 117.53905487060547, + "learning_rate": 1.1811868686868689e-05, + "loss": 0.3754, + "step": 1650 + }, + { + "epoch": 9.518518518518519, + "grad_norm": 72.46125793457031, + "learning_rate": 1.1654040404040406e-05, + "loss": 0.3906, + "step": 1675 + }, + { + "epoch": 9.660968660968662, + "grad_norm": 225.44467163085938, + "learning_rate": 1.1496212121212122e-05, + "loss": 0.5604, + "step": 1700 + }, + { + "epoch": 9.803418803418804, + "grad_norm": 27.47649383544922, + "learning_rate": 1.133838383838384e-05, + "loss": 0.2964, + "step": 1725 + }, + { + "epoch": 9.945868945868947, + "grad_norm": 58.63057327270508, + "learning_rate": 1.1180555555555557e-05, + "loss": 0.3582, + "step": 1750 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8330914368650217, + "eval_loss": 0.9443467259407043, + "eval_macro_f1": 0.832901513564988, + "eval_runtime": 1.374, + "eval_samples_per_second": 501.459, + "eval_steps_per_second": 32.023, + "eval_weighted_f1": 0.833154363158339, + "step": 1760 + }, + { + "epoch": 10.085470085470085, + "grad_norm": 275.8406677246094, + "learning_rate": 1.1022727272727275e-05, + "loss": 0.5785, + "step": 1775 + }, + { + "epoch": 10.227920227920228, + "grad_norm": 80.13965606689453, + "learning_rate": 1.086489898989899e-05, + "loss": 0.2574, + "step": 1800 + }, + { + "epoch": 10.37037037037037, + "grad_norm": 11.424076080322266, + "learning_rate": 1.0707070707070708e-05, + "loss": 0.4276, + "step": 1825 + }, + { + "epoch": 10.512820512820513, + "grad_norm": 225.17112731933594, + "learning_rate": 1.0549242424242425e-05, + "loss": 0.4567, + "step": 1850 + }, + { + "epoch": 10.655270655270655, + "grad_norm": 101.99359130859375, + "learning_rate": 1.0391414141414143e-05, + "loss": 0.3663, + "step": 1875 + }, + { + "epoch": 10.797720797720798, + "grad_norm": 231.50599670410156, + "learning_rate": 1.0233585858585859e-05, + "loss": 0.2196, + "step": 1900 + }, + { + "epoch": 10.94017094017094, + "grad_norm": 168.00267028808594, + "learning_rate": 1.0075757575757576e-05, + "loss": 0.2662, + "step": 1925 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.8374455732946299, + "eval_loss": 1.0518878698349, + "eval_macro_f1": 0.8344651818959735, + "eval_runtime": 1.285, + "eval_samples_per_second": 536.189, + "eval_steps_per_second": 34.241, + "eval_weighted_f1": 0.8346761536135385, + "step": 1936 + } + ], + "logging_steps": 25, + "max_steps": 3520, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/trainer_state.json b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..2a4aa7786c8c461a83291c310306c353b1c47d99 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-1440/trainer_state.json @@ -0,0 +1,530 @@ +{ + "best_global_step": 900, + "best_metric": 0.815764837770578, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900", + "epoch": 8.0, + "eval_steps": 500, + "global_step": 1440, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1388888888888889, + "grad_norm": 37.62110137939453, + "learning_rate": 1.3333333333333334e-06, + "loss": 4.9958, + "step": 25 + }, + { + "epoch": 0.2777777777777778, + "grad_norm": 33.463844299316406, + "learning_rate": 2.7222222222222224e-06, + "loss": 4.6188, + "step": 50 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 45.38542556762695, + "learning_rate": 4.111111111111111e-06, + "loss": 4.5391, + "step": 75 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 35.68893814086914, + "learning_rate": 5.500000000000001e-06, + "loss": 4.585, + "step": 100 + }, + { + "epoch": 0.6944444444444444, + "grad_norm": 55.31415557861328, + "learning_rate": 6.88888888888889e-06, + "loss": 4.4798, + "step": 125 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 44.58302688598633, + "learning_rate": 8.277777777777778e-06, + "loss": 4.4632, + "step": 150 + }, + { + "epoch": 0.9722222222222222, + "grad_norm": 68.91496276855469, + "learning_rate": 9.666666666666667e-06, + "loss": 4.1992, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5298507462686567, + "eval_loss": 0.9932835698127747, + "eval_macro_f1": 0.4812545627863858, + "eval_runtime": 1.1751, + "eval_samples_per_second": 570.178, + "eval_steps_per_second": 35.742, + "eval_weighted_f1": 0.4823316947081317, + "step": 180 + }, + { + "epoch": 1.1111111111111112, + "grad_norm": 43.673458099365234, + "learning_rate": 1.1055555555555557e-05, + "loss": 4.1202, + "step": 200 + }, + { + "epoch": 1.25, + "grad_norm": 33.03519821166992, + "learning_rate": 1.2444444444444446e-05, + "loss": 4.0191, + "step": 225 + }, + { + "epoch": 1.3888888888888888, + "grad_norm": 46.48714065551758, + "learning_rate": 1.3833333333333334e-05, + "loss": 3.9793, + "step": 250 + }, + { + "epoch": 1.5277777777777777, + "grad_norm": 189.76443481445312, + "learning_rate": 1.5222222222222223e-05, + "loss": 3.8691, + "step": 275 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 49.02412796020508, + "learning_rate": 1.6611111111111113e-05, + "loss": 3.9928, + "step": 300 + }, + { + "epoch": 1.8055555555555556, + "grad_norm": 35.443363189697266, + "learning_rate": 1.8e-05, + "loss": 3.499, + "step": 325 + }, + { + "epoch": 1.9444444444444444, + "grad_norm": 58.79069900512695, + "learning_rate": 1.938888888888889e-05, + "loss": 3.2991, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6850746268656717, + "eval_loss": 0.728695273399353, + "eval_macro_f1": 0.6824344063526143, + "eval_runtime": 1.1699, + "eval_samples_per_second": 572.687, + "eval_steps_per_second": 35.9, + "eval_weighted_f1": 0.6818588670899222, + "step": 360 + }, + { + "epoch": 2.0833333333333335, + "grad_norm": 74.10299682617188, + "learning_rate": 1.991358024691358e-05, + "loss": 3.4255, + "step": 375 + }, + { + "epoch": 2.2222222222222223, + "grad_norm": 79.9495620727539, + "learning_rate": 1.975925925925926e-05, + "loss": 3.434, + "step": 400 + }, + { + "epoch": 2.361111111111111, + "grad_norm": 105.76057434082031, + "learning_rate": 1.960493827160494e-05, + "loss": 2.6872, + "step": 425 + }, + { + "epoch": 2.5, + "grad_norm": 62.34187316894531, + "learning_rate": 1.945061728395062e-05, + "loss": 2.9178, + "step": 450 + }, + { + "epoch": 2.638888888888889, + "grad_norm": 72.12592315673828, + "learning_rate": 1.92962962962963e-05, + "loss": 2.5766, + "step": 475 + }, + { + "epoch": 2.7777777777777777, + "grad_norm": 46.273807525634766, + "learning_rate": 1.914197530864198e-05, + "loss": 2.6845, + "step": 500 + }, + { + "epoch": 2.9166666666666665, + "grad_norm": 103.11640930175781, + "learning_rate": 1.8987654320987657e-05, + "loss": 2.4601, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.744776119402985, + "eval_loss": 0.6502454280853271, + "eval_macro_f1": 0.741043454588696, + "eval_runtime": 1.1156, + "eval_samples_per_second": 600.594, + "eval_steps_per_second": 37.649, + "eval_weighted_f1": 0.740781179677082, + "step": 540 + }, + { + "epoch": 3.0555555555555554, + "grad_norm": 58.63907241821289, + "learning_rate": 1.8833333333333335e-05, + "loss": 2.3028, + "step": 550 + }, + { + "epoch": 3.1944444444444446, + "grad_norm": 86.794921875, + "learning_rate": 1.8679012345679013e-05, + "loss": 2.1976, + "step": 575 + }, + { + "epoch": 3.3333333333333335, + "grad_norm": 90.98960876464844, + "learning_rate": 1.8524691358024694e-05, + "loss": 2.2231, + "step": 600 + }, + { + "epoch": 3.4722222222222223, + "grad_norm": 234.18736267089844, + "learning_rate": 1.837037037037037e-05, + "loss": 2.172, + "step": 625 + }, + { + "epoch": 3.611111111111111, + "grad_norm": 49.61118698120117, + "learning_rate": 1.821604938271605e-05, + "loss": 2.1864, + "step": 650 + }, + { + "epoch": 3.75, + "grad_norm": 72.47056579589844, + "learning_rate": 1.806172839506173e-05, + "loss": 2.2088, + "step": 675 + }, + { + "epoch": 3.888888888888889, + "grad_norm": 88.57938385009766, + "learning_rate": 1.7907407407407408e-05, + "loss": 2.0463, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7776119402985074, + "eval_loss": 0.5820518136024475, + "eval_macro_f1": 0.7738326545779962, + "eval_runtime": 1.1582, + "eval_samples_per_second": 578.504, + "eval_steps_per_second": 36.264, + "eval_weighted_f1": 0.7738901764469578, + "step": 720 + }, + { + "epoch": 4.027777777777778, + "grad_norm": 67.66858673095703, + "learning_rate": 1.775308641975309e-05, + "loss": 1.7945, + "step": 725 + }, + { + "epoch": 4.166666666666667, + "grad_norm": 71.2930679321289, + "learning_rate": 1.7598765432098767e-05, + "loss": 1.7633, + "step": 750 + }, + { + "epoch": 4.305555555555555, + "grad_norm": 31.82473373413086, + "learning_rate": 1.7444444444444448e-05, + "loss": 1.574, + "step": 775 + }, + { + "epoch": 4.444444444444445, + "grad_norm": 102.59622192382812, + "learning_rate": 1.7290123456790125e-05, + "loss": 1.6594, + "step": 800 + }, + { + "epoch": 4.583333333333333, + "grad_norm": 35.59578323364258, + "learning_rate": 1.7135802469135806e-05, + "loss": 1.5991, + "step": 825 + }, + { + "epoch": 4.722222222222222, + "grad_norm": 75.3868179321289, + "learning_rate": 1.6981481481481484e-05, + "loss": 1.536, + "step": 850 + }, + { + "epoch": 4.861111111111111, + "grad_norm": 71.83612823486328, + "learning_rate": 1.6827160493827162e-05, + "loss": 1.6476, + "step": 875 + }, + { + "epoch": 5.0, + "grad_norm": 73.57019805908203, + "learning_rate": 1.667283950617284e-05, + "loss": 1.7134, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8164179104477612, + "eval_loss": 0.5686514973640442, + "eval_macro_f1": 0.815764837770578, + "eval_runtime": 1.2508, + "eval_samples_per_second": 535.641, + "eval_steps_per_second": 33.577, + "eval_weighted_f1": 0.8155710164890037, + "step": 900 + }, + { + "epoch": 5.138888888888889, + "grad_norm": 52.468299865722656, + "learning_rate": 1.651851851851852e-05, + "loss": 1.2337, + "step": 925 + }, + { + "epoch": 5.277777777777778, + "grad_norm": 187.7632293701172, + "learning_rate": 1.6364197530864198e-05, + "loss": 1.3417, + "step": 950 + }, + { + "epoch": 5.416666666666667, + "grad_norm": 46.503273010253906, + "learning_rate": 1.6209876543209876e-05, + "loss": 0.9956, + "step": 975 + }, + { + "epoch": 5.555555555555555, + "grad_norm": 60.819236755371094, + "learning_rate": 1.6055555555555557e-05, + "loss": 1.2102, + "step": 1000 + }, + { + "epoch": 5.694444444444445, + "grad_norm": 70.0938949584961, + "learning_rate": 1.5901234567901235e-05, + "loss": 1.2258, + "step": 1025 + }, + { + "epoch": 5.833333333333333, + "grad_norm": 102.03056335449219, + "learning_rate": 1.5746913580246916e-05, + "loss": 1.3799, + "step": 1050 + }, + { + "epoch": 5.972222222222222, + "grad_norm": 112.20748901367188, + "learning_rate": 1.5592592592592593e-05, + "loss": 1.4105, + "step": 1075 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7970149253731343, + "eval_loss": 0.6160737872123718, + "eval_macro_f1": 0.7962224936934712, + "eval_runtime": 1.3178, + "eval_samples_per_second": 508.41, + "eval_steps_per_second": 31.87, + "eval_weighted_f1": 0.7963763856554871, + "step": 1080 + }, + { + "epoch": 6.111111111111111, + "grad_norm": 49.92510223388672, + "learning_rate": 1.5438271604938274e-05, + "loss": 0.8145, + "step": 1100 + }, + { + "epoch": 6.25, + "grad_norm": 113.69713592529297, + "learning_rate": 1.5283950617283952e-05, + "loss": 0.821, + "step": 1125 + }, + { + "epoch": 6.388888888888889, + "grad_norm": 105.7056655883789, + "learning_rate": 1.5129629629629632e-05, + "loss": 0.9038, + "step": 1150 + }, + { + "epoch": 6.527777777777778, + "grad_norm": 104.82742309570312, + "learning_rate": 1.4975308641975311e-05, + "loss": 1.0963, + "step": 1175 + }, + { + "epoch": 6.666666666666667, + "grad_norm": 147.6919403076172, + "learning_rate": 1.4820987654320989e-05, + "loss": 0.9843, + "step": 1200 + }, + { + "epoch": 6.805555555555555, + "grad_norm": 68.3045654296875, + "learning_rate": 1.4666666666666666e-05, + "loss": 0.9752, + "step": 1225 + }, + { + "epoch": 6.944444444444445, + "grad_norm": 115.00918579101562, + "learning_rate": 1.4512345679012347e-05, + "loss": 0.8091, + "step": 1250 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8149253731343283, + "eval_loss": 0.7046055197715759, + "eval_macro_f1": 0.8139676692266414, + "eval_runtime": 1.1461, + "eval_samples_per_second": 584.611, + "eval_steps_per_second": 36.647, + "eval_weighted_f1": 0.8137634796606055, + "step": 1260 + }, + { + "epoch": 7.083333333333333, + "grad_norm": 97.88823699951172, + "learning_rate": 1.4358024691358025e-05, + "loss": 0.6555, + "step": 1275 + }, + { + "epoch": 7.222222222222222, + "grad_norm": 181.39883422851562, + "learning_rate": 1.4203703703703704e-05, + "loss": 0.6841, + "step": 1300 + }, + { + "epoch": 7.361111111111111, + "grad_norm": 117.61661529541016, + "learning_rate": 1.4049382716049384e-05, + "loss": 0.8938, + "step": 1325 + }, + { + "epoch": 7.5, + "grad_norm": 82.84748077392578, + "learning_rate": 1.3895061728395063e-05, + "loss": 0.652, + "step": 1350 + }, + { + "epoch": 7.638888888888889, + "grad_norm": 93.16009521484375, + "learning_rate": 1.3740740740740741e-05, + "loss": 0.7448, + "step": 1375 + }, + { + "epoch": 7.777777777777778, + "grad_norm": 73.9759750366211, + "learning_rate": 1.3586419753086422e-05, + "loss": 0.5702, + "step": 1400 + }, + { + "epoch": 7.916666666666667, + "grad_norm": 104.02964782714844, + "learning_rate": 1.34320987654321e-05, + "loss": 0.7234, + "step": 1425 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8059701492537313, + "eval_loss": 0.7318918704986572, + "eval_macro_f1": 0.8058259551649513, + "eval_runtime": 1.0972, + "eval_samples_per_second": 610.667, + "eval_steps_per_second": 38.281, + "eval_weighted_f1": 0.8053903878298831, + "step": 1440 + } + ], + "logging_steps": 25, + "max_steps": 3600, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/trainer_state.json b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..121f1bf9b7b3942010ce217719ea3ca78f009f5e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900/trainer_state.json @@ -0,0 +1,350 @@ +{ + "best_global_step": 900, + "best_metric": 0.815764837770578, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-05-s42-20260901T160928Z-27bda108/trainer/checkpoint-900", + "epoch": 5.0, + "eval_steps": 500, + "global_step": 900, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1388888888888889, + "grad_norm": 37.62110137939453, + "learning_rate": 1.3333333333333334e-06, + "loss": 4.9958, + "step": 25 + }, + { + "epoch": 0.2777777777777778, + "grad_norm": 33.463844299316406, + "learning_rate": 2.7222222222222224e-06, + "loss": 4.6188, + "step": 50 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 45.38542556762695, + "learning_rate": 4.111111111111111e-06, + "loss": 4.5391, + "step": 75 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 35.68893814086914, + "learning_rate": 5.500000000000001e-06, + "loss": 4.585, + "step": 100 + }, + { + "epoch": 0.6944444444444444, + "grad_norm": 55.31415557861328, + "learning_rate": 6.88888888888889e-06, + "loss": 4.4798, + "step": 125 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 44.58302688598633, + "learning_rate": 8.277777777777778e-06, + "loss": 4.4632, + "step": 150 + }, + { + "epoch": 0.9722222222222222, + "grad_norm": 68.91496276855469, + "learning_rate": 9.666666666666667e-06, + "loss": 4.1992, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5298507462686567, + "eval_loss": 0.9932835698127747, + "eval_macro_f1": 0.4812545627863858, + "eval_runtime": 1.1751, + "eval_samples_per_second": 570.178, + "eval_steps_per_second": 35.742, + "eval_weighted_f1": 0.4823316947081317, + "step": 180 + }, + { + "epoch": 1.1111111111111112, + "grad_norm": 43.673458099365234, + "learning_rate": 1.1055555555555557e-05, + "loss": 4.1202, + "step": 200 + }, + { + "epoch": 1.25, + "grad_norm": 33.03519821166992, + "learning_rate": 1.2444444444444446e-05, + "loss": 4.0191, + "step": 225 + }, + { + "epoch": 1.3888888888888888, + "grad_norm": 46.48714065551758, + "learning_rate": 1.3833333333333334e-05, + "loss": 3.9793, + "step": 250 + }, + { + "epoch": 1.5277777777777777, + "grad_norm": 189.76443481445312, + "learning_rate": 1.5222222222222223e-05, + "loss": 3.8691, + "step": 275 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 49.02412796020508, + "learning_rate": 1.6611111111111113e-05, + "loss": 3.9928, + "step": 300 + }, + { + "epoch": 1.8055555555555556, + "grad_norm": 35.443363189697266, + "learning_rate": 1.8e-05, + "loss": 3.499, + "step": 325 + }, + { + "epoch": 1.9444444444444444, + "grad_norm": 58.79069900512695, + "learning_rate": 1.938888888888889e-05, + "loss": 3.2991, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6850746268656717, + "eval_loss": 0.728695273399353, + "eval_macro_f1": 0.6824344063526143, + "eval_runtime": 1.1699, + "eval_samples_per_second": 572.687, + "eval_steps_per_second": 35.9, + "eval_weighted_f1": 0.6818588670899222, + "step": 360 + }, + { + "epoch": 2.0833333333333335, + "grad_norm": 74.10299682617188, + "learning_rate": 1.991358024691358e-05, + "loss": 3.4255, + "step": 375 + }, + { + "epoch": 2.2222222222222223, + "grad_norm": 79.9495620727539, + "learning_rate": 1.975925925925926e-05, + "loss": 3.434, + "step": 400 + }, + { + "epoch": 2.361111111111111, + "grad_norm": 105.76057434082031, + "learning_rate": 1.960493827160494e-05, + "loss": 2.6872, + "step": 425 + }, + { + "epoch": 2.5, + "grad_norm": 62.34187316894531, + "learning_rate": 1.945061728395062e-05, + "loss": 2.9178, + "step": 450 + }, + { + "epoch": 2.638888888888889, + "grad_norm": 72.12592315673828, + "learning_rate": 1.92962962962963e-05, + "loss": 2.5766, + "step": 475 + }, + { + "epoch": 2.7777777777777777, + "grad_norm": 46.273807525634766, + "learning_rate": 1.914197530864198e-05, + "loss": 2.6845, + "step": 500 + }, + { + "epoch": 2.9166666666666665, + "grad_norm": 103.11640930175781, + "learning_rate": 1.8987654320987657e-05, + "loss": 2.4601, + "step": 525 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.744776119402985, + "eval_loss": 0.6502454280853271, + "eval_macro_f1": 0.741043454588696, + "eval_runtime": 1.1156, + "eval_samples_per_second": 600.594, + "eval_steps_per_second": 37.649, + "eval_weighted_f1": 0.740781179677082, + "step": 540 + }, + { + "epoch": 3.0555555555555554, + "grad_norm": 58.63907241821289, + "learning_rate": 1.8833333333333335e-05, + "loss": 2.3028, + "step": 550 + }, + { + "epoch": 3.1944444444444446, + "grad_norm": 86.794921875, + "learning_rate": 1.8679012345679013e-05, + "loss": 2.1976, + "step": 575 + }, + { + "epoch": 3.3333333333333335, + "grad_norm": 90.98960876464844, + "learning_rate": 1.8524691358024694e-05, + "loss": 2.2231, + "step": 600 + }, + { + "epoch": 3.4722222222222223, + "grad_norm": 234.18736267089844, + "learning_rate": 1.837037037037037e-05, + "loss": 2.172, + "step": 625 + }, + { + "epoch": 3.611111111111111, + "grad_norm": 49.61118698120117, + "learning_rate": 1.821604938271605e-05, + "loss": 2.1864, + "step": 650 + }, + { + "epoch": 3.75, + "grad_norm": 72.47056579589844, + "learning_rate": 1.806172839506173e-05, + "loss": 2.2088, + "step": 675 + }, + { + "epoch": 3.888888888888889, + "grad_norm": 88.57938385009766, + "learning_rate": 1.7907407407407408e-05, + "loss": 2.0463, + "step": 700 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7776119402985074, + "eval_loss": 0.5820518136024475, + "eval_macro_f1": 0.7738326545779962, + "eval_runtime": 1.1582, + "eval_samples_per_second": 578.504, + "eval_steps_per_second": 36.264, + "eval_weighted_f1": 0.7738901764469578, + "step": 720 + }, + { + "epoch": 4.027777777777778, + "grad_norm": 67.66858673095703, + "learning_rate": 1.775308641975309e-05, + "loss": 1.7945, + "step": 725 + }, + { + "epoch": 4.166666666666667, + "grad_norm": 71.2930679321289, + "learning_rate": 1.7598765432098767e-05, + "loss": 1.7633, + "step": 750 + }, + { + "epoch": 4.305555555555555, + "grad_norm": 31.82473373413086, + "learning_rate": 1.7444444444444448e-05, + "loss": 1.574, + "step": 775 + }, + { + "epoch": 4.444444444444445, + "grad_norm": 102.59622192382812, + "learning_rate": 1.7290123456790125e-05, + "loss": 1.6594, + "step": 800 + }, + { + "epoch": 4.583333333333333, + "grad_norm": 35.59578323364258, + "learning_rate": 1.7135802469135806e-05, + "loss": 1.5991, + "step": 825 + }, + { + "epoch": 4.722222222222222, + "grad_norm": 75.3868179321289, + "learning_rate": 1.6981481481481484e-05, + "loss": 1.536, + "step": 850 + }, + { + "epoch": 4.861111111111111, + "grad_norm": 71.83612823486328, + "learning_rate": 1.6827160493827162e-05, + "loss": 1.6476, + "step": 875 + }, + { + "epoch": 5.0, + "grad_norm": 73.57019805908203, + "learning_rate": 1.667283950617284e-05, + "loss": 1.7134, + "step": 900 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8164179104477612, + "eval_loss": 0.5686514973640442, + "eval_macro_f1": 0.815764837770578, + "eval_runtime": 1.2508, + "eval_samples_per_second": 535.641, + "eval_steps_per_second": 33.577, + "eval_weighted_f1": 0.8155710164890037, + "step": 900 + } + ], + "logging_steps": 25, + "max_steps": 3600, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/trainer_state.json b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ae1d08c556b912d724295923348e02c164bd1083 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665/trainer_state.json @@ -0,0 +1,604 @@ +{ + "best_global_step": 1665, + "best_metric": 0.8381623053714656, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665", + "epoch": 9.0, + "eval_steps": 500, + "global_step": 1665, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13550135501355012, + "grad_norm": 17.273296356201172, + "learning_rate": 1.2972972972972974e-06, + "loss": 4.8612, + "step": 25 + }, + { + "epoch": 0.27100271002710025, + "grad_norm": 45.222206115722656, + "learning_rate": 2.648648648648649e-06, + "loss": 4.7234, + "step": 50 + }, + { + "epoch": 0.4065040650406504, + "grad_norm": 52.773521423339844, + "learning_rate": 4.000000000000001e-06, + "loss": 4.6747, + "step": 75 + }, + { + "epoch": 0.5420054200542005, + "grad_norm": 60.94910430908203, + "learning_rate": 5.351351351351351e-06, + "loss": 4.4934, + "step": 100 + }, + { + "epoch": 0.6775067750677507, + "grad_norm": 26.25322151184082, + "learning_rate": 6.702702702702704e-06, + "loss": 4.4177, + "step": 125 + }, + { + "epoch": 0.8130081300813008, + "grad_norm": 45.079402923583984, + "learning_rate": 8.054054054054055e-06, + "loss": 4.3802, + "step": 150 + }, + { + "epoch": 0.948509485094851, + "grad_norm": 92.62271118164062, + "learning_rate": 9.405405405405407e-06, + "loss": 4.3565, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5528455284552846, + "eval_loss": 0.9533843398094177, + "eval_macro_f1": 0.5533201568487114, + "eval_runtime": 1.2031, + "eval_samples_per_second": 613.43, + "eval_steps_per_second": 39.067, + "eval_weighted_f1": 0.5535761350421003, + "step": 185 + }, + { + "epoch": 1.08130081300813, + "grad_norm": 104.3695297241211, + "learning_rate": 1.0756756756756757e-05, + "loss": 4.1279, + "step": 200 + }, + { + "epoch": 1.2168021680216803, + "grad_norm": 61.002323150634766, + "learning_rate": 1.210810810810811e-05, + "loss": 4.0554, + "step": 225 + }, + { + "epoch": 1.3523035230352303, + "grad_norm": 39.037235260009766, + "learning_rate": 1.3459459459459461e-05, + "loss": 3.8461, + "step": 250 + }, + { + "epoch": 1.4878048780487805, + "grad_norm": 136.56031799316406, + "learning_rate": 1.4810810810810811e-05, + "loss": 3.7204, + "step": 275 + }, + { + "epoch": 1.6233062330623307, + "grad_norm": 50.59432601928711, + "learning_rate": 1.6162162162162163e-05, + "loss": 3.6472, + "step": 300 + }, + { + "epoch": 1.758807588075881, + "grad_norm": 142.0205535888672, + "learning_rate": 1.7513513513513517e-05, + "loss": 3.4228, + "step": 325 + }, + { + "epoch": 1.8943089430894309, + "grad_norm": 107.0827407836914, + "learning_rate": 1.8864864864864866e-05, + "loss": 3.1382, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6639566395663956, + "eval_loss": 0.8246460556983948, + "eval_macro_f1": 0.6570447999019429, + "eval_runtime": 1.3614, + "eval_samples_per_second": 542.09, + "eval_steps_per_second": 34.523, + "eval_weighted_f1": 0.6575824206486228, + "step": 370 + }, + { + "epoch": 2.02710027100271, + "grad_norm": 59.559391021728516, + "learning_rate": 1.9975975975975978e-05, + "loss": 2.9557, + "step": 375 + }, + { + "epoch": 2.16260162601626, + "grad_norm": 75.89033508300781, + "learning_rate": 1.9825825825825828e-05, + "loss": 2.6649, + "step": 400 + }, + { + "epoch": 2.2981029810298104, + "grad_norm": 95.62434387207031, + "learning_rate": 1.967567567567568e-05, + "loss": 2.7978, + "step": 425 + }, + { + "epoch": 2.4336043360433606, + "grad_norm": 83.58216094970703, + "learning_rate": 1.9525525525525525e-05, + "loss": 2.6414, + "step": 450 + }, + { + "epoch": 2.569105691056911, + "grad_norm": 48.62287902832031, + "learning_rate": 1.9375375375375376e-05, + "loss": 2.5494, + "step": 475 + }, + { + "epoch": 2.7046070460704605, + "grad_norm": 134.22503662109375, + "learning_rate": 1.9225225225225226e-05, + "loss": 2.4228, + "step": 500 + }, + { + "epoch": 2.8401084010840107, + "grad_norm": 82.39049530029297, + "learning_rate": 1.9075075075075076e-05, + "loss": 2.4734, + "step": 525 + }, + { + "epoch": 2.975609756097561, + "grad_norm": 57.494510650634766, + "learning_rate": 1.8924924924924926e-05, + "loss": 2.7417, + "step": 550 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7764227642276422, + "eval_loss": 0.5586827397346497, + "eval_macro_f1": 0.7739119933674031, + "eval_runtime": 1.3892, + "eval_samples_per_second": 531.228, + "eval_steps_per_second": 33.832, + "eval_weighted_f1": 0.7734635400757416, + "step": 555 + }, + { + "epoch": 3.1084010840108403, + "grad_norm": 69.49280548095703, + "learning_rate": 1.8774774774774776e-05, + "loss": 2.0024, + "step": 575 + }, + { + "epoch": 3.2439024390243905, + "grad_norm": 104.65282440185547, + "learning_rate": 1.8624624624624627e-05, + "loss": 2.0841, + "step": 600 + }, + { + "epoch": 3.3794037940379402, + "grad_norm": 122.3167724609375, + "learning_rate": 1.8474474474474477e-05, + "loss": 2.0393, + "step": 625 + }, + { + "epoch": 3.5149051490514904, + "grad_norm": 272.2197265625, + "learning_rate": 1.8324324324324324e-05, + "loss": 2.0099, + "step": 650 + }, + { + "epoch": 3.6504065040650406, + "grad_norm": 132.77316284179688, + "learning_rate": 1.8174174174174174e-05, + "loss": 2.002, + "step": 675 + }, + { + "epoch": 3.785907859078591, + "grad_norm": 53.379417419433594, + "learning_rate": 1.8024024024024024e-05, + "loss": 2.0656, + "step": 700 + }, + { + "epoch": 3.921409214092141, + "grad_norm": 83.30610656738281, + "learning_rate": 1.7873873873873874e-05, + "loss": 1.8266, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8021680216802168, + "eval_loss": 0.524685800075531, + "eval_macro_f1": 0.8037891855506656, + "eval_runtime": 1.2649, + "eval_samples_per_second": 583.452, + "eval_steps_per_second": 37.158, + "eval_weighted_f1": 0.8031503779409302, + "step": 740 + }, + { + "epoch": 4.05420054200542, + "grad_norm": 85.32197570800781, + "learning_rate": 1.7723723723723724e-05, + "loss": 1.8645, + "step": 750 + }, + { + "epoch": 4.1897018970189706, + "grad_norm": 68.92426300048828, + "learning_rate": 1.7573573573573575e-05, + "loss": 1.6606, + "step": 775 + }, + { + "epoch": 4.32520325203252, + "grad_norm": 96.4399642944336, + "learning_rate": 1.7423423423423425e-05, + "loss": 1.7402, + "step": 800 + }, + { + "epoch": 4.46070460704607, + "grad_norm": 42.06745147705078, + "learning_rate": 1.7273273273273275e-05, + "loss": 1.5336, + "step": 825 + }, + { + "epoch": 4.596205962059621, + "grad_norm": 97.70098876953125, + "learning_rate": 1.7123123123123125e-05, + "loss": 1.5495, + "step": 850 + }, + { + "epoch": 4.7317073170731705, + "grad_norm": 187.94525146484375, + "learning_rate": 1.6972972972972975e-05, + "loss": 1.6804, + "step": 875 + }, + { + "epoch": 4.867208672086721, + "grad_norm": 47.616798400878906, + "learning_rate": 1.6822822822822826e-05, + "loss": 1.7618, + "step": 900 + }, + { + "epoch": 5.0, + "grad_norm": 70.17745971679688, + "learning_rate": 1.6672672672672673e-05, + "loss": 1.6164, + "step": 925 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8279132791327913, + "eval_loss": 0.4947308897972107, + "eval_macro_f1": 0.8286053717123023, + "eval_runtime": 1.2593, + "eval_samples_per_second": 586.039, + "eval_steps_per_second": 37.322, + "eval_weighted_f1": 0.8281592424182951, + "step": 925 + }, + { + "epoch": 5.13550135501355, + "grad_norm": 60.16526794433594, + "learning_rate": 1.6522522522522523e-05, + "loss": 1.1908, + "step": 950 + }, + { + "epoch": 5.2710027100271, + "grad_norm": 83.22026062011719, + "learning_rate": 1.6372372372372373e-05, + "loss": 1.307, + "step": 975 + }, + { + "epoch": 5.40650406504065, + "grad_norm": 83.59474182128906, + "learning_rate": 1.6222222222222223e-05, + "loss": 1.3895, + "step": 1000 + }, + { + "epoch": 5.542005420054201, + "grad_norm": 83.32715606689453, + "learning_rate": 1.6072072072072073e-05, + "loss": 1.1913, + "step": 1025 + }, + { + "epoch": 5.677506775067751, + "grad_norm": 102.76142883300781, + "learning_rate": 1.5921921921921924e-05, + "loss": 1.192, + "step": 1050 + }, + { + "epoch": 5.8130081300813, + "grad_norm": 93.38829803466797, + "learning_rate": 1.5771771771771774e-05, + "loss": 1.1422, + "step": 1075 + }, + { + "epoch": 5.948509485094851, + "grad_norm": 140.5857391357422, + "learning_rate": 1.5621621621621624e-05, + "loss": 1.2216, + "step": 1100 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7886178861788617, + "eval_loss": 0.6806142926216125, + "eval_macro_f1": 0.7855888234874433, + "eval_runtime": 1.2639, + "eval_samples_per_second": 583.918, + "eval_steps_per_second": 37.187, + "eval_weighted_f1": 0.7858479984853648, + "step": 1110 + }, + { + "epoch": 6.08130081300813, + "grad_norm": 110.43148040771484, + "learning_rate": 1.5471471471471474e-05, + "loss": 1.2372, + "step": 1125 + }, + { + "epoch": 6.2168021680216805, + "grad_norm": 60.885066986083984, + "learning_rate": 1.5321321321321324e-05, + "loss": 0.8295, + "step": 1150 + }, + { + "epoch": 6.35230352303523, + "grad_norm": 291.20440673828125, + "learning_rate": 1.5171171171171173e-05, + "loss": 0.8575, + "step": 1175 + }, + { + "epoch": 6.487804878048781, + "grad_norm": 149.68145751953125, + "learning_rate": 1.5021021021021023e-05, + "loss": 0.9186, + "step": 1200 + }, + { + "epoch": 6.623306233062331, + "grad_norm": 125.19305419921875, + "learning_rate": 1.4870870870870873e-05, + "loss": 1.0152, + "step": 1225 + }, + { + "epoch": 6.7588075880758804, + "grad_norm": 108.62677764892578, + "learning_rate": 1.4720720720720722e-05, + "loss": 0.7621, + "step": 1250 + }, + { + "epoch": 6.894308943089431, + "grad_norm": 81.0872802734375, + "learning_rate": 1.4570570570570572e-05, + "loss": 0.8626, + "step": 1275 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8184281842818428, + "eval_loss": 0.6567783355712891, + "eval_macro_f1": 0.8202853520453462, + "eval_runtime": 1.2505, + "eval_samples_per_second": 590.147, + "eval_steps_per_second": 37.584, + "eval_weighted_f1": 0.820024010314149, + "step": 1295 + }, + { + "epoch": 7.02710027100271, + "grad_norm": 61.96566390991211, + "learning_rate": 1.4420420420420422e-05, + "loss": 0.8865, + "step": 1300 + }, + { + "epoch": 7.16260162601626, + "grad_norm": 112.96151733398438, + "learning_rate": 1.4270270270270272e-05, + "loss": 0.8028, + "step": 1325 + }, + { + "epoch": 7.29810298102981, + "grad_norm": 118.20665740966797, + "learning_rate": 1.4120120120120123e-05, + "loss": 0.7838, + "step": 1350 + }, + { + "epoch": 7.43360433604336, + "grad_norm": 316.36480712890625, + "learning_rate": 1.3969969969969971e-05, + "loss": 0.6848, + "step": 1375 + }, + { + "epoch": 7.569105691056911, + "grad_norm": 174.4762420654297, + "learning_rate": 1.3819819819819821e-05, + "loss": 0.7894, + "step": 1400 + }, + { + "epoch": 7.7046070460704605, + "grad_norm": 51.55758285522461, + "learning_rate": 1.3669669669669672e-05, + "loss": 0.7019, + "step": 1425 + }, + { + "epoch": 7.840108401084011, + "grad_norm": 91.97064971923828, + "learning_rate": 1.3519519519519522e-05, + "loss": 0.7346, + "step": 1450 + }, + { + "epoch": 7.975609756097561, + "grad_norm": 199.63336181640625, + "learning_rate": 1.3369369369369369e-05, + "loss": 0.6908, + "step": 1475 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8292682926829268, + "eval_loss": 0.6852610111236572, + "eval_macro_f1": 0.8297370267735968, + "eval_runtime": 1.2311, + "eval_samples_per_second": 599.442, + "eval_steps_per_second": 38.176, + "eval_weighted_f1": 0.8293397142626682, + "step": 1480 + }, + { + "epoch": 8.10840108401084, + "grad_norm": 14.679057121276855, + "learning_rate": 1.3219219219219219e-05, + "loss": 0.4381, + "step": 1500 + }, + { + "epoch": 8.24390243902439, + "grad_norm": 36.879180908203125, + "learning_rate": 1.3069069069069069e-05, + "loss": 0.4395, + "step": 1525 + }, + { + "epoch": 8.379403794037941, + "grad_norm": 93.2783203125, + "learning_rate": 1.291891891891892e-05, + "loss": 0.4395, + "step": 1550 + }, + { + "epoch": 8.51490514905149, + "grad_norm": 112.61410522460938, + "learning_rate": 1.276876876876877e-05, + "loss": 0.8745, + "step": 1575 + }, + { + "epoch": 8.65040650406504, + "grad_norm": 89.52852630615234, + "learning_rate": 1.2618618618618618e-05, + "loss": 0.6876, + "step": 1600 + }, + { + "epoch": 8.785907859078591, + "grad_norm": 55.04928207397461, + "learning_rate": 1.2468468468468468e-05, + "loss": 0.5565, + "step": 1625 + }, + { + "epoch": 8.92140921409214, + "grad_norm": 113.26346588134766, + "learning_rate": 1.2318318318318318e-05, + "loss": 0.4631, + "step": 1650 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8387533875338753, + "eval_loss": 0.8546782732009888, + "eval_macro_f1": 0.8381623053714656, + "eval_runtime": 1.1662, + "eval_samples_per_second": 632.83, + "eval_steps_per_second": 40.302, + "eval_weighted_f1": 0.8376472952472911, + "step": 1665 + } + ], + "logging_steps": 25, + "max_steps": 3700, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/trainer_state.json b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4dab0217134e4319b96799e598a1a65608e4b682 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-2220/trainer_state.json @@ -0,0 +1,791 @@ +{ + "best_global_step": 1665, + "best_metric": 0.8381623053714656, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-06-s42-20260901T161728Z-27bda108/trainer/checkpoint-1665", + "epoch": 12.0, + "eval_steps": 500, + "global_step": 2220, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13550135501355012, + "grad_norm": 17.273296356201172, + "learning_rate": 1.2972972972972974e-06, + "loss": 4.8612, + "step": 25 + }, + { + "epoch": 0.27100271002710025, + "grad_norm": 45.222206115722656, + "learning_rate": 2.648648648648649e-06, + "loss": 4.7234, + "step": 50 + }, + { + "epoch": 0.4065040650406504, + "grad_norm": 52.773521423339844, + "learning_rate": 4.000000000000001e-06, + "loss": 4.6747, + "step": 75 + }, + { + "epoch": 0.5420054200542005, + "grad_norm": 60.94910430908203, + "learning_rate": 5.351351351351351e-06, + "loss": 4.4934, + "step": 100 + }, + { + "epoch": 0.6775067750677507, + "grad_norm": 26.25322151184082, + "learning_rate": 6.702702702702704e-06, + "loss": 4.4177, + "step": 125 + }, + { + "epoch": 0.8130081300813008, + "grad_norm": 45.079402923583984, + "learning_rate": 8.054054054054055e-06, + "loss": 4.3802, + "step": 150 + }, + { + "epoch": 0.948509485094851, + "grad_norm": 92.62271118164062, + "learning_rate": 9.405405405405407e-06, + "loss": 4.3565, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5528455284552846, + "eval_loss": 0.9533843398094177, + "eval_macro_f1": 0.5533201568487114, + "eval_runtime": 1.2031, + "eval_samples_per_second": 613.43, + "eval_steps_per_second": 39.067, + "eval_weighted_f1": 0.5535761350421003, + "step": 185 + }, + { + "epoch": 1.08130081300813, + "grad_norm": 104.3695297241211, + "learning_rate": 1.0756756756756757e-05, + "loss": 4.1279, + "step": 200 + }, + { + "epoch": 1.2168021680216803, + "grad_norm": 61.002323150634766, + "learning_rate": 1.210810810810811e-05, + "loss": 4.0554, + "step": 225 + }, + { + "epoch": 1.3523035230352303, + "grad_norm": 39.037235260009766, + "learning_rate": 1.3459459459459461e-05, + "loss": 3.8461, + "step": 250 + }, + { + "epoch": 1.4878048780487805, + "grad_norm": 136.56031799316406, + "learning_rate": 1.4810810810810811e-05, + "loss": 3.7204, + "step": 275 + }, + { + "epoch": 1.6233062330623307, + "grad_norm": 50.59432601928711, + "learning_rate": 1.6162162162162163e-05, + "loss": 3.6472, + "step": 300 + }, + { + "epoch": 1.758807588075881, + "grad_norm": 142.0205535888672, + "learning_rate": 1.7513513513513517e-05, + "loss": 3.4228, + "step": 325 + }, + { + "epoch": 1.8943089430894309, + "grad_norm": 107.0827407836914, + "learning_rate": 1.8864864864864866e-05, + "loss": 3.1382, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6639566395663956, + "eval_loss": 0.8246460556983948, + "eval_macro_f1": 0.6570447999019429, + "eval_runtime": 1.3614, + "eval_samples_per_second": 542.09, + "eval_steps_per_second": 34.523, + "eval_weighted_f1": 0.6575824206486228, + "step": 370 + }, + { + "epoch": 2.02710027100271, + "grad_norm": 59.559391021728516, + "learning_rate": 1.9975975975975978e-05, + "loss": 2.9557, + "step": 375 + }, + { + "epoch": 2.16260162601626, + "grad_norm": 75.89033508300781, + "learning_rate": 1.9825825825825828e-05, + "loss": 2.6649, + "step": 400 + }, + { + "epoch": 2.2981029810298104, + "grad_norm": 95.62434387207031, + "learning_rate": 1.967567567567568e-05, + "loss": 2.7978, + "step": 425 + }, + { + "epoch": 2.4336043360433606, + "grad_norm": 83.58216094970703, + "learning_rate": 1.9525525525525525e-05, + "loss": 2.6414, + "step": 450 + }, + { + "epoch": 2.569105691056911, + "grad_norm": 48.62287902832031, + "learning_rate": 1.9375375375375376e-05, + "loss": 2.5494, + "step": 475 + }, + { + "epoch": 2.7046070460704605, + "grad_norm": 134.22503662109375, + "learning_rate": 1.9225225225225226e-05, + "loss": 2.4228, + "step": 500 + }, + { + "epoch": 2.8401084010840107, + "grad_norm": 82.39049530029297, + "learning_rate": 1.9075075075075076e-05, + "loss": 2.4734, + "step": 525 + }, + { + "epoch": 2.975609756097561, + "grad_norm": 57.494510650634766, + "learning_rate": 1.8924924924924926e-05, + "loss": 2.7417, + "step": 550 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7764227642276422, + "eval_loss": 0.5586827397346497, + "eval_macro_f1": 0.7739119933674031, + "eval_runtime": 1.3892, + "eval_samples_per_second": 531.228, + "eval_steps_per_second": 33.832, + "eval_weighted_f1": 0.7734635400757416, + "step": 555 + }, + { + "epoch": 3.1084010840108403, + "grad_norm": 69.49280548095703, + "learning_rate": 1.8774774774774776e-05, + "loss": 2.0024, + "step": 575 + }, + { + "epoch": 3.2439024390243905, + "grad_norm": 104.65282440185547, + "learning_rate": 1.8624624624624627e-05, + "loss": 2.0841, + "step": 600 + }, + { + "epoch": 3.3794037940379402, + "grad_norm": 122.3167724609375, + "learning_rate": 1.8474474474474477e-05, + "loss": 2.0393, + "step": 625 + }, + { + "epoch": 3.5149051490514904, + "grad_norm": 272.2197265625, + "learning_rate": 1.8324324324324324e-05, + "loss": 2.0099, + "step": 650 + }, + { + "epoch": 3.6504065040650406, + "grad_norm": 132.77316284179688, + "learning_rate": 1.8174174174174174e-05, + "loss": 2.002, + "step": 675 + }, + { + "epoch": 3.785907859078591, + "grad_norm": 53.379417419433594, + "learning_rate": 1.8024024024024024e-05, + "loss": 2.0656, + "step": 700 + }, + { + "epoch": 3.921409214092141, + "grad_norm": 83.30610656738281, + "learning_rate": 1.7873873873873874e-05, + "loss": 1.8266, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8021680216802168, + "eval_loss": 0.524685800075531, + "eval_macro_f1": 0.8037891855506656, + "eval_runtime": 1.2649, + "eval_samples_per_second": 583.452, + "eval_steps_per_second": 37.158, + "eval_weighted_f1": 0.8031503779409302, + "step": 740 + }, + { + "epoch": 4.05420054200542, + "grad_norm": 85.32197570800781, + "learning_rate": 1.7723723723723724e-05, + "loss": 1.8645, + "step": 750 + }, + { + "epoch": 4.1897018970189706, + "grad_norm": 68.92426300048828, + "learning_rate": 1.7573573573573575e-05, + "loss": 1.6606, + "step": 775 + }, + { + "epoch": 4.32520325203252, + "grad_norm": 96.4399642944336, + "learning_rate": 1.7423423423423425e-05, + "loss": 1.7402, + "step": 800 + }, + { + "epoch": 4.46070460704607, + "grad_norm": 42.06745147705078, + "learning_rate": 1.7273273273273275e-05, + "loss": 1.5336, + "step": 825 + }, + { + "epoch": 4.596205962059621, + "grad_norm": 97.70098876953125, + "learning_rate": 1.7123123123123125e-05, + "loss": 1.5495, + "step": 850 + }, + { + "epoch": 4.7317073170731705, + "grad_norm": 187.94525146484375, + "learning_rate": 1.6972972972972975e-05, + "loss": 1.6804, + "step": 875 + }, + { + "epoch": 4.867208672086721, + "grad_norm": 47.616798400878906, + "learning_rate": 1.6822822822822826e-05, + "loss": 1.7618, + "step": 900 + }, + { + "epoch": 5.0, + "grad_norm": 70.17745971679688, + "learning_rate": 1.6672672672672673e-05, + "loss": 1.6164, + "step": 925 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8279132791327913, + "eval_loss": 0.4947308897972107, + "eval_macro_f1": 0.8286053717123023, + "eval_runtime": 1.2593, + "eval_samples_per_second": 586.039, + "eval_steps_per_second": 37.322, + "eval_weighted_f1": 0.8281592424182951, + "step": 925 + }, + { + "epoch": 5.13550135501355, + "grad_norm": 60.16526794433594, + "learning_rate": 1.6522522522522523e-05, + "loss": 1.1908, + "step": 950 + }, + { + "epoch": 5.2710027100271, + "grad_norm": 83.22026062011719, + "learning_rate": 1.6372372372372373e-05, + "loss": 1.307, + "step": 975 + }, + { + "epoch": 5.40650406504065, + "grad_norm": 83.59474182128906, + "learning_rate": 1.6222222222222223e-05, + "loss": 1.3895, + "step": 1000 + }, + { + "epoch": 5.542005420054201, + "grad_norm": 83.32715606689453, + "learning_rate": 1.6072072072072073e-05, + "loss": 1.1913, + "step": 1025 + }, + { + "epoch": 5.677506775067751, + "grad_norm": 102.76142883300781, + "learning_rate": 1.5921921921921924e-05, + "loss": 1.192, + "step": 1050 + }, + { + "epoch": 5.8130081300813, + "grad_norm": 93.38829803466797, + "learning_rate": 1.5771771771771774e-05, + "loss": 1.1422, + "step": 1075 + }, + { + "epoch": 5.948509485094851, + "grad_norm": 140.5857391357422, + "learning_rate": 1.5621621621621624e-05, + "loss": 1.2216, + "step": 1100 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.7886178861788617, + "eval_loss": 0.6806142926216125, + "eval_macro_f1": 0.7855888234874433, + "eval_runtime": 1.2639, + "eval_samples_per_second": 583.918, + "eval_steps_per_second": 37.187, + "eval_weighted_f1": 0.7858479984853648, + "step": 1110 + }, + { + "epoch": 6.08130081300813, + "grad_norm": 110.43148040771484, + "learning_rate": 1.5471471471471474e-05, + "loss": 1.2372, + "step": 1125 + }, + { + "epoch": 6.2168021680216805, + "grad_norm": 60.885066986083984, + "learning_rate": 1.5321321321321324e-05, + "loss": 0.8295, + "step": 1150 + }, + { + "epoch": 6.35230352303523, + "grad_norm": 291.20440673828125, + "learning_rate": 1.5171171171171173e-05, + "loss": 0.8575, + "step": 1175 + }, + { + "epoch": 6.487804878048781, + "grad_norm": 149.68145751953125, + "learning_rate": 1.5021021021021023e-05, + "loss": 0.9186, + "step": 1200 + }, + { + "epoch": 6.623306233062331, + "grad_norm": 125.19305419921875, + "learning_rate": 1.4870870870870873e-05, + "loss": 1.0152, + "step": 1225 + }, + { + "epoch": 6.7588075880758804, + "grad_norm": 108.62677764892578, + "learning_rate": 1.4720720720720722e-05, + "loss": 0.7621, + "step": 1250 + }, + { + "epoch": 6.894308943089431, + "grad_norm": 81.0872802734375, + "learning_rate": 1.4570570570570572e-05, + "loss": 0.8626, + "step": 1275 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8184281842818428, + "eval_loss": 0.6567783355712891, + "eval_macro_f1": 0.8202853520453462, + "eval_runtime": 1.2505, + "eval_samples_per_second": 590.147, + "eval_steps_per_second": 37.584, + "eval_weighted_f1": 0.820024010314149, + "step": 1295 + }, + { + "epoch": 7.02710027100271, + "grad_norm": 61.96566390991211, + "learning_rate": 1.4420420420420422e-05, + "loss": 0.8865, + "step": 1300 + }, + { + "epoch": 7.16260162601626, + "grad_norm": 112.96151733398438, + "learning_rate": 1.4270270270270272e-05, + "loss": 0.8028, + "step": 1325 + }, + { + "epoch": 7.29810298102981, + "grad_norm": 118.20665740966797, + "learning_rate": 1.4120120120120123e-05, + "loss": 0.7838, + "step": 1350 + }, + { + "epoch": 7.43360433604336, + "grad_norm": 316.36480712890625, + "learning_rate": 1.3969969969969971e-05, + "loss": 0.6848, + "step": 1375 + }, + { + "epoch": 7.569105691056911, + "grad_norm": 174.4762420654297, + "learning_rate": 1.3819819819819821e-05, + "loss": 0.7894, + "step": 1400 + }, + { + "epoch": 7.7046070460704605, + "grad_norm": 51.55758285522461, + "learning_rate": 1.3669669669669672e-05, + "loss": 0.7019, + "step": 1425 + }, + { + "epoch": 7.840108401084011, + "grad_norm": 91.97064971923828, + "learning_rate": 1.3519519519519522e-05, + "loss": 0.7346, + "step": 1450 + }, + { + "epoch": 7.975609756097561, + "grad_norm": 199.63336181640625, + "learning_rate": 1.3369369369369369e-05, + "loss": 0.6908, + "step": 1475 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8292682926829268, + "eval_loss": 0.6852610111236572, + "eval_macro_f1": 0.8297370267735968, + "eval_runtime": 1.2311, + "eval_samples_per_second": 599.442, + "eval_steps_per_second": 38.176, + "eval_weighted_f1": 0.8293397142626682, + "step": 1480 + }, + { + "epoch": 8.10840108401084, + "grad_norm": 14.679057121276855, + "learning_rate": 1.3219219219219219e-05, + "loss": 0.4381, + "step": 1500 + }, + { + "epoch": 8.24390243902439, + "grad_norm": 36.879180908203125, + "learning_rate": 1.3069069069069069e-05, + "loss": 0.4395, + "step": 1525 + }, + { + "epoch": 8.379403794037941, + "grad_norm": 93.2783203125, + "learning_rate": 1.291891891891892e-05, + "loss": 0.4395, + "step": 1550 + }, + { + "epoch": 8.51490514905149, + "grad_norm": 112.61410522460938, + "learning_rate": 1.276876876876877e-05, + "loss": 0.8745, + "step": 1575 + }, + { + "epoch": 8.65040650406504, + "grad_norm": 89.52852630615234, + "learning_rate": 1.2618618618618618e-05, + "loss": 0.6876, + "step": 1600 + }, + { + "epoch": 8.785907859078591, + "grad_norm": 55.04928207397461, + "learning_rate": 1.2468468468468468e-05, + "loss": 0.5565, + "step": 1625 + }, + { + "epoch": 8.92140921409214, + "grad_norm": 113.26346588134766, + "learning_rate": 1.2318318318318318e-05, + "loss": 0.4631, + "step": 1650 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8387533875338753, + "eval_loss": 0.8546782732009888, + "eval_macro_f1": 0.8381623053714656, + "eval_runtime": 1.1662, + "eval_samples_per_second": 632.83, + "eval_steps_per_second": 40.302, + "eval_weighted_f1": 0.8376472952472911, + "step": 1665 + }, + { + "epoch": 9.05420054200542, + "grad_norm": 52.4693489074707, + "learning_rate": 1.2168168168168169e-05, + "loss": 0.404, + "step": 1675 + }, + { + "epoch": 9.18970189701897, + "grad_norm": 63.53031539916992, + "learning_rate": 1.2018018018018019e-05, + "loss": 0.5131, + "step": 1700 + }, + { + "epoch": 9.32520325203252, + "grad_norm": 54.33390426635742, + "learning_rate": 1.1867867867867869e-05, + "loss": 0.4122, + "step": 1725 + }, + { + "epoch": 9.46070460704607, + "grad_norm": 6.676888465881348, + "learning_rate": 1.1717717717717718e-05, + "loss": 0.5061, + "step": 1750 + }, + { + "epoch": 9.59620596205962, + "grad_norm": 14.658509254455566, + "learning_rate": 1.1567567567567568e-05, + "loss": 0.4456, + "step": 1775 + }, + { + "epoch": 9.731707317073171, + "grad_norm": 47.78240966796875, + "learning_rate": 1.1417417417417418e-05, + "loss": 0.5003, + "step": 1800 + }, + { + "epoch": 9.86720867208672, + "grad_norm": 80.65879821777344, + "learning_rate": 1.1267267267267268e-05, + "loss": 0.3979, + "step": 1825 + }, + { + "epoch": 10.0, + "grad_norm": 22.443307876586914, + "learning_rate": 1.1117117117117118e-05, + "loss": 0.4461, + "step": 1850 + }, + { + "epoch": 10.0, + "eval_accuracy": 0.8197831978319783, + "eval_loss": 0.9072073698043823, + "eval_macro_f1": 0.8206612205666043, + "eval_runtime": 1.486, + "eval_samples_per_second": 496.62, + "eval_steps_per_second": 31.628, + "eval_weighted_f1": 0.8199337069030147, + "step": 1850 + }, + { + "epoch": 10.13550135501355, + "grad_norm": 92.01063537597656, + "learning_rate": 1.0966966966966967e-05, + "loss": 0.2484, + "step": 1875 + }, + { + "epoch": 10.2710027100271, + "grad_norm": 95.98075866699219, + "learning_rate": 1.0816816816816817e-05, + "loss": 0.3268, + "step": 1900 + }, + { + "epoch": 10.40650406504065, + "grad_norm": 87.63033294677734, + "learning_rate": 1.0666666666666667e-05, + "loss": 0.246, + "step": 1925 + }, + { + "epoch": 10.5420054200542, + "grad_norm": 45.44370651245117, + "learning_rate": 1.0516516516516518e-05, + "loss": 0.3473, + "step": 1950 + }, + { + "epoch": 10.677506775067751, + "grad_norm": 298.974853515625, + "learning_rate": 1.0366366366366368e-05, + "loss": 0.417, + "step": 1975 + }, + { + "epoch": 10.8130081300813, + "grad_norm": 107.89459991455078, + "learning_rate": 1.0216216216216216e-05, + "loss": 0.4032, + "step": 2000 + }, + { + "epoch": 10.948509485094851, + "grad_norm": 204.2250213623047, + "learning_rate": 1.0066066066066066e-05, + "loss": 0.4096, + "step": 2025 + }, + { + "epoch": 11.0, + "eval_accuracy": 0.8319783197831978, + "eval_loss": 1.037054419517517, + "eval_macro_f1": 0.8317862968811259, + "eval_runtime": 1.2254, + "eval_samples_per_second": 602.247, + "eval_steps_per_second": 38.355, + "eval_weighted_f1": 0.8313957144091546, + "step": 2035 + }, + { + "epoch": 11.08130081300813, + "grad_norm": 328.2626647949219, + "learning_rate": 9.915915915915917e-06, + "loss": 0.3034, + "step": 2050 + }, + { + "epoch": 11.21680216802168, + "grad_norm": 100.41645812988281, + "learning_rate": 9.765765765765767e-06, + "loss": 0.1382, + "step": 2075 + }, + { + "epoch": 11.35230352303523, + "grad_norm": 2.3171396255493164, + "learning_rate": 9.615615615615617e-06, + "loss": 0.3975, + "step": 2100 + }, + { + "epoch": 11.487804878048781, + "grad_norm": 84.29185485839844, + "learning_rate": 9.465465465465467e-06, + "loss": 0.2363, + "step": 2125 + }, + { + "epoch": 11.62330623306233, + "grad_norm": 20.458799362182617, + "learning_rate": 9.315315315315316e-06, + "loss": 0.2349, + "step": 2150 + }, + { + "epoch": 11.75880758807588, + "grad_norm": 0.4830116033554077, + "learning_rate": 9.165165165165166e-06, + "loss": 0.2242, + "step": 2175 + }, + { + "epoch": 11.894308943089431, + "grad_norm": 21.482301712036133, + "learning_rate": 9.015015015015016e-06, + "loss": 0.2896, + "step": 2200 + }, + { + "epoch": 12.0, + "eval_accuracy": 0.8360433604336044, + "eval_loss": 1.2046059370040894, + "eval_macro_f1": 0.835750986050011, + "eval_runtime": 1.2246, + "eval_samples_per_second": 602.629, + "eval_steps_per_second": 38.379, + "eval_weighted_f1": 0.8354482701205196, + "step": 2220 + } + ], + "logging_steps": 25, + "max_steps": 3700, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/trainer_state.json b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..373bbba9bb7cf52e0bb910acea5c7a24c823a6f5 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110/trainer_state.json @@ -0,0 +1,417 @@ +{ + "best_global_step": 1110, + "best_metric": 0.8231322337943044, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110", + "epoch": 6.0, + "eval_steps": 500, + "global_step": 1110, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13513513513513514, + "grad_norm": 11.605118751525879, + "learning_rate": 1.2972972972972974e-06, + "loss": 4.7839, + "step": 25 + }, + { + "epoch": 0.2702702702702703, + "grad_norm": 43.74016571044922, + "learning_rate": 2.648648648648649e-06, + "loss": 4.7225, + "step": 50 + }, + { + "epoch": 0.40540540540540543, + "grad_norm": 36.49518966674805, + "learning_rate": 4.000000000000001e-06, + "loss": 4.5912, + "step": 75 + }, + { + "epoch": 0.5405405405405406, + "grad_norm": 93.14269256591797, + "learning_rate": 5.351351351351351e-06, + "loss": 4.6081, + "step": 100 + }, + { + "epoch": 0.6756756756756757, + "grad_norm": 28.402509689331055, + "learning_rate": 6.702702702702704e-06, + "loss": 4.4861, + "step": 125 + }, + { + "epoch": 0.8108108108108109, + "grad_norm": 22.84865951538086, + "learning_rate": 8.054054054054055e-06, + "loss": 4.504, + "step": 150 + }, + { + "epoch": 0.9459459459459459, + "grad_norm": 77.88905334472656, + "learning_rate": 9.405405405405407e-06, + "loss": 4.3239, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5260586319218241, + "eval_loss": 0.9716478586196899, + "eval_macro_f1": 0.51707370817878, + "eval_runtime": 1.151, + "eval_samples_per_second": 533.464, + "eval_steps_per_second": 33.884, + "eval_weighted_f1": 0.517342079836082, + "step": 185 + }, + { + "epoch": 1.0810810810810811, + "grad_norm": 55.785701751708984, + "learning_rate": 1.0756756756756757e-05, + "loss": 4.1362, + "step": 200 + }, + { + "epoch": 1.2162162162162162, + "grad_norm": 62.45780944824219, + "learning_rate": 1.210810810810811e-05, + "loss": 4.0336, + "step": 225 + }, + { + "epoch": 1.3513513513513513, + "grad_norm": 34.31761169433594, + "learning_rate": 1.3459459459459461e-05, + "loss": 4.0218, + "step": 250 + }, + { + "epoch": 1.4864864864864864, + "grad_norm": 48.01560974121094, + "learning_rate": 1.4810810810810811e-05, + "loss": 3.9971, + "step": 275 + }, + { + "epoch": 1.6216216216216215, + "grad_norm": 32.76003646850586, + "learning_rate": 1.6162162162162163e-05, + "loss": 3.7065, + "step": 300 + }, + { + "epoch": 1.7567567567567568, + "grad_norm": 134.93527221679688, + "learning_rate": 1.7513513513513517e-05, + "loss": 3.5559, + "step": 325 + }, + { + "epoch": 1.8918918918918919, + "grad_norm": 83.71576690673828, + "learning_rate": 1.8864864864864866e-05, + "loss": 3.4387, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6889250814332247, + "eval_loss": 0.7121844291687012, + "eval_macro_f1": 0.671205824739582, + "eval_runtime": 1.1163, + "eval_samples_per_second": 550.024, + "eval_steps_per_second": 34.936, + "eval_weighted_f1": 0.6711901979207707, + "step": 370 + }, + { + "epoch": 2.027027027027027, + "grad_norm": 73.79933166503906, + "learning_rate": 1.9975975975975978e-05, + "loss": 3.2071, + "step": 375 + }, + { + "epoch": 2.1621621621621623, + "grad_norm": 131.37176513671875, + "learning_rate": 1.9825825825825828e-05, + "loss": 2.9245, + "step": 400 + }, + { + "epoch": 2.2972972972972974, + "grad_norm": 63.687042236328125, + "learning_rate": 1.967567567567568e-05, + "loss": 3.175, + "step": 425 + }, + { + "epoch": 2.4324324324324325, + "grad_norm": 80.86520385742188, + "learning_rate": 1.9525525525525525e-05, + "loss": 3.0111, + "step": 450 + }, + { + "epoch": 2.5675675675675675, + "grad_norm": 78.62962341308594, + "learning_rate": 1.9375375375375376e-05, + "loss": 2.6214, + "step": 475 + }, + { + "epoch": 2.7027027027027026, + "grad_norm": 95.06387329101562, + "learning_rate": 1.9225225225225226e-05, + "loss": 2.5693, + "step": 500 + }, + { + "epoch": 2.8378378378378377, + "grad_norm": 158.118896484375, + "learning_rate": 1.9075075075075076e-05, + "loss": 2.2231, + "step": 525 + }, + { + "epoch": 2.972972972972973, + "grad_norm": 100.40279388427734, + "learning_rate": 1.8924924924924926e-05, + "loss": 2.5775, + "step": 550 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7703583061889251, + "eval_loss": 0.5711358189582825, + "eval_macro_f1": 0.7640741424673014, + "eval_runtime": 1.0881, + "eval_samples_per_second": 564.284, + "eval_steps_per_second": 35.842, + "eval_weighted_f1": 0.7642070876938322, + "step": 555 + }, + { + "epoch": 3.108108108108108, + "grad_norm": 62.59395980834961, + "learning_rate": 1.8774774774774776e-05, + "loss": 2.1234, + "step": 575 + }, + { + "epoch": 3.2432432432432434, + "grad_norm": 97.97312927246094, + "learning_rate": 1.8624624624624627e-05, + "loss": 2.346, + "step": 600 + }, + { + "epoch": 3.3783783783783785, + "grad_norm": 70.33486938476562, + "learning_rate": 1.8474474474474477e-05, + "loss": 2.0315, + "step": 625 + }, + { + "epoch": 3.5135135135135136, + "grad_norm": 73.25569152832031, + "learning_rate": 1.8324324324324324e-05, + "loss": 1.7002, + "step": 650 + }, + { + "epoch": 3.6486486486486487, + "grad_norm": 87.53681945800781, + "learning_rate": 1.8174174174174174e-05, + "loss": 2.2138, + "step": 675 + }, + { + "epoch": 3.7837837837837838, + "grad_norm": 152.4961395263672, + "learning_rate": 1.8024024024024024e-05, + "loss": 1.9726, + "step": 700 + }, + { + "epoch": 3.918918918918919, + "grad_norm": 49.09225845336914, + "learning_rate": 1.7873873873873874e-05, + "loss": 1.9897, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8078175895765473, + "eval_loss": 0.5133626461029053, + "eval_macro_f1": 0.8083530124680959, + "eval_runtime": 1.0452, + "eval_samples_per_second": 587.471, + "eval_steps_per_second": 37.315, + "eval_weighted_f1": 0.8081476121114107, + "step": 740 + }, + { + "epoch": 4.054054054054054, + "grad_norm": 40.89189529418945, + "learning_rate": 1.7723723723723724e-05, + "loss": 1.7937, + "step": 750 + }, + { + "epoch": 4.1891891891891895, + "grad_norm": 95.64726257324219, + "learning_rate": 1.7573573573573575e-05, + "loss": 1.41, + "step": 775 + }, + { + "epoch": 4.324324324324325, + "grad_norm": 76.54020690917969, + "learning_rate": 1.7423423423423425e-05, + "loss": 1.6785, + "step": 800 + }, + { + "epoch": 4.45945945945946, + "grad_norm": 51.53042221069336, + "learning_rate": 1.7273273273273275e-05, + "loss": 1.355, + "step": 825 + }, + { + "epoch": 4.594594594594595, + "grad_norm": 95.56114196777344, + "learning_rate": 1.7123123123123125e-05, + "loss": 1.7461, + "step": 850 + }, + { + "epoch": 4.72972972972973, + "grad_norm": 66.1458969116211, + "learning_rate": 1.6972972972972975e-05, + "loss": 1.609, + "step": 875 + }, + { + "epoch": 4.864864864864865, + "grad_norm": 176.03564453125, + "learning_rate": 1.6822822822822826e-05, + "loss": 1.7811, + "step": 900 + }, + { + "epoch": 5.0, + "grad_norm": 82.60294342041016, + "learning_rate": 1.6672672672672673e-05, + "loss": 1.6568, + "step": 925 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8110749185667753, + "eval_loss": 0.5527053475379944, + "eval_macro_f1": 0.809551527654976, + "eval_runtime": 1.1328, + "eval_samples_per_second": 542.019, + "eval_steps_per_second": 34.428, + "eval_weighted_f1": 0.8095463115765261, + "step": 925 + }, + { + "epoch": 5.135135135135135, + "grad_norm": 79.85108184814453, + "learning_rate": 1.6522522522522523e-05, + "loss": 1.0759, + "step": 950 + }, + { + "epoch": 5.27027027027027, + "grad_norm": 73.41077423095703, + "learning_rate": 1.6372372372372373e-05, + "loss": 1.3275, + "step": 975 + }, + { + "epoch": 5.405405405405405, + "grad_norm": 80.87350463867188, + "learning_rate": 1.6222222222222223e-05, + "loss": 1.0801, + "step": 1000 + }, + { + "epoch": 5.54054054054054, + "grad_norm": 88.7418212890625, + "learning_rate": 1.6072072072072073e-05, + "loss": 1.3417, + "step": 1025 + }, + { + "epoch": 5.675675675675675, + "grad_norm": 88.4278335571289, + "learning_rate": 1.5921921921921924e-05, + "loss": 1.2024, + "step": 1050 + }, + { + "epoch": 5.8108108108108105, + "grad_norm": 89.01246643066406, + "learning_rate": 1.5771771771771774e-05, + "loss": 0.9616, + "step": 1075 + }, + { + "epoch": 5.945945945945946, + "grad_norm": 59.05881881713867, + "learning_rate": 1.5621621621621624e-05, + "loss": 1.3088, + "step": 1100 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8241042345276873, + "eval_loss": 0.5578638315200806, + "eval_macro_f1": 0.8231322337943044, + "eval_runtime": 1.0628, + "eval_samples_per_second": 577.697, + "eval_steps_per_second": 36.694, + "eval_weighted_f1": 0.8229368138550882, + "step": 1110 + } + ], + "logging_steps": 25, + "max_steps": 3700, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/trainer_state.json b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..0b6cb3643772a7c388271d09bda65bc00a36e620 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1665/trainer_state.json @@ -0,0 +1,604 @@ +{ + "best_global_step": 1110, + "best_metric": 0.8231322337943044, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-07-s42-20260901T162936Z-27bda108/trainer/checkpoint-1110", + "epoch": 9.0, + "eval_steps": 500, + "global_step": 1665, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13513513513513514, + "grad_norm": 11.605118751525879, + "learning_rate": 1.2972972972972974e-06, + "loss": 4.7839, + "step": 25 + }, + { + "epoch": 0.2702702702702703, + "grad_norm": 43.74016571044922, + "learning_rate": 2.648648648648649e-06, + "loss": 4.7225, + "step": 50 + }, + { + "epoch": 0.40540540540540543, + "grad_norm": 36.49518966674805, + "learning_rate": 4.000000000000001e-06, + "loss": 4.5912, + "step": 75 + }, + { + "epoch": 0.5405405405405406, + "grad_norm": 93.14269256591797, + "learning_rate": 5.351351351351351e-06, + "loss": 4.6081, + "step": 100 + }, + { + "epoch": 0.6756756756756757, + "grad_norm": 28.402509689331055, + "learning_rate": 6.702702702702704e-06, + "loss": 4.4861, + "step": 125 + }, + { + "epoch": 0.8108108108108109, + "grad_norm": 22.84865951538086, + "learning_rate": 8.054054054054055e-06, + "loss": 4.504, + "step": 150 + }, + { + "epoch": 0.9459459459459459, + "grad_norm": 77.88905334472656, + "learning_rate": 9.405405405405407e-06, + "loss": 4.3239, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5260586319218241, + "eval_loss": 0.9716478586196899, + "eval_macro_f1": 0.51707370817878, + "eval_runtime": 1.151, + "eval_samples_per_second": 533.464, + "eval_steps_per_second": 33.884, + "eval_weighted_f1": 0.517342079836082, + "step": 185 + }, + { + "epoch": 1.0810810810810811, + "grad_norm": 55.785701751708984, + "learning_rate": 1.0756756756756757e-05, + "loss": 4.1362, + "step": 200 + }, + { + "epoch": 1.2162162162162162, + "grad_norm": 62.45780944824219, + "learning_rate": 1.210810810810811e-05, + "loss": 4.0336, + "step": 225 + }, + { + "epoch": 1.3513513513513513, + "grad_norm": 34.31761169433594, + "learning_rate": 1.3459459459459461e-05, + "loss": 4.0218, + "step": 250 + }, + { + "epoch": 1.4864864864864864, + "grad_norm": 48.01560974121094, + "learning_rate": 1.4810810810810811e-05, + "loss": 3.9971, + "step": 275 + }, + { + "epoch": 1.6216216216216215, + "grad_norm": 32.76003646850586, + "learning_rate": 1.6162162162162163e-05, + "loss": 3.7065, + "step": 300 + }, + { + "epoch": 1.7567567567567568, + "grad_norm": 134.93527221679688, + "learning_rate": 1.7513513513513517e-05, + "loss": 3.5559, + "step": 325 + }, + { + "epoch": 1.8918918918918919, + "grad_norm": 83.71576690673828, + "learning_rate": 1.8864864864864866e-05, + "loss": 3.4387, + "step": 350 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6889250814332247, + "eval_loss": 0.7121844291687012, + "eval_macro_f1": 0.671205824739582, + "eval_runtime": 1.1163, + "eval_samples_per_second": 550.024, + "eval_steps_per_second": 34.936, + "eval_weighted_f1": 0.6711901979207707, + "step": 370 + }, + { + "epoch": 2.027027027027027, + "grad_norm": 73.79933166503906, + "learning_rate": 1.9975975975975978e-05, + "loss": 3.2071, + "step": 375 + }, + { + "epoch": 2.1621621621621623, + "grad_norm": 131.37176513671875, + "learning_rate": 1.9825825825825828e-05, + "loss": 2.9245, + "step": 400 + }, + { + "epoch": 2.2972972972972974, + "grad_norm": 63.687042236328125, + "learning_rate": 1.967567567567568e-05, + "loss": 3.175, + "step": 425 + }, + { + "epoch": 2.4324324324324325, + "grad_norm": 80.86520385742188, + "learning_rate": 1.9525525525525525e-05, + "loss": 3.0111, + "step": 450 + }, + { + "epoch": 2.5675675675675675, + "grad_norm": 78.62962341308594, + "learning_rate": 1.9375375375375376e-05, + "loss": 2.6214, + "step": 475 + }, + { + "epoch": 2.7027027027027026, + "grad_norm": 95.06387329101562, + "learning_rate": 1.9225225225225226e-05, + "loss": 2.5693, + "step": 500 + }, + { + "epoch": 2.8378378378378377, + "grad_norm": 158.118896484375, + "learning_rate": 1.9075075075075076e-05, + "loss": 2.2231, + "step": 525 + }, + { + "epoch": 2.972972972972973, + "grad_norm": 100.40279388427734, + "learning_rate": 1.8924924924924926e-05, + "loss": 2.5775, + "step": 550 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7703583061889251, + "eval_loss": 0.5711358189582825, + "eval_macro_f1": 0.7640741424673014, + "eval_runtime": 1.0881, + "eval_samples_per_second": 564.284, + "eval_steps_per_second": 35.842, + "eval_weighted_f1": 0.7642070876938322, + "step": 555 + }, + { + "epoch": 3.108108108108108, + "grad_norm": 62.59395980834961, + "learning_rate": 1.8774774774774776e-05, + "loss": 2.1234, + "step": 575 + }, + { + "epoch": 3.2432432432432434, + "grad_norm": 97.97312927246094, + "learning_rate": 1.8624624624624627e-05, + "loss": 2.346, + "step": 600 + }, + { + "epoch": 3.3783783783783785, + "grad_norm": 70.33486938476562, + "learning_rate": 1.8474474474474477e-05, + "loss": 2.0315, + "step": 625 + }, + { + "epoch": 3.5135135135135136, + "grad_norm": 73.25569152832031, + "learning_rate": 1.8324324324324324e-05, + "loss": 1.7002, + "step": 650 + }, + { + "epoch": 3.6486486486486487, + "grad_norm": 87.53681945800781, + "learning_rate": 1.8174174174174174e-05, + "loss": 2.2138, + "step": 675 + }, + { + "epoch": 3.7837837837837838, + "grad_norm": 152.4961395263672, + "learning_rate": 1.8024024024024024e-05, + "loss": 1.9726, + "step": 700 + }, + { + "epoch": 3.918918918918919, + "grad_norm": 49.09225845336914, + "learning_rate": 1.7873873873873874e-05, + "loss": 1.9897, + "step": 725 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.8078175895765473, + "eval_loss": 0.5133626461029053, + "eval_macro_f1": 0.8083530124680959, + "eval_runtime": 1.0452, + "eval_samples_per_second": 587.471, + "eval_steps_per_second": 37.315, + "eval_weighted_f1": 0.8081476121114107, + "step": 740 + }, + { + "epoch": 4.054054054054054, + "grad_norm": 40.89189529418945, + "learning_rate": 1.7723723723723724e-05, + "loss": 1.7937, + "step": 750 + }, + { + "epoch": 4.1891891891891895, + "grad_norm": 95.64726257324219, + "learning_rate": 1.7573573573573575e-05, + "loss": 1.41, + "step": 775 + }, + { + "epoch": 4.324324324324325, + "grad_norm": 76.54020690917969, + "learning_rate": 1.7423423423423425e-05, + "loss": 1.6785, + "step": 800 + }, + { + "epoch": 4.45945945945946, + "grad_norm": 51.53042221069336, + "learning_rate": 1.7273273273273275e-05, + "loss": 1.355, + "step": 825 + }, + { + "epoch": 4.594594594594595, + "grad_norm": 95.56114196777344, + "learning_rate": 1.7123123123123125e-05, + "loss": 1.7461, + "step": 850 + }, + { + "epoch": 4.72972972972973, + "grad_norm": 66.1458969116211, + "learning_rate": 1.6972972972972975e-05, + "loss": 1.609, + "step": 875 + }, + { + "epoch": 4.864864864864865, + "grad_norm": 176.03564453125, + "learning_rate": 1.6822822822822826e-05, + "loss": 1.7811, + "step": 900 + }, + { + "epoch": 5.0, + "grad_norm": 82.60294342041016, + "learning_rate": 1.6672672672672673e-05, + "loss": 1.6568, + "step": 925 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.8110749185667753, + "eval_loss": 0.5527053475379944, + "eval_macro_f1": 0.809551527654976, + "eval_runtime": 1.1328, + "eval_samples_per_second": 542.019, + "eval_steps_per_second": 34.428, + "eval_weighted_f1": 0.8095463115765261, + "step": 925 + }, + { + "epoch": 5.135135135135135, + "grad_norm": 79.85108184814453, + "learning_rate": 1.6522522522522523e-05, + "loss": 1.0759, + "step": 950 + }, + { + "epoch": 5.27027027027027, + "grad_norm": 73.41077423095703, + "learning_rate": 1.6372372372372373e-05, + "loss": 1.3275, + "step": 975 + }, + { + "epoch": 5.405405405405405, + "grad_norm": 80.87350463867188, + "learning_rate": 1.6222222222222223e-05, + "loss": 1.0801, + "step": 1000 + }, + { + "epoch": 5.54054054054054, + "grad_norm": 88.7418212890625, + "learning_rate": 1.6072072072072073e-05, + "loss": 1.3417, + "step": 1025 + }, + { + "epoch": 5.675675675675675, + "grad_norm": 88.4278335571289, + "learning_rate": 1.5921921921921924e-05, + "loss": 1.2024, + "step": 1050 + }, + { + "epoch": 5.8108108108108105, + "grad_norm": 89.01246643066406, + "learning_rate": 1.5771771771771774e-05, + "loss": 0.9616, + "step": 1075 + }, + { + "epoch": 5.945945945945946, + "grad_norm": 59.05881881713867, + "learning_rate": 1.5621621621621624e-05, + "loss": 1.3088, + "step": 1100 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8241042345276873, + "eval_loss": 0.5578638315200806, + "eval_macro_f1": 0.8231322337943044, + "eval_runtime": 1.0628, + "eval_samples_per_second": 577.697, + "eval_steps_per_second": 36.694, + "eval_weighted_f1": 0.8229368138550882, + "step": 1110 + }, + { + "epoch": 6.081081081081081, + "grad_norm": 70.73210144042969, + "learning_rate": 1.5471471471471474e-05, + "loss": 1.1676, + "step": 1125 + }, + { + "epoch": 6.216216216216216, + "grad_norm": 42.22643280029297, + "learning_rate": 1.5321321321321324e-05, + "loss": 0.9654, + "step": 1150 + }, + { + "epoch": 6.351351351351352, + "grad_norm": 36.849666595458984, + "learning_rate": 1.5171171171171173e-05, + "loss": 0.8741, + "step": 1175 + }, + { + "epoch": 6.486486486486487, + "grad_norm": 43.80562973022461, + "learning_rate": 1.5021021021021023e-05, + "loss": 0.8245, + "step": 1200 + }, + { + "epoch": 6.621621621621622, + "grad_norm": 124.3948974609375, + "learning_rate": 1.4870870870870873e-05, + "loss": 0.9493, + "step": 1225 + }, + { + "epoch": 6.756756756756757, + "grad_norm": 68.37638092041016, + "learning_rate": 1.4720720720720722e-05, + "loss": 1.0329, + "step": 1250 + }, + { + "epoch": 6.891891891891892, + "grad_norm": 99.2005844116211, + "learning_rate": 1.4570570570570572e-05, + "loss": 0.878, + "step": 1275 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8192182410423453, + "eval_loss": 0.6424212455749512, + "eval_macro_f1": 0.8172937888504833, + "eval_runtime": 1.1316, + "eval_samples_per_second": 542.614, + "eval_steps_per_second": 34.466, + "eval_weighted_f1": 0.8171390295502668, + "step": 1295 + }, + { + "epoch": 7.027027027027027, + "grad_norm": 55.61606216430664, + "learning_rate": 1.4420420420420422e-05, + "loss": 0.7639, + "step": 1300 + }, + { + "epoch": 7.162162162162162, + "grad_norm": 10.259869575500488, + "learning_rate": 1.4270270270270272e-05, + "loss": 0.5168, + "step": 1325 + }, + { + "epoch": 7.297297297297297, + "grad_norm": 28.353843688964844, + "learning_rate": 1.4120120120120123e-05, + "loss": 0.7502, + "step": 1350 + }, + { + "epoch": 7.4324324324324325, + "grad_norm": 144.66029357910156, + "learning_rate": 1.3969969969969971e-05, + "loss": 0.641, + "step": 1375 + }, + { + "epoch": 7.5675675675675675, + "grad_norm": 124.58782196044922, + "learning_rate": 1.3819819819819821e-05, + "loss": 0.6489, + "step": 1400 + }, + { + "epoch": 7.702702702702703, + "grad_norm": 82.16483306884766, + "learning_rate": 1.3669669669669672e-05, + "loss": 0.5346, + "step": 1425 + }, + { + "epoch": 7.837837837837838, + "grad_norm": 21.13196563720703, + "learning_rate": 1.3519519519519522e-05, + "loss": 0.7089, + "step": 1450 + }, + { + "epoch": 7.972972972972973, + "grad_norm": 143.62132263183594, + "learning_rate": 1.3369369369369369e-05, + "loss": 0.6814, + "step": 1475 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8013029315960912, + "eval_loss": 0.7883321046829224, + "eval_macro_f1": 0.8026342587023673, + "eval_runtime": 1.1121, + "eval_samples_per_second": 552.127, + "eval_steps_per_second": 35.07, + "eval_weighted_f1": 0.8025539614954474, + "step": 1480 + }, + { + "epoch": 8.108108108108109, + "grad_norm": 93.11283111572266, + "learning_rate": 1.3219219219219219e-05, + "loss": 0.5499, + "step": 1500 + }, + { + "epoch": 8.243243243243244, + "grad_norm": 118.35292053222656, + "learning_rate": 1.3069069069069069e-05, + "loss": 0.4197, + "step": 1525 + }, + { + "epoch": 8.378378378378379, + "grad_norm": 241.2148895263672, + "learning_rate": 1.291891891891892e-05, + "loss": 0.5206, + "step": 1550 + }, + { + "epoch": 8.513513513513514, + "grad_norm": 79.677978515625, + "learning_rate": 1.276876876876877e-05, + "loss": 0.6288, + "step": 1575 + }, + { + "epoch": 8.64864864864865, + "grad_norm": 89.29263305664062, + "learning_rate": 1.2618618618618618e-05, + "loss": 0.4473, + "step": 1600 + }, + { + "epoch": 8.783783783783784, + "grad_norm": 89.60794830322266, + "learning_rate": 1.2468468468468468e-05, + "loss": 0.6951, + "step": 1625 + }, + { + "epoch": 8.91891891891892, + "grad_norm": 134.89093017578125, + "learning_rate": 1.2318318318318318e-05, + "loss": 0.6362, + "step": 1650 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.7980456026058632, + "eval_loss": 0.797012984752655, + "eval_macro_f1": 0.7984582398333386, + "eval_runtime": 1.1062, + "eval_samples_per_second": 555.065, + "eval_steps_per_second": 35.257, + "eval_weighted_f1": 0.7980488303998009, + "step": 1665 + } + ], + "logging_steps": 25, + "max_steps": 3700, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/trainer_state.json b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..549bbc7950cbc80982ce6dc64050f3b48498d054 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140/trainer_state.json @@ -0,0 +1,424 @@ +{ + "best_global_step": 1140, + "best_metric": 0.8243747841681272, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140", + "epoch": 6.0, + "eval_steps": 500, + "global_step": 1140, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13175230566534915, + "grad_norm": 30.87647819519043, + "learning_rate": 1.2631578947368422e-06, + "loss": 4.8055, + "step": 25 + }, + { + "epoch": 0.2635046113306983, + "grad_norm": 55.19264221191406, + "learning_rate": 2.578947368421053e-06, + "loss": 4.8098, + "step": 50 + }, + { + "epoch": 0.3952569169960474, + "grad_norm": 29.301651000976562, + "learning_rate": 3.894736842105263e-06, + "loss": 4.6563, + "step": 75 + }, + { + "epoch": 0.5270092226613966, + "grad_norm": 61.64167785644531, + "learning_rate": 5.210526315789474e-06, + "loss": 4.6241, + "step": 100 + }, + { + "epoch": 0.6587615283267457, + "grad_norm": 45.149295806884766, + "learning_rate": 6.526315789473685e-06, + "loss": 4.5889, + "step": 125 + }, + { + "epoch": 0.7905138339920948, + "grad_norm": 55.359588623046875, + "learning_rate": 7.842105263157895e-06, + "loss": 4.5611, + "step": 150 + }, + { + "epoch": 0.922266139657444, + "grad_norm": 103.41445922851562, + "learning_rate": 9.157894736842105e-06, + "loss": 4.336, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5502248875562219, + "eval_loss": 0.9538599848747253, + "eval_macro_f1": 0.48829869022613986, + "eval_runtime": 1.1679, + "eval_samples_per_second": 571.095, + "eval_steps_per_second": 35.961, + "eval_weighted_f1": 0.4909424318804821, + "step": 190 + }, + { + "epoch": 1.0527009222661396, + "grad_norm": 64.43714141845703, + "learning_rate": 1.0473684210526316e-05, + "loss": 4.1692, + "step": 200 + }, + { + "epoch": 1.1844532279314888, + "grad_norm": 67.21669006347656, + "learning_rate": 1.1789473684210527e-05, + "loss": 4.0247, + "step": 225 + }, + { + "epoch": 1.316205533596838, + "grad_norm": 23.2484073638916, + "learning_rate": 1.310526315789474e-05, + "loss": 3.9293, + "step": 250 + }, + { + "epoch": 1.447957839262187, + "grad_norm": 52.543296813964844, + "learning_rate": 1.4421052631578948e-05, + "loss": 3.6839, + "step": 275 + }, + { + "epoch": 1.5797101449275361, + "grad_norm": 58.90916442871094, + "learning_rate": 1.573684210526316e-05, + "loss": 3.503, + "step": 300 + }, + { + "epoch": 1.7114624505928853, + "grad_norm": 200.71676635742188, + "learning_rate": 1.705263157894737e-05, + "loss": 3.2837, + "step": 325 + }, + { + "epoch": 1.8432147562582344, + "grad_norm": 117.61235809326172, + "learning_rate": 1.836842105263158e-05, + "loss": 3.3762, + "step": 350 + }, + { + "epoch": 1.9749670619235835, + "grad_norm": 141.29888916015625, + "learning_rate": 1.968421052631579e-05, + "loss": 3.7328, + "step": 375 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6806596701649176, + "eval_loss": 0.6895695328712463, + "eval_macro_f1": 0.6716922851036303, + "eval_runtime": 1.1188, + "eval_samples_per_second": 596.152, + "eval_steps_per_second": 37.539, + "eval_weighted_f1": 0.6728461310490084, + "step": 380 + }, + { + "epoch": 2.1054018445322793, + "grad_norm": 103.40943908691406, + "learning_rate": 1.988888888888889e-05, + "loss": 3.131, + "step": 400 + }, + { + "epoch": 2.2371541501976284, + "grad_norm": 50.52198791503906, + "learning_rate": 1.9742690058479533e-05, + "loss": 2.9059, + "step": 425 + }, + { + "epoch": 2.3689064558629775, + "grad_norm": 64.69234466552734, + "learning_rate": 1.9596491228070177e-05, + "loss": 2.9769, + "step": 450 + }, + { + "epoch": 2.5006587615283267, + "grad_norm": 116.13067626953125, + "learning_rate": 1.945029239766082e-05, + "loss": 2.8277, + "step": 475 + }, + { + "epoch": 2.632411067193676, + "grad_norm": 41.39086151123047, + "learning_rate": 1.9304093567251464e-05, + "loss": 2.8187, + "step": 500 + }, + { + "epoch": 2.764163372859025, + "grad_norm": 106.1263427734375, + "learning_rate": 1.9157894736842108e-05, + "loss": 2.4984, + "step": 525 + }, + { + "epoch": 2.895915678524374, + "grad_norm": 129.88247680664062, + "learning_rate": 1.901169590643275e-05, + "loss": 2.9183, + "step": 550 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7931034482758621, + "eval_loss": 0.5418083071708679, + "eval_macro_f1": 0.792407856503026, + "eval_runtime": 1.1477, + "eval_samples_per_second": 581.172, + "eval_steps_per_second": 36.596, + "eval_weighted_f1": 0.7926835166673071, + "step": 570 + }, + { + "epoch": 3.02635046113307, + "grad_norm": 93.23123931884766, + "learning_rate": 1.8865497076023395e-05, + "loss": 2.2853, + "step": 575 + }, + { + "epoch": 3.158102766798419, + "grad_norm": 76.33997344970703, + "learning_rate": 1.871929824561404e-05, + "loss": 2.2193, + "step": 600 + }, + { + "epoch": 3.289855072463768, + "grad_norm": 62.839385986328125, + "learning_rate": 1.857309941520468e-05, + "loss": 2.2484, + "step": 625 + }, + { + "epoch": 3.421607378129117, + "grad_norm": 82.7411880493164, + "learning_rate": 1.8426900584795323e-05, + "loss": 2.0458, + "step": 650 + }, + { + "epoch": 3.5533596837944663, + "grad_norm": 56.64774703979492, + "learning_rate": 1.8280701754385966e-05, + "loss": 2.2254, + "step": 675 + }, + { + "epoch": 3.6851119894598154, + "grad_norm": 47.41438293457031, + "learning_rate": 1.813450292397661e-05, + "loss": 1.9588, + "step": 700 + }, + { + "epoch": 3.8168642951251646, + "grad_norm": 48.75365447998047, + "learning_rate": 1.7988304093567253e-05, + "loss": 2.0426, + "step": 725 + }, + { + "epoch": 3.9486166007905137, + "grad_norm": 72.154296875, + "learning_rate": 1.7842105263157897e-05, + "loss": 1.9697, + "step": 750 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7796101949025487, + "eval_loss": 0.528952419757843, + "eval_macro_f1": 0.7836733289235748, + "eval_runtime": 1.2131, + "eval_samples_per_second": 549.833, + "eval_steps_per_second": 34.622, + "eval_weighted_f1": 0.7837899739423578, + "step": 760 + }, + { + "epoch": 4.07905138339921, + "grad_norm": 64.87993621826172, + "learning_rate": 1.769590643274854e-05, + "loss": 1.766, + "step": 775 + }, + { + "epoch": 4.210803689064559, + "grad_norm": 147.2518310546875, + "learning_rate": 1.7549707602339184e-05, + "loss": 1.7144, + "step": 800 + }, + { + "epoch": 4.342555994729908, + "grad_norm": 140.36033630371094, + "learning_rate": 1.7403508771929828e-05, + "loss": 1.5639, + "step": 825 + }, + { + "epoch": 4.474308300395257, + "grad_norm": 126.15364837646484, + "learning_rate": 1.725730994152047e-05, + "loss": 1.595, + "step": 850 + }, + { + "epoch": 4.606060606060606, + "grad_norm": 95.8467788696289, + "learning_rate": 1.7111111111111112e-05, + "loss": 1.6152, + "step": 875 + }, + { + "epoch": 4.737812911725955, + "grad_norm": 63.30301284790039, + "learning_rate": 1.6964912280701755e-05, + "loss": 1.5819, + "step": 900 + }, + { + "epoch": 4.869565217391305, + "grad_norm": 103.65847778320312, + "learning_rate": 1.68187134502924e-05, + "loss": 1.6456, + "step": 925 + }, + { + "epoch": 5.0, + "grad_norm": 38.34540557861328, + "learning_rate": 1.6672514619883043e-05, + "loss": 1.6203, + "step": 950 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7961019490254873, + "eval_loss": 0.4937768876552582, + "eval_macro_f1": 0.7941920894743498, + "eval_runtime": 1.1091, + "eval_samples_per_second": 601.407, + "eval_steps_per_second": 37.87, + "eval_weighted_f1": 0.7946279635214719, + "step": 950 + }, + { + "epoch": 5.13175230566535, + "grad_norm": 74.3052749633789, + "learning_rate": 1.6526315789473686e-05, + "loss": 1.1875, + "step": 975 + }, + { + "epoch": 5.263504611330698, + "grad_norm": 71.36292266845703, + "learning_rate": 1.638011695906433e-05, + "loss": 1.2871, + "step": 1000 + }, + { + "epoch": 5.395256916996048, + "grad_norm": 49.50506591796875, + "learning_rate": 1.6233918128654974e-05, + "loss": 1.2786, + "step": 1025 + }, + { + "epoch": 5.5270092226613965, + "grad_norm": 98.29569244384766, + "learning_rate": 1.6087719298245614e-05, + "loss": 1.3583, + "step": 1050 + }, + { + "epoch": 5.658761528326746, + "grad_norm": 91.32157135009766, + "learning_rate": 1.5941520467836257e-05, + "loss": 1.0817, + "step": 1075 + }, + { + "epoch": 5.790513833992095, + "grad_norm": 59.23225402832031, + "learning_rate": 1.57953216374269e-05, + "loss": 1.1753, + "step": 1100 + }, + { + "epoch": 5.922266139657444, + "grad_norm": 97.48385620117188, + "learning_rate": 1.5649122807017545e-05, + "loss": 1.107, + "step": 1125 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8245877061469266, + "eval_loss": 0.5373488664627075, + "eval_macro_f1": 0.8243747841681272, + "eval_runtime": 1.1502, + "eval_samples_per_second": 579.904, + "eval_steps_per_second": 36.516, + "eval_weighted_f1": 0.8249514474109183, + "step": 1140 + } + ], + "logging_steps": 25, + "max_steps": 3800, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 0 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/special_tokens_map.json b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..d5698132694f4f1bcff08fa7d937b1701812598e --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/special_tokens_map.json @@ -0,0 +1,15 @@ +{ + "bos_token": "", + "cls_token": "", + "eos_token": "", + "mask_token": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "sep_token": "", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/tokenizer_config.json b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c2a974a0731aa55a21137e33205c0d71e03d4e24 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/tokenizer_config.json @@ -0,0 +1,55 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "250001": { + "content": "", + "lstrip": true, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "cls_token": "", + "eos_token": "", + "extra_special_tokens": {}, + "mask_token": "", + "model_max_length": 512, + "pad_token": "", + "sep_token": "", + "tokenizer_class": "XLMRobertaTokenizer", + "unk_token": "" +} diff --git a/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/trainer_state.json b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..3b395a33fe8d4b72fd5e71d010bc7b2cf611f117 --- /dev/null +++ b/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1710/trainer_state.json @@ -0,0 +1,618 @@ +{ + "best_global_step": 1140, + "best_metric": 0.8243747841681272, + "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-xlmr-source-source-08-s42-20260901T163859Z-27bda108/trainer/checkpoint-1140", + "epoch": 9.0, + "eval_steps": 500, + "global_step": 1710, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.13175230566534915, + "grad_norm": 30.87647819519043, + "learning_rate": 1.2631578947368422e-06, + "loss": 4.8055, + "step": 25 + }, + { + "epoch": 0.2635046113306983, + "grad_norm": 55.19264221191406, + "learning_rate": 2.578947368421053e-06, + "loss": 4.8098, + "step": 50 + }, + { + "epoch": 0.3952569169960474, + "grad_norm": 29.301651000976562, + "learning_rate": 3.894736842105263e-06, + "loss": 4.6563, + "step": 75 + }, + { + "epoch": 0.5270092226613966, + "grad_norm": 61.64167785644531, + "learning_rate": 5.210526315789474e-06, + "loss": 4.6241, + "step": 100 + }, + { + "epoch": 0.6587615283267457, + "grad_norm": 45.149295806884766, + "learning_rate": 6.526315789473685e-06, + "loss": 4.5889, + "step": 125 + }, + { + "epoch": 0.7905138339920948, + "grad_norm": 55.359588623046875, + "learning_rate": 7.842105263157895e-06, + "loss": 4.5611, + "step": 150 + }, + { + "epoch": 0.922266139657444, + "grad_norm": 103.41445922851562, + "learning_rate": 9.157894736842105e-06, + "loss": 4.336, + "step": 175 + }, + { + "epoch": 1.0, + "eval_accuracy": 0.5502248875562219, + "eval_loss": 0.9538599848747253, + "eval_macro_f1": 0.48829869022613986, + "eval_runtime": 1.1679, + "eval_samples_per_second": 571.095, + "eval_steps_per_second": 35.961, + "eval_weighted_f1": 0.4909424318804821, + "step": 190 + }, + { + "epoch": 1.0527009222661396, + "grad_norm": 64.43714141845703, + "learning_rate": 1.0473684210526316e-05, + "loss": 4.1692, + "step": 200 + }, + { + "epoch": 1.1844532279314888, + "grad_norm": 67.21669006347656, + "learning_rate": 1.1789473684210527e-05, + "loss": 4.0247, + "step": 225 + }, + { + "epoch": 1.316205533596838, + "grad_norm": 23.2484073638916, + "learning_rate": 1.310526315789474e-05, + "loss": 3.9293, + "step": 250 + }, + { + "epoch": 1.447957839262187, + "grad_norm": 52.543296813964844, + "learning_rate": 1.4421052631578948e-05, + "loss": 3.6839, + "step": 275 + }, + { + "epoch": 1.5797101449275361, + "grad_norm": 58.90916442871094, + "learning_rate": 1.573684210526316e-05, + "loss": 3.503, + "step": 300 + }, + { + "epoch": 1.7114624505928853, + "grad_norm": 200.71676635742188, + "learning_rate": 1.705263157894737e-05, + "loss": 3.2837, + "step": 325 + }, + { + "epoch": 1.8432147562582344, + "grad_norm": 117.61235809326172, + "learning_rate": 1.836842105263158e-05, + "loss": 3.3762, + "step": 350 + }, + { + "epoch": 1.9749670619235835, + "grad_norm": 141.29888916015625, + "learning_rate": 1.968421052631579e-05, + "loss": 3.7328, + "step": 375 + }, + { + "epoch": 2.0, + "eval_accuracy": 0.6806596701649176, + "eval_loss": 0.6895695328712463, + "eval_macro_f1": 0.6716922851036303, + "eval_runtime": 1.1188, + "eval_samples_per_second": 596.152, + "eval_steps_per_second": 37.539, + "eval_weighted_f1": 0.6728461310490084, + "step": 380 + }, + { + "epoch": 2.1054018445322793, + "grad_norm": 103.40943908691406, + "learning_rate": 1.988888888888889e-05, + "loss": 3.131, + "step": 400 + }, + { + "epoch": 2.2371541501976284, + "grad_norm": 50.52198791503906, + "learning_rate": 1.9742690058479533e-05, + "loss": 2.9059, + "step": 425 + }, + { + "epoch": 2.3689064558629775, + "grad_norm": 64.69234466552734, + "learning_rate": 1.9596491228070177e-05, + "loss": 2.9769, + "step": 450 + }, + { + "epoch": 2.5006587615283267, + "grad_norm": 116.13067626953125, + "learning_rate": 1.945029239766082e-05, + "loss": 2.8277, + "step": 475 + }, + { + "epoch": 2.632411067193676, + "grad_norm": 41.39086151123047, + "learning_rate": 1.9304093567251464e-05, + "loss": 2.8187, + "step": 500 + }, + { + "epoch": 2.764163372859025, + "grad_norm": 106.1263427734375, + "learning_rate": 1.9157894736842108e-05, + "loss": 2.4984, + "step": 525 + }, + { + "epoch": 2.895915678524374, + "grad_norm": 129.88247680664062, + "learning_rate": 1.901169590643275e-05, + "loss": 2.9183, + "step": 550 + }, + { + "epoch": 3.0, + "eval_accuracy": 0.7931034482758621, + "eval_loss": 0.5418083071708679, + "eval_macro_f1": 0.792407856503026, + "eval_runtime": 1.1477, + "eval_samples_per_second": 581.172, + "eval_steps_per_second": 36.596, + "eval_weighted_f1": 0.7926835166673071, + "step": 570 + }, + { + "epoch": 3.02635046113307, + "grad_norm": 93.23123931884766, + "learning_rate": 1.8865497076023395e-05, + "loss": 2.2853, + "step": 575 + }, + { + "epoch": 3.158102766798419, + "grad_norm": 76.33997344970703, + "learning_rate": 1.871929824561404e-05, + "loss": 2.2193, + "step": 600 + }, + { + "epoch": 3.289855072463768, + "grad_norm": 62.839385986328125, + "learning_rate": 1.857309941520468e-05, + "loss": 2.2484, + "step": 625 + }, + { + "epoch": 3.421607378129117, + "grad_norm": 82.7411880493164, + "learning_rate": 1.8426900584795323e-05, + "loss": 2.0458, + "step": 650 + }, + { + "epoch": 3.5533596837944663, + "grad_norm": 56.64774703979492, + "learning_rate": 1.8280701754385966e-05, + "loss": 2.2254, + "step": 675 + }, + { + "epoch": 3.6851119894598154, + "grad_norm": 47.41438293457031, + "learning_rate": 1.813450292397661e-05, + "loss": 1.9588, + "step": 700 + }, + { + "epoch": 3.8168642951251646, + "grad_norm": 48.75365447998047, + "learning_rate": 1.7988304093567253e-05, + "loss": 2.0426, + "step": 725 + }, + { + "epoch": 3.9486166007905137, + "grad_norm": 72.154296875, + "learning_rate": 1.7842105263157897e-05, + "loss": 1.9697, + "step": 750 + }, + { + "epoch": 4.0, + "eval_accuracy": 0.7796101949025487, + "eval_loss": 0.528952419757843, + "eval_macro_f1": 0.7836733289235748, + "eval_runtime": 1.2131, + "eval_samples_per_second": 549.833, + "eval_steps_per_second": 34.622, + "eval_weighted_f1": 0.7837899739423578, + "step": 760 + }, + { + "epoch": 4.07905138339921, + "grad_norm": 64.87993621826172, + "learning_rate": 1.769590643274854e-05, + "loss": 1.766, + "step": 775 + }, + { + "epoch": 4.210803689064559, + "grad_norm": 147.2518310546875, + "learning_rate": 1.7549707602339184e-05, + "loss": 1.7144, + "step": 800 + }, + { + "epoch": 4.342555994729908, + "grad_norm": 140.36033630371094, + "learning_rate": 1.7403508771929828e-05, + "loss": 1.5639, + "step": 825 + }, + { + "epoch": 4.474308300395257, + "grad_norm": 126.15364837646484, + "learning_rate": 1.725730994152047e-05, + "loss": 1.595, + "step": 850 + }, + { + "epoch": 4.606060606060606, + "grad_norm": 95.8467788696289, + "learning_rate": 1.7111111111111112e-05, + "loss": 1.6152, + "step": 875 + }, + { + "epoch": 4.737812911725955, + "grad_norm": 63.30301284790039, + "learning_rate": 1.6964912280701755e-05, + "loss": 1.5819, + "step": 900 + }, + { + "epoch": 4.869565217391305, + "grad_norm": 103.65847778320312, + "learning_rate": 1.68187134502924e-05, + "loss": 1.6456, + "step": 925 + }, + { + "epoch": 5.0, + "grad_norm": 38.34540557861328, + "learning_rate": 1.6672514619883043e-05, + "loss": 1.6203, + "step": 950 + }, + { + "epoch": 5.0, + "eval_accuracy": 0.7961019490254873, + "eval_loss": 0.4937768876552582, + "eval_macro_f1": 0.7941920894743498, + "eval_runtime": 1.1091, + "eval_samples_per_second": 601.407, + "eval_steps_per_second": 37.87, + "eval_weighted_f1": 0.7946279635214719, + "step": 950 + }, + { + "epoch": 5.13175230566535, + "grad_norm": 74.3052749633789, + "learning_rate": 1.6526315789473686e-05, + "loss": 1.1875, + "step": 975 + }, + { + "epoch": 5.263504611330698, + "grad_norm": 71.36292266845703, + "learning_rate": 1.638011695906433e-05, + "loss": 1.2871, + "step": 1000 + }, + { + "epoch": 5.395256916996048, + "grad_norm": 49.50506591796875, + "learning_rate": 1.6233918128654974e-05, + "loss": 1.2786, + "step": 1025 + }, + { + "epoch": 5.5270092226613965, + "grad_norm": 98.29569244384766, + "learning_rate": 1.6087719298245614e-05, + "loss": 1.3583, + "step": 1050 + }, + { + "epoch": 5.658761528326746, + "grad_norm": 91.32157135009766, + "learning_rate": 1.5941520467836257e-05, + "loss": 1.0817, + "step": 1075 + }, + { + "epoch": 5.790513833992095, + "grad_norm": 59.23225402832031, + "learning_rate": 1.57953216374269e-05, + "loss": 1.1753, + "step": 1100 + }, + { + "epoch": 5.922266139657444, + "grad_norm": 97.48385620117188, + "learning_rate": 1.5649122807017545e-05, + "loss": 1.107, + "step": 1125 + }, + { + "epoch": 6.0, + "eval_accuracy": 0.8245877061469266, + "eval_loss": 0.5373488664627075, + "eval_macro_f1": 0.8243747841681272, + "eval_runtime": 1.1502, + "eval_samples_per_second": 579.904, + "eval_steps_per_second": 36.516, + "eval_weighted_f1": 0.8249514474109183, + "step": 1140 + }, + { + "epoch": 6.05270092226614, + "grad_norm": 83.98052978515625, + "learning_rate": 1.5502923976608188e-05, + "loss": 1.0529, + "step": 1150 + }, + { + "epoch": 6.184453227931489, + "grad_norm": 79.76677703857422, + "learning_rate": 1.5356725146198832e-05, + "loss": 0.9139, + "step": 1175 + }, + { + "epoch": 6.316205533596838, + "grad_norm": 112.12026977539062, + "learning_rate": 1.5210526315789476e-05, + "loss": 0.9433, + "step": 1200 + }, + { + "epoch": 6.4479578392621875, + "grad_norm": 114.89282989501953, + "learning_rate": 1.5064327485380119e-05, + "loss": 1.0333, + "step": 1225 + }, + { + "epoch": 6.579710144927536, + "grad_norm": 38.201499938964844, + "learning_rate": 1.4918128654970761e-05, + "loss": 0.9203, + "step": 1250 + }, + { + "epoch": 6.711462450592886, + "grad_norm": 196.24383544921875, + "learning_rate": 1.4771929824561405e-05, + "loss": 0.8701, + "step": 1275 + }, + { + "epoch": 6.843214756258234, + "grad_norm": 52.879310607910156, + "learning_rate": 1.4625730994152048e-05, + "loss": 1.1404, + "step": 1300 + }, + { + "epoch": 6.974967061923584, + "grad_norm": 65.20995330810547, + "learning_rate": 1.447953216374269e-05, + "loss": 1.0777, + "step": 1325 + }, + { + "epoch": 7.0, + "eval_accuracy": 0.8170914542728636, + "eval_loss": 0.5831647515296936, + "eval_macro_f1": 0.8163885906908449, + "eval_runtime": 1.1443, + "eval_samples_per_second": 582.905, + "eval_steps_per_second": 36.705, + "eval_weighted_f1": 0.8164616320980466, + "step": 1330 + }, + { + "epoch": 7.105401844532279, + "grad_norm": 30.025270462036133, + "learning_rate": 1.4333333333333334e-05, + "loss": 0.6845, + "step": 1350 + }, + { + "epoch": 7.237154150197629, + "grad_norm": 32.1573371887207, + "learning_rate": 1.4187134502923978e-05, + "loss": 0.7555, + "step": 1375 + }, + { + "epoch": 7.3689064558629775, + "grad_norm": 92.26585388183594, + "learning_rate": 1.4040935672514621e-05, + "loss": 0.7868, + "step": 1400 + }, + { + "epoch": 7.500658761528327, + "grad_norm": 23.797170639038086, + "learning_rate": 1.3894736842105265e-05, + "loss": 0.5832, + "step": 1425 + }, + { + "epoch": 7.632411067193676, + "grad_norm": 270.5168151855469, + "learning_rate": 1.3748538011695907e-05, + "loss": 0.7426, + "step": 1450 + }, + { + "epoch": 7.764163372859025, + "grad_norm": 88.94017028808594, + "learning_rate": 1.360233918128655e-05, + "loss": 0.8071, + "step": 1475 + }, + { + "epoch": 7.895915678524374, + "grad_norm": 28.964038848876953, + "learning_rate": 1.3456140350877194e-05, + "loss": 0.5568, + "step": 1500 + }, + { + "epoch": 8.0, + "eval_accuracy": 0.8020989505247377, + "eval_loss": 0.7253847122192383, + "eval_macro_f1": 0.8028305467773208, + "eval_runtime": 1.1681, + "eval_samples_per_second": 570.997, + "eval_steps_per_second": 35.955, + "eval_weighted_f1": 0.8031750241974119, + "step": 1520 + }, + { + "epoch": 8.02635046113307, + "grad_norm": 36.436790466308594, + "learning_rate": 1.3309941520467838e-05, + "loss": 0.8615, + "step": 1525 + }, + { + "epoch": 8.15810276679842, + "grad_norm": 83.15040588378906, + "learning_rate": 1.3163742690058481e-05, + "loss": 0.5726, + "step": 1550 + }, + { + "epoch": 8.289855072463768, + "grad_norm": 144.29547119140625, + "learning_rate": 1.3017543859649125e-05, + "loss": 0.6425, + "step": 1575 + }, + { + "epoch": 8.421607378129117, + "grad_norm": 17.80103302001953, + "learning_rate": 1.2871345029239768e-05, + "loss": 0.6229, + "step": 1600 + }, + { + "epoch": 8.553359683794467, + "grad_norm": 131.2693634033203, + "learning_rate": 1.2725146198830412e-05, + "loss": 0.5472, + "step": 1625 + }, + { + "epoch": 8.685111989459816, + "grad_norm": 57.37495803833008, + "learning_rate": 1.2578947368421052e-05, + "loss": 0.4891, + "step": 1650 + }, + { + "epoch": 8.816864295125164, + "grad_norm": 131.08395385742188, + "learning_rate": 1.2432748538011696e-05, + "loss": 0.6643, + "step": 1675 + }, + { + "epoch": 8.948616600790514, + "grad_norm": 21.23287582397461, + "learning_rate": 1.228654970760234e-05, + "loss": 0.5521, + "step": 1700 + }, + { + "epoch": 9.0, + "eval_accuracy": 0.8050974512743628, + "eval_loss": 0.8861004114151001, + "eval_macro_f1": 0.8076477042648605, + "eval_runtime": 1.1766, + "eval_samples_per_second": 566.887, + "eval_steps_per_second": 35.696, + "eval_weighted_f1": 0.8079679233142613, + "step": 1710 + } + ], + "logging_steps": 25, + "max_steps": 3800, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 500, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 3, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 3 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}