| { | |
| "best_metric": 53.324968632371395, | |
| "best_model_checkpoint": "/scratch/camembertv2/runs/results/fquad/camembertv2-base-bf16-p2-17000/max_seq_length-896-doc_stride-128-max_answer_length-30-gradient_accumulation_steps-4-precision-fp32-learning_rate-5e-06-epochs-6-lr_scheduler-cosine-warmup_steps-0/SEED-25/checkpoint-3888", | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 3888, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15432098765432098, | |
| "grad_norm": 4.547507286071777, | |
| "learning_rate": 4.99184317884152e-06, | |
| "loss": 5.1604, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.30864197530864196, | |
| "grad_norm": 12.684767723083496, | |
| "learning_rate": 4.967425942351207e-06, | |
| "loss": 4.0839, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.46296296296296297, | |
| "grad_norm": 14.742673873901367, | |
| "learning_rate": 4.926907624154051e-06, | |
| "loss": 3.3159, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 12.705907821655273, | |
| "learning_rate": 4.870552624790192e-06, | |
| "loss": 2.9494, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7716049382716049, | |
| "grad_norm": 14.801329612731934, | |
| "learning_rate": 4.798728686380588e-06, | |
| "loss": 2.7635, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9259259259259259, | |
| "grad_norm": 17.233285903930664, | |
| "learning_rate": 4.711904492941644e-06, | |
| "loss": 2.6393, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 38.86449184441656, | |
| "eval_f1": 60.0036086905889, | |
| "eval_runtime": 6.9307, | |
| "eval_samples_per_second": 459.985, | |
| "eval_steps_per_second": 7.214, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.0802469135802468, | |
| "grad_norm": 14.701664924621582, | |
| "learning_rate": 4.610646612007849e-06, | |
| "loss": 2.4089, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2345679012345678, | |
| "grad_norm": 17.278104782104492, | |
| "learning_rate": 4.495615797519732e-06, | |
| "loss": 2.3405, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 11.50146770477295, | |
| "learning_rate": 4.367562678102491e-06, | |
| "loss": 2.2084, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.5432098765432098, | |
| "grad_norm": 13.203764915466309, | |
| "learning_rate": 4.22732285887122e-06, | |
| "loss": 2.1694, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6975308641975309, | |
| "grad_norm": 21.71219825744629, | |
| "learning_rate": 4.075811468725734e-06, | |
| "loss": 2.0862, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8518518518518519, | |
| "grad_norm": 12.909610748291016, | |
| "learning_rate": 3.914017188716347e-06, | |
| "loss": 2.0016, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 48.745294855708906, | |
| "eval_f1": 70.05708349304844, | |
| "eval_runtime": 6.5382, | |
| "eval_samples_per_second": 487.593, | |
| "eval_steps_per_second": 7.647, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 2.006172839506173, | |
| "grad_norm": 16.666929244995117, | |
| "learning_rate": 3.7429958004482575e-06, | |
| "loss": 1.9412, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.1604938271604937, | |
| "grad_norm": 10.462796211242676, | |
| "learning_rate": 3.5638632966241686e-06, | |
| "loss": 1.8009, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.314814814814815, | |
| "grad_norm": 13.769060134887695, | |
| "learning_rate": 3.3777885986819725e-06, | |
| "loss": 1.7928, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.4691358024691357, | |
| "grad_norm": 15.287083625793457, | |
| "learning_rate": 3.1859859290482544e-06, | |
| "loss": 1.7865, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.623456790123457, | |
| "grad_norm": 11.451101303100586, | |
| "learning_rate": 2.989706887782151e-06, | |
| "loss": 1.7489, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 17.512975692749023, | |
| "learning_rate": 2.7902322853130758e-06, | |
| "loss": 1.6978, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.932098765432099, | |
| "grad_norm": 17.15248680114746, | |
| "learning_rate": 2.5888637845674276e-06, | |
| "loss": 1.6566, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 50.47051442910916, | |
| "eval_f1": 72.25048266378954, | |
| "eval_runtime": 6.5447, | |
| "eval_samples_per_second": 487.112, | |
| "eval_steps_per_second": 7.64, | |
| "step": 1944 | |
| }, | |
| { | |
| "epoch": 3.0864197530864197, | |
| "grad_norm": 11.384383201599121, | |
| "learning_rate": 2.3869154070232346e-06, | |
| "loss": 1.6309, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.240740740740741, | |
| "grad_norm": 14.492201805114746, | |
| "learning_rate": 2.185704958119594e-06, | |
| "loss": 1.5353, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.3950617283950617, | |
| "grad_norm": 15.613585472106934, | |
| "learning_rate": 1.9865454279740452e-06, | |
| "loss": 1.5249, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.549382716049383, | |
| "grad_norm": 12.233988761901855, | |
| "learning_rate": 1.7907364235221128e-06, | |
| "loss": 1.5499, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.7037037037037037, | |
| "grad_norm": 11.811338424682617, | |
| "learning_rate": 1.5995556879882246e-06, | |
| "loss": 1.5159, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.8580246913580245, | |
| "grad_norm": 18.379695892333984, | |
| "learning_rate": 1.414250763027336e-06, | |
| "loss": 1.5072, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 53.01129234629862, | |
| "eval_f1": 74.3205610049545, | |
| "eval_runtime": 6.5679, | |
| "eval_samples_per_second": 485.39, | |
| "eval_steps_per_second": 7.613, | |
| "step": 2592 | |
| }, | |
| { | |
| "epoch": 4.012345679012346, | |
| "grad_norm": 12.669611930847168, | |
| "learning_rate": 1.2360308479456027e-06, | |
| "loss": 1.5257, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "grad_norm": 13.753548622131348, | |
| "learning_rate": 1.0660589091223854e-06, | |
| "loss": 1.4296, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.320987654320987, | |
| "grad_norm": 10.774425506591797, | |
| "learning_rate": 9.054440911232348e-07, | |
| "loss": 1.4796, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.4753086419753085, | |
| "grad_norm": 16.21649742126465, | |
| "learning_rate": 7.552344790248104e-07, | |
| "loss": 1.426, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.62962962962963, | |
| "grad_norm": 11.01417064666748, | |
| "learning_rate": 6.164102591808482e-07, | |
| "loss": 1.4245, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.783950617283951, | |
| "grad_norm": 10.40230941772461, | |
| "learning_rate": 4.898773230583353e-07, | |
| "loss": 1.4493, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 10.953381538391113, | |
| "learning_rate": 3.7646135588175676e-07, | |
| "loss": 1.404, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 53.168130489335006, | |
| "eval_f1": 74.39491719320372, | |
| "eval_runtime": 6.6406, | |
| "eval_samples_per_second": 480.08, | |
| "eval_steps_per_second": 7.529, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 5.092592592592593, | |
| "grad_norm": 13.173111915588379, | |
| "learning_rate": 2.7690244865973494e-07, | |
| "loss": 1.43, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.246913580246914, | |
| "grad_norm": 13.998867988586426, | |
| "learning_rate": 1.918502687530241e-07, | |
| "loss": 1.3968, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.401234567901234, | |
| "grad_norm": 12.186470985412598, | |
| "learning_rate": 1.2185982049813472e-07, | |
| "loss": 1.378, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 16.22747039794922, | |
| "learning_rate": 6.738782355044048e-08, | |
| "loss": 1.4347, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.709876543209877, | |
| "grad_norm": 17.75710105895996, | |
| "learning_rate": 2.878973257973955e-08, | |
| "loss": 1.422, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.864197530864198, | |
| "grad_norm": 17.476356506347656, | |
| "learning_rate": 6.317417766116829e-09, | |
| "loss": 1.3868, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 53.324968632371395, | |
| "eval_f1": 74.54839090269344, | |
| "eval_runtime": 6.5292, | |
| "eval_samples_per_second": 488.268, | |
| "eval_steps_per_second": 7.658, | |
| "step": 3888 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 3888, | |
| "total_flos": 2.0387348740618656e+16, | |
| "train_loss": 1.9457146935011624, | |
| "train_runtime": 824.1497, | |
| "train_samples_per_second": 150.926, | |
| "train_steps_per_second": 4.718 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 3888, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.0387348740618656e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |