| { | |
| "best_global_step": 400, | |
| "best_metric": 0.7742971777915955, | |
| "best_model_checkpoint": "./tinyllama-custom-quotes-fine-tuned/checkpoint-400", | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 0.7440543174743652, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 2.9746, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.447368860244751, | |
| "learning_rate": 0.0001998394219742627, | |
| "loss": 1.4501, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 0.5628308653831482, | |
| "learning_rate": 0.00019905415858090036, | |
| "loss": 1.0079, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.5305753946304321, | |
| "learning_rate": 0.0001976198547939518, | |
| "loss": 0.8519, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.7008503079414368, | |
| "learning_rate": 0.0001955459087155033, | |
| "loss": 0.798, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.3367324471473694, | |
| "learning_rate": 0.00019284590962639176, | |
| "loss": 0.9812, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 0.41013872623443604, | |
| "learning_rate": 0.00018953754894408616, | |
| "loss": 0.7821, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.3676799237728119, | |
| "learning_rate": 0.00018564250430181387, | |
| "loss": 0.8898, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 0.621743381023407, | |
| "learning_rate": 0.00018118629750849105, | |
| "loss": 0.8397, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.37409332394599915, | |
| "learning_rate": 0.00017619812732015664, | |
| "loss": 0.5815, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.8038855791091919, | |
| "eval_runtime": 48.3297, | |
| "eval_samples_per_second": 4.138, | |
| "eval_steps_per_second": 0.517, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 0.8709319829940796, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.8146, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.43820664286613464, | |
| "learning_rate": 0.00016475990575117605, | |
| "loss": 0.6003, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 0.4444628059864044, | |
| "learning_rate": 0.00015838480193391754, | |
| "loss": 0.7894, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.41406378149986267, | |
| "learning_rate": 0.00015162713876357858, | |
| "loss": 0.7565, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.5294691920280457, | |
| "learning_rate": 0.00014453119501074924, | |
| "loss": 0.8514, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.37928926944732666, | |
| "learning_rate": 0.00013714346598862166, | |
| "loss": 0.6292, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 0.4634771943092346, | |
| "learning_rate": 0.00012951235889807386, | |
| "loss": 0.9964, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.35241034626960754, | |
| "learning_rate": 0.00012168787564534078, | |
| "loss": 0.8125, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 1.2624390125274658, | |
| "learning_rate": 0.00011372128521057155, | |
| "loss": 0.9791, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.313948154449463, | |
| "learning_rate": 0.00010566478771403763, | |
| "loss": 0.6942, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.7819027900695801, | |
| "eval_runtime": 48.3939, | |
| "eval_samples_per_second": 4.133, | |
| "eval_steps_per_second": 0.517, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.3572781980037689, | |
| "learning_rate": 9.757117238115871e-05, | |
| "loss": 0.8754, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.4901013970375061, | |
| "learning_rate": 8.949347164748762e-05, | |
| "loss": 0.7883, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.5491872429847717, | |
| "learning_rate": 8.14846136700908e-05, | |
| "loss": 0.618, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.3233886659145355, | |
| "learning_rate": 7.35970755222007e-05, | |
| "loss": 0.9113, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.4254407286643982, | |
| "learning_rate": 6.588253934354039e-05, | |
| "loss": 0.9099, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.45668289065361023, | |
| "learning_rate": 5.839155369935407e-05, | |
| "loss": 0.7206, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.29698246717453003, | |
| "learning_rate": 5.117320236704697e-05, | |
| "loss": 0.7399, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.3198457658290863, | |
| "learning_rate": 4.427478272067066e-05, | |
| "loss": 0.71, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.34839925169944763, | |
| "learning_rate": 3.774149582060012e-05, | |
| "loss": 0.6937, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.362792432308197, | |
| "learning_rate": 3.161615023905265e-05, | |
| "loss": 0.7382, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.7760980129241943, | |
| "eval_runtime": 48.3842, | |
| "eval_samples_per_second": 4.134, | |
| "eval_steps_per_second": 0.517, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.29705917835235596, | |
| "learning_rate": 2.593888156209603e-05, | |
| "loss": 0.8376, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.3633488118648529, | |
| "learning_rate": 2.074688940607529e-05, | |
| "loss": 0.6221, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.3803982138633728, | |
| "learning_rate": 1.607419367162577e-05, | |
| "loss": 0.5971, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.5394597053527832, | |
| "learning_rate": 1.195141163238892e-05, | |
| "loss": 0.7052, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.32639947533607483, | |
| "learning_rate": 8.405557319029912e-06, | |
| "loss": 0.6989, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.2878645360469818, | |
| "learning_rate": 5.45986451306899e-06, | |
| "loss": 0.7779, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.5121065378189087, | |
| "learning_rate": 3.1336345103382346e-06, | |
| "loss": 0.6928, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.3159574270248413, | |
| "learning_rate": 1.4421096515749855e-06, | |
| "loss": 0.7191, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.5103066563606262, | |
| "learning_rate": 3.963734488278248e-07, | |
| "loss": 1.0234, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.32779765129089355, | |
| "learning_rate": 3.2779620843692572e-09, | |
| "loss": 0.7682, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.7742971777915955, | |
| "eval_runtime": 48.4329, | |
| "eval_samples_per_second": 4.129, | |
| "eval_steps_per_second": 0.516, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1333217407795200.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |