| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 20400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24509803921568626, | |
| "grad_norm": 2.2194137573242188, | |
| "learning_rate": 4.901960784313726e-06, | |
| "loss": 0.6351, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.49019607843137253, | |
| "grad_norm": 5.901831150054932, | |
| "learning_rate": 9.803921568627451e-06, | |
| "loss": 0.4992, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7352941176470589, | |
| "grad_norm": 5.434150695800781, | |
| "learning_rate": 1.4705882352941179e-05, | |
| "loss": 0.3933, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9803921568627451, | |
| "grad_norm": 6.714095115661621, | |
| "learning_rate": 1.9607843137254903e-05, | |
| "loss": 0.3433, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.795555184225917, | |
| "eval_f1": 0.758320987654321, | |
| "eval_precision": 0.9295399515738498, | |
| "eval_recall": 0.6403669724770642, | |
| "eval_runtime": 50.1487, | |
| "eval_samples_per_second": 238.67, | |
| "eval_steps_per_second": 3.749, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.2254901960784315, | |
| "grad_norm": 4.650435924530029, | |
| "learning_rate": 1.9501089324618738e-05, | |
| "loss": 0.2707, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 5.975987434387207, | |
| "learning_rate": 1.8956427015250548e-05, | |
| "loss": 0.2568, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.715686274509804, | |
| "grad_norm": 4.295902729034424, | |
| "learning_rate": 1.8413943355119828e-05, | |
| "loss": 0.2396, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.9607843137254903, | |
| "grad_norm": 4.352377414703369, | |
| "learning_rate": 1.7869281045751637e-05, | |
| "loss": 0.2253, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.821873172361935, | |
| "eval_f1": 0.7964094728800611, | |
| "eval_precision": 0.9314272950636587, | |
| "eval_recall": 0.69557964970809, | |
| "eval_runtime": 50.0719, | |
| "eval_samples_per_second": 239.036, | |
| "eval_steps_per_second": 3.755, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 2.2058823529411766, | |
| "grad_norm": 8.222220420837402, | |
| "learning_rate": 1.7324618736383444e-05, | |
| "loss": 0.1567, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.450980392156863, | |
| "grad_norm": 8.841856956481934, | |
| "learning_rate": 1.677995642701525e-05, | |
| "loss": 0.1406, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.696078431372549, | |
| "grad_norm": 5.413289546966553, | |
| "learning_rate": 1.6236383442265798e-05, | |
| "loss": 0.1379, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 6.019312381744385, | |
| "learning_rate": 1.5691721132897604e-05, | |
| "loss": 0.1406, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8236277049043362, | |
| "eval_f1": 0.7998103366524419, | |
| "eval_precision": 0.9268131868131868, | |
| "eval_recall": 0.7034195162635529, | |
| "eval_runtime": 50.1875, | |
| "eval_samples_per_second": 238.486, | |
| "eval_steps_per_second": 3.746, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 3.186274509803922, | |
| "grad_norm": 4.456413269042969, | |
| "learning_rate": 1.5148148148148149e-05, | |
| "loss": 0.0954, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.431372549019608, | |
| "grad_norm": 3.628704786300659, | |
| "learning_rate": 1.4603485838779959e-05, | |
| "loss": 0.0892, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.6764705882352944, | |
| "grad_norm": 2.9508347511291504, | |
| "learning_rate": 1.4058823529411765e-05, | |
| "loss": 0.0919, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.9215686274509802, | |
| "grad_norm": 6.59318208694458, | |
| "learning_rate": 1.3514161220043573e-05, | |
| "loss": 0.0873, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8282229091820537, | |
| "eval_f1": 0.808351976137211, | |
| "eval_precision": 0.9161208535812381, | |
| "eval_recall": 0.7232693911592994, | |
| "eval_runtime": 50.5224, | |
| "eval_samples_per_second": 236.905, | |
| "eval_steps_per_second": 3.721, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "grad_norm": 2.5619020462036133, | |
| "learning_rate": 1.2969498910675382e-05, | |
| "loss": 0.0696, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.411764705882353, | |
| "grad_norm": 6.823647499084473, | |
| "learning_rate": 1.242483660130719e-05, | |
| "loss": 0.0564, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.6568627450980395, | |
| "grad_norm": 7.367131233215332, | |
| "learning_rate": 1.1880174291939e-05, | |
| "loss": 0.0608, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.901960784313726, | |
| "grad_norm": 1.920448660850525, | |
| "learning_rate": 1.1335511982570806e-05, | |
| "loss": 0.0642, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8349903918456011, | |
| "eval_f1": 0.8161593595829842, | |
| "eval_precision": 0.9233361415332771, | |
| "eval_recall": 0.7312760633861551, | |
| "eval_runtime": 50.2334, | |
| "eval_samples_per_second": 238.268, | |
| "eval_steps_per_second": 3.743, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 5.147058823529412, | |
| "grad_norm": 0.24896511435508728, | |
| "learning_rate": 1.0790849673202614e-05, | |
| "loss": 0.0484, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.392156862745098, | |
| "grad_norm": 0.1454085409641266, | |
| "learning_rate": 1.0248366013071896e-05, | |
| "loss": 0.0428, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.637254901960784, | |
| "grad_norm": 0.20693035423755646, | |
| "learning_rate": 9.703703703703703e-06, | |
| "loss": 0.0422, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 7.705965518951416, | |
| "learning_rate": 9.159041394335513e-06, | |
| "loss": 0.0439, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.82078703316902, | |
| "eval_f1": 0.7939283312518013, | |
| "eval_precision": 0.9361123697326688, | |
| "eval_recall": 0.6892410341951626, | |
| "eval_runtime": 50.1119, | |
| "eval_samples_per_second": 238.846, | |
| "eval_steps_per_second": 3.752, | |
| "step": 12240 | |
| }, | |
| { | |
| "epoch": 6.127450980392156, | |
| "grad_norm": 3.6634812355041504, | |
| "learning_rate": 8.614379084967321e-06, | |
| "loss": 0.0368, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.372549019607844, | |
| "grad_norm": 0.9099088311195374, | |
| "learning_rate": 8.069716775599129e-06, | |
| "loss": 0.0299, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.617647058823529, | |
| "grad_norm": 4.98677921295166, | |
| "learning_rate": 7.5250544662309376e-06, | |
| "loss": 0.0305, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 6.862745098039216, | |
| "grad_norm": 1.047827124595642, | |
| "learning_rate": 6.9803921568627454e-06, | |
| "loss": 0.0322, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8311471300860557, | |
| "eval_f1": 0.8085630387420669, | |
| "eval_precision": 0.9355545813239807, | |
| "eval_recall": 0.7119266055045872, | |
| "eval_runtime": 50.2076, | |
| "eval_samples_per_second": 238.39, | |
| "eval_steps_per_second": 3.744, | |
| "step": 14280 | |
| }, | |
| { | |
| "epoch": 7.107843137254902, | |
| "grad_norm": 4.889837741851807, | |
| "learning_rate": 6.435729847494554e-06, | |
| "loss": 0.0278, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 7.352941176470588, | |
| "grad_norm": 0.027428582310676575, | |
| "learning_rate": 5.891067538126363e-06, | |
| "loss": 0.0242, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.598039215686274, | |
| "grad_norm": 15.176671981811523, | |
| "learning_rate": 5.34640522875817e-06, | |
| "loss": 0.0216, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 7.8431372549019605, | |
| "grad_norm": 0.10369198769330978, | |
| "learning_rate": 4.801742919389979e-06, | |
| "loss": 0.0241, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8279722616759964, | |
| "eval_f1": 0.8052586777641162, | |
| "eval_precision": 0.9298820445609436, | |
| "eval_recall": 0.710091743119266, | |
| "eval_runtime": 50.123, | |
| "eval_samples_per_second": 238.792, | |
| "eval_steps_per_second": 3.751, | |
| "step": 16320 | |
| }, | |
| { | |
| "epoch": 8.088235294117647, | |
| "grad_norm": 6.414863109588623, | |
| "learning_rate": 4.2570806100217874e-06, | |
| "loss": 0.0223, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 7.71334981918335, | |
| "learning_rate": 3.7135076252723314e-06, | |
| "loss": 0.0161, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 8.57843137254902, | |
| "grad_norm": 6.453121662139893, | |
| "learning_rate": 3.1688453159041397e-06, | |
| "loss": 0.0191, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 8.823529411764707, | |
| "grad_norm": 17.10112762451172, | |
| "learning_rate": 2.624183006535948e-06, | |
| "loss": 0.0182, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8256328849527947, | |
| "eval_f1": 0.8022738038844149, | |
| "eval_precision": 0.9285087719298246, | |
| "eval_recall": 0.706255212677231, | |
| "eval_runtime": 50.0398, | |
| "eval_samples_per_second": 239.19, | |
| "eval_steps_per_second": 3.757, | |
| "step": 18360 | |
| }, | |
| { | |
| "epoch": 9.068627450980392, | |
| "grad_norm": 0.027658773586153984, | |
| "learning_rate": 2.0795206971677563e-06, | |
| "loss": 0.0179, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 9.313725490196079, | |
| "grad_norm": 0.012034358456730843, | |
| "learning_rate": 1.5348583877995642e-06, | |
| "loss": 0.0142, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 9.558823529411764, | |
| "grad_norm": 0.053181909024715424, | |
| "learning_rate": 9.923747276688455e-07, | |
| "loss": 0.0117, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 9.803921568627452, | |
| "grad_norm": 0.0060155196115374565, | |
| "learning_rate": 4.477124183006536e-07, | |
| "loss": 0.0125, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.831397777592113, | |
| "eval_f1": 0.8103383458646617, | |
| "eval_precision": 0.9280947255113025, | |
| "eval_recall": 0.7190992493744788, | |
| "eval_runtime": 50.2235, | |
| "eval_samples_per_second": 238.315, | |
| "eval_steps_per_second": 3.743, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 20400, | |
| "total_flos": 3.433678205764608e+17, | |
| "train_loss": 0.1127255682851754, | |
| "train_runtime": 14662.5965, | |
| "train_samples_per_second": 89.004, | |
| "train_steps_per_second": 1.391 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 20400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.433678205764608e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |