| { | |
| "best_metric": 0.40061458945274353, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_60-samples_config-2_auto/checkpoint-37", | |
| "epoch": 19.82608695652174, | |
| "eval_steps": 500, | |
| "global_step": 57, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 6.809540271759033, | |
| "learning_rate": 1e-05, | |
| "loss": 3.3731, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 6.12690544128418, | |
| "learning_rate": 2e-05, | |
| "loss": 3.072, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 3.0875492095947266, | |
| "eval_runtime": 4.4181, | |
| "eval_samples_per_second": 2.716, | |
| "eval_steps_per_second": 2.716, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 6.0957536697387695, | |
| "learning_rate": 4e-05, | |
| "loss": 3.0765, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 2.5634982585906982, | |
| "eval_runtime": 4.1825, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 5.55053186416626, | |
| "learning_rate": 6e-05, | |
| "loss": 2.8666, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 3.83312726020813, | |
| "learning_rate": 8e-05, | |
| "loss": 2.0933, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.4153375625610352, | |
| "eval_runtime": 4.183, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 2.5791916847229004, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2307, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 0.7854078412055969, | |
| "eval_runtime": 4.1806, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 1.9428495168685913, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.8527, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.8739281892776489, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.5656, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 0.5589617490768433, | |
| "eval_runtime": 4.1802, | |
| "eval_samples_per_second": 2.871, | |
| "eval_steps_per_second": 2.871, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.6738723516464233, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.5325, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.4852149486541748, | |
| "eval_runtime": 4.1851, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.4890255033969879, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.3752, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.5037442445755005, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.4027, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.4474470615386963, | |
| "eval_runtime": 4.1788, | |
| "eval_samples_per_second": 2.872, | |
| "eval_steps_per_second": 2.872, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.3417052924633026, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.2861, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.43576323986053467, | |
| "eval_runtime": 4.1816, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.3297211825847626, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.3766, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 0.4267699718475342, | |
| "eval_runtime": 4.1847, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 2.868, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.37240782380104065, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.3307, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.42333266139030457, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.291, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 0.41624876856803894, | |
| "eval_runtime": 4.1819, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.3704397976398468, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.2706, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 0.4064284861087799, | |
| "eval_runtime": 4.1878, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 2.865, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.3553304672241211, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.2492, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.3354627192020416, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.2037, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 0.40273547172546387, | |
| "eval_runtime": 4.1832, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.35223323106765747, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.2057, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 0.40061458945274353, | |
| "eval_runtime": 4.1849, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.45181211829185486, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.1952, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.2943543791770935, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.1402, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 0.4081263542175293, | |
| "eval_runtime": 4.1822, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.4012012779712677, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.1817, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 0.4245516359806061, | |
| "eval_runtime": 4.1805, | |
| "eval_samples_per_second": 2.87, | |
| "eval_steps_per_second": 2.87, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.38791102170944214, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.1171, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.3992724120616913, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.119, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.4513971507549286, | |
| "eval_runtime": 4.1834, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.45827752351760864, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.1067, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 0.4774992763996124, | |
| "eval_runtime": 4.1831, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.4417066276073456, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.0954, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 0.5344513058662415, | |
| "eval_runtime": 4.1854, | |
| "eval_samples_per_second": 2.867, | |
| "eval_steps_per_second": 2.867, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.46811383962631226, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.0478, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.35958993434906006, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.0553, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 0.5769810080528259, | |
| "eval_runtime": 4.1832, | |
| "eval_samples_per_second": 2.869, | |
| "eval_steps_per_second": 2.869, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.28935864567756653, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.0604, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 0.6299728155136108, | |
| "eval_runtime": 4.1838, | |
| "eval_samples_per_second": 2.868, | |
| "eval_steps_per_second": 2.868, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "step": 57, | |
| "total_flos": 2.531751456092979e+16, | |
| "train_loss": 0.6520790398905152, | |
| "train_runtime": 1086.9346, | |
| "train_samples_per_second": 2.116, | |
| "train_steps_per_second": 0.092 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.531751456092979e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |