| { | |
| "best_metric": 0.9616682529449463, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_60-samples_config-2_full/checkpoint-46", | |
| "epoch": 22.956521739130434, | |
| "eval_steps": 500, | |
| "global_step": 66, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.43204331398010254, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4027, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.40640854835510254, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3564, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 1.3659462928771973, | |
| "eval_runtime": 4.3293, | |
| "eval_samples_per_second": 2.772, | |
| "eval_steps_per_second": 2.772, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.43465468287467957, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3765, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.3404293060302734, | |
| "eval_runtime": 4.1485, | |
| "eval_samples_per_second": 2.893, | |
| "eval_steps_per_second": 2.893, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.4115420877933502, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3832, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.35706835985183716, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3125, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.2525643110275269, | |
| "eval_runtime": 4.1478, | |
| "eval_samples_per_second": 2.893, | |
| "eval_steps_per_second": 2.893, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.24587181210517883, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2213, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.1814781427383423, | |
| "eval_runtime": 4.1533, | |
| "eval_samples_per_second": 2.889, | |
| "eval_steps_per_second": 2.889, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.24395760893821716, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.1847, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.2830147445201874, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.1432, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.1329258680343628, | |
| "eval_runtime": 4.1531, | |
| "eval_samples_per_second": 2.889, | |
| "eval_steps_per_second": 2.889, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.9978663921356201, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.0949, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.0849387645721436, | |
| "eval_runtime": 4.1475, | |
| "eval_samples_per_second": 2.893, | |
| "eval_steps_per_second": 2.893, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.2531653940677643, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.0567, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.32131874561309814, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.0142, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 1.0435309410095215, | |
| "eval_runtime": 4.1445, | |
| "eval_samples_per_second": 2.895, | |
| "eval_steps_per_second": 2.895, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.194178506731987, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.9925, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.0152888298034668, | |
| "eval_runtime": 4.1469, | |
| "eval_samples_per_second": 2.894, | |
| "eval_steps_per_second": 2.894, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.16194817423820496, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.9508, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 1.0030059814453125, | |
| "eval_runtime": 4.2479, | |
| "eval_samples_per_second": 2.825, | |
| "eval_steps_per_second": 2.825, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.16992084681987762, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.9391, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.16991503536701202, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.9191, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 0.9871246814727783, | |
| "eval_runtime": 4.1888, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 2.865, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.1509915590286255, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.9172, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 0.9778385162353516, | |
| "eval_runtime": 4.1921, | |
| "eval_samples_per_second": 2.863, | |
| "eval_steps_per_second": 2.863, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.17643801867961884, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.872, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.13660450279712677, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.892, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 0.9705256819725037, | |
| "eval_runtime": 4.1887, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 2.865, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.16165243089199066, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.8585, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 0.9655454754829407, | |
| "eval_runtime": 4.1915, | |
| "eval_samples_per_second": 2.863, | |
| "eval_steps_per_second": 2.863, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.1437292844057083, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.8647, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.14648310840129852, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.8535, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 0.9630029201507568, | |
| "eval_runtime": 4.1867, | |
| "eval_samples_per_second": 2.866, | |
| "eval_steps_per_second": 2.866, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.17346253991127014, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.8316, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 0.9617516398429871, | |
| "eval_runtime": 4.1903, | |
| "eval_samples_per_second": 2.864, | |
| "eval_steps_per_second": 2.864, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.13388116657733917, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.8255, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.14341551065444946, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.8242, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.9616682529449463, | |
| "eval_runtime": 4.1878, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 2.865, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.15474450588226318, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.7836, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 0.9622191786766052, | |
| "eval_runtime": 4.1895, | |
| "eval_samples_per_second": 2.864, | |
| "eval_steps_per_second": 2.864, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.16550879180431366, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.7962, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 0.9639542698860168, | |
| "eval_runtime": 4.1922, | |
| "eval_samples_per_second": 2.862, | |
| "eval_steps_per_second": 2.862, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.15981481969356537, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.8078, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.16021905839443207, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.7851, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 0.965728759765625, | |
| "eval_runtime": 4.1953, | |
| "eval_samples_per_second": 2.86, | |
| "eval_steps_per_second": 2.86, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.17779593169689178, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.7553, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 0.970211923122406, | |
| "eval_runtime": 4.1891, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 2.865, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 20.17391304347826, | |
| "grad_norm": 0.18282093107700348, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.7649, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 0.20358732342720032, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.7474, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "eval_loss": 0.9766749739646912, | |
| "eval_runtime": 4.1552, | |
| "eval_samples_per_second": 2.888, | |
| "eval_steps_per_second": 2.888, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 21.565217391304348, | |
| "grad_norm": 0.18090173602104187, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.7352, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "eval_loss": 0.9821853637695312, | |
| "eval_runtime": 4.148, | |
| "eval_samples_per_second": 2.893, | |
| "eval_steps_per_second": 2.893, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 22.26086956521739, | |
| "grad_norm": 0.20831148326396942, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.7279, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "grad_norm": 0.21625789999961853, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.7191, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "eval_loss": 0.9870995879173279, | |
| "eval_runtime": 4.1471, | |
| "eval_samples_per_second": 2.894, | |
| "eval_steps_per_second": 2.894, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "step": 66, | |
| "total_flos": 2.88704316833792e+16, | |
| "train_loss": 0.9510506773085305, | |
| "train_runtime": 1246.7903, | |
| "train_samples_per_second": 1.845, | |
| "train_steps_per_second": 0.08 | |
| } | |
| ], | |
| "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": 2.88704316833792e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |