| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.0744047619047619, | |
| "eval_steps": 9, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.000744047619047619, | |
| "eval_loss": 8.732146263122559, | |
| "eval_runtime": 160.2118, | |
| "eval_samples_per_second": 14.131, | |
| "eval_steps_per_second": 1.766, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002232142857142857, | |
| "grad_norm": 3.8788089752197266, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.6329, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.004464285714285714, | |
| "grad_norm": 3.3676693439483643, | |
| "learning_rate": 3e-05, | |
| "loss": 8.6812, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.006696428571428571, | |
| "grad_norm": 4.344758033752441, | |
| "learning_rate": 4.5e-05, | |
| "loss": 8.5473, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.006696428571428571, | |
| "eval_loss": 8.29433536529541, | |
| "eval_runtime": 161.9566, | |
| "eval_samples_per_second": 13.979, | |
| "eval_steps_per_second": 1.747, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.008928571428571428, | |
| "grad_norm": 4.432471752166748, | |
| "learning_rate": 4.993910125649561e-05, | |
| "loss": 7.6704, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.011160714285714286, | |
| "grad_norm": 5.1014227867126465, | |
| "learning_rate": 4.962019382530521e-05, | |
| "loss": 7.1247, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.013392857142857142, | |
| "grad_norm": 6.452256202697754, | |
| "learning_rate": 4.9031542398457974e-05, | |
| "loss": 6.7864, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.013392857142857142, | |
| "eval_loss": 6.469050884246826, | |
| "eval_runtime": 161.9588, | |
| "eval_samples_per_second": 13.979, | |
| "eval_steps_per_second": 1.747, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.015625, | |
| "grad_norm": 6.13102912902832, | |
| "learning_rate": 4.817959636416969e-05, | |
| "loss": 6.0363, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 5.863094329833984, | |
| "learning_rate": 4.707368982147318e-05, | |
| "loss": 5.4919, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.020089285714285716, | |
| "grad_norm": 7.331254959106445, | |
| "learning_rate": 4.572593931387604e-05, | |
| "loss": 5.5447, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.020089285714285716, | |
| "eval_loss": 5.072122573852539, | |
| "eval_runtime": 162.0067, | |
| "eval_samples_per_second": 13.975, | |
| "eval_steps_per_second": 1.747, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.022321428571428572, | |
| "grad_norm": 7.104370594024658, | |
| "learning_rate": 4.415111107797445e-05, | |
| "loss": 4.8921, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.024553571428571428, | |
| "grad_norm": 5.608682632446289, | |
| "learning_rate": 4.2366459261474933e-05, | |
| "loss": 4.841, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.026785714285714284, | |
| "grad_norm": 4.773255825042725, | |
| "learning_rate": 4.039153688314145e-05, | |
| "loss": 4.5234, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.026785714285714284, | |
| "eval_loss": 4.577538967132568, | |
| "eval_runtime": 162.4812, | |
| "eval_samples_per_second": 13.934, | |
| "eval_steps_per_second": 1.742, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.029017857142857144, | |
| "grad_norm": 5.47312068939209, | |
| "learning_rate": 3.824798160583012e-05, | |
| "loss": 4.5007, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.03125, | |
| "grad_norm": 5.6877641677856445, | |
| "learning_rate": 3.5959278669726935e-05, | |
| "loss": 4.0774, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.033482142857142856, | |
| "grad_norm": 4.455398082733154, | |
| "learning_rate": 3.355050358314172e-05, | |
| "loss": 4.2481, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.033482142857142856, | |
| "eval_loss": 4.341364860534668, | |
| "eval_runtime": 161.8916, | |
| "eval_samples_per_second": 13.985, | |
| "eval_steps_per_second": 1.748, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 6.039599418640137, | |
| "learning_rate": 3.104804738999169e-05, | |
| "loss": 4.1585, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.03794642857142857, | |
| "grad_norm": 5.91121244430542, | |
| "learning_rate": 2.8479327524001636e-05, | |
| "loss": 4.1427, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04017857142857143, | |
| "grad_norm": 5.7630181312561035, | |
| "learning_rate": 2.587248741756253e-05, | |
| "loss": 3.8608, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04017857142857143, | |
| "eval_loss": 4.20692253112793, | |
| "eval_runtime": 162.0191, | |
| "eval_samples_per_second": 13.974, | |
| "eval_steps_per_second": 1.747, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04241071428571429, | |
| "grad_norm": 6.365548610687256, | |
| "learning_rate": 2.3256088156396868e-05, | |
| "loss": 4.1786, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.044642857142857144, | |
| "grad_norm": 8.317145347595215, | |
| "learning_rate": 2.0658795558326743e-05, | |
| "loss": 4.2896, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.046875, | |
| "grad_norm": 6.300489902496338, | |
| "learning_rate": 1.8109066104575023e-05, | |
| "loss": 4.0623, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.046875, | |
| "eval_loss": 4.134160995483398, | |
| "eval_runtime": 161.9503, | |
| "eval_samples_per_second": 13.98, | |
| "eval_steps_per_second": 1.747, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.049107142857142856, | |
| "grad_norm": 6.4269185066223145, | |
| "learning_rate": 1.56348351646022e-05, | |
| "loss": 3.9744, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.05133928571428571, | |
| "grad_norm": 5.94874382019043, | |
| "learning_rate": 1.3263210930352737e-05, | |
| "loss": 3.9959, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 5.269709587097168, | |
| "learning_rate": 1.1020177413231334e-05, | |
| "loss": 4.0065, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "eval_loss": 4.0851731300354, | |
| "eval_runtime": 161.9017, | |
| "eval_samples_per_second": 13.984, | |
| "eval_steps_per_second": 1.748, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.05580357142857143, | |
| "grad_norm": 6.060158729553223, | |
| "learning_rate": 8.930309757836517e-06, | |
| "loss": 3.9547, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.05803571428571429, | |
| "grad_norm": 4.884081840515137, | |
| "learning_rate": 7.016504991533726e-06, | |
| "loss": 3.8446, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.060267857142857144, | |
| "grad_norm": 5.20225191116333, | |
| "learning_rate": 5.299731159831953e-06, | |
| "loss": 3.8639, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.060267857142857144, | |
| "eval_loss": 4.055640697479248, | |
| "eval_runtime": 161.812, | |
| "eval_samples_per_second": 13.992, | |
| "eval_steps_per_second": 1.749, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 6.285635948181152, | |
| "learning_rate": 3.798797596089351e-06, | |
| "loss": 4.2296, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.06473214285714286, | |
| "grad_norm": 5.423341751098633, | |
| "learning_rate": 2.5301488425208296e-06, | |
| "loss": 4.1821, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.06696428571428571, | |
| "grad_norm": 5.548636436462402, | |
| "learning_rate": 1.5076844803522922e-06, | |
| "loss": 3.9078, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06696428571428571, | |
| "eval_loss": 4.040578365325928, | |
| "eval_runtime": 161.4718, | |
| "eval_samples_per_second": 14.021, | |
| "eval_steps_per_second": 1.753, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06919642857142858, | |
| "grad_norm": 6.10594367980957, | |
| "learning_rate": 7.426068431000882e-07, | |
| "loss": 4.0287, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 6.208632946014404, | |
| "learning_rate": 2.4329828146074095e-07, | |
| "loss": 3.8063, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.07366071428571429, | |
| "grad_norm": 5.615017414093018, | |
| "learning_rate": 1.522932452260595e-08, | |
| "loss": 3.8605, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.07366071428571429, | |
| "eval_loss": 4.03863525390625, | |
| "eval_runtime": 161.5774, | |
| "eval_samples_per_second": 14.012, | |
| "eval_steps_per_second": 1.751, | |
| "step": 99 | |
| } | |
| ], | |
| "logging_steps": 3, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 9, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.483774567120896e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |