| { | |
| "best_metric": 0.3781418800354004, | |
| "best_model_checkpoint": "facial_emotions_image_detection/checkpoint-15740", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 15740, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 11.892455101013184, | |
| "learning_rate": 3.885277246653919e-06, | |
| "loss": 0.5534, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 12.262578964233398, | |
| "learning_rate": 3.7578075207138302e-06, | |
| "loss": 0.5424, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 11.160516738891602, | |
| "learning_rate": 3.630337794773741e-06, | |
| "loss": 0.5402, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8678585617798967, | |
| "eval_loss": 0.4141866862773895, | |
| "eval_runtime": 113.7657, | |
| "eval_samples_per_second": 110.622, | |
| "eval_steps_per_second": 13.835, | |
| "step": 1574 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 12.081930160522461, | |
| "learning_rate": 3.5028680688336517e-06, | |
| "loss": 0.509, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 8.023213386535645, | |
| "learning_rate": 3.375398342893563e-06, | |
| "loss": 0.4984, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 11.671684265136719, | |
| "learning_rate": 3.2479286169534735e-06, | |
| "loss": 0.4817, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8661899086213747, | |
| "eval_loss": 0.41461509466171265, | |
| "eval_runtime": 111.4496, | |
| "eval_samples_per_second": 112.921, | |
| "eval_steps_per_second": 14.123, | |
| "step": 3148 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 11.465852737426758, | |
| "learning_rate": 3.120458891013384e-06, | |
| "loss": 0.4686, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 13.02270221710205, | |
| "learning_rate": 2.992989165073295e-06, | |
| "loss": 0.451, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 7.672973155975342, | |
| "learning_rate": 2.8655194391332055e-06, | |
| "loss": 0.4459, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8698450536352801, | |
| "eval_loss": 0.40479913353919983, | |
| "eval_runtime": 111.5158, | |
| "eval_samples_per_second": 112.854, | |
| "eval_steps_per_second": 14.115, | |
| "step": 4722 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 12.763694763183594, | |
| "learning_rate": 2.7380497131931164e-06, | |
| "loss": 0.4335, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 10.964895248413086, | |
| "learning_rate": 2.6105799872530273e-06, | |
| "loss": 0.4325, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 12.650814056396484, | |
| "learning_rate": 2.4831102613129383e-06, | |
| "loss": 0.4031, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8707985697258641, | |
| "eval_loss": 0.3987765610218048, | |
| "eval_runtime": 110.9162, | |
| "eval_samples_per_second": 113.464, | |
| "eval_steps_per_second": 14.191, | |
| "step": 6296 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 7.15053653717041, | |
| "learning_rate": 2.3556405353728488e-06, | |
| "loss": 0.4112, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 17.31258773803711, | |
| "learning_rate": 2.2281708094327597e-06, | |
| "loss": 0.3886, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 7.938920497894287, | |
| "learning_rate": 2.1007010834926702e-06, | |
| "loss": 0.3804, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8742947953913389, | |
| "eval_loss": 0.39545899629592896, | |
| "eval_runtime": 110.9484, | |
| "eval_samples_per_second": 113.431, | |
| "eval_steps_per_second": 14.187, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "grad_norm": 11.187897682189941, | |
| "learning_rate": 1.973231357552581e-06, | |
| "loss": 0.3856, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 12.73786735534668, | |
| "learning_rate": 1.845761631612492e-06, | |
| "loss": 0.3732, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "grad_norm": 6.2277374267578125, | |
| "learning_rate": 1.7182919056724028e-06, | |
| "loss": 0.3733, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8767580452920143, | |
| "eval_loss": 0.3847721815109253, | |
| "eval_runtime": 113.9748, | |
| "eval_samples_per_second": 110.419, | |
| "eval_steps_per_second": 13.81, | |
| "step": 9444 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 9.49077320098877, | |
| "learning_rate": 1.5908221797323135e-06, | |
| "loss": 0.3552, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "grad_norm": 11.691384315490723, | |
| "learning_rate": 1.4633524537922245e-06, | |
| "loss": 0.355, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 7.24980354309082, | |
| "learning_rate": 1.335882727852135e-06, | |
| "loss": 0.347, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "grad_norm": 15.983756065368652, | |
| "learning_rate": 1.2084130019120457e-06, | |
| "loss": 0.3462, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.877870480731029, | |
| "eval_loss": 0.38435637950897217, | |
| "eval_runtime": 114.6747, | |
| "eval_samples_per_second": 109.745, | |
| "eval_steps_per_second": 13.726, | |
| "step": 11018 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "grad_norm": 6.091674327850342, | |
| "learning_rate": 1.0809432759719566e-06, | |
| "loss": 0.3481, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "grad_norm": 14.106987953186035, | |
| "learning_rate": 9.534735500318673e-07, | |
| "loss": 0.3373, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 7.94, | |
| "grad_norm": 8.448904991149902, | |
| "learning_rate": 8.260038240917782e-07, | |
| "loss": 0.324, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.876201827572507, | |
| "eval_loss": 0.3851085305213928, | |
| "eval_runtime": 113.668, | |
| "eval_samples_per_second": 110.717, | |
| "eval_steps_per_second": 13.847, | |
| "step": 12592 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "grad_norm": 6.959225177764893, | |
| "learning_rate": 6.98534098151689e-07, | |
| "loss": 0.3254, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 8.58, | |
| "grad_norm": 10.736397743225098, | |
| "learning_rate": 5.710643722115998e-07, | |
| "loss": 0.3179, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "grad_norm": 10.716438293457031, | |
| "learning_rate": 4.435946462715105e-07, | |
| "loss": 0.3217, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8787445371473976, | |
| "eval_loss": 0.3801954388618469, | |
| "eval_runtime": 113.1501, | |
| "eval_samples_per_second": 111.224, | |
| "eval_steps_per_second": 13.911, | |
| "step": 14166 | |
| }, | |
| { | |
| "epoch": 9.21, | |
| "grad_norm": 12.26369571685791, | |
| "learning_rate": 3.161249203314213e-07, | |
| "loss": 0.309, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 9.53, | |
| "grad_norm": 6.435930252075195, | |
| "learning_rate": 1.8865519439133203e-07, | |
| "loss": 0.3158, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 9.85, | |
| "grad_norm": 15.131240844726562, | |
| "learning_rate": 6.118546845124282e-08, | |
| "loss": 0.3105, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8789829161700436, | |
| "eval_loss": 0.3781418800354004, | |
| "eval_runtime": 112.3381, | |
| "eval_samples_per_second": 112.028, | |
| "eval_steps_per_second": 14.011, | |
| "step": 15740 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 15740, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 3.900966581033497e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |