| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5347539113428944, | |
| "eval_steps": 100, | |
| "global_step": 2100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 6.05555534362793, | |
| "learning_rate": 4.9915117562176395e-06, | |
| "loss": 1.1007, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_loss": 1.03852117061615, | |
| "eval_runtime": 454.1166, | |
| "eval_samples_per_second": 21.651, | |
| "eval_steps_per_second": 1.354, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 22.63409423828125, | |
| "learning_rate": 4.983023512435278e-06, | |
| "loss": 1.0036, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "eval_loss": 0.84719318151474, | |
| "eval_runtime": 452.9811, | |
| "eval_samples_per_second": 21.705, | |
| "eval_steps_per_second": 1.358, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 17.67691993713379, | |
| "learning_rate": 4.974535268652917e-06, | |
| "loss": 0.8356, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_loss": 0.5871103405952454, | |
| "eval_runtime": 453.5224, | |
| "eval_samples_per_second": 21.679, | |
| "eval_steps_per_second": 1.356, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 26.09956932067871, | |
| "learning_rate": 4.966047024870554e-06, | |
| "loss": 0.6282, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_loss": 0.49549001455307007, | |
| "eval_runtime": 453.1973, | |
| "eval_samples_per_second": 21.695, | |
| "eval_steps_per_second": 1.357, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 23.882291793823242, | |
| "learning_rate": 4.957558781088193e-06, | |
| "loss": 0.5428, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_loss": 0.4652431905269623, | |
| "eval_runtime": 452.8789, | |
| "eval_samples_per_second": 21.71, | |
| "eval_steps_per_second": 1.358, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 18.765167236328125, | |
| "learning_rate": 4.949070537305831e-06, | |
| "loss": 0.5124, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "eval_loss": 0.42608916759490967, | |
| "eval_runtime": 453.0516, | |
| "eval_samples_per_second": 21.702, | |
| "eval_steps_per_second": 1.357, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 19.878719329833984, | |
| "learning_rate": 4.94058229352347e-06, | |
| "loss": 0.506, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_loss": 0.4183957576751709, | |
| "eval_runtime": 452.8777, | |
| "eval_samples_per_second": 21.71, | |
| "eval_steps_per_second": 1.358, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 19.626863479614258, | |
| "learning_rate": 4.9320940497411094e-06, | |
| "loss": 0.4828, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.4115789830684662, | |
| "eval_runtime": 451.8875, | |
| "eval_samples_per_second": 21.758, | |
| "eval_steps_per_second": 1.361, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 17.65311050415039, | |
| "learning_rate": 4.923605805958748e-06, | |
| "loss": 0.4698, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "eval_loss": 0.40224820375442505, | |
| "eval_runtime": 451.4433, | |
| "eval_samples_per_second": 21.779, | |
| "eval_steps_per_second": 1.362, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 20.876745223999023, | |
| "learning_rate": 4.915117562176387e-06, | |
| "loss": 0.4572, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.41678380966186523, | |
| "eval_runtime": 452.1237, | |
| "eval_samples_per_second": 21.746, | |
| "eval_steps_per_second": 1.36, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 22.631921768188477, | |
| "learning_rate": 4.906629318394024e-06, | |
| "loss": 0.4347, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_loss": 0.39895564317703247, | |
| "eval_runtime": 452.3272, | |
| "eval_samples_per_second": 21.736, | |
| "eval_steps_per_second": 1.36, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 19.680686950683594, | |
| "learning_rate": 4.898141074611663e-06, | |
| "loss": 0.4412, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_loss": 0.3831236958503723, | |
| "eval_runtime": 452.7887, | |
| "eval_samples_per_second": 21.714, | |
| "eval_steps_per_second": 1.358, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 20.93834114074707, | |
| "learning_rate": 4.889652830829302e-06, | |
| "loss": 0.4426, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "eval_loss": 0.3780069649219513, | |
| "eval_runtime": 452.3849, | |
| "eval_samples_per_second": 21.734, | |
| "eval_steps_per_second": 1.359, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 20.318857192993164, | |
| "learning_rate": 4.88116458704694e-06, | |
| "loss": 0.4367, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_loss": 0.36752593517303467, | |
| "eval_runtime": 452.8709, | |
| "eval_samples_per_second": 21.71, | |
| "eval_steps_per_second": 1.358, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 16.789215087890625, | |
| "learning_rate": 4.8726763432645794e-06, | |
| "loss": 0.4159, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_loss": 0.37768688797950745, | |
| "eval_runtime": 453.2249, | |
| "eval_samples_per_second": 21.693, | |
| "eval_steps_per_second": 1.357, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 15.212777137756348, | |
| "learning_rate": 4.864188099482218e-06, | |
| "loss": 0.4145, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_loss": 0.3734543025493622, | |
| "eval_runtime": 453.5584, | |
| "eval_samples_per_second": 21.677, | |
| "eval_steps_per_second": 1.356, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 16.271045684814453, | |
| "learning_rate": 4.855699855699857e-06, | |
| "loss": 0.4039, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_loss": 0.35818716883659363, | |
| "eval_runtime": 453.0212, | |
| "eval_samples_per_second": 21.703, | |
| "eval_steps_per_second": 1.358, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 30.50907325744629, | |
| "learning_rate": 4.847211611917494e-06, | |
| "loss": 0.4138, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_loss": 0.38744741678237915, | |
| "eval_runtime": 452.8649, | |
| "eval_samples_per_second": 21.711, | |
| "eval_steps_per_second": 1.358, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 12.229795455932617, | |
| "learning_rate": 4.838723368135133e-06, | |
| "loss": 0.4071, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_loss": 0.3485246002674103, | |
| "eval_runtime": 453.2995, | |
| "eval_samples_per_second": 21.69, | |
| "eval_steps_per_second": 1.357, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 15.995959281921387, | |
| "learning_rate": 4.830235124352772e-06, | |
| "loss": 0.4141, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_loss": 0.34898054599761963, | |
| "eval_runtime": 453.0588, | |
| "eval_samples_per_second": 21.701, | |
| "eval_steps_per_second": 1.357, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 21.010486602783203, | |
| "learning_rate": 4.82174688057041e-06, | |
| "loss": 0.3917, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_loss": 0.36585670709609985, | |
| "eval_runtime": 453.7967, | |
| "eval_samples_per_second": 21.666, | |
| "eval_steps_per_second": 1.355, | |
| "step": 2100 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 58905, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 100, | |
| "total_flos": 1.7850550446351974e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |