learning_rate: 0.0001
train_batch_size: 4
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 100
training_steps: 4000
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
3.6678	1.8519	100	0.3965
3.5185	3.7037	200	0.3936
3.4784	5.5556	300	0.3948
3.3646	7.4074	400	0.3876
3.2878	9.2593	500	0.3897
3.3122	11.1111	600	0.3949
3.2292	12.9630	700	0.3926
3.1861	14.8148	800	0.3957
3.1694	16.6667	900	0.3938
3.1684	18.5185	1000	0.3922
3.1456	20.3704	1100	0.3926
3.0656	22.2222	1200	0.4002
3.0691	24.0741	1300	0.3984
3.0598	25.9259	1400	0.3956
3.0631	27.7778	1500	0.3967
3.0173	29.6296	1600	0.3982
2.9880	31.4815	1700	0.3948
2.9621	33.3333	1800	0.4008
3.0060	35.1852	1900	0.4013
2.8928	37.0370	2000	0.4005
2.9264	38.8889	2100	0.3962
2.9448	40.7407	2200	0.4061
2.8935	42.5926	2300	0.3983
2.9116	44.4444	2400	0.4033
3.5378	2.9317	2500	0.3946
3.4876	3.0481	2600	0.3908
3.4668	3.1654	2700	0.3852
3.4454	3.2828	2800	0.3840
3.3969	3.4001	2900	0.3822
3.3969	3.5175	3000	0.3809
3.3570	3.6348	3100	0.3804
3.3575	3.7521	3200	0.3779
3.3884	3.8695	3300	0.3784
3.4068	3.9868	3400	0.3765
3.3193	4.1033	3500	0.3756
3.3343	4.2206	3600	0.3757
3.3158	4.3379	3700	0.3744
3.3254	4.4553	3800	0.3740
3.3202	4.5726	3900	0.3735
3.3578	4.6899	4000	0.3729

truongchinh5712/chinh_speecht5_finetuned_fipilot

chinh_speecht5_finetuned_fipilot