learning_rate: 1e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
gradient_accumulation_steps: 8
total_train_batch_size: 64
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine_with_restarts
lr_scheduler_warmup_ratio: 0.1
num_epochs: 5
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
0.499	0.3504	30	0.0595
0.0364	0.7007	60	0.0291
0.0237	1.0584	90	0.0171
0.0095	1.4088	120	0.0145
0.0038	1.7591	150	0.0115
0.0097	2.1168	180	0.0110
0.0093	2.4672	210	0.0113
0.0046	2.8175	240	0.0093
0.0011	3.1752	270	0.0008
0.0001	3.5255	300	0.0003
0.001	3.8759	330	0.0004
0.0004	4.2336	360	0.0003
0.0	4.5839	390	0.0002
0.0001	4.9343	420	0.0002

yangwooko/smartmind-cyberone-20250404

smartmind-cyberone-20250404