learning_rate: 0.001
train_batch_size: 2
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 4
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
num_epochs: 5
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Cer	Wer
0.5672	0.4415	200	0.5623	0.1566	0.4468
0.5548	0.8830	400	0.5461	0.1480	0.4194
0.5165	1.3245	600	0.5330	0.1447	0.4107
0.5222	1.7660	800	0.5271	0.1426	0.3972
0.4907	2.2075	1000	0.5215	0.1432	0.4035
0.5297	2.6490	1200	0.5124	0.1358	0.3772
0.5173	3.0905	1400	0.4970	0.1331	0.3717
0.4934	3.5320	1600	0.4924	0.1338	0.3743
0.4349	3.9735	1800	0.4985	0.1317	0.3667
0.4217	4.4150	2000	0.4860	0.1311	0.3627
0.4375	4.8565	2200	0.4832	0.1308	0.3627

ctaguchi/ssc-pne-mms-model-mix-adapt-max

ssc-pne-mms-model-mix-adapt-max