learning_rate: 1e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 16
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: cosine
lr_scheduler_warmup_steps: 0.1
num_epochs: 10
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
1.1809	0.9965	285	0.5577	71.1024	26.9361
0.8395	1.9930	570	0.4625	61.6929	22.6497
0.5538	2.9895	855	0.4426	58.7402	23.2476
0.3483	3.9860	1140	0.4445	53.3858	20.1354
0.1779	4.9825	1425	0.4912	52.9528	18.9396
0.0815	5.9790	1710	0.5356	52.3622	18.8171
0.0273	6.9755	1995	0.5762	51.8504	18.7306
0.0131	7.9720	2280	0.6149	51.8110	18.7234
0.0049	8.9685	2565	0.6370	51.4173	18.6298
0.0044	9.9650	2850	0.6409	51.5748	18.7090

wandererupak/whisper-small-n-demo-ultimate-train-XX

Whisper Small N - Final