Pageee/FT-English-10maa

Whisper-Small En-10m

This model is a fine-tuned version of openai/whisper-small on the librispeech dataset. It achieves the following results on the evaluation set:

• Loss: 0.3635
• Wer: 3.7424

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-07
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 300
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.482	33.3333	100	0.7436	3.4183
0.2402	66.6667	200	0.5833	3.4448
0.0135	100.0	300	0.3881	3.5834
0.0029	133.3333	400	0.3731	3.6324
0.0019	166.6667	500	0.3685	3.6568
0.0014	200.0	600	0.3663	3.6854
0.0012	233.3333	700	0.3649	3.7098
0.0011	266.6667	800	0.3641	3.7241
0.001	300.0	900	0.3637	3.7465
0.001	333.3333	1000	0.3635	3.7424

Framework versions

• Transformers 4.41.0.dev0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1