Pageee/FT-English-1haa

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: 0.1610
• Wer: 4.1481

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
0.5827	5.1282	100	0.7468	3.4509
0.3801	10.2564	200	0.5781	3.4856
0.1166	15.3846	300	0.2330	3.8872
0.0469	20.5128	400	0.1750	4.1053
0.0249	25.6410	500	0.1637	4.1277
0.0173	30.7692	600	0.1609	4.1297
0.0119	35.8974	700	0.1604	4.1358
0.0087	41.0256	800	0.1607	4.1501
0.0074	46.1538	900	0.1609	4.1460
0.0071	51.2821	1000	0.1610	4.1481

Framework versions

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