whisper-small-te

This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4501
• Wer: 95.6522

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1715	25.0	100	1.2209	99.1304
0.0021	50.0	200	1.2032	99.1304
0.0001	75.0	300	1.2996	96.5217
0.0001	100.0	400	1.3443	97.3913
0.0001	125.0	500	1.3752	97.3913
0.0001	150.0	600	1.3996	96.5217
0.0001	175.0	700	1.4199	96.5217
0.0001	200.0	800	1.4359	95.6522
0.0001	225.0	900	1.4458	95.6522
0.0001	250.0	1000	1.4501	95.6522

Framework versions

• Transformers 4.38.0.dev0
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0