whisper-tiny-v2-ta_55

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

• Loss: 0.3972
• Wer: 62.1792

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: 1.5e-05
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 3000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.6242	0.11	300	0.5519	52.0097
0.4507	0.21	600	0.4556	55.9806
0.3999	0.32	900	0.4148	65.0363
0.3635	0.42	1200	0.3972	62.1792

Framework versions

• Transformers 4.28.0.dev0
• Pytorch 1.12.1
• Datasets 2.11.0
• Tokenizers 0.13.3