JunWorks/whisper-small-zhTW-miltilang-test2

Whisper Small multiLang

This model is a fine-tuned version of openai/whisper-small on the Common Voice 13.0 + micsell/common_voice_en(en subset) dataset. It achieves the following results on the evaluation set:

• eval_loss: 3.6185
• eval_cer: 15.7285
• eval_wer: 33.5616
• eval_both_er: 25.3465
• eval_runtime: 9.1534
• eval_samples_per_second: 3.277
• eval_steps_per_second: 3.277
• epoch: 40.0
• step: 40

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: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 15000
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1