whisper-large-v3-common_voice_19_0-zh-TW-full-1

It seems that this training has failed. Please use whisper-large-v3-turbo-common_voice_19_0-zh-TW instead.

This model is a fine-tuned version of openai/whisper-large-v3 on the JacobLinCool/common_voice_19_0_zh-TW dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.1502
• eval_wer: 253.4012
• eval_cer: 234.4286
• eval_decode_runtime: 107.4219
• eval_wer_runtime: 0.1395
• eval_cer_runtime: 0.2215
• eval_runtime: 352.6335
• eval_samples_per_second: 14.216
• eval_steps_per_second: 0.445
• epoch: 2.1825
• step: 4000

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: 0.0002
• train_batch_size: 4
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 5000

Framework versions

• PEFT 0.13.2
• Transformers 4.46.1
• Pytorch 2.4.0
• Datasets 3.0.2
• Tokenizers 0.20.1