whisper-large-v2-common_voice_19_0-zh-TW-full-1

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

whisper-large-v2-common_voice_19_0-zh-TW-full-1

This model is a fine-tuned version of openai/whisper-large-v2 on the JacobLinCool/common_voice_19_0_zh-TW dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.1192
• eval_wer: 42.0108
• eval_cer: 17.1669
• eval_decode_runtime: 91.7088
• eval_wer_runtime: 0.1238
• eval_cer_runtime: 0.1555
• eval_runtime: 334.8833
• eval_samples_per_second: 14.969
• eval_steps_per_second: 0.469
• epoch: 0.3
• step: 1500

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