M00dler/whisper-medium-malay

Whisper Medium Malay (12/6 batch size) - Gab

This model is a fine-tuned version of openai/whisper-medium on the malay-conversational-speech-corpus dataset. It achieves the following results on the evaluation set:

• eval_loss: 1.0942
• eval_wer: 52.0822
• eval_runtime: 338.3155
• eval_samples_per_second: 1.918
• eval_steps_per_second: 0.322
• epoch: 6.9444
• step: 750

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: 12
• eval_batch_size: 6
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 24
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 250
• training_steps: 1500
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1