whisper-tiny-id-v2-new

This model is a fine-tuned version of openai/whisper-tiny on an unknown dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.7813
• eval_wer: 36.7460
• eval_runtime: 166.055
• eval_samples_per_second: 4.137
• eval_steps_per_second: 0.518
• epoch: 15.4341
• step: 2500

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training Results

Step	Training Loss	Validation Loss	WER
500	1.8105	0.7662	40.45
1000	0.5350	0.7074	35.96
1500	0.1668	0.7277	36.01
2000	0.0645	0.7588	36.04
2500	0.0332	0.7813	36.75

Best checkpoint: Step 1000 with WER 35.96%.

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 500
• training_steps: 3000
• mixed_precision_training: Native AMP

Framework versions

• Transformers 5.0.0
• Pytorch 2.11.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2