ap-fEz97qWiEaKtCs943k0PtZ

This model is a fine-tuned version of openai/whisper-large-v3 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7453
• Model Preparation Time: 0.0212
• Wer: 0.2339

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: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• 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: 400
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Wer
0.2538	0.9791	41	0.2886	0.0212	0.1134
0.1828	1.9791	82	0.3033	0.0212	0.1182
0.1233	2.9791	123	0.3724	0.0212	0.1248
0.1182	3.9791	164	0.4213	0.0212	0.1399
0.1181	4.9791	205	0.4813	0.0212	0.1417
0.1273	5.9791	246	0.5741	0.0212	0.1553
0.1237	6.9791	287	0.6128	0.0212	0.1759
0.1176	7.9791	328	0.6665	0.0212	0.1823
0.1076	8.9791	369	0.7048	0.0212	0.1929
0.1357	9.9791	410	0.7453	0.0212	0.2339

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0