whisper-large-v3-turbo-16e-c

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the common_voice_16_1 dataset. It achieves the following results on the evaluation set:

• Loss: 1.2775
• Wer: 60.3430

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: 16
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• 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: 5000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0	0	7.6611	130.4102
1.3625	0.1	500	2.0087	73.6922
0.9901	0.2	1000	1.4311	58.1072
0.7534	0.3	1500	1.3291	55.8715
0.8052	0.4	2000	1.3063	50.5318
0.7024	0.5	2500	1.2896	46.6464
0.926	0.6	3000	1.2833	59.4747
0.6204	0.7	3500	1.2788	49.9674
0.7267	0.8	4000	1.2791	58.9755
0.8442	0.9	4500	1.2775	50.2930
0.9145	1.0	5000	1.2775	60.3430

Framework versions

• Transformers 4.54.0
• Pytorch 2.8.0.dev20250319+cu128
• Datasets 3.6.0
• Tokenizers 0.21.2