Sagicc/whisper-large-v2-sr-combined

Whisper Large v2 Sr Fleurs and CommonVoice

This model is a fine-tuned version of openai/whisper-large-v2 on the combined Google Fleurs and Mozilla Foundation Common Voice 13 dataset. It achieves the following results on the evaluation set:

• Loss: 0.1749
• Wer Ortho: 0.1678
• Wer: 0.0623

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: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 1500

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.0737	1.34	500	0.1735	0.1865	0.0908
0.0304	2.67	1000	0.1622	0.1670	0.0728
0.0156	4.01	1500	0.1749	0.1678	0.0623

Framework versions

• Transformers 4.33.1
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3