sassyphil/whispertest

Whisper Large Norwegian Bokmål

This model is a fine-tuned version of openai/whisper-large-v2 trained on several datasets.

It is currently in the middle of a large training. Currently it achieves the following results on the evaluation set:

• Loss: 0.2477
• Wer: 10.718635559082031

Model description

The model is trained on a large corpus of roughly 5.000 hours of voice. The sources are subtitles from the Norwegian broadcaster NRK, transcribed speeches from the Norwegian parliament and voice recordings from Norsk Språkteknologi.

Intended uses & limitations

The model will be free for everyone to use when it is finished.

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e-06
• train_batch_size: 64
• gradient_accumulation_steps: 2
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant with warmpu
• lr_scheduler_warmup_steps: 1000
• training_steps: 50.000 (currently @1.000)
• mixed_precision_training: fp16
• deepspeed: true

Live Training results

See Tensorboad Metrics