Whisper Small Norwegian Bokmål
This model is a fine-tuned version of openai/whisper-small 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: 0.3230
- Wer: 15.56
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.
The following hyperparameters were used during training:
- learning_rate: 3e-06
- train_batch_size: 128
- 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 warmup
- lr_scheduler_warmup_steps: 1000
- training_steps: 50.000 (currently @1.000)
- mixed_precision_training: fp16
- deepspeed: true
Live Training results
- Downloads last month
This model can be loaded on the Inference API on-demand.
Datasets used to train NbAiLab/whisper-small-nob
- Wer on FLEURSvalidation set self-reported15.560