Instructions to use sarvamai/shuka-1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use sarvamai/shuka-1 with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("sarvamai/shuka-1", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Is it possible to set output language other than english? If yes, how can I do that?
#3
by samruddhakf - opened
I want to get transcripts in same language as the audio language, or a specific language other than english (say Hindi). Would Shuka be able to do that? What parameters I should change to achieve this?
Since we don't finetune the decoder, it is not possible to deterministically get generations in languages other than English. However, you can try prompting the model to generate in the language you want and see what happens.
Okay, thanks!
samruddhakf changed discussion status to closed