Instructions to use sarvamai/sarvam-translate with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use sarvamai/sarvam-translate with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("translation", model="sarvamai/sarvam-translate")# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("sarvamai/sarvam-translate") model = AutoModelForMultimodalLM.from_pretrained("sarvamai/sarvam-translate", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Does this model support translation among Indian languages?
#16
by vishnu-vizz - opened
I'm running into an issue when trying to do Indian to Indian language translation.
Facing the same issue, model automatically translates it to English
Any Update on This ?
Hello all. Sorry, missed the notifications in this thread.
The model was trained only for English-to-Indic and Indic-to-Eng.
It was not trained for Indic-to-Indic.
If you have usecases for Indic-to-Indic, you can pivot it via English.
That is, you can do Indic->Eng first, and then Eng->Indic.
Thanks! GolulNC