Wissam Antoun
Wissam Antoun
We have released our re-implementation of DeBERTa pertaining code which we used to train CamemBERTa a french LM. https://gitlab.inria.fr/almanach/CamemBERTa
@BartWesthoff Yes
@ayoub-louati Thanks for your great work. I was able to successfully run the T5 conversion with ONNX. But I'm facing issues with the TensorRT conversion, from what I can read...
Anyone figured out how we can stream tokens from the generator thru fastapi back to the user?
> Thank you a lot @byshiue for your answer. Is encoder support planed? > > Our use case for encoder models are RAG linked. Vectorization is heavy on compute at...
> (Hi @WissamAntoun 👋 !) 🫡 Salut :)
@SpectralDragon Both Yippy versions, yours and the original started giving me the following error: I tried re-installing, deleting the folder under Applications Support and nothing worked. I'm on 14.6.1. Do...
Hey Mohammad, Yes but you need to train it for this task. If you don't want to train, you can search on the huggingface model repo to find pre-trained dialect...