Overwrite the `eos_token_id` for generation, avoiding the endless generation issue that happens only with the HF converted models

Update README.md

Update README.md

Update README.md

Update README.md

Update demo.ipynb with examples from the paper

Update README.md

Update relative paths in README.md

Update model name in the demo notebook

Merge modeling files into a single one to avoid relative import

Update relative import

Update relative import

Update modeling_xgenmm.py

Update README.md

upload model

update readme w/ examples

update preprocessing

Update README.md

initial commit

initial commit