David Dale
David Dale
Yes, thanks, it works!
> How load checkpoints that i got from fine-tuning. You can start by loading the original model (e.g. `seamlessM4T_v2_large`) from its card, and then use the function `load_checkpoint` ([src/seamless_communication/cli/m4t/evaluate/evaluate.py#L365](https://github.com/facebookresearch/seamless_communication/blob/main/src/seamless_communication/cli/m4t/evaluate/evaluate.py#L365)) to...
Have you tried removing the argument `do_sample=True`? This is what makes the result non-deterministic.
@rshbhrnjn We updated the dataset link to a correct one. Please check it out!
@mfcoria can you comment on the above?
Yes, adding the languages id to the dataframe would be a great first step. Another potential enhancement is to make the file `recast.py` localizeable, so that the user could provide...
Yes, currently it handles only inequality constraints. But of course implementations of other types of tasks are welcome!
Yes, it is definitely possible. Do you want to submit a pull request with this feature?
Hey, I want to continue this work! How to I reopen it?
Commenting to avoid it getting closed