Closed konstin closed 3 years ago
The easy solution is to just use an older version of pytorch to calculate the embeddings. If you need to use a newer version, the permanent solution to this is probably to use the older version of pytorch to save only the state_dict for the model (the saved models I provide are pickles of the whole model object, not just the state_dict), then load those state_dicts into the model in the new version of pytorch.
Thank you for the quick response, that made it work!
I'm trying to run
embed_sequences.py
, but I get an exception from torch code:These are the versions I installed (from
pip freeze
):I suspect that this has something to do with the torch version, so would it be possible to make the model run on more recent pytorch versions?