Open harryjulian opened 1 year ago
Likewise, having trouble setting up the model
same
Just getting the dependencies in place is a bit of a challenge :) In-progress environment.yml
, but I get an error from torch.distributed
still
channels:
- conda-forge
dependencies:
- allennlp
- ipython
- python
- python-flair
- pytorch
- torchvision
- spacy
- tokenizers
- typing_extensions[version='<4.5']
(better resolved with micromamba)
@astrojuanlu Were able to use these embeddings ? If so can you help me out as well
Hi, I'd like to use this in some of my data science work but there's not really any steps on how to reproduce.
The linked checkpoint is for a .pt file whilst the repo links to a tensorflow implementation of ELMO.
Any advice? I'm assuming the original model was trained using the AllenNLP pytorch implementation of ELMO?