search

XiangLi1999 / Diffusion-LM

Diffusion-LM
Apache License 2.0
1.03k stars 134 forks source link

Hugging-face online model not found when decoding Diffusion-LM #18

Open Junyi42 opened 2 years ago

Junyi42 commented 2 years ago

Hey there, I was just trying to run the decoding model while it says I have an error about

OSError: We couldn't connect to 'https://huggingface.co/' to load this model and it looks like predictability/diff_models/e2e-tgt_e=15_b=20_m=gpt2_wikitext-103-raw-v1_101_None is not the path to a directory conaining a config.json file.

Just 10 hours ago I was able to successfully load the model from hugging-face, I wonder if this is my personal problem?

XiangLi1999 commented 2 years ago

no... it's not your problem. I didn't post the evaluation model on hugging face hub. But you can easily train your own AR model for evaluation.

XiangLi1999 commented 2 years ago

I think the fine-tuning code is also available in the git repo.

lipingtang17 commented 2 years ago

Hi Lisa, I met the same problem. Could you please upload your evaluation model? Thanks!

YangYu-NUAA commented 2 years ago

model_name_or_path='predictability/diff_models/compress_e=5_b=60_m=gpt2_wikitext-103-raw-v1_None

Sorry for this quetion, but i really want to know the function of this evalution model so i can train a similar model to replace it. Now i'm confusing about this pretrained model.

aatmikg commented 1 year ago

Hi Lisa, would it be possible to share any information on the hyperparameters used to train the AR models / any instructions so we can replicate your training process? You noted that fine-tuning code is included in the repo, if so, could you point us to where we might find it? I tried to look but couldn't figure it out.

Any help would be greatly appreciated, thanks!