Tianshu Wang

@gkowzan Well done! Using org-protocol is really convenient!

@proofit404 Hi, the same issue here but in tramp mode. However I see there is an `~/.emacs.d/anaconda-mode/0.1.13` argument needed, should I install emacs on my server first? Updated: I just...

same issue here.

sorry, I try to delete the `anacoda-mode` folder and everything works fine now.

For discovering whether a remote tramp host has python, we can just use `(executable-find "python" t)`(t for remote). However, I don't know how to start a server on remote and...

> If this is going to be sequenced so the properties are introduced first, and then used. I would like that you open the follow-up PR already so that we...

This issue was original reported at https://github.com/huggingface/transformers/issues/14931 and It seems like this issue also occur with other AutoClass like AutoFeatureExtractor.

It should be noted that this problem also occurs in other AutoClasses, such as AutoFeatureExtractor, so I don't think handling it in Datasets is a long-term practice either.

> I think the best course of action is creating an issue in `transformers`. I can start the work on this. @Narsil Hi, I reopen this issue in `transformers` https://github.com/huggingface/transformers/issues/14931

Here is @Narsil comment from https://github.com/huggingface/transformers/issues/14931#issuecomment-1074981569 > # TL;DR > Call the function once on a dummy example beforehand will fix it. > > ```python > tokenizer("Some", "test", truncation=True) >...