Update app.py

Minor tweaks to the tensors loading

Fix the requirements versions

Allow using GPU

Add the new paper

Fix bug in leaderboard rendering

Split the method name

Tiny black edit

Update README.md

Add contact information

Enforce wide view by default

Tweak the description

Rename the leaderboard

Tweak the newly added evaluation method

Add a TODO message

Add a new evaluation function

Fix the docstring

Reorder the leaderboard columns

Fix a bug

Fix a bug

Fix the list for custom model

Show the description of the inference methods

Fix the models' URLs

Tweaks for the leaderboard

Login to the Hub first

Add a prompting-based method

Update the methods documentation rendering

Handle model evaluations staled

Apply black

Fix another bug in the results handling

Fix another tiny bug

Only show the results if they exist

Skip the results of incomplete evaluations

Make sure all the fields in the prediction json files are uploaded

Fix the commit_id bug

Implement a basic version of a queuing system

Update the model inference logging message

Log progress in percentages

Report the number of samples in the test set

Tweak the information displayed on the leaderboard

Update the leaderboard filename

Run the evaluation in background

Test using a subprocess

Tweak the progress bar

Move the citation to the outer level

Some stylistic improvements

Update the docstring for the top_p method

Fix the bug of evaluating the model twice

More descriptive text for showing progress

Do not parse the results unless they exist