Enhance model retrieval logic in read_evals.py to support dict-based model_args and improve org_and_model extraction; update .gitignore to exclude eval-results files.

try2

now with a functionning backend

added selector for model type

merge refactor

Add GPT-4 & human eval tab

update gitignore

display message after eval submission

adds revision option

creates leaderboard