Add Sagemaker support (#401)

Revert "support rate limiting based on user IP (#342)"

Reopen: Feature/disable login for n messages (#356)

support rate limiting based on user IP (#342)

Revert "Option to disable login on first N messages (#352)"

Option to disable login on first N messages (#352)

Add optional timestamps to messages (#294)

Add support for websearch retries (#318)

very basic rate limiter (#320)

Added Serper.dev API as an alternative web search provider (#302)

Refactor of websearch (#281)

Add initial support for web browsing (#237)

🐛 Vote not working with new messages (#249)

✨ add thumb up/down voting system for messages (#152)

🐛 Fix "signin with HF" within space + CSRF (#236)

Login flow (#193)

Disable models (#187)

✨ Backend preparation for "Sign in with HF" (#190)

fix chat staying pending when there is a server error (#186)

⏪

✨ Add model id to be able to switch models while keeping conversations valid (#181)

🩹 Only trim end of string & add "Assistant:" to stop words (#180)

✨ Ability to download prompt (#176)

make sure we allow user to change model if clicking on a link with mo… (#172)

🔧 Always strip <|endoftext|> (#171)

🐛 Stop bigcode prompts (#170)

Models selector (#164)

🗃️ Store model inside conversation instead (#165)

♻️ Simplify setup, prepare for multi-models (#156)

👷 Add CI to check types (#161)

Polyfill safari 14 (#149)

⬆️ @huggingface/inference 2.2.0

🐛 Make it work with default model

🥅 More helpful share error message (#148)

🥅 Different errors (#144)

✨ Retry messages (#121)

🚨 Run prettier

✨ Title edit (#135)

🔧 Change separator token to </s> (#127)

✨ Add preprompt (#124)

✨ Replace title by conversation title (#118)

Stop generation button (closes #86) (#88)

Fix share link when already created (#110)

⚡️ Load balance endpoints (#106)

✨ Switch to pure HF inference (#95)

Toast errors (#91)

🩹 Do not flash endoftext (#89)

🐛 Fix public_origin

⚡️ Limit the number of tokens sent to the backend (#93)

✨ Link to hf.co/chat from logo (#94)