35 lines
1.3 KiB
Markdown
35 lines
1.3 KiB
Markdown
# Bugs
|
|
[ ] Sometime I can redo or edit but the button is not available
|
|
[X] The SSE event that sometime fails after some times. Reconnect when sending a new message.
|
|
[X] 2 selected messages
|
|
[X] On first response, code block width are too long
|
|
[X] Change Terms of service to say that I use one cookie
|
|
[X] Change the lastSelectedLLMs, 2 users can't use the same time...
|
|
[X] CTRL + Enter not working
|
|
[ ] Add all Together AI models
|
|
|
|
# Features
|
|
[X] SSE to WebSocket
|
|
[X] Add max tokens
|
|
[X] Errors messages to know what happend
|
|
[X] Check the response status and add a message accordingly (Like 402 Payment required)
|
|
[X] Add Deepseek API
|
|
[X] Add TogetherAI API
|
|
[X] Add Nvidia NIM
|
|
[ ] Add login with email and password
|
|
[ ] Add login with AppleID
|
|
[ ] Better temperature settings. (Anthropic from 0-1, OpenAI from 0-2, DeepSeek from -2-2, ect)
|
|
[ ] Auto update the modelsInfos db (At least remove unavailable models and send me an email when new one arrive)
|
|
[ ] Encrypt the API keys
|
|
|
|
# Opti
|
|
[ ] Import all company and modelInfos at the start to not request the db
|
|
[ ] Reduce db query
|
|
[ ] Look to reduce memory
|
|
|
|
# Other
|
|
[ ] Remove all panic
|
|
[ ] Change the terms of service and enter keys page to an HTML
|
|
[X] Use the normal RequestProvider function instead of TestProvider to remove TestProvider
|
|
[ ] Implement the JADE_IN_DEV env variable to dev locally on my laptop
|