jan
jan copied to clipboard
Jan is an open source alternative to ChatGPT that runs 100% offline on your computer. Multiple engine support (llama.cpp, TensorRT-LLM)
## Describe Your Changes - Adds new Llama 3 models 8B and 80B for the Groq inference extension - Adds Gemma 7B It model for Groq inference extension
## Description It's challenging to scale the current vanilla Fastify Node.js backend. It's preferable to migrate to a scalable framework like NestJS. Fastify is much faster than Express, achieving almost...
## Describe Your Changes - Add transformMessage to core - Add transformResponse to core - Add Cohere support ## Fixes Issues - Closes # - Closes # ## Self Checklist...
## Description: All current inference engines primarily function within the browser host. We now require them to be refactored to operate on the backend. This adjustment will enable UI components...
## Description We will introduce Extensions as part of the Cortex Backend, allowing us to migrate our current remote inference engine extensions into Cortex. Some questions about how it should...
## Description Refactor the Jan business/backend logic to Cortex, a headless backend platform equivalent to OpenAI. The UI components (Desktop/Mobile/Web) should seamlessly integrate with Cortex endpoints, ensuring that server logics...
**Describe the bug:** When the model fails to load, the local API server starts, but the "API Reference" button is missing from the interface. **Steps to reproduce:** 1. Open local...
**Describe the bug** When users clear their logs using the settings in Jan, the application stops generating any new logs until the user restarts the Jan app. **Steps to reproduce**...
**Problem** Current API models (OpenAI, Groq, etc) missing some control parameters. **Success Criteria** Display all parameters like local model. **Additional context** - [ ] Have all of the need parameters...
A feature for favorite conversations is in the left panel - a dedicated tab to display these favorites for quick access. Implementation suggestions: - Add a star icon next to...