Josh Leverette
Josh Leverette
**Is your feature request related to a problem? Please describe.** Depending on screen size, there are 2 to 4 suggested prompts shown above the chatbox when starting a new conversation....
I believe the template being used for Phi-2 is incorrect. Here is an example conversation: ``` ollama run phi >>> What is the LHC? The Large Hadron Collider (LHC) is...
Mixtral OOM
I’ve been enjoying the new auto-VRAM implementation for the most part, but when trying to use Mixtral at very large context sizes (~30000) to process a 25k token document, I’m...
I’m not sure what all would be involved, but something that’s making waves is “self extend”, where it seems to be possible to make models work at larger context sizes...
**Describe the solution you'd like** There is a new code completion model available: - https://www.latent.space/p/reza-shabani - https://huggingface.co/replit/replit-code-v1-3b I have been looking for ways to try it out in a useful...
When the User and Group columns have a different value on one or more files, it would be nice to have the group column automatically printed. This would make it...
https://largeworldmodel.github.io https://huggingface.co/LargeWorldModel Seemingly a new state of the art MLLM that can also handle very large context sizes, and videos (not just images).
### What model would you like? This model looks promising: https://stability.ai/news/introducing-stable-lm-2-12b
Hello, It looks like you've invested several hundred commits into this repository, which is impressive. Can you state what benefits this version of FreeRouting has over the original? The ReadMe...
Now that SIMD has reached stable Rust, it would be great if the SIMD acceleration could be moved out of the `unstable` feature!