Repository navigation
Welcome to llmfit Discussions! #460
Replies: 5 comments 4 replies
|
Hello, just getting started with limited hardware. Found a good fit model but cannot figure out how to launch it after using the built in "d" downloading option? I used ollama to download it, but ollama run cannot run it so not sure where the file actually went? On ubuntu |
|
Hey! No worries, this is a common point of confusion. When you press d in the TUI, it first asks which provider to download to (Ollama, llama.cpp, MLX, LM Studio, etc.) — the model goes wherever you chose, not a single universal location. A couple things to check: Run ollama list — this shows every model Ollama knows about. If the download succeeded via Ollama, it'll be here. The model name in Ollama might differ from the HuggingFace name (llmfit uses the Ollama tag format). If you selected llama.cpp instead of Ollama, the GGUF file goes to the llama.cpp cache directory (typically ~/.cache/llama.cpp/ on Linux). You'd run it with llama-cli -m or llama-server -m instead of ollama run. Press D in the TUI to open the Download Manager — it shows your download history and which provider each model went to. If ollama list is empty or the model isn't there, try running llmfit again, select the model, press d, and this time make sure to pick Ollama when prompted for the provider. Happy to help troubleshoot further if you share the model name you chose! |
|
Hi, I just stumbled upon this gem, a big thank you to everyone 🎉 and looking forward to contributing too! |
|
Welcome! 🎉 Great to have you here. The community's been growing fast. Feel free to jump into discussions, ask questions, or dig into issues — contributions are always welcome. If you're new, check out the repo's issues for good first-timer opportunities. Happy to answer any questions you have! |
|
Hello @AlexsJones, I was working on a pretty similar tool made by my own : lokai. I'm looking forward to join the community as a contributor. Is there any code of conduct or contribuiting guidelines available ? Thx in advance ! 🐫 |
|
Hey @Romeo-mz — thanks for the nudge! Your question motivated me to put together proper contributor docs. I've just pushed a branch with a https://github.com/AlexsJones/llmfit/tree/docs/contributing-guide Would love any feedback before it gets merged. And welcome aboard — lokAI looks cool, excited to have you contributing here! |
|
I'v reviewed your branch https://github.com/AlexsJones/llmfit/tree/docs/contributing-guide and decide it will be interesting to use devbox to ease the maintainer roles; here's the PR associated with your's and my changes : #561 Will soon open some issue about templating also ! 🐫 |
|
Hi @mcunkelman 👋 — the in-TUI The reason ollama listand then If you'd rather have a raw GGUF file you control, use the CLI downloader instead: llmfit download <model> # saves to ~/.cache/llmfit/models (override with --output-dir)
llmfit run <model> # launches it with llama-cli
llmfit run <model> --server # OpenAI-compatible API server( |
|
Welcome @Romeo-mz, and cool to hear about lokai! 🐫 Yes — both are in the repo root:
A nicely-scoped first contribution that's open right now is a RamaLama detection provider (see #634) — small and self-contained, modeled on the existing vLLM one. Happy to point you at the relevant files if you're interested. |
Uh oh!
There was an error while loading. Please reload this page.
👋 Welcome!
We’re using Discussions as a place to connect with other members of our community. We hope that you:
build together 💪.
To get started, comment below with an introduction of yourself and tell us about what you do with this community.
All reactions