github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-11-12 03:46:14 -06:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #713] Using ollama with llm-ls #78004

New Issue
Closed
opened 2026-05-08 20:51:59 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-05-08 20:51:59 -05:00
Owner
Copy Link
Copy Source

Originally created by @noahbald on GitHub (Oct 5, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/713

I've been trying to setup ollama to use codellama with FIM in my editor with nvim.llm and llm-ls.
As suggested in the ollama docs, this is what the locally running API may expect as a FIM request.

curl -X POST http://localhost:11434/api/generate -d '{
  "model": "codellama:7b-code",
  "prompt": "<PRE> def compute_gcd(x, y): <SUF>return result <MID>"
}'

However, when using llm-ls we can't (afaik) tell it how to structure the data it sends. It ends up sending a request that looks more like this

curl -X POST http://localhost:11434/api/generate -d '{
  "inputs": "<PRE> def compute_gcd(x, y): <SUF>return result <MID>",
  "parameters": { ... },
}'

Has anyone found a better way to set this up? Is there an alternative api endpoint we can use? Something like /api/generate/codellama:7b-code?src=llm-ls?

Originally created by @noahbald on GitHub (Oct 5, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/713 I've been trying to setup ollama to use codellama with FIM in my editor with nvim.llm and llm-ls. As suggested in the ollama docs, this is what the locally running API may expect as a FIM request. ```sh curl -X POST http://localhost:11434/api/generate -d '{ "model": "codellama:7b-code", "prompt": "<PRE> def compute_gcd(x, y): <SUF>return result <MID>" }' ``` However, when using llm-ls we can't (afaik) tell it how to structure the data it sends. It ends up sending a request that looks more like this ```sh curl -X POST http://localhost:11434/api/generate -d '{ "inputs": "<PRE> def compute_gcd(x, y): <SUF>return result <MID>", "parameters": { ... }, }' ``` Has anyone found a better way to set this up? Is there an alternative api endpoint we can use? Something like `/api/generate/codellama:7b-code?src=llm-ls`?
GiteaMirror commented 2026-05-08 20:52:10 -05:00
Author
Owner
Copy Link
Copy Source

@noahbald commented on GitHub (Oct 6, 2023):

Alternatively, llm-ls could support various backends?
https://github.com/huggingface/llm-ls/issues/17

<!-- gh-comment-id:1750284937 --> @noahbald commented on GitHub (Oct 6, 2023): Alternatively, llm-ls could support various backends? https://github.com/huggingface/llm-ls/issues/17
GiteaMirror commented 2026-05-08 20:52:14 -05:00
Author
Owner
Copy Link
Copy Source

@mxyng commented on GitHub (Oct 25, 2023):

Thanks for creating this. I wasn't aware of llm-ls. I personally use copilot and created a proxy for intercepting and rerouting copilot requests to ollama requests. You can find it here.

As for implementing this in ollama, we don't normally integrate with external projects and instead rely on the integration to implement compatibility. There are so many integrations so this approach will be unsustainable.

<!-- gh-comment-id:1780088924 --> @mxyng commented on GitHub (Oct 25, 2023): Thanks for creating this. I wasn't aware of llm-ls. I personally use copilot and created a proxy for intercepting and rerouting copilot requests to ollama requests. You can find it [here](https://github.com/mxyng/proxypilot). As for implementing this in ollama, we don't normally integrate with external projects and instead rely on the integration to implement compatibility. There are so many integrations so this approach will be unsustainable.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No labels
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
GiteaMirror ninjasurge (Kaehlen Rog)
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#78004