github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 17:49:24 -05:00
Code Issues 14.6k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #2833] Running ollama on Hugging Face Spaces #1722

New Issue
Closed
opened 2026-04-12 11:41:59 -05:00 by GiteaMirror · 7 comments
GiteaMirror commented 2026-04-12 11:41:59 -05:00
Owner
Copy Link

Originally created by @jbdatascience on GitHub (Feb 29, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/2833

I want to run ollama on Hugging Face Spaces, because I run a Streamlit app there that must make use of a LLM and a embedding model served by Ollama. How can I do that?

Originally created by @jbdatascience on GitHub (Feb 29, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/2833 I want to run ollama on Hugging Face Spaces, because I run a Streamlit app there that must make use of a LLM and a embedding model served by Ollama. How can I do that?
GiteaMirror commented 2026-04-12 11:42:00 -05:00
Author
Owner
Copy Link

@Fastidious commented on GitHub (Feb 29, 2024):

What do you mean "run ollama on Hugging Face Spaces"? Do you want to consume models available in Hugging Face Spaces (HFS)? Is that your aim? If so, then that's something not available out of the box in Ollama.

If you search The Tubes, you will find a few approaches to do so. The easiest way to use HFS models is by transforming models using GPT-Generated Unified Format (gguf). Not every model on HFS is a gguf model.

<!-- gh-comment-id:1971873682 --> @Fastidious commented on GitHub (Feb 29, 2024): What do you mean "run ollama on Hugging Face Spaces"? Do you want to consume models available in Hugging Face Spaces (HFS)? Is that your aim? If so, then that's something not available out of the box in Ollama. If you search The Tubes, you will find a few approaches to do so. The easiest way to use HFS models is by transforming models using GPT-Generated Unified Format (gguf). Not every model on HFS is a gguf model.
GiteaMirror commented 2026-04-12 11:42:00 -05:00
Author
Owner
Copy Link

@jbdatascience commented on GitHub (Mar 1, 2024):

What do you mean "run ollama on Hugging Face Spaces"? Do you want to consume models available in Hugging Face Spaces (HFS)? Is that your aim? If so, then that's something not available out of the box in Ollama.

If you search The Tubes, you will find a few approaches to do so. The easiest way to use HFS models is by transforming models using GPT-Generated Unified Format (gguf). Not every model on HFS is a gguf model.

I gues you are right. Also I came to the conclusion that I oversaw the option of not using ollama in the first place! Just using a LLM model and an embedding model right from HF is the natural (and obvious!) way to go!

My plan now is to modify the existing code to implement this.

thanks for your response!

<!-- gh-comment-id:1972695258 --> @jbdatascience commented on GitHub (Mar 1, 2024): > What do you mean "run ollama on Hugging Face Spaces"? Do you want to consume models available in Hugging Face Spaces (HFS)? Is that your aim? If so, then that's something not available out of the box in Ollama. > > If you search The Tubes, you will find a few approaches to do so. The easiest way to use HFS models is by transforming models using GPT-Generated Unified Format (gguf). Not every model on HFS is a gguf model. I gues you are right. Also I came to the conclusion that I oversaw the option of not using ollama in the first place! Just using a LLM model and an embedding model right from HF is the natural (and obvious!) way to go! My plan now is to modify the existing code to implement this. thanks for your response!
GiteaMirror commented 2026-04-12 11:42:01 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (May 17, 2024):

Hey @jbdatascience I think it's OK to close this? I think you could do it through Docker Spaces, but I'm not sure if that's the best way to do it as I'm not really familiar w/ Spaces. There are lots of other places to get GPU instances on the cloud where you can run Ollama.

<!-- gh-comment-id:2118469992 --> @pdevine commented on GitHub (May 17, 2024): Hey @jbdatascience I think it's OK to close this? I think you could do it through Docker Spaces, but I'm not sure if that's the best way to do it as I'm not really familiar w/ Spaces. There are lots of other places to get GPU instances on the cloud where you can run Ollama.
GiteaMirror commented 2026-04-12 11:42:01 -05:00
Author
Owner
Copy Link

@jbdatascience commented on GitHub (Jun 14, 2024):

Hey @jbdatascience I think it's OK to close this? I think you could do it through Docker Spaces, but I'm not sure if that's the best way to do it as I'm not really familiar w/ Spaces. There are lots of other places to get GPU instances on the cloud where you can run Ollama.

I am still searching for a solution!

<!-- gh-comment-id:2167984771 --> @jbdatascience commented on GitHub (Jun 14, 2024): > Hey @jbdatascience I think it's OK to close this? I think you could do it through Docker Spaces, but I'm not sure if that's the best way to do it as I'm not really familiar w/ Spaces. There are lots of other places to get GPU instances on the cloud where you can run Ollama. I am still searching for a solution!
GiteaMirror commented 2026-04-12 11:42:02 -05:00
Author
Owner
Copy Link

@musarehmani291 commented on GitHub (Jun 14, 2024):

Hi @Fastidious and @pdevine
I wanna deploy Ollama to Huggingface space. I used the official Dockerfile of the Ollama repo but it keeps building I could never get the space to running. I'm using the free version of Huggingface space. could you look into this issue?

<!-- gh-comment-id:2168411889 --> @musarehmani291 commented on GitHub (Jun 14, 2024): Hi @Fastidious and @pdevine I wanna deploy Ollama to Huggingface space. I used the official Dockerfile of the Ollama repo but it keeps building I could never get the space to running. I'm using the free version of Huggingface space. could you look into this issue?
GiteaMirror commented 2026-04-12 11:42:02 -05:00
Author
Owner
Copy Link

@Amirjab21 commented on GitHub (Jun 24, 2024):

What do you mean "run ollama on Hugging Face Spaces"? Do you want to consume models available in Hugging Face Spaces (HFS)? Is that your aim? If so, then that's something not available out of the box in Ollama.
If you search The Tubes, you will find a few approaches to do so. The easiest way to use HFS models is by transforming models using GPT-Generated Unified Format (gguf). Not every model on HFS is a gguf model.

I gues you are right. Also I came to the conclusion that I oversaw the option of not using ollama in the first place! Just using a LLM model and an embedding model right from HF is the natural (and obvious!) way to go!

My plan now is to modify the existing code to implement this.

thanks for your response!

Did you manage to do this?

<!-- gh-comment-id:2186919372 --> @Amirjab21 commented on GitHub (Jun 24, 2024): > > What do you mean "run ollama on Hugging Face Spaces"? Do you want to consume models available in Hugging Face Spaces (HFS)? Is that your aim? If so, then that's something not available out of the box in Ollama. > > If you search The Tubes, you will find a few approaches to do so. The easiest way to use HFS models is by transforming models using GPT-Generated Unified Format (gguf). Not every model on HFS is a gguf model. > > I gues you are right. Also I came to the conclusion that I oversaw the option of not using ollama in the first place! Just using a LLM model and an embedding model right from HF is the natural (and obvious!) way to go! > > My plan now is to modify the existing code to implement this. > > thanks for your response! Did you manage to do this?
GiteaMirror commented 2026-04-12 11:42:03 -05:00
Author
Owner
Copy Link

@myyim commented on GitHub (Jun 21, 2025):

This solution works for me:
https://medium.com/p/1f5d8f871887

If you use an embedding model, when deploying the model on the client end, use "embed" instead of "generate".

For example,

curl http://localhost:11434/api/embed -d '{
"model": "all-minilm",
"input": "Why is the sky blue?"
}'

<!-- gh-comment-id:2993670172 --> @myyim commented on GitHub (Jun 21, 2025): This solution works for me: https://medium.com/p/1f5d8f871887 If you use an embedding model, when deploying the model on the client end, use "embed" instead of "generate". For example, curl http://localhost:11434/api/embed -d '{ "model": "all-minilm", "input": "Why is the sky blue?" }'
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 Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#1722
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?