[GH-ISSUE #10058] ollama run bge-m3 not support generate #53107

Closed
opened 2026-04-29 01:58:49 -05:00 by GiteaMirror · 2 comments
Owner

Originally created by @Tu1231 on GitHub (Mar 31, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10058

What is the issue?

When I execute docker run - d - v ollama:/ root/.ollama-embedding -p 9000:11434 --name ollama ollama/ollama, Then when I entered the container and executed the ollama pull bge-m3, the ollama list displayed bge-m3: latest, Then when I rolled up BGE-M3 and rolled up BGE-M3: latest, it showed not supporting generate

Image

Relevant log output


OS

Docker

GPU

No response

CPU

Intel

Ollama version

0.5.13

Originally created by @Tu1231 on GitHub (Mar 31, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10058 ### What is the issue? When I execute docker run - d - v ollama:/ root/.ollama-embedding -p 9000:11434 --name ollama ollama/ollama, Then when I entered the container and executed the ollama pull bge-m3, the ollama list displayed bge-m3: latest, Then when I rolled up BGE-M3 and rolled up BGE-M3: latest, it showed not supporting generate ![Image](https://github.com/user-attachments/assets/8b9827e2-3c1b-4bcd-bda6-0f7bab19244c) ### Relevant log output ```shell ``` ### OS Docker ### GPU _No response_ ### CPU Intel ### Ollama version 0.5.13
GiteaMirror added the bug label 2026-04-29 01:58:49 -05:00
Author
Owner

@rick-github commented on GitHub (Mar 31, 2025):

bge-m3 is an embedding model, not a text generation model. You have to use the /api/embed endpoint to get results. ollama cli doesn't use the /api/embed endpoint.

$ curl localhost:11434/api/embed -d '{"model":"bge-m3","input":"why is the sky blue?"}'
{
  "model": "bge-m3",
  "embeddings":[[-0.045661133,0.02446292, ... 0.04149892,-0.034263242]],
  "total_duration": 428879280,
  "load_duration": 382324803,
  "prompt_eval_count": 6
}

<!-- gh-comment-id:2765490748 --> @rick-github commented on GitHub (Mar 31, 2025): bge-m3 is an embedding model, not a text generation model. You have to use the `/api/embed` endpoint to get results. ollama cli doesn't use the `/api/embed` endpoint. ```console $ curl localhost:11434/api/embed -d '{"model":"bge-m3","input":"why is the sky blue?"}' { "model": "bge-m3", "embeddings":[[-0.045661133,0.02446292, ... 0.04149892,-0.034263242]], "total_duration": 428879280, "load_duration": 382324803, "prompt_eval_count": 6 } ```
Author
Owner

@sieveLau commented on GitHub (Mar 31, 2025):

Please learn about what embedding models are and what for: https://huggingface.co/blog/getting-started-with-embeddings

<!-- gh-comment-id:2766103132 --> @sieveLau commented on GitHub (Mar 31, 2025): Please learn about what embedding models are and what for: https://huggingface.co/blog/getting-started-with-embeddings
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#53107