ollama client uses the keyword 'input' for embeddings, but Ollama 0.11.4 requires 'prompt' ?
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 10.5k
- Forks
- 1.2k
- Avg merge
- 4m
- Merged PRs (30d)
- 1
Description
Could it be?
I'm using ollama client 0.5.3 and latest ollama (0.11.4) for embeddings.
But the client uses the keyword 'input' that's not recognized by ollama, that returns always empty vectors.
While changing the keyword 'input' with 'prompt' make it works...
For example, testing against Ollama 0.11.4:
Invoke-RestMethod http://127.0.0.1:11435/api/embeddings -Method Post -Body (@{ model = "mxbai-embed-large:latest" input = "This is a test" } | ConvertTo-Json) -ContentType "application/json"
returns empty vectors, while using 'prompt' makes it work.
The client 0.5.3 on the other side, sends packets with the 'input keyword.
As a test I made a shim, intercepting the calls, substituting 'input' with 'prompt' and it works.
Cannot understand if the problem arised with ollama server or client...
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the Python client's embeddings request and compare its payload with Ollama 0.11.4's /api/embeddings behavior, using the PowerShell reproduction in the issue. Confirm which field the server expects and whether the client and server versions are compatible; done when a supported call returns non-empty vectors without a shim.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- ollama, python
- Domain
- api, machine-learning
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100