filed to get any model with ollama +deepseek-r1

[hgcdanniel]

Git provider

Github Cloud

System Info

have difficult in run deepseek-r1 +ollama

Bug details

when I use ollama
setting configures like;
model="ollama/deepseek-r1"
fallback_models=["ollama/deepseek-r1"]
but I meet a bug failed to run any model with model "ollama/deepseek-r1"
while I have corrected start ollama server

[mrT23]

Try to debug from CLI.
My guess is that you need to add the model to here:
https://github.com/qodo-ai/pr-agent/blob/main/pr_agent/algo/__init__.py#L87
And it will support only the /ask feature

In any case, review also this:
https://qodo-merge-docs.qodo.ai/usage-guide/changing_a_model/#ollama

[hgcdanniel]

Thank you very much, now I can run deepseek-r1 successfully, but could you support /improve /describe and other functions?

[mrT23]

These commands are supported with qodo merge. Welcome to try them out there

https://qodo-merge-docs.qodo.ai/usage-guide/PR_agent_pro_models/#selecting-a-specific-model
#1378 (comment)

[hgcdanniel]

thank u very much
what's more, could u support https://github.com/kvcache-ai/ktransformers
which can run deepseek-r1 671b 4-bit

[mrT23]

What i shared is what we going to support currently in qodo merge

[DangHung301]

@mrT23 thank you very much, I have a the same problem as a @hgcdanniel but thanks to you I solved it.