Remote connection fails with an error message: Error: model 'mistral' not found, try pulling it first

[goshng]

My ollama.nvim setup in ubuntu works fine. Then, I want to use ollama in ubuntu from a mac. In the config of ollama.nvim,

  opts = {
    model = "mistral",
    url = "http://127.0.0.1:11434",
    ...

is edited so

url = "http://xxx.xxx.xx.xxx:11434",

where xxx.xxx.xx.xxx is the IP address of ubuntu.

Then, I did setup ollama.nvim in the mac and test a sample command of ollama.nvim, getting an error message:

Error: model 'mistral' not found, try pulling it first

In the server side at ubuntu, I ran the following command, which followed status message.

(ubuntu) $ OLLAMA_HOST=xxx.xxx.xx.xxx ollama serve
2024/01/12 15:28:28 images.go:834: total blobs: 0
2024/01/12 15:28:28 images.go:841: total unused blobs removed: 0
2024/01/12 15:28:28 routes.go:929: Listening on xxx.xxx.xx.xxx:11434 (version 0.1.18)
2024/01/12 15:28:29 shim_ext_server.go:142: Dynamic LLM variants [cuda rocm]
2024/01/12 15:28:29 gpu.go:34: Detecting GPU type
2024/01/12 15:28:29 gpu.go:53: Nvidia GPU detected
[GIN] 2024/01/12 - 15:30:05 | 404 |     472.458µs |  xxx.xxx.xx.xxx | POST     "/api/generate"
[GIN] 2024/01/12 - 15:33:14 | 404 |     262.088µs |  xxx.xxx.xx.xxx | POST     "/api/generate"

So, linux - mac setup does not work.

I went back to linux and only changed the url line above of ollama.nvim to get the same error, or I pretend that linux - linux connection even though it worked fine locally.

Any idea what might go wrong?

Thank you!

[traverseda]

Have you tried pulling the mistral model?

ollama serve

Then, in another terminal,

ollama pull mistral

This project doesn't deal with installing models or setting up ollama

[nomnivore]

Hi @goshng,

Has the above solution solved your issue?