frob

The log shows it was loading until the client was terminated.

Run htop and watch the memory usage while running the following: ``` echo 3 | sudo tee /proc/sys/vm/drop_caches ```

``` ps axho %mem,cmd | sort -n ```

https://github.com/ollama/ollama/issues/6565

Related: https://github.com/ollama/ollama/issues/7043

[Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) might give some insight.

ollama currently doesn't generate statistics for a load operation. ```console $ curl -s localhost:11434/api/generate -d '{"model":"qwen2.5:0.5b"}' | jq { "model": "qwen2.5:0.5b", "created_at": "2025-02-21T16:24:24.000933491Z", "response": "", "done": true, "done_reason": "load" }...

[Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) will aid in debugging, particularly the line that contains `OLLAMA_MODELS`. It's possible that another server is running, so the end of the server log showing recent activity to...

[Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) will aid in debugging.

https://github.com/ollama/ollama/issues/3368