Open mattly opened 6 months ago
Experienced the same issue from a fresh pipx install. Tracked down a workaround in #444 and got to a solution with this fix in https://github.com/simonw/llm-gpt4all/issues/23.
Thanks to @b421 for the noted quick fix and PR over there.
My file was located at ~/.local/pipx/venvs/llm/lib/python3.12/site-packages/llm_gpt4all.py because I used pipx, but your path will be /opt/homebrew/Cellar/llm/0.13.1/libexec/lib/python3.12/site-packages/llm_gpt4all.py as shown in your output.
def is_installed(self):
try:
GPT4All.retrieve_model(
self._details["filename"], allow_download=False, verbose=False
)
return True
- except ValueError:
+ except (FileNotFoundError, ValueError):
return False
+1, the except (FileNotFoundError, ValueError):
solved this for me. Thank you!
This error hosed my whole llm installation, but I was able to recover it by uninstalling the offending models:
$ llm uninstall llm-gpt4all
I ran
brew intstall llm
on a fresh machine. Then ranllm instal gpt4all
per the instructions in the "Other Models" section of the docs. When runningllm models
I got the following output: