nomic-ai / gpt4all

GPT4All: Run Local LLMs on Any Device. Open-source and available for commercial use.
https://nomic.ai/gpt4all
MIT License
70.09k stars 7.66k forks source link

After update to v3.1.0 it close upon loading any model #2734

Open lakhwinder108 opened 2 months ago

lakhwinder108 commented 2 months ago

Bug Report

GPT4ALL was working well before the recent update. Today I update to v3.1.0. After that when I load a model it instead of loading the model.

Steps to Reproduce

Open gpt4all, and load any model Llama 3 8b, or any other model.

Expected Behavior

It disappears when I click to load the default model. And closed.

Your Environment

cosmic-snow commented 2 months ago

Can you try the following:

lakhwinder108 commented 2 months ago

Issue fixed using C:\Users\\AppData\Roaming\nomic.ai\GPT4All.ini and set device=CPU in the [General] section. I have Nvidia graphics also, But now it's too slow. Approx 1 token per sec. Previously it was 2 tokens per second. Slow but working well.

cosmic-snow commented 2 months ago

Hm, not sure why it would crash there.

Where did you get your Llama 3? From the official download list? I haven't tried it myself.

Edit: Oh also, can you add details for your GPU? Type and drivers. Did you have CUDA or Vulkan as default before that?

RoterMilan1964 commented 2 months ago

Hallo, ist bei mir das Gleiche.. Selbst wenn ich dann das neuere Modell deinstalliere und die ein älteres Modell laufen passiert das Gleiche. (Die sind aber vorher alle wunderbar gelaufen.) Viele Grüße RoterMilan1964

lakhwinder108 commented 2 months ago

Download from the official download list inside gpt4all. Graphic NVIDIA GTX, i7.

cosmic-snow commented 2 months ago

Ok, and are your drivers up-to-date?

Did you set the device option to a specific value? CUDA or Vulkan?

You basically have 3 ways to run something with an NVIDIA graphics card: CPU, CUDA and Vulkan. They all do the processing in a different way, so having that information is important. (And having it on auto picks one for you.)