Jared Van Bortel
Jared Van Bortel
gpt4all-api has been removed, see #2314.
gpt4all-api has been removed, see #2314.
Could you describe in more detail what you are seeing? Is the GUI not opening? Does it crash or exit with a nonzero status? It is possible that our Linux...
> `No marker found, stopped after 1.00 MB.` This is the important part. Perhaps your download is corrupt/incomplete?
gpt4all-api has been removed, see #2314.
@m-roberts What is your use case for forcing metal? Metal is used whenever the model would fit into the memory that Metal is allowed to allocate. I think forcing metal...
Like I said, Metal is used by default. Forcing metal is only necessary if you want to attempt to use more than 53% of your system RAM with GPT4All. You...
> I want to use more than 53% of the RAM. How big of a model are you actually able to load with this option enabled? As far as I...
No response for over a month. Assuming that this isn't a legitimate need. If you still think it is, feel free to reopen and provide more specific information.
We don't have official builds for aarch64 Linux, so for typical Android phones you would have to build from source. llama.cpp itself has aarch64 support, and I don't know of...