search

ohmplatform / FreedomGPT

This codebase is for a React and Electron-based app that executes the FreedomGPT LLM locally (offline and private) on Mac and Windows using a chat-based interface
http://www.freedomgpt.com
GNU General Public License v3.0
2.61k stars 358 forks source link

zsh: exec format error #81

Closed ethicalhacker7192 closed 1 year ago

ethicalhacker7192 commented 1 year ago

I get this error message when I pick llama model and alpaca model (including full and speed for both models):

Child process exited with code 2 and signal null

This is run on kali linux, also when I execute the bin file, I get this error which suggests that I am having compatibility issues:

please note that this is the same error for both, just a different file name.

zsh: exec format error: ./ggml-model-q4_0.bin
ethicalhacker7192 commented 1 year ago

I got past this error but now I have a new error:

error loading model: missing tok_embedding.weight

this is after building llama.cpp using the make command, and then running ./main -i --color to test if it was anything else. I believe that the weights are missing because on second glance I saw that there was no tok_embedding.weight file inside anywhere, can you please share it with me or troubleshooting?

ethicalhacker7192 commented 1 year ago

I found the reason why, it was because the model I was using needed weights, I had to install another model (LlaMa model) and rename it to the other one, I renamed the other one to (ggml-vocab.bin) and then I renamed it to the appropriate name.