Open Cooly75 opened 11 months ago
What does "the chatgpt files for offline usage" mean? Since it uses OpenAI API, it currently does not support offline. Does that files mean model files like distributed by HuggingFace?
i do use stable diffusion on a offline computer, therefor any application who uses online ressources to work are in my case worthless, on the otherhand i do not like big tech who collects uncontrolled any data, unfortunate i do not use your tool, the idea is great, openai is also not very interessting it can not be used offline.the huggingface modell is a good idea how it can be integrated into your tool for offline usage? otherwise gpt4all or xgpt is a alternative, but it have to be trained by the user for a while.
I tried GPT4All, but it's too slow.
In the first place, it requires more specifications than stable diffusion, so it is difficult to operate comfortably. If you try the following, you will see that it is difficult because just talking uses a large amount of VRAM. https://github.com/oobabooga/text-generation-webui
It seems better to simply use the OpenAI API.
Sorry, I've changed my mind.
I tried Llama.cpp and it worked to some extent.
I have published a version under development in a branch called langchain
, so please try it out.
I checked with the model file named "openchat_3.5.Q6_K.gguf" at the URL below. https://huggingface.co/TheBloke/openchat_3.5-GGUF/tree/main
The corresponding functionality has also been reflected in the main
branch.
I would appreciate it if you could check it out.
sorry to replay that late , i was not home, i gonna test it out and come back to you in the next days .
one other good modell is also find here: https://huggingface.co/LoneStriker/openchat_3.5-8.0bpw-h6-exl2 i already used it and works good. it have almost the stand as march 2023 of ChatGPT / Openai.
smal question where to put the downloaded modell openchat_3.5.Q6_K.gguf?
smal question where to put the downloaded modell openchat_3.5.Q6_K.gguf?
As long as you specify the full path, you can place it anywhere.
mmmm. i trty so far everythink but it do not work. no reaction,
how you can see it is loading unending
but on the console nothing happens
What do i miss?
It seems that the configuration file is not loading. What happens if you restart the webui? Also, what is the version of webui etc.?
Its what i have done many times now , with no succes. version i use right now is up2date v1.7.0
I'm sorry, but this is not happening here.
What happens if you delete the files located in extensions\sd-webui-chatgpt\settings
?
is it possible that it have somethig to do with my system?
Graphic card PNY NVIDIA A16 / TCSA16M-PB 64gb & Graphic processor card Fujitsu NVIDIA Tesla A100 80GB
It seems that the files cannot be read, so the GPU does not seem to be involved.
i gonna reinstall the hole system this weekend , we will find what is the couse of it
Hello
can someone tell me where to put the chatgpt files for offline usage.
thanks in advance.