-
|
I have a model that was tuned with another program (H2O LLM Studio), that I want to now load into OobaBooga. Can anyone tell me if that is supported? I can only see how to load models from hugging face, and I like to keep my model on my own hard drive and not upload first. Thank you. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
|
Never mind, I found out how (I think). Place the model (unpacked if it is zipped up) in the correct directory: text-generation-webui-main\models<model name> In the directory you should see: THEN when you go into 'Models' and choose the 'drop down box' it will auto-populate. (I was originally looking for a way to load it by navigating to the directory from within the web-ui, I am unsure if this is possible). Thank you. |
Beta Was this translation helpful? Give feedback.
Never mind, I found out how (I think).
Place the model (unpacked if it is zipped up) in the correct directory:
text-generation-webui-main\models<model name>
In the directory you should see:
*.json, *.md, *.safetensors, etc. etc.
THEN when you go into 'Models' and choose the 'drop down box' it will auto-populate.
(I was originally looking for a way to load it by navigating to the directory from within the web-ui, I am unsure if this is possible).
In either case, this works fine.
Thank you.