To view the model weights of TIPO
, go to mirror site: 200M、500M、GGUF(QuantFactory)
Or if you can, directly visit the original site: 200M、500M、GGUF(QuantFactory)
KohakuBlueleaf's GitHub Homepage
Thank Yi-1.5 for translating this README to English.
Thank GLM4 for providing some of the code (since I am really not good
at it, so I have to use AI)
It's better to use the GPU version of llama-cpp-python
because the parameter n_gpu_layers
won't work in CPU version.
llama-cpp-python
gradio
pyperclip
- Download the
.gguf
file from the place mentioned above, and put it into themodels
folder in the same directory as the script. - Double-click the script, or run
python GUI.py
withcmd
. - Go to the "Settings" page, select your model file, then click Load.
- If you don't understand the parameters, you have two options: Don't touch them, or go to the "Tutorials" page for an explanation.
- See below for generation settings, with the same two choices as above.
- Go to the "Generate" page, select your mode, length, write your prompt, quality, then click "TIPO!".
- Click "Copy to Clipboard" to copy the results.
- Use
Language Config Manager.py
to manage language configurations. - To add a new language, follow these steps:
- Create a new
.json
file in theLocales
folder. - I recommend you copy an existing language file and then translate it; save your file after translation is complete.
- In the
Tutorials
subfolder, create a new.md
file. - Copy an existing tutorial file, translate, and save.
- Open
Language Config Manager.py
, choose2. Write new language config
, your language file name should appear in the options. - Select your file, restart Webui, and you should see the language settings take effect.
- Create a new
If you crash when double-clicking the script, try running python GUI.py
in cmd
; anyway, I have no idea why since
this was all new to me.