A simple implementation and UI of Summary Transformer models. Specifically The pszemraj/led-large-book-summary
am new to git and still learning. i have no clue what im doing. its like moving through quicksand.
- download this repo as a zip or using Git.
- download and install python 3.10.6 (recommended) OR latest python version (tested on 3.11.3) and add to PATH
- run Install.bat
- run either RunBaseModel.bat or RunLargeModel.bat depending if you want the bigger model version.
- put large amounts of text in the text box -[OR]-drag and drop text files
- (if there is a text file inserted it will disregard anything in the text box.)
- scroll to the bottom and hit generate.
all the files get stored locally and can be used offline once ran atleast one time.
all nessarry files *should* all get stored within the same file as the program
yes i know code is jank. ill polish it later
theoretically you can import summary and text-to-text models by passing the python file with the name of the huggingface repo
you can edit the venv\Scripts\python.exe Open_WebUI.py --model [hugging face repo name]
line in the batch file. look at RunLargeModel.bat
for example.
this feature is untested
- use Gradio Blocks insted of default gradio interface
- force gradio to stop making temporary files when arleady using a text file
- add ability to change summary models while open
- add ability to save input and output as text files
- add support for other models. (some models may arleady work... untested)
- look into merging as an extension with oobabooga text generator