Text-Summarization | This is an extractive based text summarization | Natural Language Processing library

Why not transfer learning? Train them on your specific texts and summaries.

I trained T5 on specific limited text over 5 epoch and got very good results. I adopted the code from here to my needs https://github.com/patil-suraj/exploring-T5/blob/master/t5_fine_tuning.ipynb

Let me know if you have a specific training questions.

Check class(dd$text). It's a factor, which is not a character.

The following works:

Try explicitly moving your model to the GPU.

It seems like at least one of your story dictionaries does not have a key-value pair for the key 'highlights'. If this is only true for certain stories, you can simply check if there is a NoneType before iterating. If this is true for all stories, there might be a discrepancy between your code and the data you are working with.

Also, I believe there is an indentation error (might just be wrong SO formatting), but I believe the code after target_text = highlight should be indented once more to the right.

The folder keras_text_summarization was outside the demo package. So as it appears the documentation is incorrect. I needed the either to:

1. Install the keras_text_summarization by running setup.py on parent folder.
2. Move keras_text_summarization inside the demo folder.

I ran step 2 and it worked (moved keras_text_summarization inside the demo folder).

This means no external modules were missing and condo installation was perfect.

Note: the following answer assumes you are on the correct version (3.6+), if not: simply, variable annotation isn't supported on your current version of Python.

The problem may seem like the type annotation is causing the SyntaxError, but another very plausible possibility is that there is an unclosed parenthesis or a unclosed something in the lines preceding. Since in the docs, it said:

The parser repeats the offending line and displays a little ‘arrow’ pointing at the earliest point in the line where the error was detected. The error is caused by (or at least detected at) the token preceding the arrow

(emphasis mine)

The parser can only detect an unclosed bracket when a token that is invalid in that context is given. Since brackets and parenthesis can carry through multiple lines (which means EOL wouldn't be raised), and text is a valid variable identifier, which leaves only that colons aren't allowed in brackets or parentheses (excepted when it's used as an parameter, which also accepts type annotation).

Here's an reproducible example of your code hosted on tio.run (the SE code-golf compiler):

https://tio.run/##K6gsycjPM/7/X4OrJLWixEqhuKRIwVZBXf3/fwA