The insight_extractor from noahfinberg

insight_extractor's People

Contributors

Stargazers

Watchers

insight_extractor's Issues

add license file

travis, readthedocs, appveyor

cut down on requirements.txt

too big for now: https://github.com/NoahFinberg/insight_extractor/blob/main/requirements.txt

Missing `models` directory when installed

Thanks for this - looks super interesting! Might be something I've done wrong, but have installed insight_extractor using Python 3.8 on OSX and when I try the example usage with some sentences I get:

>>> from insight_extractor.pipeline import extract_insights
>>> predictions = extract_insights(sentences)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/derekwillis/.virtualenvs/tensorflow-sandbox-JLWPiXbX/lib/python3.8/site-packages/insight_extractor/pipeline.py", line 25, in extract_insights
    tokenized_docs = tokenize(cleaned_docs, TOKENIZER_PATH) # tokenize docs
  File "/Users/derekwillis/.virtualenvs/tensorflow-sandbox-JLWPiXbX/lib/python3.8/site-packages/insight_extractor/preprocessors.py", line 53, in tokenize
    with open(file) as f:
FileNotFoundError: [Errno 2] No such file or directory: '/Users/derekwillis/.virtualenvs/tensorflow-sandbox-JLWPiXbX/lib/python3.8/site-packages/insight_extractor/models/combined-tokenizer.json'
>>>

Checking the virtualenv confirms that the models directory is not present.

change from array to vector

"array([[0.28326818, 0.7167318 ],
[0.3710433 , 0.6289567 ],
[0.9886193 , 0.01138071]], dtype=float32)" -> seems like we have a binary classifier. useful to only produce prob_insight.

Recommend Projects