Unable to build Docker

[eladrave]

I get an error when it get to this:

RUN playwright install --with-deps && \
    python -c "from unstructured.nlp.tokenize import download_nltk_packages; download_nltk_packages()" && \
    python -c "import nltk;nltk.download('punkt_tab'); nltk.download('averaged_perceptron_tagger_eng')" && \
    python -c "from unstructured.partition.model_init import initialize; initialize()"

Here is the error:

66.71 Webkit 18.0 (playwright build v2083) downloaded to /root/.cache/ms-playwright/webkit-2083
69.38 Traceback (most recent call last):
69.38   File "<string>", line 1, in <module>
69.38   File "/usr/local/lib/python3.11/site-packages/unstructured/nlp/tokenize.py", line 88, in download_nltk_packages
69.38     urllib.request.urlretrieve(NLTK_DATA_URL, tgz_file_path)
69.38   File "/usr/local/lib/python3.11/urllib/request.py", line 241, in urlretrieve
69.39     with contextlib.closing(urlopen(url, data)) as fp:
69.39                             ^^^^^^^^^^^^^^^^^^
69.39   File "/usr/local/lib/python3.11/urllib/request.py", line 216, in urlopen
69.39     return opener.open(url, data, timeout)
69.39            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
69.39   File "/usr/local/lib/python3.11/urllib/request.py", line 525, in open
69.39     response = meth(req, response)
69.39                ^^^^^^^^^^^^^^^^^^^
69.39   File "/usr/local/lib/python3.11/urllib/request.py", line 634, in http_response
69.39     response = self.parent.error(
69.39                ^^^^^^^^^^^^^^^^^^
69.39   File "/usr/local/lib/python3.11/urllib/request.py", line 563, in error
69.39     return self._call_chain(*args)
69.39            ^^^^^^^^^^^^^^^^^^^^^^^
69.39   File "/usr/local/lib/python3.11/urllib/request.py", line 496, in _call_chain
69.39     result = func(*args)
69.39              ^^^^^^^^^^^
69.39   File "/usr/local/lib/python3.11/urllib/request.py", line 643, in http_error_default
69.39     raise HTTPError(req.full_url, code, msg, hdrs, fp)
69.39 urllib.error.HTTPError: HTTP Error 403: Forbidden
------
failed to solve: process "/bin/sh -c playwright install --with-deps &&     python -c \"from unstructured.nlp.tokenize import download_nltk_packages; download_nltk_packages()\" &&     python -c \"import nltk;nltk.download('punkt_tab'); nltk.download('averaged_perceptron_tagger_eng')\" &&     python -c \"from unstructured.partition.model_init import initialize; initialize()\"" did not complete successfully: exit code: 1

[lmpentland]

had this problem all day yesterday, and went away this morning. The only things I've done between then and now are:

downloaded and installed node + dependencies on the host machine (chocolatey etc, all included with node),
added C:/...../megaparse/src to PYTHONPATH env variable,
a restart,
and included my username in the docker build like
docker build -t usernamehere/megaparse .

One of those things seem to fix it. or at least allowed me to build the image. I'm still getting an error, but now a ModuleNotFound error about not being able to find megaparse.api

[StanGirard]

Unstructured changed something in the way they download models. In the latest version we fixed that.