build(deps): bump unstructured inference; compile from setup.py (#176)

* bump unstructured inference; compile from setup.py

* bump version

* compile the local-inference extra

* linting, linting, linting

Author: MthwRobinson

CHANGELOG.md

@@ -1,11 +1,12 @@
-## 0.4.4-dev3
+## 0.4.4
 
 * Updated `partition_pdf` and `partition_image` to return `unstructured` `Element` objects
 * Fixed the healthcheck url path when partitioning images and PDFs via API
 * Adds an optional `coordinates` attribute to document objects
 * Adds `FigureCaption` and `CheckBox` document elements
 * Added ability to split lists detected in `LayoutElement` objects
 * Adds `partition_pptx` for partitioning PowerPoint documents
+* LayoutParser models now download from HugginfaceHub instead of DropBox
 * Fixed file type detection for XML and HTML files on Amazone Linux
 
 ## 0.4.3

Makefile

@@ -73,7 +73,7 @@ pip-compile:
 	pip-compile requirements/dev.in
 	pip-compile requirements/test.in
 	pip-compile requirements/build.in
-	pip-compile requirements/local-inference.in
+	pip-compile --extra local-inference -o requirements/local-inference.txt
 	# NOTE(robinson) - doc/requirements.txt is where the GitHub action for building
 	# sphinx docs looks for additional requirements
 	cp requirements/build.txt docs/requirements.txt

requirements/local-inference.in

@@ -2,56 +2,91 @@
 # This file is autogenerated by pip-compile with Python 3.8
 # by the following command:
 #
-#    pip-compile requirements/local-inference.in
+#    pip-compile --extra=local-inference --output-file=requirements/local-inference.txt
 #
 antlr4-python3-runtime==4.9.3
     # via omegaconf
 anyio==3.6.2
-    # via starlette
+    # via
+    #   httpcore
+    #   starlette
+argilla==1.2.1
+    # via unstructured (setup.py)
+backoff==2.2.1
+    # via argilla
 certifi==2022.12.7
-    # via requests
+    # via
+    #   httpcore
+    #   httpx
+    #   requests
+    #   unstructured (setup.py)
 cffi==1.15.1
     # via cryptography
 charset-normalizer==3.0.1
     # via
     #   pdfminer-six
     #   requests
 click==8.1.3
-    # via uvicorn
+    # via
+    #   nltk
+    #   uvicorn
 contourpy==1.0.7
     # via matplotlib
 cryptography==39.0.0
     # via pdfminer-six
 cycler==0.11.0
     # via matplotlib
+deprecated==1.2.13
+    # via argilla
 effdet==0.3.0
     # via layoutparser
+et-xmlfile==1.1.0
+    # via openpyxl
 fastapi==0.89.1
     # via unstructured-inference
 filelock==3.9.0
     # via huggingface-hub
 fonttools==4.38.0
     # via matplotlib
 h11==0.14.0
-    # via uvicorn
-huggingface-hub==0.11.1
+    # via
+    #   httpcore
+    #   uvicorn
+httpcore==0.16.3
+    # via httpx
+httpx==0.23.3
+    # via argilla
+huggingface-hub==0.12.0
     # via
     #   timm
     #   unstructured-inference
 idna==3.4
     # via
     #   anyio
     #   requests
+    #   rfc3986
 iopath==0.1.10
     # via layoutparser
+joblib==1.2.0
+    # via nltk
 kiwisolver==1.4.4
     # via matplotlib
 layoutparser[layoutmodels,tesseract]==0.3.4
     # via unstructured-inference
+lxml==4.9.2
+    # via
+    #   python-docx
+    #   python-pptx
+    #   unstructured (setup.py)
 matplotlib==3.6.3
     # via pycocotools
-numpy==1.24.1
+monotonic==1.6
+    # via argilla
+nltk==3.8.1
+    # via unstructured (setup.py)
+numpy==1.23.5
     # via
+    #   argilla
     #   contourpy
     #   layoutparser
     #   matplotlib
@@ -66,13 +101,19 @@ opencv-python==4.6.0.66
     # via
     #   layoutparser
     #   unstructured-inference
+openpyxl==3.0.10
+    # via unstructured (setup.py)
 packaging==23.0
     # via
+    #   argilla
     #   huggingface-hub
     #   matplotlib
     #   pytesseract
 pandas==1.5.3
-    # via layoutparser
+    # via
+    #   argilla
+    #   layoutparser
+    #   unstructured (setup.py)
 pdf2image==1.16.2
     # via layoutparser
 pdfminer-six==20221105
@@ -86,15 +127,19 @@ pillow==9.4.0
     #   pdf2image
     #   pdfplumber
     #   pytesseract
+    #   python-pptx
     #   torchvision
+    #   unstructured (setup.py)
 portalocker==2.7.0
     # via iopath
 pycocotools==2.0.6
     # via effdet
 pycparser==2.21
     # via cffi
 pydantic==1.10.4
-    # via fastapi
+    # via
+    #   argilla
+    #   fastapi
 pyparsing==3.0.9
     # via matplotlib
 pytesseract==0.3.10
@@ -103,8 +148,14 @@ python-dateutil==2.8.2
     # via
     #   matplotlib
     #   pandas
+python-docx==0.8.11
+    # via unstructured (setup.py)
+python-magic==0.4.27
+    # via unstructured (setup.py)
 python-multipart==0.0.5
     # via unstructured-inference
+python-pptx==0.6.21
+    # via unstructured (setup.py)
 pytz==2022.7.1
     # via pandas
 pyyaml==6.0
@@ -113,18 +164,26 @@ pyyaml==6.0
     #   layoutparser
     #   omegaconf
     #   timm
+regex==2022.10.31
+    # via nltk
 requests==2.28.2
     # via
     #   huggingface-hub
     #   torchvision
+    #   unstructured (setup.py)
+rfc3986[idna2008]==1.5.0
+    # via httpx
 scipy==1.10.0
     # via layoutparser
 six==1.16.0
     # via
     #   python-dateutil
     #   python-multipart
 sniffio==1.3.0
-    # via anyio
+    # via
+    #   anyio
+    #   httpcore
+    #   httpx
 starlette==0.22.0
     # via fastapi
 timm==0.6.12
@@ -142,8 +201,10 @@ torchvision==0.14.1
     #   timm
 tqdm==4.64.1
     # via
+    #   argilla
     #   huggingface-hub
     #   iopath
+    #   nltk
 typing-extensions==4.4.0
     # via
     #   huggingface-hub
@@ -152,11 +213,17 @@ typing-extensions==4.4.0
     #   starlette
     #   torch
     #   torchvision
-unstructured-inference==0.2.3
-    # via -r requirements/local-inference.in
+unstructured-inference==0.2.4
+    # via unstructured (setup.py)
 urllib3==1.26.14
     # via requests
 uvicorn==0.20.0
     # via unstructured-inference
 wand==0.6.11
     # via pdfplumber
+wrapt==1.14.1
+    # via
+    #   argilla
+    #   deprecated
+xlsxwriter==3.0.7
+    # via python-pptx

requirements/local-inference.txt

@@ -70,6 +70,6 @@
             "torch",
             "transformers",
         ],
-        "local-inference": ["unstructured-inference>=0.2.3"],
+        "local-inference": ["unstructured-inference>=0.2.4"],
     },
 )

setup.py

@@ -1 +1 @@
-__version__ = "0.4.4-dev3"  # pragma: no cover
+__version__ = "0.4.4"  # pragma: no cover