Feat/bump inference

CHANGELOG.md

@@ -1,15 +1,16 @@
-## 0.17.7
+## 0.17.8
 
 ### Enhancements
-- **Updated Docker file with ENV HF_HUB_OFFLINE=1 to prevent the contianer from trying to access the internet
+- **Bump `unstructured-inference` to `1.0.5`** It includes critical fix to ensure inference model initialization is thread safe
 
 ### Features
 
 ### Fixes
 
-## 0.17.7-dev0
+## 0.17.7
 
 ### Enhancements
+- **Updated Docker file with ENV HF_HUB_OFFLINE=1 to prevent the contianer from trying to access the internet
 
 ### Features
 

requirements/extra-csv.txt

@@ -8,7 +8,7 @@ numpy==2.0.2
     # via
     #   -c ./base.txt
     #   pandas
-pandas==2.2.3
+pandas==2.3.0
     # via -r ./extra-csv.in
 python-dateutil==2.9.0.post0
     # via

requirements/extra-pdf-image.in

@@ -12,5 +12,5 @@ google-cloud-vision
 effdet
 # Do not move to constraints.in, otherwise unstructured-inference will not be upgraded
 # when unstructured library is.
-unstructured-inference>=0.8.10
+unstructured-inference>=1.0.5
 unstructured.pytesseract>=0.3.12

requirements/extra-pdf-image.txt

@@ -52,7 +52,7 @@ fsspec==2025.5.1
     #   torch
 google-api-core[grpc]==2.25.0
     # via google-cloud-vision
-google-auth==2.40.2
+google-auth==2.40.3
     # via
     #   google-api-core
     #   google-cloud-vision
@@ -69,9 +69,9 @@ grpcio==1.72.1
     #   grpcio-status
 grpcio-status==1.72.1
     # via google-api-core
-hf-xet==1.1.2
+hf-xet==1.1.3
     # via huggingface-hub
-huggingface-hub==0.32.3
+huggingface-hub==0.32.4
     # via
     #   accelerate
     #   timm
@@ -139,7 +139,7 @@ packaging==25.0
     #   pikepdf
     #   transformers
     #   unstructured-pytesseract
-pandas==2.2.3
+pandas==2.3.0
     # via unstructured-inference
 pdf2image==1.17.0
     # via -r ./extra-pdf-image.in
@@ -184,7 +184,7 @@ pyasn1==0.6.1
     #   rsa
 pyasn1-modules==0.4.2
     # via google-auth
-pycocotools==2.0.9
+pycocotools==2.0.10
     # via effdet
 pycparser==2.22
     # via
@@ -253,14 +253,14 @@ tokenizers==0.21.1
     # via
     #   -c ././deps/constraints.txt
     #   transformers
-torch==2.7.0
+torch==2.7.1
     # via
     #   accelerate
     #   effdet
     #   timm
     #   torchvision
     #   unstructured-inference
-torchvision==0.22.0
+torchvision==0.22.1
     # via
     #   effdet
     #   timm
@@ -280,7 +280,7 @@ typing-extensions==4.14.0
     #   torch
 tzdata==2025.2
     # via pandas
-unstructured-inference==1.0.2
+unstructured-inference==1.0.5
     # via -r ./extra-pdf-image.in
 unstructured-pytesseract==0.3.15
     # via -r ./extra-pdf-image.in

requirements/extra-xlsx.txt

@@ -14,7 +14,7 @@ numpy==2.0.2
     #   pandas
 openpyxl==3.1.5
     # via -r ./extra-xlsx.in
-pandas==2.2.3
+pandas==2.3.0
     # via -r ./extra-xlsx.in
 python-dateutil==2.9.0.post0
     # via

requirements/huggingface.txt

@@ -25,9 +25,9 @@ fsspec==2025.5.1
     # via
     #   huggingface-hub
     #   torch
-hf-xet==1.1.2
+hf-xet==1.1.3
     # via huggingface-hub
-huggingface-hub==0.32.3
+huggingface-hub==0.32.4
     # via
     #   tokenizers
     #   transformers
@@ -90,7 +90,7 @@ tokenizers==0.21.1
     # via
     #   -c ././deps/constraints.txt
     #   transformers
-torch==2.7.0
+torch==2.7.1
     # via -r ./huggingface.in
 tqdm==4.67.1
     # via

test_unstructured_ingest/expected-structured-output-html/local-single-file-with-pdf-infer-table-structure/layout-parser-paper-with-table.jpg.html

@@ -26,7 +26,7 @@ <h1 class="Title" id="5d45a28d875e403c7294a15f22a0162f">
       Large Model
      </th>
      <th style="border: 1px solid black;">
-      Notes
+      | Notes
      </th>
     </tr>
    </thead>

test_unstructured_ingest/expected-structured-output-html/local-single-file-with-pdf-infer-table-structure/layout-parser-paper.pdf.html

@@ -168,33 +168,21 @@ <h1 class="Title" id="5a1838a8f40b4523094652cf14ab974c">
       Dataset
      </th>
      <th style="border: 1px solid black;">
-      |
+      | Base Model'|
      </th>
      <th style="border: 1px solid black;">
-      Base Model'|
-     </th>
-     <th style="border: 1px solid black;">
-      Large Model |
-     </th>
-     <th style="border: 1px solid black;">
-      Notes
+      | Notes
      </th>
     </tr>
    </thead>
    <tbody>
     <tr style="border: 1px solid black;">
      <td style="border: 1px solid black;">
-      PubLayNet
-     </td>
-     <td style="border: 1px solid black;">
-      B8]|
+      PubLayNet B8]|
      </td>
      <td style="border: 1px solid black;">
       F/M
      </td>
-     <td style="border: 1px solid black;">
-      M
-     </td>
      <td style="border: 1px solid black;">
       Layouts of modern scientific documents
      </td>
@@ -203,14 +191,9 @@ <h1 class="Title" id="5a1838a8f40b4523094652cf14ab974c">
      <td style="border: 1px solid black;">
       PRImA
      </td>
-     <td style="border: 1px solid black;">
-     </td>
      <td style="border: 1px solid black;">
       M
      </td>
-     <td style="border: 1px solid black;">
-      -
-     </td>
      <td style="border: 1px solid black;">
       Layouts of scanned modern magazines and scientific report
      </td>
@@ -219,14 +202,9 @@ <h1 class="Title" id="5a1838a8f40b4523094652cf14ab974c">
      <td style="border: 1px solid black;">
       Newspaper
      </td>
-     <td style="border: 1px solid black;">
-     </td>
      <td style="border: 1px solid black;">
       F
      </td>
-     <td style="border: 1px solid black;">
-      -
-     </td>
      <td style="border: 1px solid black;">
       Layouts of scanned US newspapers from the 20th century
      </td>
@@ -235,11 +213,6 @@ <h1 class="Title" id="5a1838a8f40b4523094652cf14ab974c">
      <td style="border: 1px solid black;">
       TableBank
      </td>
-     <td style="border: 1px solid black;">
-     </td>
-     <td style="border: 1px solid black;">
-      F
-     </td>
      <td style="border: 1px solid black;">
       F
      </td>
@@ -251,14 +224,9 @@ <h1 class="Title" id="5a1838a8f40b4523094652cf14ab974c">
      <td style="border: 1px solid black;">
       HJDataset
      </td>
-     <td style="border: 1px solid black;">
-     </td>
      <td style="border: 1px solid black;">
       F/M
      </td>
-     <td style="border: 1px solid black;">
-      -
-     </td>
      <td style="border: 1px solid black;">
       Layouts of history Japanese documents
      </td>
@@ -348,7 +316,10 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
    <thead>
     <tr style="border: 1px solid black;">
      <th style="border: 1px solid black;">
-      block.pad(top, bottom, right,
+      block.pad(top, bottom,
+     </th>
+     <th style="border: 1px solid black;">
+      right,
      </th>
      <th style="border: 1px solid black;">
       left)
@@ -365,6 +336,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Scale the current block given the ratio in x and y direction
      </td>
@@ -375,6 +348,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Move the current block with the shift distances in x and y direction
      </td>
@@ -385,6 +360,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Whether block] is inside of block2
      </td>
@@ -395,6 +372,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Return the intersection region of blockl and block2. Coordinate type to be determined based on the inputs
      </td>
@@ -405,6 +384,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Return the union region of blockl and block2. Coordinate type to be determined based on the inputs
      </td>
@@ -415,6 +396,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Convert the absolute coordinates of block to relative coordinates to block2
      </td>
@@ -425,6 +408,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Calculate the absolute coordinates of blockl given the canvas block2’s absolute coordinates
      </td>
@@ -435,6 +420,8 @@ <h1 class="Title" id="2b81bd7a3f21b84379bfcd4bb175c5d1">
      </td>
      <td style="border: 1px solid black;">
      </td>
+     <td style="border: 1px solid black;">
+     </td>
      <td style="border: 1px solid black;">
       Obtain the image segments in the block region
      </td>

test_unstructured_ingest/expected-structured-output/local-single-file-with-pdf-infer-table-structure/layout-parser-paper-with-table.jpg.json

@@ -48,7 +48,7 @@
     "element_id": "dddac446da6c93dc1449ecb5d997c423",
     "text": "Dataset | Base Model\" Large Model | Notes PubLayNet [38] P/M M Layouts of modern scientific documents PRImA [3) M - Layouts of scanned modern magazines and scientific reports Newspaper [17] P - Layouts of scanned US newspapers from the 20th century ‘TableBank (18) P P Table region on modern scientific and business document HJDataset (31) | F/M - Layouts of history Japanese documents",
     "metadata": {
-      "text_as_html": "<table><thead><tr><th>Dataset</th><th>| Base Model!|</th><th>Large Model</th><th>Notes</th></tr></thead><tbody><tr><td>PubLayNet [33]</td><td>P/M</td><td>M</td><td>Layouts of modern scientific documents</td></tr><tr><td>PRImA [3]</td><td>M</td><td></td><td>Layouts of scanned modern magazines and scientific reports</td></tr><tr><td>Newspaper [17]</td><td>P</td><td></td><td>Layouts of scanned US newspapers from the 20th century</td></tr><tr><td>TableBank [18]</td><td>P</td><td></td><td>Table region on modern scientific and business document</td></tr><tr><td>HIDataset [31]</td><td>P/M</td><td></td><td>Layouts of history Japanese documents</td></tr></tbody></table>",
+      "text_as_html": "<table><thead><tr><th>Dataset</th><th>| Base Model!|</th><th>Large Model</th><th>| Notes</th></tr></thead><tbody><tr><td>PubLayNet [33]</td><td>P/M</td><td>M</td><td>Layouts of modern scientific documents</td></tr><tr><td>PRImA [3]</td><td>M</td><td></td><td>Layouts of scanned modern magazines and scientific reports</td></tr><tr><td>Newspaper [17]</td><td>P</td><td></td><td>Layouts of scanned US newspapers from the 20th century</td></tr><tr><td>TableBank [18]</td><td>P</td><td></td><td>Table region on modern scientific and business document</td></tr><tr><td>HIDataset [31]</td><td>P/M</td><td></td><td>Layouts of history Japanese documents</td></tr></tbody></table>",
       "filetype": "image/jpeg",
       "languages": [
         "eng"

test_unstructured_ingest/expected-structured-output/local-single-file-with-pdf-infer-table-structure/layout-parser-paper.pdf.json

@@ -1459,7 +1459,7 @@
           "start_index": 65
         }
       ],
-      "text_as_html": "<table><thead><tr><th>Dataset</th><th>|</th><th>Base Model'|</th><th>Large Model |</th><th>Notes</th></tr></thead><tbody><tr><td>PubLayNet</td><td>B8]|</td><td>F/M</td><td>M</td><td>Layouts of modern scientific documents</td></tr><tr><td>PRImA</td><td></td><td>M</td><td>-</td><td>Layouts of scanned modern magazines and scientific report</td></tr><tr><td>Newspaper</td><td></td><td>F</td><td>-</td><td>Layouts of scanned US newspapers from the 20th century</td></tr><tr><td>TableBank</td><td></td><td>F</td><td>F</td><td>Table region on modern scientific and business document</td></tr><tr><td>HJDataset</td><td></td><td>F/M</td><td>-</td><td>Layouts of history Japanese documents</td></tr></tbody></table>",
+      "text_as_html": "<table><thead><tr><th>Dataset</th><th>| Base Model'|</th><th>| Notes</th></tr></thead><tbody><tr><td>PubLayNet B8]|</td><td>F/M</td><td>Layouts of modern scientific documents</td></tr><tr><td>PRImA</td><td>M</td><td>Layouts of scanned modern magazines and scientific report</td></tr><tr><td>Newspaper</td><td>F</td><td>Layouts of scanned US newspapers from the 20th century</td></tr><tr><td>TableBank</td><td>F</td><td>Table region on modern scientific and business document</td></tr><tr><td>HJDataset</td><td>F/M</td><td>Layouts of history Japanese documents</td></tr></tbody></table>",
       "filetype": "application/pdf",
       "languages": [
         "eng"
@@ -2153,7 +2153,7 @@
     "element_id": "64bc79d1132a89c71837f420d6e4e2dc",
     "text": "Operation Name Description block.pad(top, bottom, right, left) Enlarge the current block according to the input block.scale(fx, fy) Scale the current block given the ratio in x and y direction block.shift(dx, dy) Move the current block with the shift distances in x and y direction block1.is in(block2) Whether block1 is inside of block2 block1.intersect(block2) Return the intersection region of block1 and block2. Coordinate type to be determined based on the inputs. block1.union(block2) Return the union region of block1 and block2. Coordinate type to be determined based on the inputs. block1.relative to(block2) Convert the absolute coordinates of block1 to relative coordinates to block2 block1.condition on(block2) Calculate the absolute coordinates of block1 given the canvas block2’s absolute coordinates block.crop image(image) Obtain the image segments in the block region",
     "metadata": {
-      "text_as_html": "<table><thead><tr><th>block.pad(top, bottom, right,</th><th>left)</th><th>Enlarge the current block according to the input</th></tr></thead><tbody><tr><td>block.scale(fx, fy)</td><td></td><td>Scale the current block given the ratio in x and y direction</td></tr><tr><td>block.shift(dx, dy)</td><td></td><td>Move the current block with the shift distances in x and y direction</td></tr><tr><td>block1.is_in(block2)</td><td></td><td>Whether block] is inside of block2</td></tr><tr><td>block1. intersect (block2)</td><td></td><td>Return the intersection region of blockl and block2. Coordinate type to be determined based on the inputs</td></tr><tr><td>block1.union(block2)</td><td></td><td>Return the union region of blockl and block2. Coordinate type to be determined based on the inputs</td></tr><tr><td>block1.relative_to(block2)</td><td></td><td>Convert the absolute coordinates of block to relative coordinates to block2</td></tr><tr><td>block1.condition_on(block2)</td><td></td><td>Calculate the absolute coordinates of blockl given the canvas block2’s absolute coordinates</td></tr><tr><td>block. crop_image (image)</td><td></td><td>Obtain the image segments in the block region</td></tr></tbody></table>",
+      "text_as_html": "<table><thead><tr><th>block.pad(top, bottom,</th><th>right,</th><th>left)</th><th>Enlarge the current block according to the input</th></tr></thead><tbody><tr><td>block.scale(fx, fy)</td><td></td><td></td><td>Scale the current block given the ratio in x and y direction</td></tr><tr><td>block.shift(dx, dy)</td><td></td><td></td><td>Move the current block with the shift distances in x and y direction</td></tr><tr><td>block1.is_in(block2)</td><td></td><td></td><td>Whether block] is inside of block2</td></tr><tr><td>block1. intersect (block2)</td><td></td><td></td><td>Return the intersection region of blockl and block2. Coordinate type to be determined based on the inputs</td></tr><tr><td>block1.union(block2)</td><td></td><td></td><td>Return the union region of blockl and block2. Coordinate type to be determined based on the inputs</td></tr><tr><td>block1.relative_to(block2)</td><td></td><td></td><td>Convert the absolute coordinates of block to relative coordinates to block2</td></tr><tr><td>block1.condition_on(block2)</td><td></td><td></td><td>Calculate the absolute coordinates of blockl given the canvas block2’s absolute coordinates</td></tr><tr><td>block. crop_image (image)</td><td></td><td></td><td>Obtain the image segments in the block region</td></tr></tbody></table>",
       "filetype": "application/pdf",
       "languages": [
         "eng"

unstructured/__version__.py

@@ -1 +1 @@
-__version__ = "0.17.7"  # pragma: no cover
+__version__ = "0.17.8"  # pragma: no cover