Update PDF extraction and OCR options for hybrid chunking

.github/workflows/test.yml

@@ -107,6 +107,9 @@ jobs:
       - name: Run unit and functional tests with tox
         run: |
           tox
+        env:
+          # Increase from 1.7 to a greater value to avoid the PyTorch MPS backend running OOM.
+          PYTORCH_MPS_HIGH_WATERMARK_RATIO: 2.0
 
       - name: Remove llama-cpp-python from cache
         if: always()

requirements.txt

@@ -1,9 +1,9 @@
 # SPDX-License-Identifier: Apache-2.0
 click>=8.1.7,<9.0.0
 datasets>=2.18.0,<3.0.0
-docling[tesserocr]>=2.4.2,<=2.8.3; sys_platform != 'darwin'
-docling>=2.4.2,<=2.8.3; sys_platform == 'darwin'
-docling-parse>=2.0.0,<3.0.0
+docling-core[chunking]>=2.9.0
+docling[tesserocr]>=2.9.0; sys_platform != 'darwin'
+docling>=2.9.0; sys_platform == 'darwin'
 GitPython>=3.1.42,<4.0.0
 gguf>=0.6.0
 httpx>=0.25.0,<1.0.0