sparrow-parse 0.4.7__tar.gz → 0.4.8__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (25) hide show
  1. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/PKG-INFO +1 -1
  2. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/setup.py +1 -1
  3. sparrow-parse-0.4.8/sparrow_parse/__init__.py +1 -0
  4. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/extractors/vllm_extractor.py +31 -31
  5. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/vllm/mlx_inference.py +1 -0
  6. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse.egg-info/PKG-INFO +1 -1
  7. sparrow-parse-0.4.7/sparrow_parse/__init__.py +0 -1
  8. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/README.md +0 -0
  9. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/setup.cfg +0 -0
  10. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/__main__.py +0 -0
  11. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/extractors/__init__.py +0 -0
  12. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/helpers/__init__.py +0 -0
  13. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/helpers/pdf_optimizer.py +0 -0
  14. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/processors/__init__.py +0 -0
  15. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/processors/table_structure_processor.py +0 -0
  16. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/vllm/__init__.py +0 -0
  17. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/vllm/huggingface_inference.py +0 -0
  18. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/vllm/inference_base.py +0 -0
  19. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/vllm/inference_factory.py +0 -0
  20. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse/vllm/local_gpu_inference.py +0 -0
  21. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse.egg-info/SOURCES.txt +0 -0
  22. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse.egg-info/dependency_links.txt +0 -0
  23. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse.egg-info/entry_points.txt +0 -0
  24. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse.egg-info/requires.txt +0 -0
  25. {sparrow-parse-0.4.7 → sparrow-parse-0.4.8}/sparrow_parse.egg-info/top_level.txt +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: sparrow-parse
3
- Version: 0.4.7
3
+ Version: 0.4.8
4
4
  Summary: Sparrow Parse is a Python package (part of Sparrow) for parsing and extracting information from documents.
5
5
  Home-page: https://github.com/katanaml/sparrow/tree/main/sparrow-data/parse
6
6
  Author: Andrej Baranovskij
@@ -8,7 +8,7 @@ with open("requirements.txt", "r", encoding="utf-8") as fh:
8
8
 
9
9
  setup(
10
10
  name="sparrow-parse",
11
- version="0.4.7",
11
+ version="0.4.8",
12
12
  author="Andrej Baranovskij",
13
13
  author_email="andrejus.baranovskis@gmail.com",
14
14
  description="Sparrow Parse is a Python package (part of Sparrow) for parsing and extracting information from documents.",
@@ -0,0 +1 @@
1
+ __version__ = '0.4.8'
@@ -152,34 +152,34 @@ if __name__ == "__main__":
152
152
 
153
153
  extractor = VLLMExtractor()
154
154
 
155
- # export HF_TOKEN="hf_"
156
- config = {
157
- "method": "mlx", # Could be 'huggingface', 'mlx' or 'local_gpu'
158
- "model_name": "mlx-community/Qwen2-VL-7B-Instruct-8bit",
159
- # "hf_space": "katanaml/sparrow-qwen2-vl-7b",
160
- # "hf_token": os.getenv('HF_TOKEN'),
161
- # Additional fields for local GPU inference
162
- # "device": "cuda", "model_path": "model.pth"
163
- }
164
-
165
- # Use the factory to get the correct instance
166
- factory = InferenceFactory(config)
167
- model_inference_instance = factory.get_inference_instance()
168
-
169
- input_data = [
170
- {
171
- "file_path": "/Users/andrejb/Work/katana-git/sparrow/sparrow-ml/llm/data/invoice_1.jpg",
172
- "text_input": "retrieve document data. return response in JSON format"
173
- }
174
- ]
175
-
176
- # Now you can run inference without knowing which implementation is used
177
- results_array, num_pages = extractor.run_inference(model_inference_instance, input_data, tables_only=True,
178
- generic_query=False,
179
- debug_dir="/Users/andrejb/Work/katana-git/sparrow/sparrow-ml/llm/data/",
180
- debug=True,
181
- mode=None)
182
-
183
- for i, result in enumerate(results_array):
184
- print(f"Result for page {i + 1}:", result)
185
- print(f"Number of pages: {num_pages}")
155
+ # # export HF_TOKEN="hf_"
156
+ # config = {
157
+ # "method": "mlx", # Could be 'huggingface', 'mlx' or 'local_gpu'
158
+ # "model_name": "mlx-community/Qwen2-VL-7B-Instruct-8bit",
159
+ # # "hf_space": "katanaml/sparrow-qwen2-vl-7b",
160
+ # # "hf_token": os.getenv('HF_TOKEN'),
161
+ # # Additional fields for local GPU inference
162
+ # # "device": "cuda", "model_path": "model.pth"
163
+ # }
164
+ #
165
+ # # Use the factory to get the correct instance
166
+ # factory = InferenceFactory(config)
167
+ # model_inference_instance = factory.get_inference_instance()
168
+ #
169
+ # input_data = [
170
+ # {
171
+ # "file_path": "/Users/andrejb/Work/katana-git/sparrow/sparrow-ml/llm/data/invoice_1.jpg",
172
+ # "text_input": "retrieve document data. return response in JSON format"
173
+ # }
174
+ # ]
175
+ #
176
+ # # Now you can run inference without knowing which implementation is used
177
+ # results_array, num_pages = extractor.run_inference(model_inference_instance, input_data, tables_only=True,
178
+ # generic_query=False,
179
+ # debug_dir="/Users/andrejb/Work/katana-git/sparrow/sparrow-ml/llm/data/",
180
+ # debug=True,
181
+ # mode=None)
182
+ #
183
+ # for i, result in enumerate(results_array):
184
+ # print(f"Result for page {i + 1}:", result)
185
+ # print(f"Number of pages: {num_pages}")
@@ -4,6 +4,7 @@ from mlx_vlm.utils import load_image
4
4
  from sparrow_parse.vllm.inference_base import ModelInference
5
5
  import os
6
6
  import json
7
+ from rich import print
7
8
 
8
9
 
9
10
  class MLXInference(ModelInference):
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: sparrow-parse
3
- Version: 0.4.7
3
+ Version: 0.4.8
4
4
  Summary: Sparrow Parse is a Python package (part of Sparrow) for parsing and extracting information from documents.
5
5
  Home-page: https://github.com/katanaml/sparrow/tree/main/sparrow-data/parse
6
6
  Author: Andrej Baranovskij
@@ -1 +0,0 @@
1
- __version__ = '0.4.7'
File without changes
File without changes