vlm4ocr 0.4.1__tar.gz → 0.4.2__tar.gz

{vlm4ocr-0.4.1 → vlm4ocr-0.4.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vlm4ocr
-Version: 0.4.1
+Version: 0.4.2
 Summary: Python package and Web App for OCR with vision language models.
 License: MIT
 Author: Enshuo (David) Hsu
@@ -12,7 +12,7 @@ Classifier: Programming Language :: Python :: 3.12
 Provides-Extra: tesseract
 Requires-Dist: colorama (>=0.4.4)
 Requires-Dist: json-repair (>=0.30.0)
-Requires-Dist: llm-inference-engine (>=0.1.1,<0.2.0)
+Requires-Dist: llm-inference-engine (>=0.1.5)
 Requires-Dist: pdf2image (>=1.16.0)
 Requires-Dist: pillow (>=10.0.0)
 Requires-Dist: pytesseract (>=0.3.13) ; extra == "tesseract"

{vlm4ocr-0.4.1 → vlm4ocr-0.4.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "vlm4ocr"
-version = "0.4.1"
+version = "0.4.2"
 description = "Python package and Web App for OCR with vision language models."
 authors = ["Enshuo (David) Hsu"]
 license = "MIT"
@@ -18,7 +18,7 @@ pdf2image = ">=1.16.0"
 colorama = ">=0.4.4"
 pillow = ">=10.0.0"
 json-repair = ">=0.30.0"
-llm-inference-engine = "^0.1.1"
+llm-inference-engine = ">=0.1.5"
 pytesseract = { version = ">=0.3.13", optional = true }
 
 [tool.poetry.scripts]

{vlm4ocr-0.4.1 → vlm4ocr-0.4.2}/vlm4ocr/ocr_engines.py

@@ -126,9 +126,8 @@ class OCREngine:
                                                             few_shot_examples=few_shot_examples)
                 
                 # Stream response
-                response_stream = self.vlm_engine.chat(
-                    messages,
-                    stream=True
+                response_stream = self.vlm_engine.chat_stream(
+                    messages
                 )
                 for chunk in response_stream:
                     if chunk["type"] == "response":
@@ -163,9 +162,8 @@ class OCREngine:
                                                         image=image,
                                                         few_shot_examples=few_shot_examples)
             # Stream response
-            response_stream = self.vlm_engine.chat(
-                    messages,
-                    stream=True
+            response_stream = self.vlm_engine.chat_stream(
+                    messages
                 )
             for chunk in response_stream:
                 if chunk["type"] == "response":
@@ -295,7 +293,6 @@ class OCREngine:
                     response = self.vlm_engine.chat(
                         messages,
                         verbose=verbose,
-                        stream=False,
                         messages_logger=messages_logger
                     )
                     ocr_text = response["response"]