parallex 0.3.1__tar.gz → 0.3.3__tar.gz

{parallex-0.3.1 → parallex-0.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: parallex
-Version: 0.3.1
+Version: 0.3.3
 Summary: PDF to markdown using Azure OpenAI batch processing
 Home-page: https://github.com/Summed-AI/parallex
 Author: Jeff Hostetler

{parallex-0.3.1 → parallex-0.3.3}/parallex/ai/open_ai_client.py

@@ -18,6 +18,7 @@ class OpenAIClient:
             azure_endpoint=os.getenv("AZURE_API_BASE"),
             api_key=os.getenv("AZURE_API_KEY"),
             api_version=os.getenv("AZURE_API_VERSION"),
+            timeout=3000,
         )
 
     async def upload(self, file_path: str) -> FileObject:

{parallex-0.3.1 → parallex-0.3.3}/parallex/parallex.py

@@ -106,6 +106,9 @@ async def _prompts_execute(
         process_semaphore = asyncio.Semaphore(concurrency)
         prompt_tasks = []
         for batch in batch_jobs:
+            logger.info(
+                f"waiting for batch to complete - {batch.id} - {batch.trace_id}"
+            )
             prompt_task = asyncio.create_task(
                 _wait_and_create_prompt_responses(batch=batch, client=open_ai_client, semaphore=process_semaphore)
             )
@@ -174,7 +177,7 @@ async def _execute(
 
         pages = [page for batch_pages in page_groups for page in batch_pages]
         logger.info(f"pages done. total pages- {len(pages)} - {trace_id}")
-        sorted_pages = sorted(pages, key=lambda x: x.prompt_index)
+        sorted_pages = sorted(pages, key=lambda x: x.page_number)
 
         # TODO add combined version of MD to output / save to file system
         callable_output = ParallexCallableOutput(

{parallex-0.3.1 → parallex-0.3.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "parallex"
-version = "0.3.1"
+version = "0.3.3"
 description = "PDF to markdown using Azure OpenAI batch processing"
 authors = ["Jeff Hostetler <jeff@summed.ai>", "Kevin Bao <kevin@summed.ai>"]
 repository = "https://github.com/Summed-AI/parallex"