Merge pull request #798 from runarmod/fix-tqdm-total

Fix: show tqdm total iteration count

Author: VikParuchuri

marker/processors/llm/__init__.py

@@ -156,7 +156,9 @@ def rewrite_blocks(self, document: Document):
             return
 
         pbar = tqdm(
-            desc=f"{self.__class__.__name__} running", disable=self.disable_tqdm
+            total=total_blocks,
+            desc=f"{self.__class__.__name__} running",
+            disable=self.disable_tqdm
         )
         with ThreadPoolExecutor(max_workers=self.max_concurrency) as executor:
             for future in as_completed(

marker/processors/llm/llm_mathblock.py

@@ -142,7 +142,9 @@ def rewrite_blocks(self, document: Document):
             return
 
         pbar = tqdm(
-            desc=f"{self.__class__.__name__} running", disable=self.disable_tqdm
+            total=total_blocks,
+            desc=f"{self.__class__.__name__} running",
+            disable=self.disable_tqdm
         )
         with ThreadPoolExecutor(max_workers=self.max_concurrency) as executor:
             for future in as_completed(

marker/processors/llm/llm_table_merge.py

@@ -158,8 +158,7 @@ def rewrite_blocks(self, document: Document):
         if self.no_merge_tables_across_pages:
             logger.info("Skipping table merging across pages due to --no_merge_tables_across_pages flag")
             return
-            
-        pbar = tqdm(desc=f"{self.__class__.__name__} running", disable=self.disable_tqdm)
+
         table_runs = []
         table_run = []
         prev_block = None
@@ -221,6 +220,17 @@ def rewrite_blocks(self, document: Document):
         if table_run:
             table_runs.append(table_run)
 
+        # Don't show progress if there is nothing to process
+        total_table_runs = len(table_runs)
+        if total_table_runs == 0:
+            return
+
+        pbar = tqdm(
+            total=total_table_runs,
+            desc=f"{self.__class__.__name__} running",
+            disable=self.disable_tqdm,
+        )
+
         with ThreadPoolExecutor(max_workers=self.max_concurrency) as executor:
             for future in as_completed([
                 executor.submit(self.process_rewriting, document, blocks)