Merge pull request #2151 from myhloli/dev

refactor(ocr): improve OCR score precision to three decimal places

magic_pdf/model/batch_analyze.py

@@ -241,7 +241,7 @@ class BatchAnalyze:
                     for index, layout_res_item in enumerate(need_ocr_lists_by_lang[lang]):
                         ocr_text, ocr_score = ocr_res_list[index]
                         layout_res_item['text'] = ocr_text
-                        layout_res_item['score'] = float(round(ocr_score, 2))
+                        layout_res_item['score'] = float(f"{ocr_score:.3f}")
 
                     total_processed += len(img_crop_list)
 

magic_pdf/pdf_parse_union_core_v2.py

@@ -997,7 +997,7 @@ def pdf_parse_union(
         for index, span in enumerate(need_ocr_list):
             ocr_text, ocr_score = ocr_res_list[index]
             span['content'] = ocr_text
-            span['score'] = float(round(ocr_score, 2))
+            span['score'] = float(f"{ocr_score:.3f}")
         # rec_time = time.time() - rec_start
         # logger.info(f'ocr-dynamic-rec time: {round(rec_time, 2)}, total images processed: {len(img_crop_list)}')