nv-ingest-api 2025.6.8.dev20250608__py3-none-any.whl → 2025.6.9.dev20250609__py3-none-any.whl

nv_ingest_api/interface/transform.py

@@ -207,7 +207,7 @@ def transform_image_create_vlm_caption(
         "api_key": api_key,
         "prompt": prompt,
         "endpoint_url": endpoint_url,
-        "image_caption_model_name": model_name,
+        "model_name": model_name,
     }
     filtered_task_config: Dict[str, str] = {k: v for k, v in task_config.items() if v is not None}
 

nv_ingest_api/internal/extract/pptx/engines/pptx_helper.py

@@ -105,7 +105,7 @@ def _finalize_images(
             logger.warning(f"Unable to process shape image: {e}")
 
     # If you want table/chart detection for these images, do it now
-    # (similar to docx approach). This might use your YOLO or other method:
+    # (similar to docx approach). This might use your YOLO or another method:
     detection_map = defaultdict(list)  # image_idx -> list of CroppedImageWithContent
     if extract_tables or extract_charts:
         try:
@@ -407,7 +407,7 @@ def python_pptx(
                     page_nearby_blocks["text"]["bbox"].append(get_bbox(shape_object=shape))
 
                 # Image processing (deferred)
-                if extract_images:
+                if extract_images or extract_tables or extract_charts:
                     try:
                         process_shape(
                             shape,

nv_ingest_api/internal/schemas/meta/ingest_job_schema.py

@@ -76,7 +76,7 @@ class IngestTaskCaptionSchema(BaseModelNoExt):
     api_key: Optional[str] = None
     endpoint_url: Optional[str] = None
     prompt: Optional[str] = None
-    caption_model_name: Optional[str] = None
+    model_name: Optional[str] = None
 
 
 class IngestTaskFilterParamsSchema(BaseModelNoExt):

nv_ingest_api/internal/schemas/transform/transform_image_caption_schema.py

@@ -10,6 +10,6 @@ class ImageCaptionExtractionSchema(BaseModel):
     api_key: str = "api_key"
     endpoint_url: str = "https://ai.api.nvidia.com/v1/gr/meta/llama-3.2-11b-vision-instruct/chat/completions"
     prompt: str = "Caption the content of this image:"
-    image_caption_model_name: str = "meta/llama-3.2-11b-vision-instruct"
+    model_name: str = "meta/llama-3.2-11b-vision-instruct"
     raise_on_failure: bool = False
     model_config = ConfigDict(extra="forbid")

nv_ingest_api/internal/transform/caption_image.py

@@ -173,7 +173,7 @@ def transform_image_create_vlm_caption_internal(
     api_key: str = task_config.get("api_key") or transform_config.api_key
     prompt: str = task_config.get("prompt") or transform_config.prompt
     endpoint_url: str = task_config.get("endpoint_url") or transform_config.endpoint_url
-    model_name: str = task_config.get("image_caption_model_name") or transform_config.image_caption_model_name
+    model_name: str = task_config.get("model_name") or transform_config.model_name
 
     # Create a mask for rows where the content type is "image".
     df_mask: pd.Series = df_transform_ledger["metadata"].apply(

{nv_ingest_api-2025.6.8.dev20250608.dist-info → nv_ingest_api-2025.6.9.dev20250609.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: nv-ingest-api
-Version: 2025.6.8.dev20250608
+Version: 2025.6.9.dev20250609
 Summary: Python module with core document ingestion functions.
 Author-email: Jeremy Dyer <jdyer@nvidia.com>
 License:                                  Apache License

{nv_ingest_api-2025.6.8.dev20250608.dist-info → nv_ingest_api-2025.6.9.dev20250609.dist-info}/RECORD

@@ -3,7 +3,7 @@ nv_ingest_api/interface/__init__.py,sha256=ltWlfmtCewHSRK4B7DF__QvlSUPuliz58JEcE
 nv_ingest_api/interface/extract.py,sha256=GyBfXKKTGwSb-y0k0nMiTf4HcCT2E-lxLY4aMYAPeOI,38815
 nv_ingest_api/interface/mutate.py,sha256=eZkd3sbHEJQiEPJyMbhewlPxQNMnL_Xur15icclnb-U,5934
 nv_ingest_api/interface/store.py,sha256=aR3Cf19lq9Yo9AHlAy1VVcrOP2dgyN01yYhwxyTprkQ,8207
-nv_ingest_api/interface/transform.py,sha256=NtRmS36LqL4P4vivuiDe1MqKaELeuh-rlO-9kb6skAE,15829
+nv_ingest_api/interface/transform.py,sha256=g6YnFR7TpEU0xNtzCvv6kqnFbuCwQ6vRMjjBxz3G4n4,15815
 nv_ingest_api/interface/utility.py,sha256=oXHV2Miz2BKviQg5vOVfiGSvPs2fKJsPDmnxe3fJL9c,7857
 nv_ingest_api/internal/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nv_ingest_api/internal/enums/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
@@ -39,7 +39,7 @@ nv_ingest_api/internal/extract/pdf/engines/pdf_helpers/__init__.py,sha256=Jk3wrQ
 nv_ingest_api/internal/extract/pptx/__init__.py,sha256=HIHfzSig66GT0Uk8qsGBm_f13fKYcPtItBicRUWOOVA,183
 nv_ingest_api/internal/extract/pptx/pptx_extractor.py,sha256=o-0P2dDyRFW37uQi_lKk6-eFozTcZvbq-2Y4I0EBMIY,7749
 nv_ingest_api/internal/extract/pptx/engines/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nv_ingest_api/internal/extract/pptx/engines/pptx_helper.py,sha256=IZu0c_RHDSJwwclOZD3_tDu5jg4AEEfumbwKB78dUE0,29716
+nv_ingest_api/internal/extract/pptx/engines/pptx_helper.py,sha256=p9-77JoG-ztMBrAFfvveSx94y0OE-Q4oCJfzw2iMVgI,29754
 nv_ingest_api/internal/mutate/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
 nv_ingest_api/internal/mutate/deduplicate.py,sha256=hmvTTGevpCtlkM_wVZSoc8-Exr6rUJwqLjoEnbPcPzY,3849
 nv_ingest_api/internal/mutate/filter.py,sha256=H-hOTBVP-zLpvQr-FoGIJKxkhtj4l_sZ9V2Fgu3rTEM,5183
@@ -82,7 +82,7 @@ nv_ingest_api/internal/schemas/message_brokers/request_schema.py,sha256=LZX_wXDx
 nv_ingest_api/internal/schemas/message_brokers/response_schema.py,sha256=4b275HlzBSzpmuE2wdoeaGKPCdKki3wuWldtRIfrj8w,727
 nv_ingest_api/internal/schemas/meta/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
 nv_ingest_api/internal/schemas/meta/base_model_noext.py,sha256=8hXU1uuiqZ6t8EsoZ8vlC5EFf2zSZrKEX133FcfZMwI,316
-nv_ingest_api/internal/schemas/meta/ingest_job_schema.py,sha256=JqlK66fg_eRQE9kwzAX4b15FqOSbpgrzLWkcWrWbKdE,8124
+nv_ingest_api/internal/schemas/meta/ingest_job_schema.py,sha256=szDvgc2A_JetD2Jyewyl4ac4lwpy3NiLxD9dOYz42sM,8116
 nv_ingest_api/internal/schemas/meta/metadata_schema.py,sha256=KqCUMeBNNgPtpBWzzNhsMtlfr_XvUxnALBbYkI-kfT4,6673
 nv_ingest_api/internal/schemas/mutate/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
 nv_ingest_api/internal/schemas/mutate/mutate_image_dedup_schema.py,sha256=k1JOdlPPpsipc0XhHf-9YxJ_-W0HvpVE1ZhYmr7fzj0,395
@@ -90,7 +90,7 @@ nv_ingest_api/internal/schemas/store/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQu
 nv_ingest_api/internal/schemas/store/store_embedding_schema.py,sha256=tdKeiraim9CDL9htgp4oUSCoPMoO5PrHBnlXqDyCpMw,956
 nv_ingest_api/internal/schemas/store/store_image_schema.py,sha256=p2LGij9i6sG6RYmsfdiQOiWIc2j-POjxYrNuMrp3ELU,1010
 nv_ingest_api/internal/schemas/transform/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/schemas/transform/transform_image_caption_schema.py,sha256=xLxXJsm8QeaL7KPe7m5sP2rd_AuNRMX29rdeVdoei3Y,582
+nv_ingest_api/internal/schemas/transform/transform_image_caption_schema.py,sha256=ORXAowdjxBUyfkw95eg2F82DRFqEsuV9PwNKVBulcmY,568
 nv_ingest_api/internal/schemas/transform/transform_image_filter_schema.py,sha256=31ThI5fr0yyENeJeE1xMAA-pxk1QVJLwM842zMate_k,429
 nv_ingest_api/internal/schemas/transform/transform_text_embedding_schema.py,sha256=ongmHkJA2953f9_RI7ZYzf5BUnFzVL6Al5E8WKyfgw4,885
 nv_ingest_api/internal/schemas/transform/transform_text_splitter_schema.py,sha256=D9K8tvu-tkEBQkZo7uuRzgrHdGyM3ZcNycHbHy5HV2E,791
@@ -98,7 +98,7 @@ nv_ingest_api/internal/store/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8y
 nv_ingest_api/internal/store/embed_text_upload.py,sha256=maxb4FPsBvWgvlrjAPEBlRZEFdJX5NxPG-p8kUbzV7I,9898
 nv_ingest_api/internal/store/image_upload.py,sha256=GNlY4k3pfcHv3lzXxkbmGLeHFsf9PI25bkBn6Xn9h3I,9654
 nv_ingest_api/internal/transform/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
-nv_ingest_api/internal/transform/caption_image.py,sha256=RYL_b26zfaRlbHz0XvLw9HwaMlXpNhr7gayjxGzdALQ,8545
+nv_ingest_api/internal/transform/caption_image.py,sha256=0ILCG2F8ESqKtZiPUM-6F1BHUflFZ76Dzi2GNzkE-lU,8517
 nv_ingest_api/internal/transform/embed_text.py,sha256=F8kg-WXihtuUMwDQUUYjnfGDCdQp1Mkd-jeThOiJT0s,16507
 nv_ingest_api/internal/transform/split_text.py,sha256=DlVoyHLqZ-6_FiWwZmofPcq7TX8Ta23hIE0St9tw1IY,6822
 nv_ingest_api/util/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
@@ -150,8 +150,8 @@ nv_ingest_api/util/service_clients/rest/rest_client.py,sha256=dZ-jrk7IK7oNtHoXFS
 nv_ingest_api/util/string_processing/__init__.py,sha256=mkwHthyS-IILcLcL1tJYeF6mpqX3pxEw5aUzDGjTSeU,1411
 nv_ingest_api/util/system/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nv_ingest_api/util/system/hardware_info.py,sha256=ORZeKpH9kSGU_vuPhyBwkIiMyCViKUX2CP__MCjrfbU,19463
-nv_ingest_api-2025.6.8.dev20250608.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-nv_ingest_api-2025.6.8.dev20250608.dist-info/METADATA,sha256=gNSa5IP1pXAw5PlHmyQU7K1RMJZWdsvaIqszQfvevIM,13918
-nv_ingest_api-2025.6.8.dev20250608.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-nv_ingest_api-2025.6.8.dev20250608.dist-info/top_level.txt,sha256=abjYMlTJGoG5tOdfIB-IWvLyKclw6HLaRSc8MxX4X6I,14
-nv_ingest_api-2025.6.8.dev20250608.dist-info/RECORD,,
+nv_ingest_api-2025.6.9.dev20250609.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+nv_ingest_api-2025.6.9.dev20250609.dist-info/METADATA,sha256=u6ShxLAz31PXFhJkQGiBLGZ_c_N2HEalSTuy2oN24WA,13918
+nv_ingest_api-2025.6.9.dev20250609.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+nv_ingest_api-2025.6.9.dev20250609.dist-info/top_level.txt,sha256=abjYMlTJGoG5tOdfIB-IWvLyKclw6HLaRSc8MxX4X6I,14
+nv_ingest_api-2025.6.9.dev20250609.dist-info/RECORD,,