mirrors/docling
1
0
Fork 0
mirror of https://github.com/DS4SD/docling.git synced 2025-12-08 12:48:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add VLM task interpreters

Browse Source 
Signed-off-by: Christoph Auer <cau@zurich.ibm.com>
This commit is contained in:
Christoph Auer
2025-08-14 20:48:10 +02:00
parent c8737f71da
commit c4de11bdb3
1 changed files with 0 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
docling/pipeline/threaded_multistage_vlm_pipeline.py
View File

@@ -366,8 +366,6 @@ class VlmProcessingStage(ThreadedPipelineStage):
one output item (success, empty, or failed). Non-cluster payloads are ignored.
"""
print("VLM stage: ", self.task_name)
groups: dict[int, list[ThreadedItem]] = defaultdict(list)
for itm in batch:
if (
@@ -456,8 +454,6 @@ class VlmProcessingStage(ThreadedPipelineStage):
# 3. Interpret predictions back into page/cluster structures
predictions_list: List[VlmPrediction] = list(vlm_predictions)
print("Predictions: ", predictions_list)
for i, (item, prediction) in enumerate(zip(item_mapping, predictions_list)):
assert item.payload is not None
assert isinstance(item.payload, ClusterItem)