upstream-archive/fastembed
1
0
Fork 0
mirror of https://github.com/qdrant/fastembed.git synced 2026-07-23 11:20:51 -05:00
Code Issues Packages Projects Releases Wiki Activity

fix: Pass providers and cuda to multimodal models

Browse Source
This commit is contained in:
hh-space-invader
2025-03-13 08:24:04 +02:00
parent 4bc5fbedf9
commit 9e21de21d6
1 changed files with 6 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
fastembed/late_interaction_multimodal/onnx_multimodal_model.py
View File

@@ -151,7 +151,9 @@ class OnnxMultimodalModel(OnnxModel[T]):
if not hasattr(self, "model") or self.model is None:
self.load_onnx_model()
for batch in iter_batch(documents, batch_size):
yield from self._post_process_onnx_text_output(self.onnx_embed_text(batch))
yield from self._post_process_onnx_text_output(
self.onnx_embed_text(batch, cuda=cuda, providers=providers)
)
else:
if parallel == 0:
parallel = os.cpu_count()
@@ -228,7 +230,9 @@ class OnnxMultimodalModel(OnnxModel[T]):
self.load_onnx_model()
for batch in iter_batch(images, batch_size):
yield from self._post_process_onnx_image_output(self.onnx_embed_image(batch))
yield from self._post_process_onnx_image_output(
self.onnx_embed_image(batch, cuda=cuda, providers=providers)
)
else:
if parallel == 0:
parallel = os.cpu_count()