rhysdg/vision-at-a-clip
Low-latency ONNX and TensorRT based zero-shot classification and detection with contrastive language-image pre-training based prompts
Jupyter Notebook
Pinned issues
Issues
- 3
how convert mm-groudingDino.pt to ONNX
#29 opened by wuqingzhou828 - 0
ONNX Conversion Script for GroundingDino
#30 opened by Vektor284 - 1
Grounding DINO For zero-shot object detection go ahead and build from the following example:
#25 opened by lfreee - 22
Runtime Exception Error
#27 opened by Masrur02 - 1
Shape Inference Error
#26 opened by Masrur02 - 0
Feat- add a Whisper (onnxruntime-genai) option for realtime prompts, robotics etc
#24 opened by rhysdg - 0
feat - add a vision at a clip universe subfolder
#23 opened by rhysdg - 0
Add BLIP-2, FLAVA and Ray Serve examples
#22 opened by rhysdg - 0
- 1
- 1
Evaluate lower onnx latency for Gdino
#18 opened by rhysdg - 0
Grounding DINO - ONNX inference class
#7 opened by rhysdg - 0
- 0
Evaluate kernl/ triton for an ONNX alternative
#15 opened by rhysdg - 0
Bug - Unhashable type AddedToken
#14 opened by rhysdg - 0
Strip of reliance on the transformer library
#8 opened by rhysdg - 0
Consolidate CLIP and SigLIP image preprocessing
#11 opened by rhysdg - 0
Add X-Clip
#10 opened by rhysdg - 0
Debug TensorRT support
#9 opened by rhysdg - 4
Request for ONNX Conversion Script
#6 opened by godiclee