rtmo

3 contributors

History: 47 commits

Luigi

Add method to free buffers

b79a8af 5 months ago

video
Add a sample video 8 months ago
.gitattributes

1.66 kB

Add a sample video 8 months ago
README.md

1.47 kB

Update README.md 8 months ago
constant_folding.sh

337 Bytes

Apply constant folding 8 months ago
convert_to_fp16.py

1.25 kB

Re-quantize models in FP16 in keep positional encoding in FP32 to avoid bad accuracy 7 months ago
convert_to_fp16_all.py

1.36 kB

Bugfix: correct batch dimension fixation then regenerate ONNX models 8 months ago
convert_to_mixed.py

5.26 kB

Improve Debug Message Layout in Validate Funciton 7 months ago
demo.sh

56 Bytes

Remove models with fixed batch size as dynamic batch is generally supported 8 months ago
demo_batch.sh

71 Bytes

Unify batched and non-batched versions 6 months ago
onnx_to_engine.py

7.05 kB

Use dynamic batch size by default 6 months ago
rtmo-l.fp16.onnx

88 MB
LFS

Re-quantize models in FP16 in keep positional encoding in FP32 to avoid bad accuracy 7 months ago
rtmo-l.onnx

176 MB
LFS

Apply constant folding 8 months ago
rtmo-m.fp16.onnx

44.7 MB
LFS

Re-quantize models in FP16 in keep positional encoding in FP32 to avoid bad accuracy 7 months ago
rtmo-m.onnx

89.3 MB
LFS

Apply constant folding 8 months ago
rtmo-s.fp16.onnx

19.9 MB
LFS

Re-quantize models in FP16 in keep positional encoding in FP32 to avoid bad accuracy 7 months ago
rtmo-s.onnx

39.6 MB
LFS

Apply constant folding 8 months ago
rtmo-t.fp16.onnx

13.7 MB
LFS

Re-quantize models in FP16 in keep positional encoding in FP32 to avoid bad accuracy 7 months ago
rtmo-t.onnx

27.4 MB
LFS

Apply constant folding 8 months ago
rtmo_demo.py

2.66 kB

Return and show bounding box confidence 6 months ago
rtmo_demo_batch.py

3.64 kB

Return and show bounding box confidence 6 months ago
rtmo_gpu.py

31.5 kB

Add method to free buffers 5 months ago
symbolic_shape_infer.py

138 kB

Make ONNX models compatible with ONNXruntime's TensorrtExecutionProvider 8 months ago
symbolic_shape_infer_all.sh

135 Bytes

Make ONNX models compatible with ONNXruntime's TensorrtExecutionProvider 8 months ago