43a616a955
PyTorch 1.6.0 compatability updates
5 years ago
fe094333e8
Update utils.py strip_optimizer() ( #509 )
...
Follow-on update that I missed adding into PR 509.
5 years ago
02445d176d
improved model.yaml source tracking
6 years ago
78c9420e09
--classes bug fix #387
6 years ago
094079bd26
update detect.py --class argument #387
6 years ago
e8cf24b6c8
Initial model ensemble capability #318
6 years ago
a40f615c6f
.half() bug fix
6 years ago
04bdbe4104
fuse update
6 years ago
2541f77946
update detect.py
6 years ago
1fdaa4987b
update detect.py
6 years ago
659ad748c8
update get_voc.sh
6 years ago
5323ad224d
export.py update
6 years ago
a62a1c2c67
export.py update
6 years ago
a9dc0c2c29
update check_img_size() for model strides
6 years ago
0d5065d3b9
Remove non-ascii that breaks pdb single-stepping
6 years ago
8c5b6220f5
Update detect.py
6 years ago
496ec33a33
Update detect.py
...
Added some recent updates that were missing, and updated the filename with an if else.
6 years ago
68f63616b3
Fix after suggestion
6 years ago
6c1b87a42e
update google_utils import
6 years ago
bbd12c7cfe
detect.py fast bug fix
6 years ago
2368603484
fix refrence bug
...
In torch==1.5, the import of the API has changed. Although it does not interrupt the operation of the program, it seems to me to be an implicit error and may throw an exception in later versions.
6 years ago
76ca367a01
FP16 to FP32 ckpt load
6 years ago
111994956e
switch default inference to FP16 on GPU
6 years ago
260b1729f0
FP16 inference update
6 years ago
099e6f5ebd
--img-size stride-multiple verification
6 years ago
ad71d2d513
update times to end-to-end latency (inference + NMS)
6 years ago
c14368d768
new CSP model release
6 years ago
c672bef10f
model fuse
6 years ago
dbdee3a4a3
multi-gpu test bug fix #15
6 years ago
ba9ab663a2
model.names for dataParallel
6 years ago
eb97b2e413
NMS fast mode
6 years ago
391492ee5b
updates
6 years ago
df7988d8d0
onnx_export.py
6 years ago
1e84a23f38
initial commit
6 years ago
Glenn Jocher
Russ Ferriday
Anish Hiranandani
Lornatang