yolov7_VitisAI

This is the implementation of YOLOv7 on Vitis AI (KV260).

See here for the overall tutorial.

Branches

main: train a new yolov7 model on PC
quantize: inspect, quantize and compile a model in docker environment
execute_v2: launch the model with compiled xmodel file on edge device (PetaLinux on KV260)

Name		Name	Last commit message	Last commit date
Latest commit History 155 Commits
MTV2		MTV2
cfg		cfg
data		data
deploy/triton-inference-server		deploy/triton-inference-server
figure		figure
inference/images		inference/images
models		models
paper		paper
scripts		scripts
tools		tools
utils		utils
.gitignore		.gitignore
DPU_list.txt		DPU_list.txt
Implement_of_Yolov7_on_Vitis_AI_KV260.pdf		Implement_of_Yolov7_on_Vitis_AI_KV260.pdf
LICENSE.md		LICENSE.md
MNIST_execute.py		MNIST_execute.py
README.md		README.md
detect.py		detect.py
export.py		export.py
hubconf.py		hubconf.py
requirements.txt		requirements.txt
test.py		test.py
test.sh		test.sh
test_docker.sh		test_docker.sh
test_quant.py		test_quant.py
test_quant.sh		test_quant.sh
torchhub.py		torchhub.py
train.py		train.py
train.sh		train.sh
train_aux.py		train_aux.py