TrojanXu/yolov5-tensorrt
Fork: 46 Star: 190 (更新于 2024-11-02 15:15:44)
license: Apache-2.0
Language: Python .
A tensorrt implementation of yolov5: https://github.com/ultralytics/yolov5
最后发布版本: v3.0 ( 2020-10-04 15:51:07)
yolov5-tensorrt
A tensorrt implementation of yolov5: https://github.com/ultralytics/yolov5
requirement
Please use torch>=1.6.0 + onnx>=1.6.0 + TRT 7.1+ (fix upsample issue) to run the sample code
onnx-simplifier-0.2.16
The code
Add newly implemented upsample to get this working with current combination of onnx and tensorrt.
0. prepare above mentioned environment.
- git clone && git submodule update --init
- download weights file (use yolov5/models/export.py)
- python main.py to run the benchmark
- Generally, for image of size 640*640, using batchsize=1, the speedup is 4x on V100.
Updates
- 20201004 update to track yolov5 - v3.0 release. download model file from official websites please.
TODO
- NMS support
- dynamic shape or dynamic batchsize support (won't implement soon because onnx-simplifier only supports fixed shape)
- FP16 numerical issue and performance investigation
- Benchmark
最近版本更新:(数据更新于 2024-09-11 21:15:05)
2020-10-04 15:51:07 v3.0
TrojanXu/yolov5-tensorrt同语言 Python最近更新仓库
2024-11-22 19:26:55 ultralytics/ultralytics
2024-11-22 08:12:43 jxxghp/MoviePilot
2024-11-22 06:12:44 dagster-io/dagster
2024-11-22 02:39:01 goauthentik/authentik
2024-11-22 00:15:39 jumpserver/jumpserver
2024-11-22 00:03:47 comfyanonymous/ComfyUI