{"id":13643330,"url":"https://github.com/lewes6369/TensorRT-Yolov3","last_synced_at":"2025-04-21T01:32:18.748Z","repository":{"id":41519397,"uuid":"160379243","full_name":"lewes6369/TensorRT-Yolov3","owner":"lewes6369","description":"TensorRT for Yolov3","archived":false,"fork":false,"pushed_at":"2022-09-26T19:26:41.000Z","size":74,"stargazers_count":487,"open_issues_count":42,"forks_count":165,"subscribers_count":19,"default_branch":"master","last_synced_at":"2024-11-09T15:42:27.791Z","etag":null,"topics":["caffe","tensorrt","yolov3"],"latest_commit_sha":null,"homepage":null,"language":"C++","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/lewes6369.png","metadata":{"files":{"readme":"readme.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2018-12-04T15:32:28.000Z","updated_at":"2024-10-10T02:15:19.000Z","dependencies_parsed_at":"2023-01-18T10:51:21.412Z","dependency_job_id":null,"html_url":"https://github.com/lewes6369/TensorRT-Yolov3","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lewes6369%2FTensorRT-Yolov3","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lewes6369%2FTensorRT-Yolov3/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lewes6369%2FTensorRT-Yolov3/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lewes6369%2FTensorRT-Yolov3/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/lewes6369","download_url":"https://codeload.github.com/lewes6369/TensorRT-Yolov3/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":249982607,"owners_count":21355732,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["caffe","tensorrt","yolov3"],"created_at":"2024-08-02T01:01:45.984Z","updated_at":"2025-04-21T01:32:18.383Z","avatar_url":"https://github.com/lewes6369.png","language":"C++","funding_links":[],"categories":["Lighter and Deployment Frameworks","Applications"],"sub_categories":[],"readme":"# TRTForYolov3\n\n## Desc\n\n    tensorRT for Yolov3\n\n### Test Enviroments\n\n    Ubuntu  16.04\n    TensorRT 5.0.2.6/4.0.1.6\n    CUDA 9.2\n\n### Models\n\nDownload the caffe model converted by official model:\n\n+ Baidu Cloud [here](https://pan.baidu.com/s/1VBqEmUPN33XrAol3ScrVQA) pwd: gbue\n+ Google Drive [here](https://drive.google.com/open?id=18OxNcRrDrCUmoAMgngJlhEglQ1Hqk_NJ)\n\n\nIf run model trained by yourself, comment the \"upsample_param\" blocks, and modify the prototxt the last layer as:\n```\nlayer {\n    #the bottoms are the yolo input layers\n    bottom: \"layer82-conv\"\n    bottom: \"layer94-conv\"\n    bottom: \"layer106-conv\"\n    top: \"yolo-det\"\n    name: \"yolo-det\"\n    type: \"Yolo\"\n}\n```\n\nIt also needs to change the yolo configs in \"YoloConfigs.h\" if different kernels.\n\n### Run Sample\n\n```bash\n#build source code\ngit submodule update --init --recursive\nmkdir build\ncd build \u0026\u0026 cmake .. \u0026\u0026 make \u0026\u0026 make install\ncd ..\n\n#for yolov3-608\n./install/runYolov3 --caffemodel=./yolov3_608.caffemodel --prototxt=./yolov3_608.prototxt --input=./test.jpg --W=608 --H=608 --class=80\n\n#for fp16\n./install/runYolov3 --caffemodel=./yolov3_608.caffemodel --prototxt=./yolov3_608.prototxt --input=./test.jpg --W=608 --H=608 --class=80 --mode=fp16\n\n#for int8 with calibration datasets\n./install/runYolov3 --caffemodel=./yolov3_608.caffemodel --prototxt=./yolov3_608.prototxt --input=./test.jpg --W=608 --H=608 --class=80 --mode=int8 --calib=./calib_sample.txt\n\n#for yolov3-416 (need to modify include/YoloConfigs for YoloKernel)\n./install/runYolov3 --caffemodel=./yolov3_416.caffemodel --prototxt=./yolov3_416.prototxt --input=./test.jpg --W=416 --H=416 --class=80\n```\n\n\n### Performance\n\nModel | GPU | Mode | Inference Time\n-- | -- | -- | -- \nYolov3-416 |  GTX 1060 | Caffe | 54.593ms\nYolov3-416 |  GTX 1060 | float32 | 23.817ms\nYolov3-416 |  GTX 1060 | int8 | 11.921ms\nYolov3-608 |  GTX 1060 | Caffe | 88.489ms\nYolov3-608 | GTX 1060 | float32 | 43.965ms\nYolov3-608 |  GTX 1060 | int8 | 21.638ms\nYolov3-608 | GTX 1080 Ti | float32 | 19.353ms\nYolov3-608 | GTX 1080 Ti | int8 | 9.727ms\nYolov3-416 |  GTX 1080 Ti | float32 | 9.677ms\nYolov3-416 |  GTX 1080 Ti | int8 | 6.129ms  | li\n\n### Eval Result\n\nrun above models with appending ```--evallist=labels.txt```\n\nint8 calibration data made from 200 pics selected in val2014 (see scripts dir)\n\nModel | GPU | Mode | dataset | MAP(0.50) | MAP(0.75)\n-- | -- | -- | -- | -- | --\nYolov3-416 | GTX 1060 | Caffe(fp32) | COCO val2014 | 50.33 | 33.00\nYolov3-416 | GTX 1060 | float32 | COCO val2014 | 50.27 | 32.98\nYolov3-416 | GTX 1060 | int8 | COCO val2014 | 44.15 | 30.24\nYolov3-608 | GTX 1060 | Caffe(fp32) | COCO val2014 | 52.89 | 35.31\nYolov3-608 | GTX 1060 | float32 | COCO val2014 |  52.84 | 35.26\nYolov3-608 | GTX 1060 | int8 | COCO val2014 |  48.55 | 35.53 | li\n\n\nNotice: \n+ caffe implementation is little different in yolo layer and nms, and it should be the similar result compared to tensorRT fp32. \n\n### Details About Wrapper\n\nsee link [TensorRTWrapper](https://github.com/lewes6369/tensorRTWrapper)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flewes6369%2FTensorRT-Yolov3","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flewes6369%2FTensorRT-Yolov3","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flewes6369%2FTensorRT-Yolov3/lists"}