CVU: New Deployment Tool For Yolov5 #4291
shivamswarnkar
started this conversation in
Show and tell
Replies: 2 comments 3 replies
-
@shivamswarnkar nice work on the documentation and profiling especially! |
Beta Was this translation helpful? Give feedback.
2 replies
-
@shivamswarnkar nice work!
|
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi Everyone !
@aditya-dl and I have been working on a pip package named CVU-Python that can help you deploy Yolov5 models in various frameworks and devices, without much hassle. We recently released its alpha-version and we would love to hear your feedback and suggestions on it.
Currently, we are giving support for GPU, CPU, and TPU devices through TensorRT, Torch-Script, ONNX, TensorFlow, and TFLite. And to switch between these devices and frameworks, all you need to do is to set
device
andbackend
parameters. CVU even tries to take care of most of the installation requirements (even for TensorRT, tho we have only tested on Colab yet).We are designing CVU to be used by both the experts and the novices. CVU aims at making CV pipelines easier to build and consistent around platforms, devices, and models.
We plan to extend CVU to include other common computer vision tasks (such as tracking with deep sort) in the upcoming days!
We are also making submissions in the Yolov5 Export challenge, so please also check that out as well (submission links).
Thank You all! Looking forward to hearing from you guys!
GPU (Colab-NVIDIA T4)
Based on 5000 inference iterations after 50 iterations of warmups. Includes Image Preprocessing (letterboxing etc.), Model Inference and Output Postprocessing (NMS, Scale-Coords, etc.) time.
CPU (Colab)
Note: We are performing more benchmarks, we will update info later on.
Based on 500 inference iterations after 10 iterations of warmups. Includes Image Preprocessing (letterboxing etc.), Model Inference and Output Postprocessing (NMS, Scale-Coords, etc.) time.
Beta Was this translation helpful? Give feedback.
All reactions