Comments (5)
I know these three files, I have compared these three training files and there is a difference in calculating the LOSS, but loading and saving the model is the same. I mean, if I use train_dual.py for training and save the weights and model with auxiliary branches, how do I remove the auxiliary branches in detect?
from yolov9.
Take a look at this from yolov7 https://github.com/WongKinYiu/yolov7/blob/main/tools/reparameterization.ipynb
Load yolov9 model (model with auxiliary branches) and a new gelan model. This is equivalent to the train and deploy respectively in the above code. Move the weights and biases from the lead detection head to the new gelan model in the correct position. Ffor yolov9-c you want to move index 38 to index 22 and for yolov9-e you want to move index 49 to index 42.
When moving the weights and biases of everything you have to take into account that there is a Silence module added only in the yolov9 models. The Silence module is basically nn.Identity, it can be removed so rebase everything in the trained model by subtracting all indices by 1.
from yolov9.
Use train.py for gelan for gelan models (without aux branchs) and train_dual.py for yolov9 models (with aux branch).
from yolov9.
#1 (comment)
from yolov9.
https://github.com/WongKinYiu/yolov9?tab=readme-ov-file#re-parameterization
from yolov9.
Related Issues (20)
- How to improve yolov9 network to achieve dual data input?
- Is validation loss calculation available now ?
- Corresponding converted code of the yolov9u-pose model (reparameterization.ipynb) HOT 1
- How can I train with models/hub/yolov3.yaml
- real-time RTSP streams HOT 1
- Facing "CUDA out of memory" when running the "detect_dual.py" HOT 3
- IndexError: index 1 is out of bounds for dimension 0 with size 1 while using detect.py in yolov9 HOT 1
- Error about detect.py
- training won't start
- Combine Segmentation and Detection Model (YOLOV7 & YOLOV9) HOT 1
- Yolov9 Train AttributeError: 'FreeTypeFont' object has no attribute 'getsize and nan values HOT 2
- Some questions about YOLOv9.yaml, YOLOv9-c.yaml and YOLOv9-e.yaml HOT 1
- Target can't be detected
- Letterbox padding, `auto` option
- export openvino error
- ques
- question about val.py HOT 1
- about wandb HOT 1
- TensorRT - End2End support for Object Detection and Instance Segmentation
- save-period or save_period ? HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from yolov9.