Comments (1)
I am not the author itself but I think I can answer some of your questions.
There are certainly some issues in exporting the network to ONNX even with op-level 11 at least with Pytorch version 1.3 I have not tried with 1.4 yet. But the ONNX model exported is pretty much unusable even though it loads & runs, it produces no output and the problem seem to be in some of the later processing layers of the network that create the final prediction outputs they seem to be not really exportable as they use functionality that can not be probably traced by the PyTorch JIT. So you probably would have to modify the network structure to take this post-processing bit out of the network itself and just output the last layer of the feature extraction before and then try reexporting and move the post-processing inside the code executing your ONNX model.
from efficientdet.pytorch.
Related Issues (20)
- Training results in detections at all anchorboxes
- When I trained the code on my dataset, the mAP always kept 0.0. HOT 3
- zero confidence value
- is it possible to port official weights to your repo? HOT 1
- No boxes to NMS
- Can anyone train coco from scratch? HOT 1
- CUDA Out of Memory. The RTX 2080 TI 8GB can only afford the efficientdet-d3 with batch size 2. Why?
- 已经解决了win10下关于pytorch-EfficientDet下训练自己数据集的问题 HOT 2
- First time training, HTTP Error 403: Forbidden?! HOT 5
- threshold/iou_threshold
- KeyError: 'effcientdet-d0' How to solve this error? HOT 1
- train my datasets , IndexError: too many indices for array: array is 1-dimensional, but 2 were indexed HOT 1
- When is TorchScript planed? HOT 1
- urllib.error.HTTPError: HTTP Error 403: Forbidden HOT 2
- how many gpu and u guys using for batch 64?
- --multiprocessing-distributed --multiprocessing-distributed
- How to download checkpoint_VOC_efficientdet-d1_34.pth HOT 1
- Size mismatch problem.
- Gradient computation error HOT 2
- EfficientDet-D0(with weight) doesn't exist, can't be loaded.
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 efficientdet.pytorch.