Comments (2)
Hi,
We provide training scripts for usage with the command line, see https://github.com/facebookresearch/detr#training
For more information on how to train on your own dataset, check #9 and #28
As such, I'm closing this issue as a duplicate of #9
from detr.
Hi there,
I have been using DETR on my own dataset and it works very well. I get a good mAP and Recall on the validation set. My question is, how to I run cocoEval to give me the same or similar results to what it got during model training. For example the model achieved and mAP of 0.89 on the validation set. I then decided to see if I could produce the same results again. I ran the model in eval mode on the dataset and set a confidence threshold > 0.8 and saved the results in a json file. I then used cocoeval and gave the validation set json and my new resFile as inputs and the evaluation results gave me an mAP of 0.6, which isnβt right. How do I go about getting the same or similar results as to what the model achieved originally and consequently how do I adjust the confidence threshold and get the precision-recall curves for these different thresholds. I should have said that I did try the cocoeval with all predictions (no filter on the confidences) and my mAP result was still a lot lower than what the model showed me. I guess my real question is, what code / steps should I take so that I can get the same results as what the model gave on the validation set? How to I replicate these figures ? What do I need to do to achieve this?
So far I ran the model in eval mode, ran predictions on my dataset and saved them into a json file. I then ran the code as shown in the attached image, but the results I got were no where near the same as what the model gave.
thank you.
from detr.
Related Issues (20)
- Keyerror: image_id (training detr on custom dataset
- How to train with a custom dataset on mac m2?
- continuously growing memory
- Question about object queries. HOT 4
- I want to train the DETR model on a CPU. How can I make it possible on a small computer, 8gb RAM HOT 3
- Why positional encoding is added to different role in encoder and decoder. HOT 1
- π Bug: Architecture diagram in README.md renders incorrectly when using dark mode
- continue training with chekckpoint
- How to finetune DETR for semantic segmentation task?
- I do not understand what the mask meaning in "samlpes"
- Process finished with exit code 137 (interrupted by signal 9: SIGKILL)Please read & provide the following
- Very low performance for segmentation task.
- box_cxcywh_to_xyxy
- ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -9) local_rank: 6 (pid: 257736) of binary: /home/public/anaconda3/envs/DL/bin/python
- Average Precision of each class for best epoch and then it's mean HOT 1
- the mAP is chage
- I think there are some errors in the posted code HOT 6
- Queries for images with low number of objects HOT 2
- RuntimeError: Error(s) in loading state_dict for DETRsegm: HOT 2
- Map metrics anomalies after backbone replacement
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
π 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. πππ
-
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 β€οΈ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from detr.