Comments (2)
We are sorry that we will only provide original checkpoints reported in the paper since as an intern in MYbank I only transferred checkpoints in GPU servers to the company internal storage.
But we are now also reproducing models ourself in the university lab. We
provide the reproduced adamixer_r50_1x_coco.py
(42.5AP on 8 Titan Xp cards) log file 20220412_001456.log.
The validation mAP per epoch is
2022-04-12 01:27:55,144 - mmdet - INFO - Epoch(val) [1][7330] bbox_mAP: 0.2500, bbox_mAP_50: 0.4100, bbox_mAP_75: 0.2590, bbox_mAP_s: 0.1220, bbox_mAP_m: 0.2700, bbox_mAP_l: 0.3530, bbox_mAP_copypaste: 0.250 0.410 0.259 0.122 0.270 0.353
2022-04-12 02:40:50,004 - mmdet - INFO - Epoch(val) [2][7330] bbox_mAP: 0.3030, bbox_mAP_50: 0.4780, bbox_mAP_75: 0.3190, bbox_mAP_s: 0.1570, bbox_mAP_m: 0.3340, bbox_mAP_l: 0.4210, bbox_mAP_copypaste: 0.303 0.478 0.319 0.157 0.334 0.421
2022-04-12 03:53:27,222 - mmdet - INFO - Epoch(val) [3][7330] bbox_mAP: 0.3330, bbox_mAP_50: 0.5110, bbox_mAP_75: 0.3540, bbox_mAP_s: 0.1890, bbox_mAP_m: 0.3560, bbox_mAP_l: 0.4580, bbox_mAP_copypaste: 0.333 0.511 0.354 0.189 0.356 0.458
2022-04-12 05:05:57,427 - mmdet - INFO - Epoch(val) [4][7330] bbox_mAP: 0.3470, bbox_mAP_50: 0.5230, bbox_mAP_75: 0.3720, bbox_mAP_s: 0.1930, bbox_mAP_m: 0.3830, bbox_mAP_l: 0.4770, bbox_mAP_copypaste: 0.347 0.523 0.372 0.193 0.383 0.477
2022-04-12 06:18:39,438 - mmdet - INFO - Epoch(val) [5][7330] bbox_mAP: 0.3640, bbox_mAP_50: 0.5460, bbox_mAP_75: 0.3870, bbox_mAP_s: 0.2110, bbox_mAP_m: 0.3910, bbox_mAP_l: 0.5100, bbox_mAP_copypaste: 0.364 0.546 0.387 0.211 0.391 0.510
2022-04-12 07:34:31,302 - mmdet - INFO - Epoch(val) [6][7330] bbox_mAP: 0.3760, bbox_mAP_50: 0.5600, bbox_mAP_75: 0.4030, bbox_mAP_s: 0.2160, bbox_mAP_m: 0.4060, bbox_mAP_l: 0.5140, bbox_mAP_copypaste: 0.376 0.560 0.403 0.216 0.406 0.514
2022-04-12 08:47:14,789 - mmdet - INFO - Epoch(val) [7][7330] bbox_mAP: 0.3830, bbox_mAP_50: 0.5630, bbox_mAP_75: 0.4090, bbox_mAP_s: 0.2110, bbox_mAP_m: 0.4170, bbox_mAP_l: 0.5280, bbox_mAP_copypaste: 0.383 0.563 0.409 0.211 0.417 0.528
2022-04-12 09:59:53,450 - mmdet - INFO - Epoch(val) [8][7330] bbox_mAP: 0.3930, bbox_mAP_50: 0.5780, bbox_mAP_75: 0.4230, bbox_mAP_s: 0.2380, bbox_mAP_m: 0.4240, bbox_mAP_l: 0.5400, bbox_mAP_copypaste: 0.393 0.578 0.423 0.238 0.424 0.540
2022-04-12 11:12:28,376 - mmdet - INFO - Epoch(val) [9][7330] bbox_mAP: 0.4190, bbox_mAP_50: 0.6060, bbox_mAP_75: 0.4500, bbox_mAP_s: 0.2500, bbox_mAP_m: 0.4450, bbox_mAP_l: 0.5750, bbox_mAP_copypaste: 0.419 0.606 0.450 0.250 0.445 0.575
2022-04-12 12:25:17,758 - mmdet - INFO - Epoch(val) [10][7330] bbox_mAP: 0.4220, bbox_mAP_50: 0.6090, bbox_mAP_75: 0.4540, bbox_mAP_s: 0.2450, bbox_mAP_m: 0.4510, bbox_mAP_l: 0.5800, bbox_mAP_copypaste: 0.422 0.609 0.454 0.245 0.451 0.580
2022-04-12 13:37:56,010 - mmdet - INFO - Epoch(val) [11][7330] bbox_mAP: 0.4250, bbox_mAP_50: 0.6120, bbox_mAP_75: 0.4560, bbox_mAP_s: 0.2460, bbox_mAP_m: 0.4510, bbox_mAP_l: 0.5830, bbox_mAP_copypaste: 0.425 0.612 0.456 0.246 0.451 0.583
2022-04-12 14:50:40,213 - mmdet - INFO - Epoch(val) [12][7330] bbox_mAP: 0.4250, bbox_mAP_50: 0.6140, bbox_mAP_75: 0.4570, bbox_mAP_s: 0.2500, bbox_mAP_m: 0.4510, bbox_mAP_l: 0.5820, bbox_mAP_copypaste: 0.425 0.614 0.457 0.250 0.451 0.582
from adamixer.
Many thanks to you. That's very helpful. I do have not enough GPUs to reproduce the baseline in the University Lab. Therefore, the baseline log file is largely helpful.
Regards,
from adamixer.
Related Issues (20)
- The Conversion between (x, y, z, r) and (c, c, w, h) in code is not same with the paper? HOT 2
- work_dirs files HOT 2
- RuntimeError: output with shape [16, 300, 300] doesn't match the broadcast shape [1, 16, 300, 300] HOT 5
- [BUG] when training with pure background images, loss tend to be inf.
- RuntimeError: output with shape [16, 100, 100] doesn't match the broadcast shape [1, 16, 100, 100] HOT 1
- Release the Trained Models HOT 4
- Excuse me, does COCO minival mean coco as a whole dataset or as part of it? HOT 1
- [BUG] params M's dimention shape mismatch HOT 1
- matrix contains invalid numeric entries HOT 2
- How to compute adamixer's FLOPs? HOT 3
- ValueError: matrix contains invalid numeric entries HOT 4
- code for Visualization HOT 1
- Pytorch version HOT 2
- Why cannot I find Adamixer in the mmdetection? HOT 2
- Welcome update to OpenMMLab 2.0
- 公式5
- 推理gpu资源
- 问个问题, HOT 1
- Visualization about sampling points HOT 3
- About the "debug" and "postfix" in config file HOT 2
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 adamixer.