Comments (7)
I will double check tonight or tomorrow and get back to you.
from anomaly-seg.
Yes.. Thanks..
If you have some opinions for me, please tell me...
Thanks.
from anomaly-seg.
Could you maybe provide some more context to your setup?
I downloaded the code and data from scratch. Then followed the instructions on the README. Had to install yacs which as a dependency I forgot to mention, but I believe is mentioned in the submodule requirements then started the training.
python3 train.py --gpus 0-1
[2022-01-06 23:14:56,706 INFO train.py line 240 1846380] Loaded configuration file config/ade20k-resnet50dilated-ppm_deepsup.yaml
[2022-01-06 23:14:56,706 INFO train.py line 241 1846380] Running with config:
DATASET:
imgMaxSize: 1000
imgSizes: (300, 375, 450, 525, 600)
list_train: ./data/training.odgt
list_val: ./data/validation.odgt
num_class: 150
padding_constant: 8
random_flip: True
root_dataset: ./data/
segm_downsampling_rate: 8
DIR: ckpt/ade20k-resnet50dilated-ppm_deepsup
MODEL:
arch_decoder: ppm_deepsup
arch_encoder: resnet50dilated
fc_dim: 2048
weights_decoder:
weights_encoder:
OOD:
exclude_back: False
ood: msp
out_labels: (13,)
TEST:
batch_size: 1
checkpoint: epoch_20.pth
result: ./
TRAIN:
batch_size_per_gpu: 2
beta1: 0.9
deep_sup_scale: 0.4
disp_iter: 20
epoch_iters: 5000
fix_bn: False
lr_decoder: 0.02
lr_encoder: 0.02
lr_pow: 0.9
num_epoch: 20
optim: SGD
seed: 304
start_epoch: 0
weight_decay: 0.0001
workers: 16
VAL:
batch_size: 1
checkpoint: epoch_20.pth
visualize: False
[2022-01-06 23:14:56,706 INFO train.py line 246 1846380] Outputing checkpoints to: ckpt/ade20k-resnet50dilated-ppm_deepsup
# samples: 5125
1 Epoch = 5000 iters
Epoch: [1][0/5000], Time: 9.62, Data: 2.50, lr_encoder: 0.020000, lr_decoder: 0.020000, Accuracy: 0.66, Loss: 7.690948
Epoch: [1][20/5000], Time: 1.21, Data: 0.16, lr_encoder: 0.019996, lr_decoder: 0.019996, Accuracy: 70.52, Loss: 2.431588
Epoch: [1][40/5000], Time: 0.96, Data: 0.10, lr_encoder: 0.019993, lr_decoder: 0.019993, Accuracy: 76.97, Loss: 1.634942
It seems to be training successfully.
from anomaly-seg.
Will close soon unless I get updated with more information. Otherwise I cannot reproduce your issue.
from anomaly-seg.
Facing the same issue, have u BEBUG it yet?
from anomaly-seg.
@LT1st can you describe your steps or what you did?
from anomaly-seg.
It seems the issue is with training with one GPU. I'll update the Readme and previous issue. The solution is something along these lines CSAILVision/semantic-segmentation-pytorch#58
But I haven't tested single GPU support and not sure when I'll be able to test it. Maybe sometime next week.
from anomaly-seg.
Related Issues (20)
- Is there any restriction regarding classes definition? HOT 5
- Question: Paper Section 3 HOT 4
- pretrained model HOT 4
- Issues with evaluate the model on OOD test set HOT 1
- Couldn't match the BDD100K file paths HOT 3
- StreetHazards color encoding for visualization HOT 6
- Some questions about the CAOS benchmark HOT 2
- Some points to confirm about the CAOS benchmark HOT 4
- Two more stages in resnet HOT 2
- how about the class define? HOT 2
- how to show the anomaly scores? HOT 1
- Evaluation based on each image vs entire dataset HOT 1
- What is the true label of -1 in seg_label HOT 1
- Anomaly detecting and visualization HOT 2
- About metrics AUPR, AUROC, and FPR95 HOT 2
- About ground truth Semantic Segmentation images HOT 4
- Broken link to BDD-Anomaly dataset HOT 1
- Updating the google drive link of the Species OOD dataset HOT 4
- Question about CAOS dataset training HOT 13
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 anomaly-seg.