cosmiq / cosmiq_sn6_baseline Goto Github PK

View Code? Open in Web Editor NEW

35.0 35.0 16.0 81.25 MB

Baseline algorithm for the SpaceNet 6 Challenge

License: Apache License 2.0

Dockerfile 5.57% Python 91.58% Shell 2.85%

cosmiq_sn6_baseline's People

Contributors

Stargazers

Watchers

Forkers

nmoran zhuoyuwei derrickwang005 sachin9390 ternaus tatigabru antoinecollas hbpeters rkolyan getawork rituyadav92 tandaman03 kevytong sumanth181099 singhdavinderpa1 kratya1223

cosmiq_sn6_baseline's Issues

Training with 3 channels images

The current solution forces the use of 4 channels images. Trying to use 3 channel images I have the following error even having specified in the YAML configuration image type = 24 bit and channels = 3

Traceback (most recent call last):
  File "./baseline.py", line 883, in <module>
    train(args)
  File "./baseline.py", line 599, in train
    trainer.train()
  File "/opt/conda/envs/solaris/lib/python3.7/site-packages/solaris/nets/train.py", line 139, in train
    output = self.model(data)
  File "/opt/conda/envs/solaris/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__
    result = self.forward(*input, **kwargs)
  File "/root/model.py", line 95, in forward
    conv1 = self.relu(self.conv1(x))
  File "/opt/conda/envs/solaris/lib/python3.7/site-packages/torch/nn/modules/module.py", line 532, in __call__
    result = self.forward(*input, **kwargs)
  File "/opt/conda/envs/solaris/lib/python3.7/site-packages/torch/nn/modules/conv.py", line 345, in forward
    return self.conv2d_forward(input, self.weight)
  File "/opt/conda/envs/solaris/lib/python3.7/site-packages/torch/nn/modules/conv.py", line 342, in conv2d_forward
    self.padding, self.dilation, self.groups)
RuntimeError: Given groups=1, weight of size 64 4 3 3, expected input[8, 3, 512, 512] to have 4 channels, but got 3 channels instead

Problem loading the included weights - best_epoch19_0.6000000238418579.model

Hi,

I tried loading the included model via torch.load() and got the following error:

model = torch.load(path, map_location=torch.device('cpu') )

RuntimeError: unexpected EOF, expected 10696272 more bytes. The file might be corrupted.

I'm using torch 1.5.0.

Thanks !

Dockerfile not working

When trying to build a docker image from the root folder with

docker build .

Fails with the following messages

Executing transaction: failed                                                                                           
ERROR conda.core.link:_execute(502): An error occurred while installing package 'conda-forge::affine-2.3.0-py_0'.
FileNotFoundError(2, "No such file or directory: '/opt/conda/envs/solaris/bin/python3.6'")
Attempting to roll back.

Rolling back transaction: done

FileNotFoundError(2, "No such file or directory: '/opt/conda/envs/solaris/bin/python3.6'")

Appears that it is attempting to install the affine package before installing the python interpreter itself. Have found a work around which involves first creating the environment with

conda create -n solaris python=3.6.7

And then installing the packages

conda install --file environment-gpu.txt

Where envrionment-gpu.txt is a flat list of the packages from the environment-gpu.yml file without the channel information or other structural elements.

cosmiq / cosmiq_sn6_baseline Goto Github PK

cosmiq_sn6_baseline's People

Contributors

Stargazers

Watchers

Forkers

cosmiq_sn6_baseline's Issues

Training with 3 channels images

Problem loading the included weights - best_epoch19_0.6000000238418579.model

Dockerfile not working

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent