cwerner/deadtrees

Issues

• Checkpoint file on S3?

  #97 opened 5 months ago by zexuan-everbright
  2

• pull the data

  #78 opened 3 years ago by haozidede
  3

• Investigate how to implement overlapping tiles

  #95 opened 2 years ago by cwerner
  1

• Geopandas dependency missing in setup.py

  #94 opened 2 years ago by cwerner
  0

• Add an alivetrees class

  #64 opened 2 years ago by selschwarz
  2

• Inference of RGB models broken

  #83 opened 2 years ago by cwerner
  2

• Add extra test sets for 2018 and 2020

  #91 opened 2 years ago by cwerner
  1

• Pre-train encoder weights with an autoencoder on our datasets

  #90 opened 2 years ago by cwerner
  1

• SWA causing crash at end of training

  #69 opened 3 years ago by cwerner
  2

• Problem in handling of CLI-passed models in inference

  #87 opened 3 years ago by cwerner
  1

• Ensemble inference cannot handle RGB models

  #88 opened 3 years ago by cwerner
  1

• GDice loss unstable for multi-class training

  #84 opened 3 years ago by cwerner
  1

• Multi-class setting leads to corrupted generalised dice loss

  #81 opened 3 years ago by cwerner
  0

• Add dedicated 2017 test set

  #55 opened 3 years ago by cwerner
  0

• Update UI inference interface

  #77 opened 3 years ago by cwerner
  0

• Bump dependencies

  #67 opened 3 years ago by cwerner
  1

• Remove torchvision

  #74 opened 3 years ago by cwerner
  1

• Duplicate log messages

  #75 opened 3 years ago by cwerner
  0

• Adapt to newest project template base settings

  #71 opened 3 years ago by cwerner
  0

• Clean up tensor transformations in segmodel

  #51 opened 3 years ago by cwerner
  0

• Consider boundary loss

  #65 opened 3 years ago by cwerner
  0

• Enhance data loading and transformation performance

  #68 opened 3 years ago by cwerner
  0

• Bump dependencies

  #48 opened 3 years ago by cwerner
  1

• Move to poetry

  #66 opened 3 years ago by cwerner
  0

• Add data for years 2018 and 2020

  #60 opened 3 years ago by cwerner
  0

• Update installation instructions

  #54 opened 3 years ago by selschwarz
  2

• Refactor inference script

  #57 opened 3 years ago by cwerner
  0

• Shard sizes currently have wrong number of samples

  #56 opened 3 years ago by cwerner
  0

• Check that the setup still works in single-dataset mode

  #52 opened 3 years ago by cwerner
  0

• Check if the normalisation is correct

  #19 opened 3 years ago by cwerner
  0

• Use 2+bg class setup

  #32 opened 3 years ago by cwerner
  0

• Cleanup preprocessing pipeline and move to smaller tile_size

  #49 opened 3 years ago by cwerner
  0

• Extent dataset to include NIR

  #33 opened 3 years ago by cwerner
  0

• Label smoothing

  #47 opened 3 years ago by cwerner
  0

• Add SLURM W&B sweep job

  #44 opened 3 years ago by cwerner
  0

• Remove MONAI from requirements

  #41 opened 3 years ago by cwerner
  0

• Move to generic model framework

  #39 opened 3 years ago by cwerner
  0

• Create up-to-date environment.yaml file

  #36 opened 3 years ago by cwerner
  0

• Include external UNet models

  #31 opened 3 years ago by cwerner
  0

• Warning: Leaking Caffe2 thread-pool after fork. (function pthreadpool)

  #34 opened 3 years ago by cwerner
  0

• Bug: Crash if IPython not installed

  #35 opened 3 years ago by cwerner
  0

• Handle incomplete tiles

  #24 opened 3 years ago by cwerner
  0

• Add automatic subtiling to inference mode

  #17 opened 3 years ago by cwerner
  0

• Add coverage and more tests

  #21 opened 4 years ago by cwerner
  0

• Move DVC preprocessing scripts to scripts directory

  #15 opened 4 years ago by cwerner
  0

• Source inference model from S3 bucket and make it selectable

  #18 opened 4 years ago by cwerner
  0

• Create batch inference module

  #11 opened 4 years ago by cwerner
  0

• Change y_hat map tile to multi-colour F1 result map

  #6 opened 4 years ago by cwerner
  0

• Move wandb sample generation to dedicated callback

  #14 opened 4 years ago by cwerner
  0

• Compute image stats from dataset shards

  #7 opened 4 years ago by cwerner
  0