ML4GW/DeepClean

Issues

• Running deepclean of MDC (pseudo) replay at LHO

  #67 opened a year ago
  0

• Correlate test metrics with parameter recovery

  #65 opened 2 years ago
  3

• in `gwftools.find`, add logic for finding files that contain superset of data needed

  #60 opened 2 years ago
  0

• Use `tritonserve` library instead of `deepclean.serve`

  #58 opened 2 years ago
  1

• Add filters for CI

  #57 opened 2 years ago
  1

• Fix probabilistic behavior in PSD loss test in `deepclean.trainer`

  #56 opened 2 years ago
  0

• Create sandbox project for generating replays

  #55 opened 2 years ago
  0

• Create CI testing for `sandbox/analyze` project

  #54 opened 2 years ago
  0

• Create CI testing for `sandbox/infer` project

  #53 opened 2 years ago
  0

• Create CI testing for `sandbox/export` project

  #52 opened 2 years ago
  0

• Add CI testing to `sandbox/train`

  #51 opened 2 years ago
  0

• Build a dedicated visualization library

  #50 opened 2 years ago
  0

• Build dedicated model analysis library

  #49 opened 2 years ago
  0

• Versioning data used for training

  #48 opened 2 years ago
  0

• Add testing to sandbox projects

  #47 opened 2 years ago
  1

• Add filters to CI

  #44 opened 2 years ago
  1

• Update CI to use project-wide user

  #42 opened 2 years ago
  0

• Architecture library won't work for more than one architecture

  #41 opened 2 years ago
  0

• How to name online framefiles when there is non-integer aggregation latency?

  #40 opened 2 years ago
  9

• Images for the wiki

  #38 opened 2 years ago
  0

• Build experiment tracking framework

  #35 opened 2 years ago
  0

• Parameterize autoencoder architecture

  #34 opened 2 years ago
  1

• Can we make transformers work here?

  #33 opened 2 years ago
  0

• Using convolutional kernels with sizes that are multiples of 8

  #32 opened 2 years ago
  0

• Relationship between receptive field and learnable bandwidth

  #31 opened 2 years ago
  0

• DeepClean receptive field

  #30 opened 2 years ago
  0

• Saliency map analysis

  #29 opened 2 years ago
  1

• Shuffling produces empirically worse results

  #28 opened 2 years ago
  1

• Train stride tradeoffs

  #27 opened 2 years ago
  0

• Performance on data at the edge of kernels

  #26 opened 2 years ago
  1

• Why does DeepClean take so long to get to an ASD of 1?

  #25 opened 2 years ago
  1

• Alternative loss functions

  #24 opened 2 years ago
  0

• How many FFTs required for stable ASD loss?

  #23 opened 2 years ago
  0

• Move normalization/denormalization onto the server

  #22 opened 2 years ago
  0

• Replacing No-data channels with dummy dc timeseries

  #20 opened 3 years ago
  6

• NaN in the recent offline training

  #19 opened 2 years ago
  9

• Issue templates

  #14 opened 3 years ago
  0

• Unit testing

  #11 opened 2 years ago
  0

• Newer TensorRT and Triton containers

  #10 opened 3 years ago
  1

• Branching and release cycles

  #9 opened 3 years ago
  0

• Containerize applications

  #8 opened 3 years ago
  1

• Hyperparameter search

  #7 opened 3 years ago
  0

• Documentation

  #6 opened 3 years ago
  1

• Create CI for unit testing

  #5 opened 2 years ago
  2

• Implement production inference code

  #4 opened 2 years ago
  0

• Implement retraining pipeline

  #3 opened 3 years ago
  0

• Reorganized `projects` directory

  #2 opened 3 years ago
  1

• Automate existing projects

  #1 opened 3 years ago
  0