jpata/particleflow

Issues

• pytorch evaluation sometimes segfaults

  #317 opened 2 months ago by jpata
  1

• run CMSSW workflows with CHS as well as PUPPI

  #335 opened 2 months ago by jpata
  1

• unify Key4HEP postprocessing scripts

  #333 opened 3 months ago by jpata
  0

• Develop visualizations for attention matrices

  #329 opened 4 months ago by jpata
  0

• Run CMSSW validation of the 2024 attention model

  #321 opened 4 months ago by jpata
  5

• Ray workflow fails due to new ray version

  #331 opened 4 months ago by jpata
  0

• Integrate new pytorch attention model in CMSSW via ONNX

  #216 opened 4 months ago by jpata
  19

• redo JEC in CMSSW

  #328 opened 4 months ago by jpata
  0

• Investigate and learn from outliers in CMS

  #327 opened 4 months ago by jpata
  0

• CMSSW validation on data

  #326 opened 4 months ago by jpata
  0

• add convert_int8 option

  #315 opened 4 months ago by jpata
  1

• Document CMSSW workflows

  #318 opened 4 months ago by jpata
  1

• implement a per-particle weight to predict a MET correction

  #299 opened 7 months ago by jpata
  1

• investigate HEPT

  #303 opened 6 months ago by jpata
  1

• CLIC samples: ddsim MCParticles parent/daughter indices are not correct

  #313 opened 6 months ago by jpata
  1

• implement simple JEC in our validation code

  #308 opened 6 months ago by jpata
  1

• try PerceiverIO

  #275 opened a year ago by jpata
  0

• huggingface model card, run model via huggingface

  #241 opened a year ago by jpata
  1

• investigate set-conditional set generation

  #247 opened a year ago by jpata
  0

• create a short documentation on training the CLIC hit-based model on the 10k event dataset

  #230 opened a year ago by jpata
  0

• Comet times out on long trainings

  #311 opened 6 months ago by jpata
  0

• try CrossAttention / TreeCrossAttention

  #276 opened a year ago by jpata
  0

• investigate Mamba

  #281 opened 10 months ago by jpata
  0

• ECAL cluster shape variables are empty

  #307 opened 6 months ago by jpata
  0

• try optimized kNN kernels

  #218 opened 6 months ago by jpata
  1

• Separate the regression "MLP" into "NUM_CLASSES*MLP"

  #242 opened 6 months ago by farakiko
  1

• separate MLP embedding for tracks and clusters before the graph layers

  #235 opened 6 months ago by jpata
  1

• Separate training of Classification and Regression

  #243 opened 6 months ago by farakiko
  1

• add pt-dependent jet and MET response plots for pytorch

  #295 opened 6 months ago by jpata
  1

• add HCAL depth feature

  #244 opened 7 months ago by jpata
  1

• bug when updating from uproot 5.0.12 -> 5.2.1

  #286 opened 8 months ago by jpata
  1

• regressing multimodal distributions accurately

  #236 opened 8 months ago by jpata
  2

• Implement HPO in PyTorch pipeline.

  #251 opened 8 months ago by erwulff
  2

• pytorch occasional tkinter error

  #260 opened 8 months ago by jpata
  1

• investigate PELICAN

  #271 opened a year ago by jpata
  0

• pytorch opens too many data loader worker processes

  #259 opened a year ago by jpata
  1

• GNN-LSH in pytorch is too slow

  #239 opened a year ago by jpata
  1

• pytorch different multi-GPU runs on the same node don't work

  #258 opened a year ago by jpata
  0

• try fast feedforward

  #214 opened a year ago by jpata
  0

• try HyperTrack for graph construction

  #221 opened a year ago by jpata
  0

• try contrastive event-based loss

  #223 opened a year ago by jpata
  0

• try FlatFormer

  #226 opened a year ago by jpata
  0

• try HyperGraph

  #227 opened a year ago by jpata
  0

• try HyperAttention

  #231 opened a year ago by jpata
  0

• add an ongoing absolute runtime benchmark

  #212 opened a year ago by jpata
  1

• readme for using the pretrained model

  #217 opened a year ago by jpata
  1

• re-enable ONNX export in tensorflow

  #225 opened a year ago by jpata
  1

• CUDA optimization prevents running on older cards

  #220 opened a year ago by jpata
  1

• data preprocessing fails with latest awkward

  #222 opened a year ago by jpata
  0

• try quantization, both post-training and quantization-aware training

  #224 opened a year ago by jpata
  1