v2.5.0

What's Changed

• Update README.md by @jpata in #414
• Fix lepton fake rate spikes by @farakiko in #416
• some additional plot fixes by @jpata in #415
• Remove (Preliminary) from plots by @farakiko in #417
• Add option to match particle type when making eff / fake rate plots by @farakiko in #420
• Add eta<2.5 cut for particle-level stacked distribution plots by @farakiko in #421
• Update final version of particle-plots in DP note by @farakiko in #422
• Add 2025 DP note to README by @jpata in #423
• Update particle notebook to use miniaod by @farakiko in #424
• more updates to plots, PU studies, dijet asymmetry, 13p6TeV validation by @jpata in #418
• added postprocessing validation notebook by @jpata in #426
• Add delta_r selection to caloparticle splitting and fix phi wrap around issue in cms postprocessing by @kawaho in #427
• fix dataloading in ptkl notebook by @farakiko in #425
• Update particle-level plots (AN v9) by @farakiko in #429
• Update plots, new postprocessing by @jpata in #428
• update notebook for paper v6 (in progress) by @farakiko in #430
• PU analysis notebook by @jpata in #433
• fix PU loss, pad to fixed sizes, JME plot fixes by @jpata in #432
• run CMS validation notebooks via papermill by @jpata in #434
• Clear up the CLI parameters by @jpata in #436
• update plots towards the post-approval freeze by @jpata in #437
• disable T2_EE_Estonia configuration in the cmssw validation script by @jpata in #438
• notebook for pu threshold calibration by @kawaho in #439
• consolidate onnx export by @jpata in #441

Full Changelog: v2.4.0...v2.5.0

v2.4.0

This is the release corresponding to the CMS AN v6, paper v1, preapproval freeze.

• Training: log, checkpoint
• Validation results: lxplus
• CMSSW tag: jp_mlpf_CMSSW_15_0_5_mlpf_v2.5.0_p01_603dc5_ANv6_Paperv2_preappfreeze

What's Changed

• CLIC loss plots by @jpata in #401
• Switch to binary classification for PU by @kawaho in #402
• remove confusing outdir option by @jpata in #399
• Update README.md by @jpata in #404
• Add postprocessing for hits by @Z0zzz in #403
• feat: LAMB optimizer & reduce lr on plateau schedule by @erwulff in #409
• add particle eval notebook by @farakiko in #411
• ANv6, paperv1, CMS dataset 2.6.0, model 2.5.0: PU prediction integration, full CMSSW_15_0_5 datasets by @jpata in #407
• notebook for pileup id performance related plots by @kawaho in #412
• update data plots for AN v6, paper v1 freeze by @jpata in #413

New Contributors

• @Z0zzz made their first contribution in #403

Full Changelog: v2.3.0...v2.4.0

v2.3.0

What's Changed

• Update README.md by @jpata in #386
• Finetuning by @farakiko in #387
• Update README.md by @jpata in #388
• Implement PU prediction by @kawaho in #389
• fix: multi-gpu hang by @erwulff in #391
• added lxplus scripts by @jpata in #392
• Clarify citation policy by @jpata in #394
• process gun samples for CMS, CLIC switch to flashattn and bf16 by @jpata in #397

Full Changelog: v2.2.0...v2.3.0

v2.2.0

What's Changed

• Update onnx export and validation for cms by @erwulff in #378
• Unify key4hep by @farakiko in #374
• allow missing layers in models when loading checkpoint by @kawaho in #380
• Plots for AN V1 by @jpata in #381
• Refactor training code by @jpata in #384
• Compute and track detailed evaluation metrics on each epoch by @jpata in #385
• Use per-particle pt weight in loss by @jpata in #383

New Contributors

• @kawaho made their first contribution in #380

Full Changelog: v2.1.0...v2.2.0

v2.1.0

What's Changed

• fix: remove command line argument --dataset by @erwulff in #358
• Update README.md by @jpata in #360
• Fix inference race condition by @erwulff in #359
• datasets 2.5.0, CMS plotting code by @jpata in #350
• final fixes for v2.1.0 by @jpata in #375

Full Changelog: v2.0.0...v2.1.0

v2.0.0

What's Changed

• add torch to requirements by @jpata in #353
• Cleanup tensorflow and pyg by @jpata in #354
• add compatibility table by @jpata in #355
• CLIC dataset v2.3.0: fix target/truth momentum, st=1 to target more inclusive by @jpata in #352

Full Changelog: v1.9.0...v2.0.0

v1.9.0

What's Changed

• fix CMS instructions by @jpata in #334
• MLPF datasets v2.0.0: track pythia-level genjets, genmet in datasets; add per-particle ispu flag by @jpata in #332
• CMS training instructions by @jpata in #336
• Retraining with CMS samples v2.0.0 by @jpata in #337
• Fix use of deprecated Ray Tune environment variable by @erwulff in #338
• CMS dataset relabel, generate v2.1.0 with more stats, separate binary classifier by @jpata in #340
• Pre-layernorm by @erwulff in #339
• Switch to datasets v2.2.0 by @jpata in #341
• try to improve val loss stability by @jpata in #342
• Regression of log-transformed energy and pt, training checkpoints by @jpata in #343
• CLIC dataset v2.2.0, CMS dataset 2.4.0 by @jpata in #345
• Update dataset validation plot notebooks by @jpata in #347

Full Changelog: v1.8.0...v1.9.0

v1.8.0

What's Changed

The focus of this release is training on CMS datasets. The model has been retrained on high-statistics CMS samples and outperforms the baseline in our MLPF samples. The export of the transformer model to ONNX works with Flash Attention, and it can be integrated with CMSSW 14 and run on GPU. We have run the first physics validations in CMSSW, and we find that the performance with respect to the previous CMS version of MLPF is improved, but it does not yet outperform the baseline PF in CMSSW.

Some slides from the CMS progress:

• https://indico.cern.ch/event/1399778/#2-cms-status
• https://indico.cern.ch/event/1415765/#2-cms-status-and-plans
• https://indico.cern.ch/event/1421798/#2-cms-status-and-plans-virtual
• https://indico.cern.ch/event/1426959/#2-cms-status-and-plans

The full list of PRs:

• Remove pytorch geometric by @jpata in #310
• add new paper to README by @jpata in #312
• Add Ray Train training to GitHub actions CI/CD test by @erwulff in #314
• CMSSW documentation by @jpata in #319
• Full CMS pytorch training in May 2024 by @jpata in #316
• update CMSSW validation scripts and documentation by @jpata in #322
• onnx export with dynamic shapes, fast attention by @jpata in #324
• switch ONNX model to full float for CMSSW compatibility by @jpata in #325
• Update validation scripts to CMSSW_14_1_0 by @jpata in #323
• update cmssw plots, add ttbar sample to valid, add multiparticlegun and vbf to training by @jpata in #330

Full Changelog: v1.7.0...v1.8.0

v1.7.0

What's Changed

The primary feature of the new release is that pytorch is now the main mode of training.
The CMS status was presented at https://indico.cern.ch/event/1399688/#1-ml-for-pf.

• switch pytorch training to tfds array-record datasets by @farakiko in #228
• Timing the ONNX model, retrain CMS-GNNLSH-TF by @jpata in #229
• fixes for pytorch, CMS t1tttt dataset, update response plots by @jpata in #232
• fix pytorch multi-GPU training hang by @farakiko in #233
• feat: specify number of samples as cmd line arg in pytorch training and testing by @erwulff in #237
• Automatically name training dir in pytorch pipeline by @erwulff in #238
• pytorch backend major update by @farakiko in #240
• Update dist.barrier() and fix stale epochs for torch backend by @farakiko in #249
• multi-bin loss in TF, plot fixes by @jpata in #234
• PyTorch distributed num-workers>0 fix by @farakiko in #252
• speedup of the pytorch GNN-LSH model by @jpata in #245
• Implement HPO for PyTorch pipeline. by @erwulff in #246
• fix tensorboard error by @farakiko in #254
• fix config files by @erwulff in #255
• making the 3d-padded models more efficient in pytorch by @jpata in #256
• Fix pytorch inference after #256 by @jpata in #257
• Update training.py by @jpata in #261
• Reduce the number of data loader workers per dataset in pytorch by @farakiko in #262
• fix inference by @farakiko in #264
• Implementing configurable checkpointing. by @erwulff in #263
• restore onnx export in pytorch by @jpata in #265
• remove outdated forward_batch from pytorch by @jpata in #266
• Separate multiparticlegun samples from singleparticle gun samples by @farakiko in #267
• compare all three models in pytorch by @jpata in #268
• Allows testing on a given --load-checkpoint by @farakiko in #269
• added clic evaluation notebook by @jpata in #272
• Fix --load-checkpoint bug by @farakiko in #270
• Implement CometML logging to PyTorch training pipeline. by @erwulff in #273
• Add command line argument to choose experiments dir in PyTorch training pipeline by @erwulff in #274
• Implement multi-gpu training in HPO with Ray Tune and Ray Train by @erwulff in #277
• Better CometML logging + Ray Train vs DDP comparison by @erwulff in #278
• Fix checkpoint loading by @erwulff in #280
• Learning rate schedules and Mamba layer by @erwulff in #282
• use modern optimizer, revert multi-bin loss in TF by @jpata in #253
• track individual particle loss components, speedup inference by @jpata in #284
• Update the jet pt threshold to be the same as the PF paper by @farakiko in #283
• towards v1.7: new CMS datasets, CLIC hit-based datasets, TF backward-compat optimizations by @jpata in #285
• fix torch no grad by @jpata in #290
• pytorch regression output layer configurability by @jpata in #291
• Implement resume-from-checkpoint in HPO by @erwulff in #293
• enable FlashAttention in pytorch, update to torch 2.2.0 by @jpata in #292
• fix pad_power_of_two by @jpata in #296
• Feat val freq by @erwulff in #298
• normalize loss, reparametrize network by @jpata in #297
• fix up configs by @jpata in #300
• clean up loading by @jpata in #301
• Fix unpacking for 3d padded batch, update plot style by @jpata in #306

Full Changelog: v1.6...v1.7.0

v1.6.2

What's Changed

• residual paper fixes from the paper review by @jpata in #304
  Full Changelog: v1.6.1...v1.6.2