update readme

README.md

@@ -9,5 +9,6 @@ tags:
 
 Please see the linked model cards below for more details.
   - [clic/clusters/v1.6](clic/clusters/v1.6/README.md)
+  - [clic/clusters/v1.9.0](clic/clusters/v1.9.0/README.md)
   - [cms/2024_04_05](cms/2024_04_05/pyg-cms_20240324_235743_208080/README.md)
   - [cms/2024_05_16_attn_model21M](cms/2024_05_16_attn_model21M/README.md)

clic/clusters/v1.9.0/README.md

@@ -6,7 +6,7 @@ This model reconstructs particles in a detector, based on the tracks and calorim
 
 ### Model Description
 
-- **Developed by:** Joosep Pata, Eric Wulff, Farouk Mokhtar, Mengke Zhang, David Southwick, Maria Girone, David Southwick, Javier Duarte
+- **Developed by:** Joosep Pata, Eric Wulff, Farouk Mokhtar, Mengke Zhang, David Southwick, Maria Girone, David Southwick, Javier Duarte, Michael Kagan
 - **Model type:** transformer
 - **License:** Apache License
 
@@ -26,15 +26,20 @@ This model is not intended for physics measurements on real data.
 ## Bias, Risks, and Limitations
 
 The model has only been trained on simulation data and has not been validated against real data.
-The model has not been peer reviewed or published in a journal.
+The model has not been peer reviewed or published in a peer-reviewed journal.
 
 ## How to Get Started with the Model
 
 Use the code below to get started with the model.
 
 ```
-git clone https://github.com/jpata/particleflow/releases/tag/v1.9.0
+#get the code
+git clone https://github.com/jpata/particleflow
 cd particleflow
+git checkout v1.9.0
+
+#get the models
+git clone https://huggingface.co/jpata/particleflow models
 ```
 
 ## Training Details
@@ -44,13 +49,21 @@ The training was continued twice from a checkpoint due to the 24h time limit.
 ### Training Data
 The following datasets were used:
 ```
-clic_edm_qq_pf:2.2.0
-clic_edm_ttbar_pf:2.2.0
-clic_edm_ww_fullhad_pf:2.2.0
+/eos/user/j/jpata/mlpf/tensorflow_datasets/clic/clic_edm_qq_pf/2.2.0
+/eos/user/j/jpata/mlpf/tensorflow_datasets/clic/clic_edm_ttbar_pf/2.2.0
+/eos/user/j/jpata/mlpf/tensorflow_datasets/clic/clic_edm_ww_fullhad_pf/2.2.0
 ```
-They have an updated truth and target definition: https://github.com/jpata/particleflow/pull/345
 
-### Training Procedure 
+They have an updated truth and target definition with respect to [Pata, J. et al, Commun Phys 7, 124 (2024)](https://doi.org/10.1038/s42005-024-01599-5): https://github.com/jpata/particleflow/pull/345
+
+In particular, target particles for MLPF reconstruction are based on status=1 particles.
+For non-interacting status=1, nearby (dR<0.2) interacting status=0 are used instead.
+
+The datasets were generated using Key4HEP with the following scripts:
+- https://github.com/HEP-KBFI/key4hep-sim/releases/tag/v1.0.0
+- https://github.com/HEP-KBFI/key4hep-sim/blob/v1.0.0/clic/run_sim.sh
+
+## Training Procedure 
 
 ```bash
 #!/bin/bash
@@ -109,7 +122,7 @@ singularity exec \
 IMG=/home/software/singularity/pytorch.simg:2024-08-18
 cd ~/particleflow
 
-WEIGHTS=experiments/pyg-clic_20241001_215132_345408/checkpoints/checkpoint-26-2.004527.pth
+WEIGHTS=models/clic/clusters/v1.9.0/checkpoints/checkpoint-26-2.004527.pth
 singularity exec -B /scratch/persistent --nv \
      --env PYTHONPATH=hep_tfds \
      --env KERAS_BACKEND=torch \
@@ -124,10 +137,9 @@ singularity exec -B /scratch/persistent --nv \
 ## Glossary
 
 PF - particle flow reconstruction
+MLPF - machine learning for particle flow
+CLIC - Compact Linear Collider
 
 ## Model Card Contact
 
 Joosep Pata, [email protected]
-
-
-