added May 2024 model

cms/2024_05_16_attn_model21M/checkpoint-25-17.631161.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d8a593d279ca3fd58fe3836d49145f634238fc4a7cf808c5df580c4a0bdefb5
+size 255895346

cms/2024_05_16_attn_model21M/model_fp32.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fddd236376ab0a119f2db393b19e5fd20269f03ab60525b961ec5157d2a94ce5
+size 85479699

cms/2024_05_16_attn_model21M/overridden_config.yaml

@@ -0,0 +1,123 @@
+backend: pytorch
+checkpoint_freq: 1
+comet: true
+comet_name: particleflow-pt
+comet_offline: false
+comet_step_freq: 10
+config: parameters/pytorch/pyg-cms.yaml
+conv_type: attention
+data_dir: /scratch/persistent/joosep/tensorflow_datasets
+dataset: cms
+dtype: bfloat16
+gpu_batch_multiplier: 40
+gpus: 1
+load: experiments/pyg-cms_20240430_094836_751206/checkpoints/checkpoint-12-17.861217.pth
+lr: 5.0e-05
+lr_schedule: cosinedecay
+lr_schedule_config:
+  onecycle:
+    pct_start: 0.3
+model:
+  attention:
+    activation: relu
+    attention_type: flash
+    conv_type: attention
+    dropout_conv_id_ff: 0.0
+    dropout_conv_id_mha: 0.0
+    dropout_conv_reg_ff: 0.0
+    dropout_conv_reg_mha: 0.0
+    dropout_ff: 0.0
+    head_dim: 16
+    num_convs: 6
+    num_heads: 32
+  cos_phi_mode: linear
+  energy_mode: linear
+  eta_mode: linear
+  gnn_lsh:
+    activation: elu
+    bin_size: 640
+    conv_type: gnn_lsh
+    distance_dim: 128
+    dropout_ff: 0.0
+    embedding_dim: 512
+    ffn_dist_hidden_dim: 128
+    ffn_dist_num_layers: 2
+    layernorm: true
+    max_num_bins: 200
+    num_convs: 3
+    num_node_messages: 2
+    width: 512
+  input_encoding: joint
+  learned_representation_mode: last
+  mamba:
+    activation: elu
+    conv_type: mamba
+    d_conv: 4
+    d_state: 32
+    dropout_ff: 0.0
+    embedding_dim: 1024
+    expand: 2
+    num_convs: 4
+    width: 1024
+  pt_mode: linear
+  sin_phi_mode: linear
+  trainable: all
+ntest: null
+ntrain: null
+num_epochs: 100
+num_workers: 4
+nvalid: null
+patience: 20
+prefetch_factor: 50
+ray_train: false
+raytune:
+  asha:
+    brackets: 1
+    grace_period: 10
+    max_t: 200
+    reduction_factor: 4
+  default_metric: val_loss
+  default_mode: min
+  hyperband:
+    max_t: 200
+    reduction_factor: 4
+  hyperopt:
+    n_random_steps: 10
+  local_dir: null
+  nevergrad:
+    n_random_steps: 10
+  sched: asha
+  search_alg: hyperopt
+sort_data: true
+test_dataset:
+  cms_pf_qcd:
+    version: 1.7.1
+  cms_pf_ttbar:
+    version: 1.7.1
+  cms_pf_ztt:
+    version: 1.7.1
+test_datasets: []
+train: true
+train_dataset:
+  cms:
+    physical:
+      batch_size: 1
+      samples:
+        cms_pf_qcd:
+          version: 1.7.1
+        cms_pf_ttbar:
+          version: 1.7.1
+        cms_pf_ztt:
+          version: 1.7.1
+val_freq: null
+valid_dataset:
+  cms:
+    physical:
+      batch_size: 1
+      samples:
+        cms_pf_qcd:
+          version: 1.7.1
+        cms_pf_ttbar:
+          version: 1.7.1
+        cms_pf_ztt:
+          version: 1.7.1