Hugging Face's logo

jpata
/
particleflow

TensorBoard
ONNX
physics
reconstruction
Model card Files Metrics Community
particleflow
File size: 2,442 Bytes 
652e089
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
 
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: 20
gpus: 1
load: null
lr: 0.0001
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
  gravnet:
    activation: elu
    conv_type: gravnet
    dropout_ff: 0.1
    embedding_dim: 512
    k: 16
    num_convs: 3
    propagate_dimensions: 32
    space_dimensions: 4
    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: 1.7.0
  cms_pf_qcd_high_pt:
    version: 1.7.0
  cms_pf_sms_t1tttt:
    version: 1.7.0
  cms_pf_ttbar:
    version: 1.7.1
  cms_pf_ztt:
    version: 1.7.0
test_datasets: []
train: true
train_dataset:
  cms:
    physical:
      batch_size: 1
      samples:
        cms_pf_ttbar:
          version: 1.7.1
val_freq: null
valid_dataset:
  cms:
    physical:
      batch_size: 1
      samples:
        cms_pf_ttbar:
          version: 1.7.1