Hugging Face's logo

king159
/
SimpleP2G

PyTorch
Safetensors
English
biology
spatial-transcriptomics
computational-pathology
gene-expression
Model card Files
xet
 Community
SimpleP2G
File size: 2,010 Bytes
30e5a24
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
{
  "format_version": 1,
  "architecture": "SimpleP2G",
  "setting": "single-patch",
  "dataset": "ccrcc",
  "panel": "hvg",
  "fold": 0,
  "seed": 42,
  "model_config": {
    "decoder_depth": 12,
    "num_genes": 200,
    "hidden_size": 384,
    "cond_size": 2560,
    "num_cond_tokens": 4,
    "num_heads": 6,
    "loss_name": "gaussian_nll",
    "dropout": 0.1
  },
  "input": {
    "representation": "frozen Virchow2 CLS token concatenated with mean patch tokens",
    "embedding_dimension": 2560,
    "image_encoder_weights_included": false
  },
  "output": {
    "num_genes": 200,
    "gene_order": "genes.json",
    "expression_scale": "log2(raw_count + 1)",
    "returns": [
      "expression_mean",
      "log_variance"
    ]
  },
  "checkpoint": {
    "epoch": 7,
    "weights": "ema",
    "selection": "highest validation-slide PCC-200; earliest epoch breaks ties"
  },
  "training": {
    "epochs": 100,
    "batch_size": 256,
    "optimizer": {
      "name": "adamw",
      "lr": 3e-05,
      "weight_decay": 0.1,
      "betas": [
        0.9,
        0.999
      ]
    },
    "loss": "gaussian_nll",
    "precision": "fp16",
    "ema_decay": 0.999,
    "sampling": "pooled spots"
  },
  "split": {
    "training_slides": [
      "INT1",
      "INT10",
      "INT11",
      "INT12",
      "INT13",
      "INT14",
      "INT15",
      "INT16",
      "INT17",
      "INT18",
      "INT19",
      "INT20",
      "INT21",
      "INT22",
      "INT23",
      "INT24",
      "INT3",
      "INT4",
      "INT5",
      "INT6",
      "INT7",
      "INT8"
    ],
    "validation_slide": "INT9",
    "test_slide": "INT2"
  },
  "validation_metrics": {
    "PCC10": 0.5082438063563466,
    "PCC50": 0.2864946574885263,
    "PCC200": 0.11651308085853335,
    "MSE": 0.5486487746238708,
    "MAE": 0.4507647454738617
  },
  "test_metrics": {
    "PCC10": 0.3779823666623128,
    "PCC50": 0.24105083678052494,
    "PCC200": 0.10978539095367261,
    "MSE": 0.5488066077232361,
    "MAE": 0.4156813621520996
  }
}