Im2Mesh/docs/yaml_configs/inference_Probabilistic_Unet.yaml at main · JairMathAI/Im2Mesh · GitHub

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
mode: inference

data:
  inference_input:
    dir: "/path/to/data/inference"
    data_type: .tif
    reader_params:
      dimension_order_out: "ZYX"
      C: 0
      T: 0
  inference_output:
    path: "/path/to/output/folder"

  preprocess:
    - module_name: monai.transforms
      func_name: NormalizeIntensity
      params:
        channel_wise: True

    - module_name: mmv_im2im.utils.custom_transforms
      func_name: RecordShapeAndPad
      params:
        k: 16
        mode: "constant"
        constant_value: 0.0

  postprocess:
    - module_name:  mmv_im2im.utils.custom_transforms
      func_name: RemovePadFromPrediction
      params:
        k: 16
        n_coord_dims: 3

model:
  framework: ProbUnet
  net:
    module_name: mmv_im2im.models.nets.ProbUnet
    func_name: ProbabilisticUNet
    params:
      in_channels: 1 # nuber of channels in the input IM
      out_channels: 2503 # Vector size on this case [v1,...,vn]-> n
      spatial_dims: 3 # 2d o 3d for convolutions
      latent_dim: 10
      channels: [32, 64, 128, 256, 512]
      strides: [1, 2, 2, 2, 2]
      dropout: 0.2
      task: "regression"
  checkpoint: "path/to/weights"
  model_extra:
    cpu_only: False

trainer:
  verbose: True
  params:
    gpus: 1
    precision: 16