GLM/dvc.lock at master · Emvlt/GLM · 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
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
schema: '2.0'
stages:
  prepare_demo:
    cmd: uv run python src/glm/preprocess_2detect.py mode2 demo
    deps:
    - path: src/glm/datasets/raw/2detect
      hash: md5
      md5: 67e7e6227a3c83214c1c50fb338e4eea.dir
      size: 19026176
      nfiles: 6
    - path: src/glm/preprocess_2detect.py
      hash: md5
      md5: ccfb656291e0f83eff067e63aece625d
      size: 2538
    params:
      params.yaml:
        demo:
          data:
            raw_path: src/glm/datasets/raw/2detect
            processed_path: src/glm/datasets/processed/2detect
    outs:
    - path: src/glm/datasets/processed/2detect
      hash: md5
      md5: 714c75e736e98e6bc7d0dc4f76f59684.dir
      size: 31727360
      nfiles: 2
  run_demo:
    cmd: uv run python src/glm/run_demo.py
    deps:
    - path: src/glm/datasets/processed/2detect
      hash: md5
      md5: 714c75e736e98e6bc7d0dc4f76f59684.dir
      size: 31727360
      nfiles: 2
    - path: src/glm/run_demo.py
      hash: md5
      md5: e3c7c54289abaddc857cde7017478d21
      size: 1768
    params:
      params.yaml:
        demo:
          data:
            raw_path: src/glm/datasets/raw/2detect
            processed_path: src/glm/datasets/processed/2detect
    outs:
    - path: src/glm/images/demo/reconstruction.jpg
      hash: md5
      md5: 4d094a10b9b06e0a7a763ec31622441e
      size: 17981
    - path: src/glm/images/demo/sinogram.jpg
      hash: md5
      md5: 9564192459134caa2f0e79440ff057fc
      size: 41795
    - path: src/glm/images/demo/target.jpg
      hash: md5
      md5: 3d40d439da4a3c4e4f7cf1ce7a749f74
      size: 20319
  prepare_training:
    cmd: uv run python src/glm/preprocess_2detect.py mode2 train
    deps:
    - path: /home/emilien/datasets/raw/2detect
      hash: md5
      md5: e4ac8fdf7a5d8002a946f8b988c777d5.dir
      size: 316154324027
      nfiles: 92001
    - path: src/glm/preprocess_2detect.py
      hash: md5
      md5: 55e400fc88c0fa959bb7d81cbc86ba39
      size: 2717
    params:
      params.yaml:
        data:
          raw_path: /home/emilien/datasets/raw/2detect
          processed_path: /home/emilien/datasets/processed/2detect
    outs:
    - path: /home/emilien/datasets/processed/2detect
      hash: md5
      md5: c9b5188af6c5685168104b496816ee67.dir
      size: 182432320000
      nfiles: 11500
  pretraining:
    cmd: torchrun --nproc_per_node=2 src/glm/pretrain.py
    deps:
    - path: /home/emilien/datasets/processed/2detect
      hash: md5
      md5: c9b5188af6c5685168104b496816ee67.dir
      size: 182432320000
      nfiles: 11500
    - path: src/glm/pretrain.py
      hash: md5
      md5: 8c6981c206d7f0a1abcbf442e2359cdc
      size: 7730
    params:
      params.yaml:
        pretrain_parameters:
          hyperparameters:
            learning_rate: 0.0005
            epochs: 1
            batch_size: 8
            num_workers: 2
            downsampling: 1
          active_model: GLM
          models:
            GLM:
              kernel_size: 7
              kernel_dimension: 1
              padding: 3
              n_channels: 16
              normalize: true
              improved: true
              cached: false
              add_self_loops: true
              aggr: add
            sinogram_CNN:
              kernel_size: 7
              kernel_dimension: 2
              padding: 3
              n_channels: 24
    outs:
    - path: dvclive
      hash: md5
      md5: 732ec4c99b55719f3c0a84c4f8400731.dir
      size: 46760
      nfiles: 7
    - path: src/glm/saved_models/pretrained_sinogram_model.pt
      hash: md5
      md5: 1c95c22ca7d997538bd46fcc1873995a
      size: 28291