DECOMP/install.txt at main · DeepMicroscopy/DECOMP · 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
conda create -n hactnet python=3.9
conda activate hactnet
conda install pip
pip install dgl -f https://data.dgl.ai/wheels/torch-2.2/cu121/repo.html
pip install torch==2.2.1 torchvision==0.17.1 --index-url https://download.pytorch.org/whl/cu121
pip install mlflow matplotlib h5py seaborn histocartography opencv-python scikit-image scikit-learn


Bug:------------------------------------------
Traceback (most recent call last):
  File "code/experiments.py", line 117, in <module>
    pred_dict, train_conf = predict_train_val(cfg=cfg,
  File "code/decompose.py", line 77, in predict_train_val
    for batch in tqdm(dataloader):
  File "conda/envs/hactnet/lib/python3.9/site-packages/tqdm/std.py", line 1181, in __iter__
    for obj in iterable:
  File "conda/envs/hactnet/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 631, in __next__
    data = self._next_data()
  File "conda/envs/hactnet/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 675, in _next_data
    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
  File "conda/envs/hactnet/lib/python3.9/site-packages/torch/utils/data/_utils/fetch.py", line 51, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "conda/envs/hactnet/lib/python3.9/site-packages/torch/utils/data/_utils/fetch.py", line 51, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "code/dataloader.py", line 136, in __getitem__
    cg = set_graph_on_cuda(cg) if IS_CUDA else cg
  File "conda/envs/hactnet/lib/python3.9/site-packages/histocartography/utils/graph.py", line 125, in set_graph_on_cuda
    cuda_graph.ndata[key_graph] = tmp.cuda()
  File "conda/envs/hactnet/lib/python3.9/site-packages/dgl/view.py", line 99, in __setitem__
    self._graph._set_n_repr(self._ntid, self._nodes, {key: val})
  File "conda/envs/hactnet/lib/python3.9/site-packages/dgl/heterograph.py", line 4349, in _set_n_repr
    raise DGLError(
dgl._ffi.base.DGLError: Cannot assign node feature "centroid" on device cuda:0 to a graph on device cpu. Call DGLGraph.to() to copy the graph to the same device.

--> solution:
    change function set_graph_on_cuda in miniconda3/envs/hactnet/lib/python3.9/site-packages/histocartography/utils/graph.py:
    def set_graph_on_cuda(graph):
      cuda_graph = dgl.DGLGraph()
      cuda_graph.add_nodes(graph.number_of_nodes())
      cuda_graph.add_edges(graph.edges()[0], graph.edges()[1])
      cuda_graph = cuda_graph.to('cuda:0')   ####################################### added ############
      for key_graph, val_graph in graph.ndata.items():
          tmp = graph.ndata[key_graph].clone()
          cuda_graph.ndata[key_graph] = tmp.cuda()
      for key_graph, val_graph in graph.edata.items():
          cuda_graph.edata[key_graph] = graph.edata[key_graph].clone().cuda()
      return cuda_graph

Bug:--------------------------------------------------------------------------
    load_backend(get_preferred_backend())
  File "conda/envs/hactnet/lib/python3.9/site-packages/dgl/backend/__init__.py", line 51, in load_backend
    from .._ffi.base import load_tensor_adapter  # imports DGL C library
  File "conda/envs/hactnet/lib/python3.9/site-packages/dgl/_ffi/base.py", line 50, in <module>
    _LIB, _LIB_NAME, _DIR_NAME = _load_lib()
  File "conda/envs/hactnet/lib/python3.9/site-packages/dgl/_ffi/base.py", line 39, in _load_lib
    lib = ctypes.CDLL(lib_path[0])
  File "conda/envs/hactnet/lib/python3.9/ctypes/__init__.py", line 374, in __init__
    self._handle = _dlopen(self._name, mode)
OSError: /lib64/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by conda/envs/hactnet/lib/python3.9/site-packages/dgl/libdgl.so)
--> solution:
    module load gcc/11.2.0
    module load cuda/12.1

Bug:------------------------------------------------------------------------------
    device: torch.device = torch.device(torch._C._get_default_device()),  # torch.device('cpu'),
conda/envs/hactnet/lib/python3.9/site-packages/torch/nn/modules/transformer.py:20: UserWarning: Failed to initialize NumPy: _ARRAY_API not found (Triggered internally at ../torch/csrc/utils/tensor_numpy.cpp:84.)
  device: torch.device = torch.device(torch._C._get_default_device()),  # torch.device('cpu'),
--> solution:
  pip uninstall numpy
  pip install "numpy<2"