search

HannesStark / EquiBind

EquiBind: geometric deep learning for fast predictions of the 3D structure in which a small molecule binds to a protein
MIT License
473 stars 109 forks source link

dgl issue #4

Closed mariolovric closed 2 years ago

mariolovric commented 2 years ago

Dear all,

I created the given environment on a windows computer and experience a dgl error.

(equibind) PS C:\Users\reps\EquiBind> python inference.py --config=configs_clean/inference_file_for_reproduce.yml

Using backend: pytorch [2022-02-13 09:04:14.926225] [ Using Seed : 1 ] Traceback (most recent call last): File "inference.py", line 459, in inference(args) File "inference.py", line 121, in inference seed_all(args.seed) File "C:\Users\reps\EquiBind\commons\utils.py", line 62, in seed_all dgl.random.seed(seed) File "C:\Users.conda\envs\equibind\lib\site-packages\dgl\random.py", line 18, in seed _CAPI_SetSeed(val) File "C:\Users.conda\envs\equibind\lib\site-packages\dgl_ffi_ctypes\function.py", line 190, in call ctypes.byref(ret_val), ctypes.byref(ret_tcode))) File "C:\Users.conda\envs\equibind\lib\site-packages\dgl_ffi\base.py", line 64, in check_call raise DGLError(py_str(_LIB.DGLGetLastError())) dgl._ffi.base.DGLError: [09:04:14] C:\Users\Administrator\dgl-0.5\src\random\random.cc:34: Check failed: e == CURAND_STATUS_SUCCESS: CURAND Error: CURAND_STATUS_INITIALIZATION_FAILED at C:\Users\Administrator\dgl-0.5\src\random\random.cc:34

HannesStark commented 2 years ago

Hi, Could it be that you do not have a CUDA GPU?

the env.yml installs

mariolovric commented 2 years ago

Hi, Could it be that you do not have a CUDA GPU?

the env.yml installs

  • cudatoolkit=10.2
  • dgl-cuda10.2

I will clarify that in the readme.

I did it on a windows machine to test. Ok CUDA it is. Thanks