You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
`No other smiles inputs
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
ImportError: numpy.core._multiarray_umath failed to import
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | distributed init (rank 0): tcp://localhost:18393
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | distributed init (rank 1): tcp://localhost:18393
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Added key: store_based_barrier_key:1 to store for rank: 1
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Added key: store_based_barrier_key:1 to store for rank: 0
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Rank 0: Completed store-based barrier for key:store_based_barrier_key:1 with 2 nodes.
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | initialized host 27d257810fc0 as rank 0
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Rank 1: Completed store-based barrier for key:store_based_barrier_key:1 with 2 nodes.
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | initialized host 27d257810fc0 as rank 1
Traceback (most recent call last):
File "./Uni-Mol/unimol/unimol/infer.py", line 120, in
cli_main()
File "./Uni-Mol/unimol/unimol/infer.py", line 116, in cli_main
distributed_utils.call_main(args, main)
File "/opt/conda/lib/python3.8/site-packages/unicore-0.0.1-py3.8-linux-x86_64.egg/unicore/distributed/utils.py", line 180, in call_main
torch.multiprocessing.spawn(
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 240, in spawn
return start_processes(fn, args, nprocs, join, daemon, start_method='spawn')
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 198, in start_processes
while not context.join():
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 160, in join
raise ProcessRaisedException(msg, error_index, failed_process.pid)
torch.multiprocessing.spawn.ProcessRaisedException:
-- Process 0 terminated with the following error:
Traceback (most recent call last):
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 69, in _wrap
fn(i, *args)
File "/opt/conda/lib/python3.8/site-packages/unicore-0.0.1-py3.8-linux-x86_64.egg/unicore/distributed/utils.py", line 158, in distributed_main
args.distributed_rank = distributed_init(args)
File "/opt/conda/lib/python3.8/site-packages/unicore-0.0.1-py3.8-linux-x86_64.egg/unicore/distributed/utils.py", line 138, in distributed_init
dist.all_reduce(torch.zeros(1).cuda())
RuntimeError: CUDA error: out of memory
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
`
It seems to be related to a numpy version problem. I am encountering this issue while trying to run unimol_binding_pose_demo from Uni-Mol. I would appreciate any guidance on how to resolve this.
The text was updated successfully, but these errors were encountered:
@ZhouGengmo Thank you very much for your attention and response to the problem I am having.
I've tried all of the solutions you've provided and encountered the same error about numpy version, and as for the CUDA issue, I ran the code on a higher capacity GPU and no longer get the error.
I am facing the issue of
`No other smiles inputs
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
ImportError: numpy.core._multiarray_umath failed to import
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xe . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | distributed init (rank 0): tcp://localhost:18393
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | distributed init (rank 1): tcp://localhost:18393
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Added key: store_based_barrier_key:1 to store for rank: 1
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Added key: store_based_barrier_key:1 to store for rank: 0
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Rank 0: Completed store-based barrier for key:store_based_barrier_key:1 with 2 nodes.
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | initialized host 27d257810fc0 as rank 0
2024-08-28 14:01:18 | INFO | torch.distributed.distributed_c10d | Rank 1: Completed store-based barrier for key:store_based_barrier_key:1 with 2 nodes.
2024-08-28 14:01:18 | INFO | unicore.distributed.utils | initialized host 27d257810fc0 as rank 1
Traceback (most recent call last):
File "./Uni-Mol/unimol/unimol/infer.py", line 120, in
cli_main()
File "./Uni-Mol/unimol/unimol/infer.py", line 116, in cli_main
distributed_utils.call_main(args, main)
File "/opt/conda/lib/python3.8/site-packages/unicore-0.0.1-py3.8-linux-x86_64.egg/unicore/distributed/utils.py", line 180, in call_main
torch.multiprocessing.spawn(
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 240, in spawn
return start_processes(fn, args, nprocs, join, daemon, start_method='spawn')
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 198, in start_processes
while not context.join():
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 160, in join
raise ProcessRaisedException(msg, error_index, failed_process.pid)
torch.multiprocessing.spawn.ProcessRaisedException:
-- Process 0 terminated with the following error:
Traceback (most recent call last):
File "/opt/conda/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 69, in _wrap
fn(i, *args)
File "/opt/conda/lib/python3.8/site-packages/unicore-0.0.1-py3.8-linux-x86_64.egg/unicore/distributed/utils.py", line 158, in distributed_main
args.distributed_rank = distributed_init(args)
File "/opt/conda/lib/python3.8/site-packages/unicore-0.0.1-py3.8-linux-x86_64.egg/unicore/distributed/utils.py", line 138, in distributed_init
dist.all_reduce(torch.zeros(1).cuda())
RuntimeError: CUDA error: out of memory
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
`
It seems to be related to a numpy version problem. I am encountering this issue while trying to run unimol_binding_pose_demo from Uni-Mol. I would appreciate any guidance on how to resolve this.
The text was updated successfully, but these errors were encountered: