How to install esmfold for A100 on cuda 10.1

[linda5mith]

Can someone please help me to install esmfold?

nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Sun_Jul_28_19:07:16_PDT_2019
Cuda compilation tools, release 10.1, V10.1.243

nvidia-smi
Thu Jul 27 15:17:42 2023       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.125.06   Driver Version: 525.125.06   CUDA Version: 12.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA A100-SXM...  Off  | 00000000:00:07.0 Off |                    0 |
| N/A   32C    P0    55W / 500W |      7MiB / 81920MiB |      0%      Default |
|                               |                      |             Disabled |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A       975      G   /usr/lib/xorg/Xorg                  4MiB |
+-----------------------------------------------------------------------------+

This is what I'm doing:

1. conda create --name esmfold python=3.7
2. conda activate esmfold
   Installing cuda compatible version of pytorch for cuda 10.1:
3. pip3 install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html
4. pip install "fair-esm[esmfold]"

Attempting uninstall: torch
    Found existing installation: torch 1.9.0+cu102
    Uninstalling torch-1.9.0+cu102:
      Successfully uninstalled torch-1.9.0+cu102
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
torchaudio 0.9.0 requires torch==1.9.0, but you have torch 2.0.1 which is incompatible.
torchvision 0.10.0+cu102 requires torch==1.9.0, but you have torch 2.0.1 which is incompatible.

6. pip install 'dllogger @ git+https://github.com/NVIDIA/dllogger.git'
7. pip install 'openfold @ git+https://github.com/aqlaboratory/openfold.git@4b41059694619831a7db195b7e0988fc4ff3a307'

RuntimeError:
      The detected CUDA version (10.1) mismatches the version that was used to compile
      PyTorch (11.7). Please make sure to use the same CUDA versions.
      
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
error: legacy-install-failure

× Encountered error while trying to install package.
╰─> openfold

Ok so I try to resolve the torch cuda version conflict by running step 4 again:
8. pip3 install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html

Attempting uninstall: torch
    Found existing installation: torch 1.13.1
    Uninstalling torch-1.13.1:
      Successfully uninstalled torch-1.13.1
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
pytorch-lightning 1.9.5 requires torch>=1.10.0, but you have torch 1.9.0+cu111 which is incompatible.
Successfully installed torch-1.9.0+cu111

9. pip install pytorch_lightning==1.5.10
10. pip install 'openfold @ git+https://github.com/aqlaboratory/openfold.git@4b41059694619831a7db195b7e0988fc4ff3a307'

File "/home/administrator/programs/miniconda3/envs/esmfold/lib/python3.7/site-packages/torch/utils/cpp_extension.py", line 1682, in _run_ninja_build
          raise RuntimeError(message) from e
      RuntimeError: Error compiling objects for extension
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
error: legacy-install-failure

× Encountered error while trying to install package.
╰─> openfold

note: This is an issue with the package mentioned above, not pip.
hint: See above for output from the failure.

Woudl really appreciate any help. I'm losing my mind. Thanks so much!

[evgeni-nikolaev-bio]

Hello Linda I also need help with the installation. Could you let me know if you solved your problem? Can you share your experience? I appreciate any help you can provide. -Evgeni.

…

On Jul 27, 2023, at 11:04 AM, Linda Smith ***@***.***> wrote: Can someone please help me to install esmfold? nvcc --version nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2019 NVIDIA Corporation Built on Sun_Jul_28_19:07:16_PDT_2019 Cuda compilation tools, release 10.1, V10.1.243 Thu Jul 27 15:17:42 2023 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 525.125.06 Driver Version: 525.125.06 CUDA Version: 12.0 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 NVIDIA A100-SXM... Off | 00000000:00:07.0 Off | 0 | | N/A 32C P0 55W / 500W | 7MiB / 81920MiB | 0% Default | | | | Disabled | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================| | 0 N/A N/A 975 G /usr/lib/xorg/Xorg 4MiB | +-----------------------------------------------------------------------------+ This is what I'm doing: conda create --name esmfold python=3.7 conda activate esmfold Installing cuda compatible version of pytorch for cuda 10.1 pip3 install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html pip install "fair-esm[esmfold]" Attempting uninstall: torch Found existing installation: torch 1.9.0+cu102 Uninstalling torch-1.9.0+cu102: Successfully uninstalled torch-1.9.0+cu102 ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. torchaudio 0.9.0 requires torch==1.9.0, but you have torch 2.0.1 which is incompatible. torchvision 0.10.0+cu102 requires torch==1.9.0, but you have torch 2.0.1 which is incompatible. pip install 'dllogger @ git+https://github.com/NVIDIA/dllogger.git' pip install 'openfold @ ***@***.***' RuntimeError: The detected CUDA version (10.1) mismatches the version that was used to compile PyTorch (11.7). Please make sure to use the same CUDA versions. [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. error: legacy-install-failure × Encountered error while trying to install package. ╰─> openfold Ok so I try to resolve the torch version conflict by running step 4 again: 8. pip3 install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html Attempting uninstall: torch Found existing installation: torch 1.13.1 Uninstalling torch-1.13.1: Successfully uninstalled torch-1.13.1 ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. pytorch-lightning 1.9.5 requires torch>=1.10.0, but you have torch 1.9.0+cu111 which is incompatible. Successfully installed torch-1.9.0+cu111 pip install pytorch_lightning==1.5.10 pip install 'openfold @ ***@***.***' File "/home/administrator/programs/miniconda3/envs/esmfold/lib/python3.7/site-packages/torch/utils/cpp_extension.py", line 1682, in _run_ninja_build raise RuntimeError(message) from e RuntimeError: Error compiling objects for extension [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. error: legacy-install-failure × Encountered error while trying to install package. ╰─> openfold note: This is an issue with the package mentioned above, not pip. hint: See above for output from the failure. Woudl really appreciate any help. I'm losing my mind. Thanks so much! — Reply to this email directly, view it on GitHub <#594>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AZS3CJCDT4BLCYDV44JYO4LXSJ7PRANCNFSM6AAAAAA22HFLKA>. You are receiving this because you are subscribed to this thread.

[imSeaton]

same problem with you when executing pip install 'openfold @ git+https://github.com/aqlaboratory/openfold.git@4b41059694619831a7db195b7e0988fc4ff3a307'

[linda5mith]

Hey so there are error lines in the output which are important that I didn't include which were relevant to my issue:

/data0/miniconda3/envs/esmfold/lib/python3.7/site-packages/torch/utils/cpp_extension.py:813: UserWarning: The detected CUDA version (11.5) has a minor version mismatch with the version that was used to compile PyTorch (11.3). Most likely this shouldn't be a problem.

warnings.warn(CUDA_MISMATCH_WARN.format(cuda_str_version, torch.version.cuda))
/data0/miniconda3/envs/esmfold/lib/python3.7/site-packages/torch/utils/cpp_extension.py:820: UserWarning: There are no g++ version bounds defined for CUDA version 11.5

Ultimately my issue was resolved by downgrading my gcc++ version by following this tutorial. My working configuration:

nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Sun_Mar_21_19:15:46_PDT_2021
Cuda compilation tools, release 11.3, V11.3.58
Build cuda_11.3.r11.3/compiler.29745058_0

gcc --version
gcc (Ubuntu 10.5.0-1ubuntu1~22.04) 10.5.0
Copyright (C) 2020 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

I followed this table to find CUDA driver and gcc ++ version compatability.

Hope this helps :)