Error Running the eval_samples.sh Script

[vecos1990]

Dear all, I encountered the following error at step 3: I have tried severally to install different versions of the conflicting packages but the error gets more complicated. This is the error from a fresh installation with no modifications as per the README file. Wondering is anyone is facing the same

-------------------------------- Step3 : Get Pocket ESM Embedding ----------------------------
/homes/cveranso/.local/lib/python3.10/site-packages/pandas/core/computation/expressions.py:21: UserWarning: Pandas requires version '2.8.4' or newer of 'numexpr' (version '2.7.3' currently installed).
from pandas.core.computation.check import NUMEXPR_INSTALLED
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 18.24it/s]
have done ,just skip!: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 3139.45it/s]
100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 215.55it/s]
------------------------ Step3 : Start Sampling Ligand Confromers ----------------------------
_/homes/cveranso/.local/lib/python3.10/site-packages/pandas/core/computation/expressions.py:21: UserWarning: Pandas requires version '2.8.4' or newer of 'numexpr' (version '2.7.3' currently installed).
from pandas.core.computation.check import NUMEXPR_INSTALLED
Traceback (most recent call last):
File "/home/disk6/homes/cveranso/SurfDock/inference_accelerate.py", line 19, in
from datasets.process_mols import write_mol_with_coords
File "/home/disk6/homes/cveranso/SurfDock/datasets/process_mols.py", line 16, in
from torch_cluster import radius_graph
File "/homes/cveranso/miniconda3/envs/SurfDock/lib/python3.10/site-packages/torch_cluster/init.py", line 18, in
torch.ops.load_library(spec.origin)
File "/homes/cveranso/miniconda3/envs/SurfDock/lib/python3.10/site-packages/torch/_ops.py", line 1032, in load_library
ctypes.CDLL(path)
File "/homes/cveranso/miniconda3/envs/SurfDock/lib/python3.10/ctypes/init.py", line 374, in init
self._handle = _dlopen(self._name, mode)
OSError: /home/disk6/homes/cveranso/miniconda3/envs/SurfDock/lib/python3.10/site-packages/torch_cluster/_version_cuda.so: undefined symbol: ZN5torch3jit17parseSchemaOrNameERKSsb
E0113 21:48:07.351000 22437543982912 torch/distributed/elastic/multiprocessing/api.py:826] failed (exitcode: 1) local_rank: 0 (pid: 2146838) of binary: /homes/cveranso/miniconda3/envs/SurfDock/bin/python3.10

[CLG68]

I made an installation script and added:
conda install -y -c conda-forge numexpr=2.8.x

It solved the panda/numexpr problem but I'm still having other problems when running eval_samples.sh

[CAODH]

I made an installation script and added: conda install -y -c conda-forge numexpr=2.8.x

It solved the panda/numexpr problem but I'm still having other problems when running eval_samples.sh

It seems that this issue is caused by installing pyg-related packages, particularly torch_cluster.

[vecos1990]

Is there a turnaround for this?

Because I tried playing with different torch versions and pyg related packages.
However, I only move from one decency issue to another?

[CLG68]

Everything works for me now...

I did:
conda install -y -c conda-forge numexpr=2.8.x, solve the numexpr problem.

I made the activation of the environment dynamic in the screening script:

if [[ -z "$CONDA_PREFIX" || "$(basename $CONDA_PREFIX)" != "SurfDock" ]]; then
source $(conda info --base)/etc/profile.d/conda.sh
conda activate SurfDock
fi

All my other problems were indirectly related to: from where I was running the script. I created a screening script in the root of SurfDock that can be ran from there and it solved all the remaining problems.

In my environment, I have:

pyg 2.5.2 py310_torch_2.2.0_cu121 pyg
pyg-lib 0.4.0+pt22cu121 pypi_0 pypi

torch-cluster 1.6.3+pt22cu121 pypi_0 pypi
torch-ema 0.3 pyhd8ed1ab_1 conda-forge
torch-scatter 2.1.2+pt22cu121 pypi_0 pypi
torch-sparse 0.6.18+pt22cu121 pypi_0 pypi
torch-spline-conv 1.2.2+pt22cu121 pypi_0 pypi
torchmetrics 1.2.1 pyhd8ed1ab_0 conda-forge
torchtriton 2.2.0 py310 pytorch