search

uw-ipd / RoseTTAFold2NA

RoseTTAFold2 protein/nucleic acid complex prediction
MIT License
310 stars 69 forks source link

_spherical_harmonics.py error when running example #24

Open fafnir31 opened 1 year ago

fafnir31 commented 1 year ago

After complete installation according to instructions and testing the example, I have the following error Running on GPU plddt best /home/xxxx/miniconda3_4.6.14/envs/RF2NA/lib/python3.8/site-packages/e3nn/o3/_spherical_harmonics.py:82: UserWarning: FALLBACK path has been taken inside: compileCudaFusionGroup. This is an indication that codegen Failed for some reason. To debug try disable codegen fallback path via setting the env variable export PYTORCH_NVFUSER_DISABLE=fallback To report the issue, try enable logging via setting the envvariable export PYTORCH_JIT_LOG_LEVEL=manager.cpp (Triggered internally at /opt/conda/conda-bld/pytorch_1659484810403/work/torch/csrc/jit/codegen/cuda/manager.cpp:237.) sh = _spherical_harmonics(self._lmax, x[..., 0], x[..., 1], x[..., 2])

However the program continues and generates models

Phage-structure-geek commented 1 year ago

Hi,

If you read carefully, this is not an error. It's a warning. I see the same thing (see below, the path is different). A good-looking model is created nevertheless.

It would be good, of course, if the github distro contained an example output in addition to the input so that we could compare our models with the author-generated model.

Petr

./runRF2NA.sh t001 P:protein.fa R:RNA.fa Running HHblits Running PSIPRED Running hhsearch Running rMSA (lite) Running RoseTTAFold2NA to predict structures Running on GPU plddt best /home/leiman/anaconda3/envs/RF2NA/lib/python3.8/site-packages/e3nn/o3/_spherical_harmonics.py:82: UserWarning: FALLBACK path has been taken inside: compileCudaFusionGroup. This is an indication that codegen Failed for some reason. To debug try disable codegen fallback path via setting the env variable export PYTORCH_NVFUSER_DISABLE=fallback To report the issue, try enable logging via setting the envvariable export PYTORCH_JIT_LOG_LEVEL=manager.cpp (Triggered internally at ../torch/csrc/jit/codegen/cuda/manager.cpp:237.) sh = _spherical_harmonics(self._lmax, x[..., 0], x[..., 1], x[..., 2]) RECYCLE 0 0.836 -1.000 RECYCLE 1 0.861 0.836 RECYCLE 2 0.866 0.861 RECYCLE 3 0.870 0.866 RECYCLE 4 0.871 0.870 RECYCLE 5 0.873 0.871 RECYCLE 6 0.873 0.873 RECYCLE 7 0.873 0.873 RECYCLE 8 0.873 0.873 RECYCLE 9 0.874 0.873 Done