search

RosettaCommons / RFdiffusion

Code for running RFdiffusion
Other
1.63k stars 311 forks source link

RuntimeError: CUDA out of memory. Tried to allocate 638.00 MiB (GPU 0; 6.00 GiB total capacity; 3.23 GiB already allocated; 117.44 MiB free; 4.13 GiB reserved in total by PyTorch) Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace. #198

Open Hanoriega opened 7 months ago

Hanoriega commented 7 months ago

Hello,

I have been trying to recreate other oligomers, with an example of the code below:

python .\scripts\run_inference.py --config-name=symmetry inference.symmetry="C4" inference.num_designs=1 inference.output_prefix="example_outputs/C4_oligo" "potentials.guiding_potentials=['type:olig_contacts,weight_intra:1,weight_inter:0.1']" potentials.olig_intra_all=True potentials.olig_inter_all=True potentials.guide_scale=2.0 potentials.guide_decay="quadratic" "contigmap.contigs=[660-660]"

Why does it keep giving me this error?

RuntimeError: CUDA out of memory. Tried to allocate 638.00 MiB (GPU 0; 6.00 GiB total capacity; 3.23 GiB already allocated; 117.44 MiB free; 4.13 GiB reserved in total by PyTorch) Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.

Is there a way to correct this or extend the memory?

Thank you,

Heather

KoningCheems commented 6 months ago

I think the problem is you're trying to diffuse a protein that is too large for your GPU to handle. If that is not crucial to your design, try reducing the "contigmap.contigs=[660-660]"