Repository navigation
Troubleshooting
The first run of an instance writes a .vnnlib.compiled cache next to its .vnnlib file in benchmarks/, and on a GPU the first solve compiles GPU kernels (Triton). Run one small instance first (Quick Start) and do not time it. The benchmarks/ directory must therefore be writable. The .compiled files are not part of SHA256SUMS.
- The status
MEMORY_LIMITin the Python API (Result.termination) means the search stopped on memory. - Lower the number of boxes per batch with
batch=gr.Batch(domains=...)(solver.batch_size) and setsolver.auto_enlarge_batch_sizetofalseso the runner does not grow it again (Configuration). - Lower
bab.domain_storage_gpu_mb, the memory in which stored boxes stay on the GPU; boxes beyond it are kept in host memory. - Lower
solver.batched_crown_auto_enlarge_max_vram_ratioto leave memory free. - Close other processes on the GPU.
Result.stats.peak_gpu_mbshows the peak of the last solve. - The largest instances of the benchmark need the most memory; on a smaller GPU than the one used for the paper, some instances may stop earlier.
- The tested environment is PyTorch 2.8.0 built for CUDA 12.8 (
torch==2.8.0from thecu128index, see Installation). The check at the end of Installation prints the PyTorch version and whether CUDA is available. - A Triton version that does not match the PyTorch build fails when the first kernel is compiled. Use the Triton version that
torch==2.8.0installs. - The fused kernel is used only on CUDA in float64. If it fails, GEAR switches it off and continues with the standard implementation, which is slower.
Use --device cpu (or device="cpu") for small tests. Expect much longer run times, and do not compare CPU results with the GPU results in logs/.
Compare certified flags and instance counts, not times (Reproducing the Paper). Check Result.provenance for the preset, device and commit, and the Configurations: note in Output and Logs.
The default preset calls an LP solver on the CPU and IPOPT in parts of the search (Python API). The CPU solvers a solve used are listed in Result.provenance.