The objective is to compare the performances of the mlir sparse tensor
reference implementation, numpy matrix multiplication, and pytaco
multiplication.
Details
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
I have copied pytaco tools directory from tests directory to benchmark to use python bindings from benchmark. We can decide on the next course of action now that we have a concrete revision.
mlir/benchmark/python/sparse_tensor/benchmark_sparse.py | ||
---|---|---|
128–134 | I am choosing these matrix sizes here. Are they appropriate for benchmarking? I also moved this code to setup because it seemed like it was taking a long time so I thought it appropriate to move it under compiler. Let me know your thoughts. |
mlir/benchmark/python/sparse_tensor/tools/mlir_pytaco.py | ||
---|---|---|
1869 | That is a non-trivial amount of code here duplicated in the repo isn't it? |
mlir/benchmark/python/sparse_tensor/tools/mlir_pytaco.py | ||
---|---|---|
1869 | Yeah +1 I thought you would just move a single *small* routine over, and that we could consolidate that later. |
mlir/benchmark/python/sparse_tensor/tools/mlir_pytaco.py | ||
---|---|---|
1869 | I will work on trimming this a bit more. |
I couldn't trim it in obvious ways. I am closing this patch for now and approach it another way in a new patch. Sorry about that.
No worries, we are looking forward to your next patch. Please know that we appreciate your efforts a lot!
I am choosing these matrix sizes here. Are they appropriate for benchmarking?
I also moved this code to setup because it seemed like it was taking a long time so I thought it appropriate to move it under compiler. Let me know your thoughts.