mlx/benchmarks/python/comparative
2025-06-15 17:34:10 +08:00
..
bench_mlx.py feat: Add benchmarking for ReLUSquared activation function 2025-06-15 17:34:10 +08:00
bench_torch.py feat: Add benchmarking for ReLUSquared activation function 2025-06-15 17:34:10 +08:00
compare.py feat: Add benchmarking for ReLUSquared activation function 2025-06-15 17:34:10 +08:00
README.md awni's commit files 2023-11-29 10:30:41 -08:00

Microbenchmarks comparing MLX to PyTorch

Implement the same microbenchmarks in MLX and PyTorch to compare and make a list of the biggest possible performance improvements and/or regressions.

Run with python bench_mlx.py sum_axis --size 8x1024x128 --axis 2 --cpu for instance to measure the times it takes to sum across the 3rd axis of the above tensor on the cpu.

compare.py runs several benchmarks and compares the speed-up or lack thereof in comparison to PyTorch.

Each bench script can be run with --print-pid to print the PID and wait for a key in order to ease attaching a debugger.