loop_tool

I've done some work in the past on NN representations and you actually can represent Conv and MatMul in more primitive ways. I ended up writing an IR called loop_tool that exposes this stuff: https://github.com/facebookresearch/loop_tool/blob/main/pyth... The idea is basically this: https://news.ycombinator.com/item?id=28883086

I just finished adding a basic WASM[1] backend + basic JavaScript frontend[2]. I'm in the process of adding in-browser optimization[3] and will hope to have a demo some time this week! [1] https://github.com/facebookresearch/loop_tool/blob/main/src/... [2] https://github.com/facebookresearch/loop_tool/blob/main/java... [3] https://github.com/facebookresearch/loop_tool/blob/main/test...

In an ideal world, absolutely! It's a hard problem and there are many attempts to make that happen automatically including polyhedral optimization (Polly[1]) and tensor compiler libraries (XLA[2] and TVM[3]). I work on a project called LoopTool[4] which is researching ways to dramatically reduce the representations of the other projects to simplify optimization scope. [1] https://polly.llvm.org [2] https://www.tensorflow.org/xla [3] https://tvm.apache.org [4] https://github.com/facebookresearch/loop_tool