Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs · HackerPapers discussions

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — mentioned by KenoFischer

Fri, 19 Jul 2019 20:23:26 GMT

JAX is a sophisticated and well implemented high level tracer, combined with a backend compiler that makes use of XLA. High level tracing has some fundamental draw backs in key features we want (efficient scalar AD for example), so the approach being advocated here is doing AD as a compiler transform on the original source language. Mike has some details in an earlier paper: https://arxiv.org/abs/1810.07951. As a shameless plug, you can plug (heh) Julia into XLA as well, e.g. for targeting TPUs (https://arxiv.org/abs/1810.09868). In fact, Julia and JAX are probably better at that than…

Paper and discussion history. A mention is not an endorsement.

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — mentioned by KenoFischer

Wed, 27 Feb 2019 18:36:10 GMT

There's very little TensorFlow here. We're using it as a glorified grpc client basically to ship XLA to the TPU. The actual mechanism is described in https://arxiv.org/abs/1810.09868. The model definition itself uses the layers from Flux, but there's a couple assumptions that Flux makes that don't hold for the TPU so we don't get to use everything from Flux (e.g. the training loop and the optimizers). Luckily everything is julia, so we can get TPU-compatiable versions in just a few lines of code. Unifying the abstractions is an active work in progress, but this shows off that Julia runs on…

Paper and discussion history. A mention is not an endorsement.

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — mentioned by chrispeel

Sat, 09 Feb 2019 01:52:16 GMT

As I recall, after the publication of [1] on Julia+TPUs there was some talk about doing a comparison of some ML models on the CPU, GPU, and TPU. Is this something you're working on? [1] https://arxiv.org/pdf/1810.09868.pdf

Paper and discussion history. A mention is not an endorsement.

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — mentioned by byt143

Thu, 10 Jan 2019 23:14:18 GMT

>Julia's secret sauce is LLVM That most certainly is not true. See this for example: https://arxiv.org/pdf/1810.09868.pdf

Paper and discussion history. A mention is not an endorsement.

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — HN submission

Mon, 29 Oct 2018 12:22:50 GMT

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — HN submission

Thu, 25 Oct 2018 03:22:29 GMT

Automatic Compilation of Julia Programs and Generic ML Models to Cloud TPUs

7 points; 2 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — HN submission

Wed, 24 Oct 2018 19:13:50 GMT

Compiling Julia to TPUs [pdf]

7 points; 5 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs — HN submission

Wed, 24 Oct 2018 07:14:02 GMT

Automatic Full Compilation of Julia Programs and ML Models to Cloud TPUs

7 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.