FluxML/Flux.jl

Benchmark optimizer latency on GPU

Open

#1699 aperta il 24 ago 2021

Vedi su GitHub
 (0 commenti) (0 reazioni) (0 assegnatari)Julia (619 fork)batch import
cudahelp wantedperformance

Metriche repository

Star
 (4725 star)
Metriche merge PR
 (Merge medio 4h 27m) (2 PR mergiate in 30 g)

Descrizione

Ref. https://github.com/JuliaGPU/GPUCompiler.jl/issues/65#issuecomment-903155469. We should do this for both the current mutating optimizers and for Optimisers.jl. It may be that TTFG is only part of the problem.

Guida contributor