Matthew Schlegel
Matthew Schlegel
When dashboard-refresh-buffer is called all the recent files are removed, even though the recentf-list is still set with files in it. I have a really minimal config with use-package, and...
This changes the RNN implementation such that it stores all its parameters in a block. We include helper views to access specific parts of the weight matrix more easily. This...
While we were discussing #1675 and #1671 several improvements/updates to the recurrent network API came up. Instead of taking over #1675, @ToucheSir and myself thought it would be best to...
Add function to seed the RNG of the underlying python/gym.
I was wondering if there were any plans to specialize when sparse arrays are loaded/being used. For example: ```julia using SparseArrays, LinearAlgebra, Tullio, BenchmarkTools, LoopVectorization n = 10000 n2 =...
I've been benchmarking packages for a tensor operation package to use w/ Flux. I noticed that OMEinsum was faster for a particular operation than Tullio, but Tullio was faster in...
This adds the NewRecur functionality from Fluxperimental.jl. There are some road bumps that need to be ironed out. 1. The rrule for `scan_full` needs to be made type stable. I...
### Motivation and description Given #2185 and other issues caused by the current mutability of the recur interface, we should move to a more standard blocked (i.e. 3D for simple...
When having Flux v0.13.15 installed in a cluster environment I am facing an issue where loading the package through `using Flux` fails the first time with the error: ```julia julia>...
Just working through my upgrade to Flux v0.13. I'm seeing a difference in `batchseq`'s behavior. Specifically, in the Flux 0.12 `batchseq` I can do the following: ```julia Flux.batchseq([[rand(3), rand(3), rand(3)],...