v0.13.12

Flux v0.13.12

Diff since v0.13.11

Closed issues:

• Delta neural networks inference (#2129)
• [Bug] Embedding forward pass breaks for onehotbatch with multiple batch dimensions (#2160)
• MethodError: no method matching when training LSTMs even when loss function is working corrently (#2168)
• Type instability with Flux.update! when loss function involves extra arguments (#2175)

Merged pull requests:

• Un-deprecate track_stats for InstanceNorm (#2149) (@ToucheSir)
• Move dropout to NNlib (#2150) (@mcabbott)
• Use NNlib's within_gradient (#2152) (@mcabbott)
• Export rand32 and friends (#2157) (@mcabbott)
• Remove piratical array conversion rule (#2167) (@ToucheSir)
• update: actions node 12 => node 16 (#2173) (@skyleaworlder)
• cuda 4.0 compat (#2177) (@CarloLucibello)