issues
search
FluxML
/
Optimisers.jl
Optimisers.jl defines many standard optimisers and utilities for learning loops.
https://fluxml.ai/Optimisers.jl
MIT License
72
stars
20
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
GPU kernels for optimizers
#178
vpuri3
opened
2 weeks ago
2
Restructure is not type stable but could be made stable?
#177
Red-Portal
closed
4 weeks ago
4
Grokfast exponential moving average Optimizer
#176
vpuri3
opened
1 month ago
0
Add `trainables_nt`
#175
CarloLucibello
opened
3 months ago
6
add `path` option to `trainables`
#174
CarloLucibello
closed
3 months ago
2
add trainables_with_path
#173
CarloLucibello
closed
3 months ago
3
fix broken documentation
#172
CarloLucibello
closed
3 months ago
0
add `trainables`
#171
CarloLucibello
closed
3 months ago
5
fix broken doc links
#170
CarloLucibello
closed
3 months ago
0
Documenter CI is failing
#169
CarloLucibello
closed
3 months ago
0
add missing SignDecay doc reference
#168
CarloLucibello
closed
4 months ago
0
Adam optimizer can produce NaNs with Float16 due to small epsilon
#167
pevnak
opened
5 months ago
3
Don't load Yota at all
#166
mcabbott
closed
5 months ago
3
Add in-place `destructure!`
#165
mcabbott
opened
8 months ago
9
CompatHelper: add new compat entry for Statistics at version 1, (keep existing compat)
#164
github-actions[bot]
closed
9 months ago
0
`reset!(optimiser_state)`
#163
Vilin97
opened
9 months ago
2
Type instability in `Flux.setup`
#162
Vilin97
opened
9 months ago
7
Document `destructure` handling shared parameters differently to ComponentArrays.jl
#161
mcabbott
opened
9 months ago
0
Add all-keyword constructors, much like `@kwdef`
#160
mcabbott
closed
5 months ago
3
WeightDecay for L1 norm
#159
mcabbott
closed
5 months ago
20
Fix Float64 beta in Adam etc.
#158
mcabbott
closed
10 months ago
0
Add Lion to docs
#157
ToucheSir
closed
11 months ago
0
Implement Lion, up to 5x faster than Adam, and more accurate
#156
PallHaraldsson
closed
11 months ago
7
`destructure` doesn't work on Dictionaries
#154
mcabbott
opened
11 months ago
1
How to handle long compile times?
#153
DrChainsaw
opened
12 months ago
4
Rule for mixed precision training
#152
CarloLucibello
opened
1 year ago
9
Use `eltype(x)` everywhere, ignore `typeof(η)`
#151
mcabbott
closed
11 months ago
6
Error in `update!` for Metal arrays and Adam optimiser
#150
CarloLucibello
closed
10 months ago
4
Add a new optimizer PAdam
#149
4SAnalyticsnModelling
opened
1 year ago
0
Add a variant of Adam called "PAdam"
#148
4SAnalyticsnModelling
closed
1 year ago
3
Update Optimisers.jl
#147
4SAnalyticsnModelling
closed
1 year ago
0
update readme
#145
mcabbott
opened
1 year ago
4
Make ClipNorm work on GPU Broadcasted
#144
mcabbott
closed
1 year ago
0
Restructure makes a copy
#146
linusheck
opened
1 year ago
4
Utility for walking a tree (e.g. gradients) w.r.t. a model
#143
darsnack
opened
1 year ago
6
Update TagBot config
#142
ToucheSir
closed
1 year ago
0
Optimisers.update fails with gradient of type `CUDA.CUSPARSE.CuSparseMatrixCSC`
#141
hsseung
closed
1 year ago
5
`nothing` does not correspond to updating the state with a zero gradient.
#140
CarloLucibello
opened
1 year ago
0
write `>>` as infix notation for `OptimiserChain`
#139
mcabbott
opened
1 year ago
12
Use `OptChain` as an alias for `OptimiserChain`?
#138
CarloLucibello
opened
1 year ago
1
Rule for gradient accumulation
#137
CarloLucibello
closed
1 year ago
0
Don't use `state` anywhere for the whole state tree
#136
mcabbott
closed
1 year ago
0
some minor style changes
#135
CarloLucibello
opened
1 year ago
6
fix: rm Functors@0.3
#134
ven-k
closed
1 year ago
3
docs on freezing layers reworded
#133
ghost
closed
1 year ago
1
fix typo
#132
bicycle1885
closed
1 year ago
0
Documentation error
#131
erlebach
closed
1 year ago
1
Interface for gradient accumulation
#130
chengchingwen
closed
1 year ago
7
Add implementation of Lion optimiser
#129
mashu
closed
1 year ago
3
Discourage use of `trainable`
#128
mcabbott
closed
1 year ago
0
Next