Keno Fischer
16369769ab
Handle various cases of multiplying transpose-wrapped matrices
...
See test cases. I hit these while taking third-order derivatives of
matrix multiplies (whose gradient definitions use transpose).
2019-02-06 10:41:32 -05:00
Keno Fischer
d36d7879df
Add higher order autodiff for getindex
...
This fixed higher order autodiff or getindex (and by extension vcat, since
it's backward pass uses getindex). This makes tracker able to handle
the third order derivatives needed to implement [1].
[1] Physics-Informed Generative Adversarial Networks for Stochastic Differential Equations. https://arxiv.org/abs/1811.02033
2019-02-06 10:41:23 -05:00
Mike J Innes
ecc55ec9e1
Revert "Fix OneHotVector/Matrix performance on GPU"
2019-02-06 14:31:15 +00:00
Mike J Innes
e8b2ec6f67
Merge pull request #311 from tejank10/conv_transpose
...
2D Conv transpose support
2019-02-06 14:14:14 +00:00
Mike J Innes
7fc920240d
Merge pull request #591 from dhairyagandhi96/onehot
...
Fix OneHotVector/Matrix performance on GPU
2019-02-04 13:53:55 +00:00
Dhairya Gandhi
2f916f9763
better tests
2019-02-04 18:43:25 +05:30
Dhairya Gandhi
6654ebfc90
added onecold broadcast test
2019-02-04 17:57:34 +05:30
Mike J Innes
cfe6859186
auto-collect in forward
2019-02-04 10:37:02 +00:00
Mike J Innes
838070968e
vcat with scalars
2019-02-04 00:05:16 +00:00
Tejan Karmali
e54df2de06
Merge branch 'master' into conv_transpose
2019-02-02 10:20:45 +05:30
Mike J Innes
0469394715
Merge pull request #576 from mcabbott/patch-1
...
PermutedDimsArray
2019-01-29 14:55:55 +00:00
Anand Bisen
3670fabbe6
add tests for stack and unstack
2019-01-29 01:41:15 -08:00
Michael Abbott
55a7359f67
PermutedDimsArray test
2019-01-28 18:19:06 +01:00
Mike J Innes
0f2975d905
update -> apply
2019-01-28 13:59:23 +00:00
Mike J Innes
013b421b08
Merge pull request #570 from avik-pal/ap/batchnorm_fixes
...
Patches for default initializers
2019-01-28 10:40:55 +00:00
Mike J Innes
58ac415f6b
forward mode
2019-01-25 16:14:24 +00:00
Mike J Innes
791939709b
numeric precision utilities
2019-01-25 10:06:37 +00:00
Avik Pal
2f3ad56166
Add test for Depthwise Conv
2019-01-24 18:53:04 +05:30
Mike Innes
0142d89943
test onecold-of-tracked-gpu-vector
...
see #556
2019-01-24 10:40:52 +00:00
Mike J Innes
152ce4a164
conversions for dual numbers
2019-01-22 10:07:42 +00:00
Mike J Innes
f6397e7358
Merge pull request #517 from FluxML/fix_adamw
...
Fix decay argument in ADAMW
2019-01-18 10:06:23 +00:00
Mike J Innes
4d79f499bf
fixes #549
2019-01-15 15:49:37 +00:00
Mike J Innes
a3e0de1ee5
fixes #516
2019-01-15 15:49:18 +00:00
Mike J Innes
67d9016319
Merge pull request #538 from KristofferC/kc/promote
...
fix promotion by avoiding integer division in mse and crossentropy
2019-01-15 13:20:46 +00:00
Kristoffer Carlsson
c74aa67c5d
fix promotion by avoiding integer division in mse and crossentropy
...
oops
add tests
2019-01-15 14:15:05 +01:00
Mike J Innes
735b970c12
fix update for scalars
2019-01-10 10:19:05 +00:00
Mike J Innes
6b11c552f3
better h/vcat, fixes #378
2018-12-19 11:19:01 +00:00
Dhairya Gandhi
e48268ff06
fix argument name in ADAMW
2018-12-12 16:47:42 +05:30
Tejan Karmali
1648414a5d
fixes for layer and test
2018-12-04 11:08:40 -05:00
Tejan Karmali
95e490a2c5
merge conflict resolved
2018-11-28 11:10:22 -05:00
Tejan Karmali
a71ee386d0
1.0 fix for conv transpose
2018-11-28 10:55:21 -05:00
Mike J Innes
1c36504768
fixup
2018-11-27 18:44:07 -05:00
Avik Pal
dfd680646c
Fix conflict
2018-11-14 22:18:57 +05:30
Mike J Innes
b3331205d1
faster default gradient performance
2018-11-12 23:39:25 +00:00
Mike J Innes
903db70673
float32 param initialisers
2018-11-12 20:10:47 +00:00
Avik Pal
9f12e8ec68
Make the test more reliable
2018-11-10 14:00:25 +05:30
Avik Pal
4df9e10516
Add test for 2D inputs
2018-11-10 11:52:23 +05:30
Avik Pal
564518e448
Merge branch 'master' of https://github.com/FluxML/Flux.jl into cudnn_batchnorm
2018-11-08 19:13:34 +05:30
Mike J Innes
30486f9c03
Merge pull request #441 from Paethon/rm_initn
...
Removes initn initialization
2018-11-08 13:25:02 +00:00
Mike J Innes
5e572df557
Merge pull request #485 from dhairyagandhi96/master
...
Add call back
2018-11-08 13:18:17 +00:00
Dhairya Gandhi
392c3c942b
re-add removed call function
2018-11-08 18:44:57 +05:30
Mike J Innes
d0e4fbb1e0
Merge branch 'master' into ed/diagm-pair
2018-11-05 11:51:29 +00:00
Mike J Innes
43c5f90d93
Merge pull request #379 from dhairyagandhi96/master
...
New optimisers interface
2018-10-31 16:38:40 +00:00
Mike J Innes
bffaceee02
tweaks
2018-10-31 14:58:55 +00:00
Mike J Innes
70283e1971
Merge pull request #465 from FluxML/mji/once
...
Destroy AD graph when doing in-place gradients
2018-10-31 14:14:38 +00:00
Mike J Innes
77178b7d67
remove old-style definition and test
2018-10-30 14:21:22 +00:00
Dhairya Gandhi
bebf4eb95f
fixed ExpDecay update! rule
2018-10-29 23:12:24 +05:30
Dhairya Gandhi
32ce2d78b8
fixed ExpDecay test
2018-10-27 19:53:06 +05:30
Dhairya Gandhi
ea508a79b0
use explicit update! rule
2018-10-27 19:39:56 +05:30
Dhairya Gandhi
815e8c206d
decay fixes
2018-10-27 19:26:42 +05:30
