index
:
nerv
bugfix-matrixfree
fastnn
gh-pages
master
rnn
tnn
wrapped-handles
Lua-based toolkit for high-performance deep learning
about
summary
refs
log
tree
commit
diff
log msg
author
committer
range
path:
root
/
nerv
/
lib
Commit message (
Collapse
)
Author
Age
*
add relu layer
Qi Liu
2016-04-29
|
*
support SeqBuffer decoding; fix bugs in finding params
alpha-4
Determinant
2016-04-18
|
*
Merge branch 'master' of github.com:liuq901/nerv into liuq901-master
Determinant
2016-03-15
|
\
|
*
add total time on profile
Qi Liu
2016-03-15
|
|
|
*
add mask on softmax_ce
Qi Liu
2016-03-15
|
|
*
|
ready for alpha-2 release
Determinant
2016-03-12
|
/
*
fix bug
Qi Liu
2016-03-11
|
*
update diagonlal lstm
Qi Liu
2016-03-11
|
\
|
*
update LSTM layer
Qi Liu
2016-02-24
|
|
*
|
add GPU auto selection
Determinant
2016-03-02
|
|
*
|
add options to main entry `nerv`; fix a bug in handling `cudaSetDevice` error
Determinant
2016-03-02
|
|
*
|
add network config for timit baseline
Determinant
2016-02-29
|
|
*
|
clean up code
Determinant
2016-02-24
|
|
*
|
add CuContext/MContext
wrapped-handles
Determinant
2016-02-24
|
/
*
...
tnn
Determinant
2016-02-17
|
*
try to merge manually
Determinant
2016-02-17
|
*
pass cflags to nvcc correctly; add omitted func decl
Determinant
2016-02-16
|
*
move curand_gen out of the matrix layout
Determinant
2016-02-15
|
*
try a basic merge
Determinant
2016-02-15
|
*
added copy_rows_fromd_by_colidx for clearer naming
txh18
2015-12-22
|
*
give a update_select_rows a more proper name
txh18
2015-12-22
|
*
bug fix: added check in the select_linear kernel
txh18
2015-12-11
|
*
merge conflict resolve
txh18
2015-12-10
|
\
|
*
fix the bug in freeing submatrix; add `offset` to Matrix structure
bugfix-matrixfree
Determinant
2015-12-10
|
|
*
|
...
txh18
2015-12-09
|
|
*
|
changed thres_mask function of matrix to a more standard api
txh18
2015-12-02
|
|
*
|
added rand_uniform and thres_mask for cumatrix
txh18
2015-12-01
|
|
*
|
bug fix: tanh implementation could cause nan
txh18
2015-11-29
|
|
*
|
added tanh operation for matrix
txh18
2015-11-25
|
/
*
add cflag __NERV_FUTURE_CUDA_7
Determinant
2015-11-23
|
*
small coding style change
txh18
2015-11-18
|
*
added atomicAdd for select_linear update, however, the result still seems ↵
txh18
2015-11-17
|
|
|
|
unreproducable, I changed select_linear layer update back to line-by-line
*
using atomicAdd for select_linear update
txh18
2015-11-17
|
*
changed copy_rows_fromd_by_idx a bit to make it clearer
txh18
2015-10-28
|
*
added update_select_rows for select_linear:update speed-up
txh18
2015-10-27
|
*
move sequence related layers to kaldi_seq
Yimmon Zhuang
2015-10-10
|
*
kaldi mpe training support
Yimmon Zhuang
2015-09-18
|
*
support gpu buffering
Determinant
2015-08-28
|
*
add profiling for copy_rows_fromh_by_idx
Determinant
2015-08-14
|
*
add embedding_example
Determinant
2015-08-04
|
*
added matrix:clip, TODO:testing
cloudygoose
2015-07-08
|
*
bug fix: Matrix space did not free
cloudygoose
2015-07-02
|
*
...
Determinant
2015-06-26
|
*
[a little big]matrix dim added, now use m:dim() to determine whether vector ↵
cloudygoose
2015-06-25
|
|
|
|
or matrix
*
use the standard nerv error reporting in io lib
Determinant
2015-06-25
|
*
separate non-Lua part of matrix code to a dedicated dir
Determinant
2015-06-24
|
*
separate non-Lua part of io code to a dedicated dir; code clean-up
Determinant
2015-06-23