Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
K
kaldi-commonvoice
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Container Registry
Model registry
Operate
Environments
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Simon Will
kaldi-commonvoice
Graph
2d6bd05bcc82d2eafa0f7f90db516d551f2630ee
Select Git revision
Branches
1
master
default
protected
1 result
You can move around the graph by using the arrow keys.
Begin with the selected commit
Created with Raphaël 2.2.0
11
Sep
10
9
8
7
6
5
4
3
2
1
31
Aug
30
29
28
27
26
24
23
22
21
20
19
17
16
15
14
13
11
10
9
8
6
5
4
2
1
31
Jul
30
29
27
26
25
24
23
22
21
20
19
18
17
16
14
12
11
10
9
7
6
5
4
3
2
1
30
Jun
29
28
27
26
25
24
21
20
19
18
17
16
15
14
13
7
6
5
4
2
30
May
29
28
25
24
23
22
21
20
19
18
17
16
15
14
11
10
9
8
6
30
Apr
29
28
27
25
24
23
22
21
19
18
17
16
14
13
12
11
10
9
8
7
6
5
4
3
2
1
31
Mar
30
29
28
27
26
25
24
23
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
5
4
3
2
sandbox/dan2: wrote comments for Splice and Copy in cu-math.h, still need to finish unit tests for those functions
sandbox/dan2: Fixed a bug in AddMatMat function in cu-matrix.cc, which doesn't check 0-dimensionality for CPU case
Adding description for CompObjfAndDeriv
sandbox/dan2: CUDA-lize calculation in NnetUpdater::ComputeObjfAndDeriv() function. Ugly pointers are replaced by cleaner codes
trunk/nnet : fixing memory leak in destructor of CuRand (big thanks goes to Dan)
(trunk/egs/wsj) Fix the train_sgmm2_group.sh to correctly handle the ci_phones options
trunk:egs/hkust Update with GPU-based DNN experiments
trunk:egs/hkust Update with GPU-based DNN experiments and results
sandbox/dan2: CUDA-lize calculation in NnetUpdater::ComputeObjfAndDeriv() function. Documentation and Tests added. There are a few lines of REALLY ugly codes (lots of pointers) to be changed
sandbox/dan2: Rework CuBlockMatrix to make initialization more efficient; add memory checking code; fix memory leaks in packed-matrix copy; compilation fix in nnet2/nnet-update.cc
sandbox/dan2: CUDA-lize calculation in NnetUpdater::ComputeObjfAndDeriv() function. Documentation and Tests coming soon...
sandbox/dan2: Committing initial version of CuBlockMatrix code (for use in MixtureProbComponent)
(BABEL): follow the release of the IndusDB for the new langs (102 and 103), fixes, include example configs with FFV features
trunk/nnetbin : Introducing length constrain of 1 min on segments that are used for mini-batch frame cross-entropy DNN training. This will avoid possible out-of-memory issues by discarding the very long segments which are infrequent.
sandbon/dan2: Added documentation to Randomize in cu-math.h
fix a bug - the last argument was checked by ClassifyRSpecifier, but it should be the first argument, since the last is a W specifier
fix to babel/s5/local/make_ffv.sh
sandbox/dan2: rework of CuStlVector.
sandbox/dan2: cudamatrix bug-fixes and more tests; modifying scripts to use single-threaded program for GPU (multi-threaded code is not easily compatible with CUDA).
sandbox/dan2: add test for cuda-matrix copy across types; fix bugs in cross-type cuda-matrix copying, and streamline that code.
sandbox/dan2: minor improvements to the code; bug fixes in matrix copying RE zero dim
sandbox/dan2: add speed test for Sum() of CuVector
trunk: small fix to nnet training script (remove call to deprecated script)
sandbox/dan2: various cudamatrix optimizations and fixes. neural net training job now running with cudamatrix (training script not yet fully tested).
sandbox/dan2: fixes to tests for random-number generators, cosmetic changes; mov cu-rand-inl.h to cu-rand.cc.
sandbox/dan2: Adding SetRandUniform() function for matrix; adding tests for SetRandn() and SetRandUniform(), cudamatrix SetRandUniform() test failing currently.
sandbox/dan2: Added efficient CUDA implementation for one case of CopyRowsFromVec.
trunk: code change that should fix memory leak in bin/latgen-faster-mapped-parallel [fix other branch of the code, omitted first time]
trunk: code change that should fix memory leak in bin/latgen-faster-mapped-parallel
sandbox/dan2: Fixing more errors; nnet-component-test now passes.
sandbox/dan2: various fixes.
sandbox/dan2: Fixes and extension to CopyRows()/CopyCols() for matrix/cudamatrix
sandbox/dan2: some fairly minor changes.
sandbox/dan2: fixing the bug where src.dim() is passed into cuda kernel instead of this->dim()
sandbox/dan2: cudamatrix bug fix; improve test programs to test with and without GPU (would have caught the bug I just fixed).
sandbox/dan2: Various cudamatrix extensions. nnet-component test code currently failing.
sandbox/dan2:add cu-vector-speed-test.cc
opt softmax in cu-vector
trunk/nnet : alignment script can optionally generate output in lattice format (handy for getting word alignments)
sandbox/dan2: fix the previous equal-stride constraints. Not completed yet, might be bugs
Loading