Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/marian-nmt/marian.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
AgeCommit message (Expand)Author
2018-08-04avoid cudaMalloc in CopyRowsWenyong Huang
2018-08-01Fix compiler warningsRoman Grundkiewicz
2018-07-31Autoformat files, part 4Roman Grundkiewicz
2018-07-27fix mergeMarcin Junczys-Dowmunt
2018-07-02Fix #262 PoolingWrapper dependencyRoman Grundkiewicz
2018-06-29another attempt at fixing gradients for transpose etcMarcin Junczys-Dowmunt
2018-06-29resolve merge conflictMarcin Junczys-Dowmunt
2018-06-28make cols operator non-constantMarcin Junczys-Dowmunt
2018-06-28add proper gradient summation to shift operatorMarcin Junczys-Dowmunt
2018-06-27fix transpose operatorMarcin Junczys-Dowmunt
2018-06-22further small refactoring in transformer; renamed layer_norm to layerNormFrank Seide
2018-06-22renamed logit() to sigmoid(), logit() is a different functionFrank Seide
2018-06-21bug fix: History::Add() should obey the actual EOS symbol from the given voca...Frank Seide
2018-06-08Merge branch 'master' into jonathac/windows_buildMarcin Junczys-Dowmunt
2018-05-25Pile of truly horrible hacks to *almost* get things building on windowsJon Clark
2018-05-12add gradient to clipping functionMarcin Junczys-Dowmunt
2018-05-12clipping gemmMarcin Junczys-Dowmunt
2018-04-12working memoizationMarcin Junczys-Dowmunt
2018-04-11add int16 operators, attempt at memoizationMarcin Junczys-Dowmunt
2018-03-13Clean includesRoman Grundkiewicz
2018-03-12Autoformat filesRoman Grundkiewicz
2018-03-03more coda separationMarcin Junczys-Dowmunt
2018-03-01simplified interface, get slowly rid of keywordsMarcin Junczys-Dowmunt
2018-02-25add cudnn backMarcin Junczys-Dowmunt
2018-02-22prototype cpu versionMarcin Junczys-Dowmunt
2018-02-16pass through backendMarcin Junczys-Dowmunt
2018-02-16more clean up in backendMarcin Junczys-Dowmunt
2017-11-23Merge branch 'master' of https://github.com/marian-nmt/marian-dev into charS2STomasz Dwojak
2017-11-21working batched translationMarcin Junczys-Dowmunt
2017-11-16Refactor convolution and and poolingWithMaskingTomasz Dwojak
2017-11-13Move pooling code to cudnn filesTomasz Dwojak
2017-11-03cleaned up functional templatesMarcin Junczys-Dowmunt
2017-11-02farewell thrustMarcin Junczys-Dowmunt
2017-11-02new functional templatesMarcin Junczys-Dowmunt
2017-11-02commented out offending code make it compile, needs to be fixed when USE_CUDN...Marcin Junczys-Dowmunt
2017-11-02merge with developMarcin Junczys-Dowmunt
2017-11-02clean up permutationMarcin Junczys-Dowmunt
2017-11-01more gpu-specific TMPMarcin Junczys-Dowmunt
2017-10-31exception on rowsMarcin Junczys-Dowmunt
2017-10-29adjust rnn to work with new shapeMarcin Junczys-Dowmunt
2017-10-29Use PReLU in LeakyReLURoman Grundkiewicz
2017-10-29Add PReLURoman Grundkiewicz
2017-10-29Add LeakyReLURoman Grundkiewicz
2017-10-29Add comments to ReLU i SwishRoman Grundkiewicz
2017-10-29Add a comment for SwishNodeOpRoman Grundkiewicz
2017-10-29replace TimeStepNode with more general StepNodeMarcin Junczys-Dowmunt
2017-10-29merge rnn_test.cppMarcin Junczys-Dowmunt
2017-10-28AutoformatRoman Grundkiewicz
2017-10-28Replace UTIL_THROWNs with ABORTsRoman Grundkiewicz
2017-10-27fixed bug in transpose for ndarrayMarcin Junczys-Dowmunt