commits  

Messages by Thread

• systemml git commit: [MINOR] update notice file for python assembly from 2017 to 2018 reinwald
• systemml git commit: [MINOR] update notice files for all assemblies to change year from 2017 to 2018 reinwald
• systemml git commit: [SYSTEMML-2199] Fix special cases IPA function inlining (unique hop IDs) mboehm7
• systemml git commit: [SYSTEMML-2187] Disable single-precision MKL conv2d_bias_add mboehm7
• systemml git commit: [SYSTEMML-2187] Fix robustness native mkl single/double conv2d ops mboehm7
• systemml git commit: [SYSTEMML-2185, 2186] Fix performance sparse sumByKeyStable w/ corr mboehm7
• systemml git commit: [HOTFIX] Fix literal replacement test (for new rewrite phase order) mboehm7
• [1/4] systemml git commit: [SYSTEMML-2185] Improved dag-split rewrite (avoid redundant cuts) mboehm7
  • [3/4] systemml git commit: [SYSTEMML-2185] Shuffle tuning spark binary matrix-matrix operations mboehm7
  • [2/4] systemml git commit: [SYSTEMML-2185] Phase ordering of statement block rewrites (merge/split) mboehm7
  • [4/4] systemml git commit: [SYSTEMML-2188] Fix unnecessary evictions to local fs on rdd collect mboehm7
• systemml git commit: [SYSTEMML-2184] Apply basic wdivmm conservatively (sparse/dense inputs) mboehm7
• [1/2] systemml git commit: [SYSTEMML-2178, 2183] Revert ANTLR version upgrade (back to 4.5.3) mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2182] Fine-tuning parfor optimizer (avoid spark ctx creation) mboehm7
• systemml git commit: [HOTFIX] Extended codegen pruning of invalid fusion plans mboehm7
• systemml git commit: [HOTFIX] Exclude new ANTLR-generated *.interp files from rat check mboehm7
• systemml git commit: [SYSTEMML-2179] Update janino version from 3.0.0 to 3.0.8 mboehm7
• systemml git commit: [SYSTEMML-2178] Update ANTLR version and include codgenalg tests mboehm7
• [1/3] systemml git commit: [SYSTEMML-2077] Extended IPA w/ conditional removal of unused functions mboehm7
  • [2/3] systemml git commit: [SYSTEMML-2077] Extended split-dag rewrite for new eval function mboehm7
  • [3/3] systemml git commit: [SYSTEMML-2077] New second-order eval builtin function mboehm7
• [1/2] systemml git commit: [SYSTEMML-1491] Add an ELU activation function. dusenberrymw
  • [2/2] systemml git commit: Improve stability of the ELU grad check test. dusenberrymw
• systemml git commit: [SYSTEMML-2130] Extended sparse block validation for MCSR format mboehm7
• [1/3] systemml git commit: [SYSTEMML-2159] Extended codegen tests for FA and FNR heuristics mboehm7
  • [2/3] systemml git commit: [SYSTEMML-2159] Extended codegen tests for FA and FNR heuristics mboehm7
  • [3/3] systemml git commit: [SYSTEMML-2180] Fix codegen row opening condition for right indexing mboehm7
• systemml git commit: [SYSTEMML-2177] Performance spark cpmm (set join parallelism) mboehm7
• systemml git commit: [SYSTEMML-2176] Performance spark reshape ultra-sparse matrices mboehm7
• systemml git commit: [SYSTEMML-2174] Performance spark cpmm over ultra-sparse inputs mboehm7
• [1/2] systemml git commit: [SYSTEMML-2171] Improved caching of ultra-sparse matrices mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2173] Propagation of upper bound nnz over spark instructions mboehm7
• systemml git commit: [SYSTEMML-658] Test for JMLC scalar string value types deron
• systemml git commit: [SYSTEMML-2170] Fix parfor optimizer exec type selection (valid dims) mboehm7
• systemml git commit: [SYSTEMML-2130] Fix pack correct sparse/dense format and nnzs, II mboehm7
• [1/3] systemml git commit: [MINOR] Cleanup unnecessary indirections in dml/pydml parsers mboehm7
  • [3/3] systemml git commit: [SYSTEMML-1444] Fix function call hoisting in nested control flow mboehm7
  • [2/3] systemml git commit: [SYSTEMML-1444] Simplify nn-lstm layer with UDFs in expressions mboehm7
• systemml git commit: [SYSTEMML-1444] Additional tests for external UDFs in expressions mboehm7
• systemml git commit: [HOTFIX] Update function call tests wrt UDF support in expressions mboehm7
• systemml git commit: [SYSTEMML-1444, 1759, 1785] Support for UDFs in expressions mboehm7
• systemml git commit: [SYSTEMML-2130] Extended sparse block validation for COO format mboehm7
• systemml git commit: [SYSTEMML-2130] Fix pack correct sparse/dense representations and nnzs mboehm7
• systemml git commit: [SYSTEMML-2164] Improved serialization of compressed matrix blocks mboehm7
• systemml git commit: [SYSTEMML-2166] Fix scrambled print order after function inlining mboehm7
• systemml git commit: [SYSTEMML-445] Cleanup GPU memory management niketanpansare
• systemml git commit: [SYSTEMML-2163] Performance large partitioned spark broadcasts >2GB mboehm7
• systemml git commit: [SYSTEMML-2162] Fix deserialization of compressed blocks w/ shared dict mboehm7
• systemml git commit: [SYSTEMML-445] Allow users to pass the file paths to the binary blocked, csv and ijv datasets to mllearn classes. niketanpansare
• systemml git commit: [SYSTEMML-1872] Added average pooling and upsampling layers niketanpansare
• [1/2] systemml git commit: [SYSTEMML-1872] Added average pooling and upsampling layers niketanpansare
  • [2/2] systemml git commit: [SYSTEMML-1872] Added average pooling and upsampling layers niketanpansare
• [1/3] systemml git commit: [SYSTEMML-819] Fix missing matrix-matrix mult over compressed matrices mboehm7
  • [3/3] systemml git commit: [MINOR] Refactoring matrix block slice and append operations mboehm7
  • [2/3] systemml git commit: [MINOR] Refactoring matrix block slice and append operations mboehm7
• systemml git commit: [SYSTEMML-2161] Fix codegen optimizer cost vector memoization mboehm7
• [1/2] systemml git commit: [SYTEMML-2160] Fix codegen optimizer sparsity-aware cost-based pruning mboehm7
  • [2/2] systemml git commit: [MINOR] Improved analysis for sparse-safe codegen cell operations mboehm7
• [1/2] systemml git commit: [SYSTEMML-2157] Fix codegen optimizer (suboptimal plans after row2cell) mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2158] Fix size propagation nary rbind (incorrect rows/cols) mboehm7
• systemml git commit: [HOTFIX] Fix minor java doc issue (breaking distribution build) mboehm7
• systemml git commit: [MINOR] Extended UDF framework (access to execution context) mboehm7
• [1/3] systemml git commit: [SYSTEMML-2154] Fix misleading explain recompile hops output mboehm7
  • [2/3] systemml git commit: [SYSTEMML-2155] Recompute nnz on spark checkpoints for large matrices mboehm7
  • [3/3] systemml git commit: [SYSTEMML-2157] Fix parfor optimizer side effect on codegen plans mboehm7
• systemml git commit: [SYSTEMML-2130] Sparse block API extension for validity checks mboehm7
• systemml git commit: [MINOR] Fine-tuning codegen row ops (load balance, nnz recompute) mboehm7
• systemml git commit: [MINOR] Extended JMLC API (pass global dml config before prepareScript) mboehm7
• systemml git commit: [SYSTEMML-2153] Fix robustness parfor check for block partitioning mboehm7
• systemml git commit: [SYSTEMML-2152] Fix spark mapmm selection (missing dims constraint) mboehm7
• [1/4] systemml git commit: [MINOR] Additional debug info on matrix compression (column classify) mboehm7
  • [3/4] systemml git commit: [SYSTEMML-2150] Fix codegen optimizer cost model (incorrect input size) mboehm7
  • [4/4] systemml git commit: [SYSTEMML-2151] Fix correctness of codegen row powAdd for sparse mboehm7
  • [2/4] systemml git commit: [SYSTEMML-2149] New simplification rewrite for replace zero w/ scalar mboehm7
• systemml git commit: [SYSTEMML-2068] Fix unnecessary meta data in sparse block csr mboehm7
• systemml git commit: [SYSTEMML-2148] Performance codegen cell ops over compressed matrices mboehm7
• systemml git commit: [SYSTEMML-2110] New tests for codegen relu backward operations mboehm7
• svn commit: r1824275 - in /systemml/site: ./ release-notes/ deron
• systemml-website git commit: [MINOR] Update footer copyright to 2018 deron
• systemml-website git commit: [MINOR] Update website NOTICE copyright to 2018 deron
• [1/2] systemml git commit: [SYSTEMML-2135] Extended removeEmpty with outputs of zero rows/cols deron
  • [2/2] systemml git commit: [MINOR] Update doc version to match pom.xml version deron
• systemml git commit: [MINOR] Update doc version to match pom.xml version deron
• systemml git commit: [MINOR] Update NOTICE copyright to 2018 deron
• systemml git commit: [MINOR] Improved handling of commutative binary codegen operations mboehm7
• systemml git commit: [MINOR] Fix codegen cost model (missing ifelse, warn on stats overflow) mboehm7
• systemml git commit: [SYSTEMML-2147] Performance read of sparse short-wide binary matrices mboehm7
• systemml git commit: [SYSTEMML-2135] I/O operation support for matrices with zero rows/cols mboehm7
• systemml git commit: [MINOR] Cleanup and simplification of l2svm algorithm script mboehm7
• systemml git commit: [MINOR] Extended JMLC API (obtain current config of prepared scripts) mboehm7
• systemml git commit: [HOTFIX][SYSTEMML-2068] Fix corrupted merge of cell bitwAnd operations mboehm7
• systemml git commit: [SYSTEMML-2068] Codegen support for bitwise AND operations mboehm7
• systemml git commit: [MINOR] Fix robustness of mr aggregation code path (valid dims) mboehm7
• systemml git commit: [SYSTEMML-2067] Refactoring of im2col operations into functional API mboehm7
• [1/4] systemml git commit: [HOTFIX] Fix initialization of compressed matrix blocks (dim check) mboehm7
  • [3/4] systemml git commit: [SYSTEMML-2145] Fix codegen row tmpl compiler/runtime for kmeans mboehm7
  • [2/4] systemml git commit: [SYSTEMML-2144] Disable codegen for config w/ forced gpu operations mboehm7
  • [4/4] systemml git commit: [SYSTEMML-2143] Fix missing sprop/sigmoid codegen support in row templ mboehm7
• [1/2] systemml git commit: [SYSTEMML-2041] Performance/memory-efficiency ultra-sparse conv2d mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2142] Performance sparse im2col (constraint handling) mboehm7
• systemml git commit: [SYSTEMML-2140] Performance spark matrix-vector reshape operations mboehm7
• systemml git commit: [MINOR] Fix long to int truncation of dims in spark reshape operations mboehm7
• [1/3] systemml git commit: [SYSTEMML-2135] Extended removeEmpty with outputs of zero rows/cols mboehm7
  • [2/3] systemml git commit: [SYSTEMML-2139] Fix codegen with boolean literal inputs (e.g., ifelse) mboehm7
  • [3/3] systemml git commit: [SYSTEMML-2135] Fix aggregate correctness over matrices w/ zero dims mboehm7
• systemml git commit: [SYSTEMML-445] Bugfix in Caffe2DML/Keras2DML's concat layer for sentence CNN niketanpansare
• [1/2] systemml git commit: [SYSTEMML-2135] Compiler/runtime support for matrices w/ zero rows/cols mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2135] Compiler/runtime support for matrices w/ zero rows/cols mboehm7
• systemml git commit: [SYSTEMML-2136] Fix int overflow in spark/mr text to binary conversion mboehm7
• [1/2] systemml git commit: [SYSTEMML-2133] Performance cp maxpooling operations (load balance) mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2082] Codegen support for ternary ifelse in cell/magg tmpls mboehm7
• [1/2] systemml git commit: [SYSTEMML-2132] Fix intermediate memory estimate matrix multiply hops mboehm7
  • [2/2] systemml git commit: [MINOR] Cleanup codegen candidate exploration (preserve all candidates) mboehm7
• systemml git commit: [MINOR] Fix naming sparse-sparse relu_backward mboehm7
• [1/2] systemml git commit: [SYSTEMML-2120] New plan cache for cost-based codegen optimizer mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2068] Fix sparse-dense vector xor operations mboehm7
• systemml git commit: [SYSTEMML-2068] Codegen support for xor operations mboehm7
• systemml git commit: [SYSTEMML-2129] Performance sparse relu_backward operations mboehm7
• systemml git commit: [SYSTEMML-2108] Performance CP ternary +* and -* operations mboehm7
• systemml git commit: [SYSTEMML-2106] New single-precision native matrix multiply mboehm7
• systemml git commit: [SYSTEMML-2105] Support single-precision conv2d with MKL mboehm7
• systemml git commit: [SYSTEMML-2105] New single-precision native conv2d/conv2d_bias_add ops mboehm7
• systemml git commit: [SYSTEMML-445] Added load_keras_weights flag in Keras2DML to avoid transfering randomly initialized weights niketanpansare
• systemml git commit: [SYSTEMML-445] Disable ternary aggregate rewrite on GPU backend niketanpansare
• systemml git commit: [SYSTEMML-445] Guard JCudaKernels with sysml.gpu.sync.postProcess flag niketanpansare
• systemml git commit: [SYSTEMML-445] Update algorithm selection logic on GPU for conv2d_backward_data niketanpansare
• [1/2] systemml git commit: [SYSTEMML-1437] Factorization Machines: Core Layer Module dusenberrymw
  • [2/2] systemml git commit: Minor cleanup and formatting of the new Factorization Machines code dusenberrymw
• [1/2] systemml git commit: [SYSTEMML-2100] Extended JMLC API (configuration handling) mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2101] Fix JMLC robustness for missing fs impl classes mboehm7
• systemml git commit: [HOTFIX] Fix sparse binary inplace operations (empty row handling) mboehm7
• systemml git commit: [SYSTEMML-445] Support newer as well as older Keras API niketanpansare
• systemml git commit: [MINOR] Data preparation script for recoding user/product ratings mboehm7
• systemml git commit: [SYSTEMML-445] Display time taken for eviction as well as two-staged fragmentation in the statistics. niketanpansare
• systemml git commit: [SYSTEMML-445] Added two-step strategy to deal with potential fragmentation on GPU niketanpansare
• systemml git commit: [SYSTEMML-2004] Covariance Kernels niketanpansare
• systemml git commit: [SYSTEMML-540] Improved the script generated by Caffe2DML in allreduce setting niketanpansare
• [1/2] systemml git commit: [SYSTEMML-2098] Fix robustness CSR resizing w/ initial size zero mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2099] Performance sparse result merge accum w/o compare mboehm7
• systemml git commit: [HOTFIX][SYSTEMML-2095] Fix runtime integration parfor result variables mboehm7
• systemml git commit: [MINOR] Performance binary inplace ops for result merge, cleanups mboehm7
• [1/3] systemml git commit: [SYSTEMML-2095] Extended local parfor result merge w/ accumulators mboehm7
  • [2/3] systemml git commit: [SYSTEMML-2096] Extended spark/mr parfor result merge w/ accumulators mboehm7
  • [3/3] systemml git commit: [SYSTEMML-2097] Minor simplifications of nn library w/ += and ifelse mboehm7
• [1/2] systemml git commit: [SYSTEMML-2093] Parser and language support for new += operator mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2094] Extended parfor dependency analysis (accumulators) mboehm7
• systemml git commit: [SYSTEMML-2091] Performance sparse-sparse reshape (shallow for CSR) mboehm7
• [1/2] systemml git commit: [MINOR] Cleanups of unchecked generics and improved list concatenation mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2081] Support for nary cbind in codegen row operations mboehm7
• systemml git commit: [SYSTEMML-2080] Fix correctness sparse relu-backward CP operations mboehm7
• systemml git commit: [SYSTEMML-1931] New tests for matrix-scalar logical operations mboehm7
• systemml git commit: [SYSTEMML-2075] Fix invalid constant folding of matrix-scalar and/or mboehm7
• systemml git commit: [SYSTEMML-540] Low-rank fully connected layer niketanpansare
• systemml git commit: [SYSTEMML-540] Bugfix for fused ReLU-maxpooling and ReLU-maxpooling backward operators niketanpansare
• systemml git commit: [SYSTEMML-1992] Implemenation of Mode finding for Gaussian Process niketanpansare
• [1/2] systemml git commit: [MINOR] Fix license header formatting lresende
  • [2/2] systemml git commit: [MINOR] Update rat-plugin to release 0.12 lresende
• systemml git commit: [SYSTEMML-2070] New codegen algorithm tests (linreg/logreg datagen) mboehm7
• [1/2] systemml git commit: [SYSTEMML-2074] Fix codegen row operations over large dense blocks mboehm7
  • [2/2] systemml git commit: [MINOR] Fix flaky tests (larger eps GLM[0], disabled jmlc input steam) mboehm7
• systemml git commit: [SYSTEMML-540] Added support for GPU relu, scalar min and max operations niketanpansare
• systemml git commit: [SYSTEMML-540] Added an assert builtin function niketanpansare
  • systemml git commit: [SYSTEMML-540] Added an assert builtin function niketanpansare
• systemml git commit: [SYSTEMML-540] Bugfix in GPU axpy instruction due to recent ternary changes niketanpansare
• systemml git commit: [MINOR] Limit the size of intermediate im2col matrix to MAX_WORKSPACE_LIMIT_BYTES niketanpansare
• [1/3] systemml git commit: [SYSTEMML-540] Add support for sparse filter dense image conv2d niketanpansare
  • [2/3] systemml git commit: [SYSTEMML-540] Add support for sparse filter dense image conv2d niketanpansare
  • [3/3] systemml git commit: [SYSTEMML-540] Add support for sparse filter dense image conv2d niketanpansare
• [1/2] systemml git commit: [SYSTEMML-2050] Performance ifelse w/ constant expression (shallow copy) mboehm7
  • [2/2] systemml git commit: [SYSTEMML-2050] New rewrites for removing unnecessary ifelse operations mboehm7
• systemml git commit: [SYSTEMML-2050] Minor simplifications of GLM, Kmeans, Linreg, MLogreg mboehm7
• systemml git commit: [MINOR] Cleanup jmlc build lite jar tests, incl remove temp files mboehm7
• [1/5] systemml git commit: [MINOR] Refactoring agg-binary, ternary, and quaternary instructions mboehm7
  • [3/5] systemml git commit: [MINOR] Refactoring agg-binary, ternary, and quaternary instructions mboehm7
  • [5/5] systemml git commit: [SYSTEMML-2055] New ifelse builtin function, ternary inst framework mboehm7
  • [2/5] systemml git commit: [MINOR] Refactoring agg-binary, ternary, and quaternary instructions mboehm7
  • [4/5] systemml git commit: [SYSTEMML-2055] New ifelse builtin function, ternary inst framework mboehm7
• systemml git commit: [HOTFIX][SYSTEMML-1570] Fix rewrite for relu (selp) expressions mboehm7
• [1/2] systemml git commit: [SYSTEMML-1570] Removed unnecessary sel+ operator, improved max/min mboehm7
  • [2/2] systemml git commit: [SYSTEMML-1175] Fix robustness toString for scalars, tests mboehm7
• [1/3] systemml git commit: [SYSTEMML-2046] Large dense blocks in rotate180, relu, incl cleanups mboehm7
  • [2/3] systemml git commit: [SYSTEMML-2073] Fix lost update on lock-free statistics maintenance mboehm7
  • [3/3] systemml git commit: [MINOR] Performance nrow/ncol runtime instructions (parsing overhead) mboehm7
• systemml git commit: [HOTFIX][SYSTEMML-2070] Fix compressed LinregCG (R script location) mboehm7
• systemml git commit: [MINOR] Improved recompilation of size expressions and checkpoints mboehm7
• [1/6] systemml git commit: [SYSTEMML-2070] Refactoring codegen tests, incl new stepwise tests mboehm7
  • [3/6] systemml git commit: [SYSTEMML-2070] Refactoring codegen tests, incl new stepwise tests mboehm7
  • [6/6] systemml git commit: [SYSTEMML-2072] Fix codegen integration w/ parfor inplace recompile mboehm7
  • [4/6] systemml git commit: [SYSTEMML-2070] Refactoring codegen tests, incl new stepwise tests mboehm7
  • [2/6] systemml git commit: [SYSTEMML-2070] Refactoring codegen tests, incl new stepwise tests mboehm7
  • [5/6] systemml git commit: [SYSTEMML-2071] Fix stepwise linregDS script (incorrect indexing range) mboehm7
• systemml git commit: [HOTFIX] Fix use of unmodifiable lists in compression planning mboehm7
• [1/2] systemml git commit: [SYSTEMML-2069] Lock-free maintenance of heavy hitter statistics mboehm7
  • [2/2] systemml git commit: [MINOR] Reduced recompilation overhead and various cleanups mboehm7
• systemml git commit: [SYSTEMML-2057] New builtin functions for bitwise binary operators mboehm7
• [1/2] systemml git commit: [SYSTEMML-540] Add support for RNN and LSTM in Caffe2DML and Keras2DML niketanpansare
  • [2/2] systemml git commit: [SYSTEMML-540] Added optimizer support in Keras2DML niketanpansare
• systemml git commit: [SYSTEMML-540] Added optimizer support in Keras2DML niketanpansare
• [1/6] systemml git commit: [MINOR] Cleanup tensor index computation in convolution operations mboehm7
  • [6/6] systemml git commit: [MINOR] Refactoring and cleanup cp convolution operations mboehm7

• Earlier messages
• Later messages