commits  

Messages by Thread

• • [madlib] 02/02: Use uname to determine architecture okislal
  • [madlib] 01/02: Revert "Use uname to determine architecture" okislal
• [madlib] branch master updated: Use uname to determine architecture okislal
• [madlib] branch master updated: indicate optional param in elastic net fmcquillan
• [madlib] branch master updated: Check -mno-sse2 is available for GNU CXX okislal
• [madlib] branch master updated: indicate optional params in DR and RF fmcquillan
• [madlib] branch master updated: DL: Avoid constant folding of weights in GPDB6 plan khannaekta
• [madlib] branch master updated: Kmeans: Convert python inf and nan values to postgres okislal
• svn commit: r37665 - /dev/madlib/KEYS okislal
• svn commit: r37664 - /release/madlib/KEYS okislal
• [madlib] branch master updated (515dc25 -> 5ddec47) domino
  • [madlib] 01/02: DL: Fix metrics_elapsed_time for fit multi model domino
  • [madlib] 02/02: DL: Drop leftover tables domino
• [madlib] branch master updated: misc user doc clarifications fmcquillan
• [madlib] branch master updated (7625ae0 -> 273301e) fmcquillan
  • [madlib] 02/02: Update Apache Copyright date fmcquillan
  • [madlib] 01/02: Decrease the learning rate for transfer learning test fmcquillan
• [madlib] branch master updated: DL: Fix failure on GPDB6 for preprocessor khannaekta
• [madlib] branch master updated: DL: Improve performance of mini-batch preprocessor (#467) domino
• [madlib] branch master updated: Build: Add support for pg 12 khannaekta
• [madlib] branch master updated: DL: Update tests calling trap_error khannaekta
• [madlib] branch master updated: clarify warm start with model selection in user docs fmcquillan
• [madlib] branch master updated: DL: Modify multi-fit warm start to accept non-matching mst&model tables okislal
• [madlib] branch master updated: DL: Fix transfer learning dev-check test khannaekta
• [madlib] branch master updated: clarify warm start vs transfer learning in user docs fmcquillan
• [madlib] branch master updated: correct fit function call in user docs for multi fit fmcquillan
• [madlib] branch master updated: misc user doc updates for 1dot17 fmcquillan
• [madlib] branch master updated (9d2558f -> ca17da4) okislal
  • [madlib] 02/02: DL: Update user docs for multi-model okislal
  • [madlib] 01/02: DL: Add asymmetric cluster check for predict okislal
• [madlib] branch master updated: DL: Add asymmetric cluster support for fit and evaluate okislal
• [madlib] branch master updated: DL: Add optional parameters for multi model training okislal
• [madlib] branch master updated: DL: Add support for asymmetric segment distribution to preprocessor okislal
• [madlib-site] branch automl updated: hyperband diagonal E2E update fmcquillan
• [madlib-site] branch automl updated: hyperband diagonal E2E still in work... fmcquillan
• [madlib] branch master updated: LDA: Turn off create table notices nkak
• [madlib-site] branch automl created (now 0c8e677) fmcquillan
• [madlib] branch master updated: MADLIB-1351 : Added stopping criteria on perplexity to LDA khannaekta
• [madlib] branch master updated: DL: Add new helper function for gpu_configuration (#455) nkak
• [madlib] branch master updated: DL: Make model_id naming consistent khannaekta
• [madlib] branch master updated: DL: Enable transfer learning for multi model training okislal
• [madlib] branch master updated (22fabc8 -> ef2f0e9) khannaekta
• [madlib] branch master updated (1e7aba3 -> 22fabc8) khannaekta
• [madlib] branch master updated: Add pull request template khannaekta
• [madlib] branch master updated (35e959d -> 72dfd30) fmcquillan
  • [madlib] 01/02: Add keras version to the docs and release notes fmcquillan
  • [madlib] 02/02: Address review comments fmcquillan
• [madlib] branch master updated: DL: Remove quote_ident to allow tables on schemas okislal
• [madlib] branch master updated (610cf6d -> 913b66e) nkak
• [madlib] branch master updated: DL: Add training for multiple models okislal
• [madlib] branch master updated: Pivot: Use output_table instead of out_table okislal
• [madlib] branch master updated: Kmeans: Use squared dist norm2 as a default for silh functions nkak
• [madlib] branch master updated: SVM: Lower bound the default for n_components fmcquillan
• [madlib] branch master updated: updated DL preprocessor docs for bytea (#445) fmcquillan
• [madlib] branch master updated: DL: Update jenkins to install tensorflow 1.14. nkak
• [madlib] branch master updated (dae72a0 -> c416cd7) nkak
• [madlib] branch master updated (72df650 -> dae72a0) okislal
  • [madlib] 01/02: Kmeans: Add simple silhouette score for every point okislal
  • [madlib] 02/02: update user docs for auto-k and per-point silh and generally reorganize okislal
• [madlib] branch master updated: Use -p while creating tmp dir for jenkins build. nkak
• [madlib] branch master updated: Kmeans: Add automatic optimal cluster estimation okislal
• [madlib] branch master updated: DL: Update training_preprocessor_dl to use bytea okislal
• [madlib] branch master updated: DL: Add support in preprocessor to evenly distribute data for GPDB khannaekta
• [madlib] branch master updated (19ea7f7 -> 5bc8ea1) khannaekta
• [madlib] branch master updated (a4f8ae1 -> 19ea7f7) khannaekta
• [madlib] branch master updated (f4e2ab6 -> a4f8ae1) nkak
• [madlib] branch load_mst_user_docs deleted (was 2059ddb) nkak
• [madlib] branch load_mst_user_docs created (now 2059ddb) fmcquillan
  • [madlib] 01/01: user docs for setting up model selection table fmcquillan
• [madlib] branch master updated (33988ce -> f4e2ab6) domino
• [madlib] branch master updated (e0c2bad -> 33988ce) domino
• [madlib] branch master updated: Add generate mst table utility. This commit adds a new function to generate the model selection table. It generates such table as the naive combinations of the three input lists of parameter choices. This utility also validates and de-duplicates the inputs. domino
• [madlib] branch master updated (205bdba -> 264a4d8) nkak
• [madlib-site] branch asf-site updated: updated 3 notebooks with minor changes fmcquillan
• [madlib-site] branch asf-site updated (a283758 -> 3b665af) fmcquillan
  • [madlib-site] 01/02: Disable --no-temp-files|-m option, since it doesn't work fmcquillan
  • [madlib-site] 02/02: Update demo notebook fmcquillan
• [madlib-site] branch asf-site updated: update Load-model-architecture-v1.ipynb with faster way to load model weights - minor fix fmcquillan
• [madlib-site] branch asf-site updated: update Load-model-architecture-v1.ipynb with faster way to load model weights fmcquillan
• [madlib] branch master updated: updated user docs for madlib-keras with BYOM inference okislal
• [madlib] branch keras_byom deleted (was 49cf31f) okislal
• [madlib] branch keras_byom created (now 49cf31f) fmcquillan
  • [madlib] 01/01: updated user docs for madlib-keras with BYOM inference fmcquillan
• [madlib] branch master updated: KNN: Add distances to the output table okislal
• [madlib] branch delme1 deleted (was 96885c8) njayaram
• [madlib] branch delme1 created (now 96885c8) njayaram
• [madlib] branch delme deleted (was 96885c8) njayaram
• [madlib] branch delme created (now 96885c8) njayaram
• [madlib] branch master updated: DL: Split madlib_keras devcheck file khannaekta
• [madlib] branch master updated: Quick fix for bison issue domino
• [madlib] branch master updated: DL: Add function for predict byom nkak
• [madlib] branch master updated (640392d -> ed202b1) njayaram
  • [madlib] 02/02: Assoc Rules: Minor updates to user docs for new params njayaram
  • [madlib] 01/02: Association Rules: Improve performance njayaram
• [madlib-site] branch asf-site updated: update jupyter notebooks for new image loader script fmcquillan
• [madlib-site] branch asf-site updated: MADlib image loader: load dataset from disk okislal
• [madlib-site] branch asf-site updated: Add .gitignore nkak
• [madlib] branch master updated: Correlation: Process deconstruction in chunks for grouping okislal
• [madlib-site] branch asf-site updated: add links to deep learning notes and Jupyter notebooks fmcquillan
• [madlib-site] branch asf-site updated: Rebuilt design doc to include 1.16 changes okislal
• [madlib] branch master updated: Add input params to last part of WCC impl okislal
• [madlib] branch latest_release updated (fa02339 -> 13aa1cb) okislal
• [madlib] branch master updated: Update version from 1.16 to 1.17-dev njayaram
• svn commit: r34821 - /release/madlib/1.15.1/ njayaram
• [madlib] tag untagged-f712ca5b385bc9915dc1 created (now e2af003) njayaram
• [madlib] annotated tag archive/feature/svm_minibatch_1 created (now 986e211) njayaram
• [madlib] annotated tag rel/v1.16 created (now a5c193c) njayaram
• [madlib] annotated tag archive/bugfix/dt_null_rows created (now bb2a608) njayaram
• [madlib-site] branch asf-site updated: update website for 1.16 release fmcquillan
• [madlib-site] branch asf-site updated: Move v1.16 RC2 to latest released njayaram
• svn commit: r34800 - /dev/madlib/1.16.RC1/ njayaram
• svn commit: r34799 - /dev/madlib/1.16.RC2/ /release/madlib/1.16/ njayaram
• svn commit: r34756 - /dev/madlib/1.16.RC2/ okislal
• svn commit: r34718 - /dev/madlib/1.16.RC2/ okislal
• [madlib] annotated tag rc/1.16-rc2 created (now 0d82c92) okislal
• svn commit: r34669 - in /dev/madlib: 1.16.RC1/ 1.16/ okislal
• svn commit: r34668 - /release/madlib/KEYS okislal
• svn commit: r34667 - /dev/madlib/1.16/ okislal
• [madlib] annotated tag rc/1.16-rc1 created (now bf430f2) okislal
• svn commit: r34649 - /dev/madlib/KEYS okislal
• [madlib] branch master updated: Add changelist files & update version okislal
• [madlib] branch master updated: Add release notes for 1.16 okislal
• [madlib] branch master updated: DL: Add preprocessor specific error message njayaram
• [madlib-site] branch asf-site updated: Image loader python module, and demo notebook okislal
• [madlib] branch master updated: DL: Reduce plpy info message frequency okislal
• [madlib] branch master updated: DL: Standardize the data types used for keras and sql nkak
• [madlib] branch master updated: Pivot: Fix array_agg + distinct scaling issue on gpdb okislal
• [madlib] branch master updated: Fix svm temp table bug by dropping it first okislal
• [madlib] branch master updated: DL:Suppress unnecessary notices and misc cleanup okislal
• [madlib] branch master updated (bb27374 -> 0ed2df5) okislal
  • [madlib] 01/02: Add load model weights examples okislal
  • [madlib] 02/02: DL: Add online docs for madlib_keras functions okislal
• [madlib] branch master updated (f6a7ddb -> bb27374) jingyimei
  • [madlib] 02/02: DL: refactor image count related code jingyimei
  • [madlib] 01/02: DL: Create one keras session per iteration for pg in fit and evaluate jingyimei
• [madlib] branch master updated: DL: Fix predict with 'NULL' string class values njayaram
• [madlib] branch master updated: DL: Fix tensorflow crashes njayaram
• [madlib] branch master updated: DL: Update user docs for madlib-keras deep learning okislal
• [madlib] branch master updated: SVM: Fix class weights when specified as a mapping njayaram
• [madlib] branch bugfix-svm-class-weights deleted (was dec12cd) njayaram
• [madlib] branch bugfix-svm-class-weights created (now dec12cd) njayaram
  • [madlib] 01/01: SVM: Fix class weights when specified as a mapping njayaram
• [madlib] branch master updated: DL: Reduce dev check assert threshold while comparing loss njayaram
• [madlib] branch master updated (2479a6a -> 9933c1d) nkak
  • [madlib] 01/02: DL: Enable warm start nkak
  • [madlib] 02/02: DL: Add dev check test for loading model weights nkak
• [madlib] branch master updated (6c8643c -> 2479a6a) njayaram
  • [madlib] 03/03: get_cols() only uses 1 paramter, so get rid of all the unused ones. njayaram
  • [madlib] 01/03: DL: Refactor gp_segment_id_col njayaram
  • [madlib] 02/03: DL: Replace old evaluate1 function with madlib_keras_evaluate() njayaram
• [madlib] branch master updated: DL: fix default_buffer_size calculation in input_data_preprocessor njayaram
• [madlib] branch master updated: DL: modify the logic to get number of output units from model_arch table njayaram
• [madlib] branch master updated: MLP: Fix convergence based on threshold njayaram
• [madlib] branch master updated (d1fa5ab -> b7ff51b) nkak
  • [madlib] 02/02: DL: Rename Format class to be more meaningful. nkak
  • [madlib] 01/02: DL: Add model_weights, name and desc to the model arch table. nkak
• [madlib] branch master updated (5f1e376 -> d1fa5ab) njayaram
  • [madlib] 01/02: DL: Update content of the model summary table njayaram
  • [madlib] 02/02: DL: Record metrics_elapsed_time in model summary table njayaram
• [madlib] branch master updated (374145f -> 5f1e376) jingyimei
  • [madlib] 03/03: DL: Remove loss and metric calculations from fit transition function jingyimei
  • [madlib] 01/03: DL: Support metrics_compute_frequency for training data jingyimei
  • [madlib] 02/03: DL: Add validation for metrics in compile_params jingyimei
• [madlib] branch master updated: DL: Add model_arch column to model data table nkak
• [madlib] branch master updated (874d189 -> 642e5a5) nkak
  • [madlib] 01/02: DL: Reorder arguments for fit and evaluate UDA nkak
  • [madlib] 02/02: DL: Improve performance for predict nkak
• [madlib] branch master updated: add comment to graph user docs to distribute edge table by source vertex id fmcquillan
• [madlib] branch master updated: user doc updates to multiple modules fmcquillan
• [madlib] branch master updated: DL: Refactor code to move some constants to helper file njayaram
• [madlib] branch master updated (a16dffd -> b9453b4) njayaram
  • [madlib] 02/03: DL: Add dev-check and unit tests for metrics_compute_frequency njayaram
  • [madlib] 03/03: DL: Change how we create tables in dev-check njayaram
  • [madlib] 01/03: DL: Add new param metrics_compute_frequency to madlib_keras_fit() njayaram
• [madlib] branch master updated: DL: Improve minibatch performance by using array_cat jingyimei
• [madlib] branch master updated: DL: Improve parameter parsing and add unit tests okislal
• [madlib] branch master updated (d9c2025 -> 6f19d90) njayaram
  • [madlib] 01/02: DL: Convert the keras_eval function from UDF to UDA njayaram
  • [madlib] 02/02: DL: Update input table for fit dev check njayaram
• [madlib] branch master updated: DL: Naming improvements for dependent and independent varname jingyimei
• [madlib] branch master updated: Add contribution guideline page link in README jingyimei
• [madlib] branch master updated (01bfbcb -> 983296f) njayaram
  • [madlib] 02/04: DL: Dev-check for validation_preprocessor_dl njayaram
  • [madlib] 01/04: DL: Add input preprocessor for validation data njayaram
  • [madlib] 04/04: DL: Update validation preprocessor examples and description njayaram
  • [madlib] 03/04: DL: Add documentation for input preprocessor njayaram
• [madlib] branch master updated: MLP: Fix column names in prediction output table njayaram
• [madlib] branch master updated (e4b53a7 -> 6c1f737) nkak
  • [madlib] 02/02: DL: Fix pg bug for getting device name nkak
  • [madlib] 01/02: DL: Replace use_gpu flag with gpus_per_host integer nkak
• [madlib] branch master updated: add examples for generalize cross validation fmcquillan
• [madlib] branch master updated: DL: Add negative unit tests for unsupported compile and fit params okislal
• [madlib] branch master updated (27ddd27 -> 56efedf) njayaram
  • [madlib] 03/03: DL: Refactor computation of images per segment njayaram
  • [madlib] 01/03: DL: Update madlib_keras_fit code to weight by images instead of buffers. njayaram
  • [madlib] 02/03: DL: Add new unit tests and update existing one's njayaram
• [madlib] branch master updated (6cae627 -> 27ddd27) njayaram
  • [madlib] 02/03: DL: Add GPU support for predict njayaram
  • [madlib] 01/03: DL: Do not compile params in predict njayaram
  • [madlib] 03/03: DL: Handle NULL value for optional pred_type param in predict njayaram

• Earlier messages
• Later messages