Commit Graph

  • d84c61c29b Fixes cuDNN version for CUDA 9.0 build environment (#15001) perdasilva 2019-05-20 12:19:54 +02:00
  • 96b1cde15f [MXNET-1403] Disable numpy's writability of NDArray once it is zero-copied to MXNet (#14948) Junru Shao 2019-05-19 22:18:36 -07:00
  • 720aa29a97 add an operator for computing the likelihood of a Hawkes self-exciting process (#14683) Caner Turkmen 2019-05-20 07:32:17 +03:00
  • 07a8903343 clean up duplicate cudnn installation (#14996) Jake Lee 2019-05-19 19:58:27 -07:00
  • bd44ff48c4 Update mkldnn window build instructions in MKLDNN_README.md (#14952) yinghu5 2019-05-19 15:09:06 +08:00
  • 0e570a7431 Disables flaky test_operator_gpu.test_activation (#14969) perdasilva 2019-05-19 03:52:19 +02:00
  • 8c4a0e62b3 Disables test_bulking due to flakyness (#14971) perdasilva 2019-05-19 03:52:03 +02:00
  • b37713012a implementation for equivalence of tf.moments (#14842) Hao Jin 2019-05-18 18:27:14 -07:00
  • a7e7cdc0c4 add ctx for rand_ndarray and rand_sparse_ndarray (#14966) Hao Jin 2019-05-18 11:13:28 -07:00
  • 9f451fb6f4 [MXNET-1333] Estimator and Fit API (#14629) Lai Wei 2019-05-18 11:03:18 -07:00
  • 3d6646a498 Silence excessive mkldnn logging output on tests. (#14947) Pedro Larroy 2019-05-18 00:44:35 -07:00
  • e11a23d1dd [Doc] Add MKL-DNN operator list (#14891) Tao Lv 2019-05-18 15:28:37 +08:00
  • c8ee1b31ec Fix test memory with ResourceScope (#14666) Zach Kimberg 2019-05-17 19:01:19 -07:00
  • 379d11f112 Disables large tensor size cpu test step (#14982) perdasilva 2019-05-18 02:09:11 +02:00
  • 5aa62d8ae2 Use env var to enforce safe accumulation in ReduceAxesCompute (#14830) Hao Jin 2019-05-17 13:48:19 -07:00
  • 5b4e866353 Sync Horovod distributed training examples with latest changes (#14748) Yuxi Hu 2019-05-17 11:24:55 -07:00
  • 016b3b25c7 [Dependency Update] Upgrade CI to use latest cuDNN (#14950) Jake Lee 2019-05-17 10:34:01 -07:00
  • 0d7794730a [MKLDNN] Enable more convolution + activation fusion (#14819) Zhennan Qin 2019-05-17 12:26:06 +08:00
  • 8d6ac4a5f7 Support 3D input for MKL-DNN softmax operator (#14818) Tao Lv 2019-05-17 12:06:42 +08:00
  • d87bd2aba1 Fix the incorrect MKLDNN/MKL logic in cmake (#14877) yinghu5 2019-05-16 21:27:29 +08:00
  • 60342d6ddb add KEY for zachgk (#14965) Zach Kimberg 2019-05-15 22:19:51 -07:00
  • eecf07a93c Update MKL-DNN submodule to v0.19 (#14783) Tao Lv 2019-05-16 11:20:07 +08:00
  • c5265fbf63 Add primitive cache for MKL-DNN sum(elemwise_add operator (#14914) ciyong 2019-05-16 06:30:06 +08:00
  • 2ee7f4e2bf Improve order of execution of install scripts. (#14867) Pedro Larroy 2019-05-15 15:29:32 -07:00
  • 8b7e374d38 Fixes call to build ubuntu gpu in nightly tests (#14964) perdasilva 2019-05-15 22:14:54 +02:00
  • f7b7163253 Fixed and re-enables TensorRT steps (#14960) perdasilva 2019-05-15 20:07:26 +02:00
  • 99f5f668e2 Disables TensorRT build step (#14958) perdasilva 2019-05-15 13:31:39 +02:00
  • 1eba37a8c3 Re-enable static cached_op optimization (#14931) Zhennan Qin 2019-05-15 07:01:28 +08:00
  • 31225485c3 use mx.context.num_gpus instead of mx.test_utils.list_gpus in MF recommender example (#14926) Pedro Larroy 2019-05-14 07:55:13 -07:00
  • 13d6ee6bda Fix reshape to add in-place back (#14903) Tao Lv 2019-05-14 08:25:32 +08:00
  • f67d067539 Add numpy linspace (#14927) Wang Jiajun 2019-05-12 22:30:35 -05:00
  • d577b6ff82 [MXNET-1352] Allow dynamic shape in while_loop and if conditionals (#14393) Junru Shao 2019-05-12 16:01:51 -07:00
  • 669ab2c0d0 Updates to cudnn package installation (#14923) perdasilva 2019-05-12 07:28:57 +02:00
  • 8a4ad9fe2c upgrade the version to 2.0.2 (#14621) Jake Lee 2019-05-11 20:09:56 -07:00
  • b22ee951ae [MXNET-857] Add initial NVTX profiler implementation (#12328) Kellen Sunderland 2019-05-10 22:15:12 -07:00
  • 4796851b2f Deprecate NDArrayCollector and instead use ResourceScope (#14780) Zach Kimberg 2019-05-10 17:12:09 -07:00
  • 6feadf4e72 Pins version of scikit-learn for python2 due to drop in support (#14928) (#14929) perdasilva 2019-05-10 15:52:54 +02:00
  • f4598e743a Pins version of scikit-learn for python2 due to drop in support (#14928) perdasilva 2019-05-10 12:54:32 +02:00
  • 874fb89cd3 Add API documentation for upsampling operator with examples (#14919) Sandeep Krishnamurthy 2019-05-09 16:12:18 -07:00
  • b9ea0fa11a [1.4.x] Adds additional build envs (#14920) perdasilva 2019-05-09 20:23:35 +02:00
  • 13f81a03ec Improve dev_menu virtualenv handling (#14788) Pedro Larroy 2019-05-09 06:47:55 -04:00
  • da024881e8 upgrade the libpng to 1.6.35 (#14620) Jake Lee 2019-05-09 03:46:12 -07:00
  • 8dddac0ccd [MXNET-1400] adding tests cases to verify large tensor support for depth_to_space and space_to_depth (#14797) Rohit Kumar Srivastava 2019-05-08 21:50:11 -07:00
  • 2e03e9ff07 Adds additional CUDA build environments (#14909) perdasilva 2019-05-08 17:20:41 +02:00
  • aafe3dfe40 Backports static build scripts from master (#14906) perdasilva 2019-05-08 13:27:46 +02:00
  • e7fe640192 Updates tolerances for test_layer_bidirectional (#14682) (#14908) perdasilva 2019-05-08 08:16:06 +02:00
  • 0ddef13bd2 Revert "[Dependency Update] Upgrade cuDNN & NCCL (#14884)" (#14910) perdasilva 2019-05-08 07:57:44 +02:00
  • 527573ec2b Add support for fast variable-length LSTM (#14208) stephenrawls 2019-05-07 13:52:09 -07:00
  • 08895b7111 Fix the return type of sparse.clip operator (#14856) Anirudh 2019-05-07 05:10:33 -07:00
  • 5bda980c5e fix add_n bug: when input mem overlap with output mem, results is wrong (#14889) rongzha1 2019-05-07 13:30:12 +08:00
  • fdd45cf5a0 Add mkldnn_version.h to pip package (#14899) Yuxi Hu 2019-05-06 18:38:17 -07:00
  • 0255dd6d84 [Dependency Update] Upgrade cuDNN & NCCL (#14884) Jake Lee 2019-05-07 06:55:10 +08:00
  • a722db45c6 [Dependency Update] Upgrade openssl to 1.1.1b (#14837) Jake Lee 2019-05-07 01:09:02 +08:00
  • 42ede5016a rewrite test_custom_op_exc (#14878) Wang Jiajun 2019-05-06 08:30:40 -05:00
  • f6ef206a16 [Bugfix] Fix layer norm for large input shape (#14870) Xingjian Shi 2019-05-06 14:26:45 +08:00
  • 2113cb7fb1 [Clojure] Add Fine Tuning Sentence Pair Classification BERT Example (#14769) Carin Meier 2019-05-05 21:34:01 -04:00
  • b30949f3bf [MXNet-1211] Factor and "Like" modes in BilinearResize2D operator (#13226) Mikhail Lobanov 2019-05-05 12:53:14 +03:00
  • 25ba1d1d75 Prevent crashes for opencv exception and std::exception (#14433) Anirudh Subramanian 2019-05-04 21:12:51 -07:00
  • 621b391c14 Refactor ImageRecordIter (#14824) Zhennan Qin 2019-05-05 10:13:26 +08:00
  • 4d7bae1e07 Add the Gluon Implementation of Deformable Convolution (#14810) suyz526 2019-05-05 03:51:47 +02:00
  • 5ba285bec1 Fix sample_multinomial number of outputs bug (#14873) reminisce 2019-05-03 23:28:04 -07:00
  • d09f68af36 Update lstm_crf.py (#14865) Maoge 2019-05-03 11:10:46 -07:00
  • 204f3f2de8 Revert "Improve cached_op performance for static mode (#14785)" (#14868) Anirudh 2019-05-03 10:40:05 -07:00
  • 372f53100b [DEV] update code owner (#14862) Haibin Lin 2019-05-02 21:11:03 -07:00
  • 1540a84f1e [Clojure] Remove unneeded test files (#14813) Carin Meier 2019-05-02 14:38:56 -04:00
  • 36c3306745 Update base CUDA image for CI to v10.0 cuDNN 7.3.1 (#14513) perdasilva 2019-05-02 16:57:50 +02:00
  • e17b7e2947 [MXNET-13578] Fix cmake installation failed (#14692) Gustavo Alvarez 2019-05-02 04:14:12 +02:00
  • 381a9da4b9 Print reproduction command on CI failure (#14815) Pedro Larroy 2019-05-01 19:03:52 -07:00
  • 977e5587be Make docblocks for Gluon BatchNorm and SyncBatchNorm consistent with the code (#14840) Ziyi Mu 2019-05-01 19:02:28 -07:00
  • 5e5a59ec6b Upgrade Pylint version to 2.3.1 (#14807) Jake Lee 2019-05-02 07:58:10 +08:00
  • 1c874cfc80 reformat trt to use subgraph API, add fp16 support (#14040) Caenorst 2019-05-01 18:36:46 +02:00
  • bde1b8498f [int8] Add MobileNetV2_1.0 & ResNet18 Quantization (#14823) Xinyu Chen 2019-05-01 05:59:34 +08:00
  • 84c1635adc [MKLDNN] add quantized sum (#14614) rongzha1 2019-05-01 05:56:04 +08:00
  • cdd708745e Fix Clojure BERT example's context argument (#14843) Dave Liepmann 2019-04-30 22:37:15 +02:00
  • 488fad2c5c Fix iterator over symbol when multiple children have the same name (#14597) Przemyslaw Tredak 2019-04-30 21:18:11 +02:00
  • 5fda0a53f6 add clojure tutorials to index (#14814) Aaron Markham 2019-04-30 11:34:54 -07:00
  • c7577e5984 added extraction/generation of diagonal and triangonal matrices to linalg (#14501) moin 2019-04-30 18:28:23 +02:00
  • 1238aa09f0 Revert "use mkl sparse matrix to improve performance (#14492)" (#14806) PatricZhao 2019-04-30 21:37:08 +08:00
  • 07aef13e87 Add unpooled gpu memory type (#14716) vlado 2019-04-29 20:24:52 -06:00
  • 3a46980588 Use USE_SIGNAL_HANDLER by default set to ON in CMakeLists.txt (#14599) Pedro Larroy 2019-04-29 19:22:03 -07:00
  • 1a7199691f Bump version number (#14841) 1.4.1 Junru Shao 2019-04-29 17:51:26 -07:00
  • 69515c2f9b [v1.4.1] Java bug-fix cherry pick (#14834) Lanking 2019-04-29 16:03:36 -07:00
  • 1af29e9c06 Fixes for wine detection tutorial (#13886) Pedro Larroy 2019-04-29 15:47:13 -07:00
  • 64287dd23d Speed up SequenceReverse (#14627) Przemyslaw Tredak 2019-04-29 18:49:53 +02:00
  • c18381d336 [MXNET-1398] Enable zero-copy from numpy to MXNet NDArray (#14733) Junru Shao 2019-04-28 16:12:01 -07:00
  • 6c600250f2 [MKLDNN]Refactor requantize to speed up execution (#14608) Zhennan Qin 2019-04-28 09:19:37 +08:00
  • 003800c50b clean up submodule (#14645) Lanking 2019-04-26 20:40:17 -07:00
  • 6cbc273454 Scala/Java Predict API fix #14756 (#14804) Lanking 2019-04-26 20:40:02 -07:00
  • 40e3d7c8e8 data preparation file moved in example (#14781) Roshani Nagmote 2019-04-26 18:39:00 -07:00
  • 369b66d0f1 Improve cached_op performance for static mode (#14785) Zhennan Qin 2019-04-27 01:52:38 +08:00
  • 5dd9fa27d8 [clojure][generator] ndarray/symbol api random merged (#14800) Arthur Caillau 2019-04-26 19:30:41 +02:00
  • 680badef6e Use correct stash name when running nightly tests (#14809) Edison Gustavo Muenz 2019-04-26 17:18:45 +02:00
  • 6aeb97e62f change mxnet_option behavior (#14743) Hu Shiwen 2019-04-26 12:06:03 +08:00
  • 97e09f23cf Improve CMake handling of sse2 and sse3 (#14757) Pedro Larroy 2019-04-25 15:39:48 -07:00
  • 3d3803eb1e Updates python setup.py for recent license changes (#14778) perdasilva 2019-04-25 23:36:37 +01:00
  • 19d78e5585 Use DEFAULT macro in C APIs (#14767) (#14789) Yuxi Hu 2019-04-25 11:52:31 -07:00
  • acf53fd54d Mention additional language bindings and add links (#14798) Yuan (Terry) Tang 2019-04-25 13:26:16 -04:00
  • 22377edef3 Fix GELU backward possible NaN (#14782) Hao Jin 2019-04-25 00:20:42 -07:00
  • 587d480372 Use DEFAULT macro in C APIs (#14767) Yuxi Hu 2019-04-24 15:46:43 -07:00
  • 8cae72eead julia/ndarray: fix flaky test cases for clamp (#14776) Iblis Lin 2019-04-25 06:44:34 +08:00