This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
Tags: apache/mxnet
Tags
[FEATURE] Add g5 instance to CI (#20876) * Add g5 instance to jenkinsfiles where both p3 and g4 are mentioned * Remove reference to non-existent restricted-mxnetlinux-gpu-g5 * Enable unittest job on g5 * Fix Jenkinsfile_unix_gpu syntax * Include A10G arch 86 in build for g5 * Update is_TF32_enabled() for SM arch > 80 * Remove gpu arch 86 from centos builds on cuda 10 * Fix test_convolution_{grouping,dilated_impulse_response}, test_np_linalg_qr * Fix test_convolution_grouping on A100 * Fix test_rnn_unroll_variant_length * Fix test_convolution_dilated_impulse_response * Skip test_np_standard_binary_funcs test of 0-dim array broadcast * Temporarily add '-s' to pytest cpu tests * Revert "Temporarily add '-s' to pytest cpu tests" This reverts commit 4a9056a. * Improve test_rnn_layers_fp{16,32} invocation * Pin MarkupSafe==2.0.1 to avoid soft_unicode import failure * Run test_rnn_layers_fp32 only when cuDNN is present * Fix potential out-of-bounds write in count_sketch.cu * Revert "Pin MarkupSafe==2.0.1 to avoid soft_unicode import failure" This reverts commit ae17b1f.
[FEATURE] Add g5 instance to CI (#20876) * Add g5 instance to jenkinsfiles where both p3 and g4 are mentioned * Remove reference to non-existent restricted-mxnetlinux-gpu-g5 * Enable unittest job on g5 * Fix Jenkinsfile_unix_gpu syntax * Include A10G arch 86 in build for g5 * Update is_TF32_enabled() for SM arch > 80 * Remove gpu arch 86 from centos builds on cuda 10 * Fix test_convolution_{grouping,dilated_impulse_response}, test_np_linalg_qr * Fix test_convolution_grouping on A100 * Fix test_rnn_unroll_variant_length * Fix test_convolution_dilated_impulse_response * Skip test_np_standard_binary_funcs test of 0-dim array broadcast * Temporarily add '-s' to pytest cpu tests * Revert "Temporarily add '-s' to pytest cpu tests" This reverts commit 4a9056a. * Improve test_rnn_layers_fp{16,32} invocation * Pin MarkupSafe==2.0.1 to avoid soft_unicode import failure * Run test_rnn_layers_fp32 only when cuDNN is present * Fix potential out-of-bounds write in count_sketch.cu * Revert "Pin MarkupSafe==2.0.1 to avoid soft_unicode import failure" This reverts commit ae17b1f.
Optimize 'take' operator for CPU (#20745) * Improve performance of take operator * remove comment * Fix build * fix sanity * Add comment * review * Update src/operator/tensor/indexing_op.h Co-authored-by: bartekkuncer <[email protected]> Co-authored-by: Sheng Zha <[email protected]> Co-authored-by: bartekkuncer <[email protected]>
Use cuDNN for conv bias and bias grad (#20771) * Use cuDNN for conv bias and bias grad * Environment variables to use native add-bias and bias-grad * Handle 3D tensors in cuDNN legacy API * Fix AMP for ndarray.numpy * Remove env vars, used for benchmarking Co-authored-by: Vladimir Cherepanov <[email protected]>
[CI] Add Simple GitHub-Action Based License Checker (#20617) Remove ASF license headers on 3rdparty files based on discussion: > It depends on what you mean by significant, usually changes made to 3rd party files stay under the original license. See [1] on this, note that "Do not add the standard Apache License header to the top of third-party source files.”. If the license was to change, and the headers replaced then that needs to be noted in NOTICE. Please make sure that the license is noted in LICENSE. https://lists.apache.org/thread.html/r4010c53f976819c94f3450469db8b936f7c15cb046a61f0d488674d3%40%3Cgeneral.incubator.apache.org%3E
[CI] Add Simple GitHub-Action Based License Checker (#20617) Remove ASF license headers on 3rdparty files based on discussion: > It depends on what you mean by significant, usually changes made to 3rd party files stay under the original license. See [1] on this, note that "Do not add the standard Apache License header to the top of third-party source files.”. If the license was to change, and the headers replaced then that needs to be noted in NOTICE. Please make sure that the license is noted in LICENSE. https://lists.apache.org/thread.html/r4010c53f976819c94f3450469db8b936f7c15cb046a61f0d488674d3%40%3Cgeneral.incubator.apache.org%3E
PreviousNext