1. 15 Feb, 2023 2 commits
  2. 14 Feb, 2023 6 commits
  3. 13 Feb, 2023 4 commits
  4. 10 Feb, 2023 8 commits
  5. 09 Feb, 2023 2 commits
    • rocking5566's avatar
      Gemm+layernorm instance, ckProfiler, client example (#568) · f7d28f3e
      rocking5566 authored
      * Add gemm + layernorm instance
      
      * Add ckProfiler
      
      * Add test
      
      * Add client example
      
      * Detect if user forger to set the workrspace
      
      * Use literal in the example
      
      * [What] use builtin function for sqrt
      [Why] compiler will not use v_sqrt_f64_e64 if we use ::sqrt()
      
      * check gemm vaildity in IsSupportedArgument
      
      * Add more testcases
      
      * Merge duplicated folder in client example
      
      * Print more infomation
      
      * Use better kernel parameter for MS problem size
      
      * clang format
      
      * Add constexpr for if condition and remove redundant include
      
      * Remove cstdlib and add constexpr
      f7d28f3e
    • guangzlu's avatar
      Add instance for elementwise normlization (#573) · 76d144fa
      guangzlu authored
      * added instances for large N
      
      * add instance for elementwise normlization
      
      * added supported restrict in device_elementwise_normalization_impl.hpp
      76d144fa
  6. 08 Feb, 2023 4 commits
    • Illia Silin's avatar
      adding the first draft of changelog (#571) · b63accee
      Illia Silin authored
      * adding the first draft of changelog
      
      * second draft of changelog
      b63accee
    • ltqin's avatar
      Add GemmAddSoftmaxGemm support for MSFT ORT (instances and client API) (#576) · 332ccc33
      ltqin authored
      * add instance for gemm bias softmax gemm
      
      * add client example
      
      * change CGridDesc_G_M_N to CGridDesc_G_M_O
      
      * add gridwise
      
      * change c grid name
      
      * device add d0s data
      
      * fix 08 client_example
      
      * add example 47_fused_attention
      
      * example output correct
      
      * add d0 to example
      
      * add d0 element op
      
      * rechange instance code
      
      * change Acc0ElementwiseOperation to C0DEElementwiseOperation
      
      * change example name
      
      * update instance for cdeelementwiseop
      
      * add bhalf_t ScaleAdd
      
      * add test
      
      * not surport geem1 bias
      
      * remove some ignore
      
      * fix test bug
      332ccc33
    • rocking's avatar
      Separate sweeponce flow and optimize the flow · 1a38e362
      rocking authored
      1a38e362
    • Illia Silin's avatar
      Fix a couple more CI issues. (#578) · bb3d9546
      Illia Silin authored
      * test the QA cron parameter for compiler commit
      
      * create separate dockers for latest and fixed amd-stg-open compiler versions
      
      * change groovy syntax
      
      * apply cron timers back to develop branch
      bb3d9546
  7. 07 Feb, 2023 2 commits
  8. 06 Feb, 2023 3 commits
  9. 01 Feb, 2023 1 commit
  10. 31 Jan, 2023 1 commit
  11. 30 Jan, 2023 1 commit
  12. 26 Jan, 2023 1 commit
  13. 25 Jan, 2023 1 commit
    • Qianfeng's avatar
      Batchnorm inference instances, external API, client examples and gtests (#531) · a1b2441f
      Qianfeng authored
      * File renaming and class renaming for device element-wise operation
      
      * Add batchnorm-infer instances, external API and client example
      
      * Add batchnorm-infer profiler module and gtests
      
      * Remove file device_elementwise_extension.hpp and move NormalizeInInfer operation to element_wise_operation.hpp
      
      * Remove the using of class aliasing for DeviceElementwiseForBatchNormInfer
      
      * Rename class and file due to conflict from device_elementwise_2d.hpp
      
      * Fix namespace in batcnnorm_infer_nhwc client example
      a1b2441f
  14. 18 Jan, 2023 4 commits
    • Qianfeng's avatar
      Use double for all scaling values and float-point constant values at the Device Op API (#557) · 52abc2f3
      Qianfeng authored
      * Use double as alpha/beta values type in reduce device op api
      
      * Use double as alpha/beta values type in softmax device op api
      
      * Use double as alpha/beta values type in multiple-reduce device op api
      
      * Use double as epsilon value type in normalization/elementwise-normalization device op api
      52abc2f3
    • Raman R jana's avatar
      Wavelet (inter-wave consumer-producer) GEMM (#310) · 1cfa8760
      Raman R jana authored
      
      
      * wavelet gemm programming model support for CK
      
      * GEMM pipeline update for wavelet progrmmaing model
      
      * Updated wavelet programming pipeline
      
      * fixes for global-write for math-wave
      
      * fixed bug in global writes
      
      * Updated comments for better readability
      
      * fixed clang format errors
      
      * added block_lds without barrier sync
      
      * clean
      
      * clean
      
      * clean
      
      * clean
      
      * refactor
      
      * prototype
      
      4 layouts
      
      fix default stride
      
      all problem sizes
      
      tidy
      
      move file
      
      update build script
      
      restore old file
      
      fix build
      
      * refactor standalone test to use gemm test harness
      
      * simplify gemm test
      
      * update build script
      
      * remove redundant
      
      * early return when cmd arg doesn't match
      
      * tidy
      
      * report failure when result not validated
      
      * tidy
      
      * Add comment depicting B2C mapping pattern.
      
      * Formatting & comments.
      
      * Comparison with custom B2C mapping pattern.
      
      * Example for wavelet gemm.
      
      * Add wavelet to Gemm standalone test.
      
      * Remove debug code.
      
      * Remove dangling #endif directive.
      
      Co-authored-by: root <Raman Jana>
      Co-authored-by: default avatarChao Liu <chao.liu2@amd.com>
      Co-authored-by: default avatarAdam Osewski <aosewski@amd.com>
      Co-authored-by: default avatarAnthony Chang <ac.chang@outlook.com>
      Co-authored-by: default avatarAdam Osewski <19374865+aosewski@users.noreply.github.com>
      1cfa8760
    • ltqin's avatar
      Add multiD Gemm client APIs (#534) · d66421fe
      ltqin authored
      
      
      * start add example
      
      * fix config
      
      * fix showinfo bug
      
      * add an elementop
      
      * change to padding
      
      * add xdl example
      
      * change elementwiseop
      
      * add instance
      
      * add instance to profiler
      
      * change file name
      
      * fix deive not support issue
      
      * add client example
      
      * fix client gemm_add_multiply name
      
      * change AddMultiply elementwiseop
      
      * fix elementwiseop
      
      * fix client example
      
      * fix addmultiply op
      
      * fix comments and fun name
      Co-authored-by: default avatarletaoqin <letaoqin@amd.com>
      d66421fe
    • Illia Silin's avatar
      fix a bug for 6-dim kernels (#555) · 00ff30af
      Illia Silin authored
      00ff30af