Skip to content
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.

CudnnFind() usage improvements (v1.3.x) #13123

Merged
merged 14 commits into from
Nov 7, 2018
Merged

CudnnFind() usage improvements (v1.3.x) #13123

merged 14 commits into from
Nov 7, 2018

Commits on Nov 5, 2018

  1. Add mx.context.gpu_memory_info() to python api for flexible tests.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    003aeda View commit details
    Browse the repository at this point in the history

  2. Add test_gluon_gpu.py:test_large_models to show cudnnFind headroom is… 

    …sue.
    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    6962199 View commit details
    Browse the repository at this point in the history

  3. Output model sizes tried by test_gluon_gpu.py:test_large_models.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    8af7d0e View commit details
    Browse the repository at this point in the history

  4. Fix perl interface to MXGetGPUMemoryInformation.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    f08d585 View commit details
    Browse the repository at this point in the history

  5. Increase difficulty of test_gluon_gpu.py:test_large_models.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    d561e68 View commit details
    Browse the repository at this point in the history

  6. Forgot a file in fix for perl.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    b1b25a2 View commit details
    Browse the repository at this point in the history

  7. Modify test to pass on no-cudnn CI runner.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    e0892d2 View commit details
    Browse the repository at this point in the history

  8. Mutex algo reg updates, serialize cudnnFind calls.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    0fcc3b2 View commit details
    Browse the repository at this point in the history

  9. Fix for cudnnFind memory headroom issue.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    c3becef View commit details
    Browse the repository at this point in the history

  10. Fix cpplint.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    2dd37b3 View commit details
    Browse the repository at this point in the history

  11. Respond to reviewers comments.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    56c20ed View commit details
    Browse the repository at this point in the history

  12. Guard against improper MXNET_GPU_MEM_LARGE_ALLOC_ROUND_SIZE values.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    e226477 View commit details
    Browse the repository at this point in the history

  13. Fix potentially unassigned var.

    @DickJC123 @lebeg
    DickJC123 authored and lebeg committed Nov 5, 2018
    Configuration menu
    Copy the full SHA
    a1719c7 View commit details
    Browse the repository at this point in the history

Commits on Nov 7, 2018

  1. Merge branch 'v1.3.x' into v1.3.x_cudnnfind_usage_improvements

    @szha
    szha authored Nov 7, 2018
    Configuration menu
    Copy the full SHA
    434e453 View commit details
    Browse the repository at this point in the history