Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update from upstream repo Microsoft/CNTK@master #3

Merged
merged 186 commits into from
Feb 26, 2018
Merged

Update from upstream repo Microsoft/CNTK@master #3

merged 186 commits into from
Feb 26, 2018

Commits on Nov 28, 2017

  1. Adding ResNet101_ImageNet_CNTK/ResNet152_ImageNet_CNTK in 

    download_model.py.
    @MingyuJoo
    MingyuJoo committed Nov 28, 2017
    Configuration menu
    Copy the full SHA
    7c08325 View commit details
    Browse the repository at this point in the history

Commits on Dec 31, 2017

  1. add missing ops (logical, reduceL1, reduceL2, etc.) with ONNX support

    @liqunfu
    liqunfu committed Dec 31, 2017
    Configuration menu
    Copy the full SHA
    348773e View commit details
    Browse the repository at this point in the history

Commits on Jan 2, 2018

  1. fix bugs in crosstalkcaffe

    @yuxiaoguo
    yuxiaoguo committed Jan 2, 2018
    Configuration menu
    Copy the full SHA
    0e069fa View commit details
    Browse the repository at this point in the history

  2. Update current_iteration.md

    @ebarsoumMS
    ebarsoumMS committed Jan 2, 2018
    Configuration menu
    Copy the full SHA
    3889d0a View commit details
    Browse the repository at this point in the history

Commits on Jan 3, 2018

  1. Update current_iteration.md

    @ebarsoumMS @yuxiaoguo
    ebarsoumMS authored and yuxiaoguo committed Jan 3, 2018
    Configuration menu
    Copy the full SHA
    14105a2 View commit details
    Browse the repository at this point in the history

  2. Integrate v-yuxgu/crosstalkcaffe-bugfix into master

    Project Philly authored and CNTK Team committed Jan 3, 2018
    Configuration menu
    Copy the full SHA
    5214e2e View commit details
    Browse the repository at this point in the history

  3. gather over axis

    @n17s
    n17s committed Jan 3, 2018
    Configuration menu
    Copy the full SHA
    6477d5b View commit details
    Browse the repository at this point in the history

  4. Expose StopGradient node in BS

    @vmazalov
    vmazalov committed Jan 3, 2018
    Configuration menu
    Copy the full SHA
    8bb6864 View commit details
    Browse the repository at this point in the history

  5. transforms with composite source work correctly. 

    closes #2309
closes #2332
closes #2636
    @n17s
    n17s committed Jan 3, 2018
    Configuration menu
    Copy the full SHA
    6ee82cc View commit details
    Browse the repository at this point in the history

Commits on Jan 5, 2018

  1. Bump up model version to 29

    @vmazalov
    vmazalov committed Jan 5, 2018
    Configuration menu
    Copy the full SHA
    9bfe525 View commit details
    Browse the repository at this point in the history

  2. Integrate vadimma/StopGradientBS into master

    Project Philly authored and CNTK Team committed Jan 5, 2018
    Configuration menu
    Copy the full SHA
    13ab006 View commit details
    Browse the repository at this point in the history

Commits on Jan 6, 2018

  1. Integrate nikosk/synced-xforms3 into master

    Project Philly authored and CNTK Team committed Jan 6, 2018
    Configuration menu
    Copy the full SHA
    ae2678f View commit details
    Browse the repository at this point in the history

Commits on Jan 7, 2018

  1. Grammar fix. 

    you -> your
    @tsbertalan
    tsbertalan committed Jan 7, 2018
    Configuration menu
    Copy the full SHA
    5424fa6 View commit details
    Browse the repository at this point in the history

Commits on Jan 8, 2018

  1. Update current_iteration.md

    @n17s
    n17s committed Jan 8, 2018
    Configuration menu
    Copy the full SHA
    21c4d20 View commit details
    Browse the repository at this point in the history

Commits on Jan 9, 2018

  1. Adding error message for Convolution layer when groups > 1.

    Spandan Tiwari committed Jan 9, 2018
    Configuration menu
    Copy the full SHA
    a04c1c9 View commit details
    Browse the repository at this point in the history

  2. Allow cntk.times operator over tensors: [shape, free_axis] x [free_ax… 

    …is, shape] and [shape, batch_axis] x [batch_axis, shape].
    @tangyuq
    tangyuq committed Jan 9, 2018
    Configuration menu
    Copy the full SHA
    7142995 View commit details
    Browse the repository at this point in the history

Commits on Jan 10, 2018

  1. Integrate sptiwari/group_conv_layer_bugfix into master

    Project Philly authored and CNTK Team committed Jan 10, 2018
    Configuration menu
    Copy the full SHA
    8b261b8 View commit details
    Browse the repository at this point in the history

Commits on Jan 12, 2018

  1. Fixes #2704 (#2805)

    @nglee @manikjindal
    nglee authored and manikjindal committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    959f1c8 View commit details
    Browse the repository at this point in the history

  2. Introduce lattice reader

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    0b76ebc View commit details
    Browse the repository at this point in the history

  3. Lattice reader, cont

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    2758a8c View commit details
    Browse the repository at this point in the history

  4. Lattice reader, cont

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    e1f1101 View commit details
    Browse the repository at this point in the history

  5. Lattice deserializer - index

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    31c8c1e View commit details
    Browse the repository at this point in the history

  6. Lattice reader, add lattice index

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    e43e23f View commit details
    Browse the repository at this point in the history

  7. make the initial version compile

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    f9539f8 View commit details
    Browse the repository at this point in the history

  8. Refine lattice reader changes

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    b88cac1 View commit details
    Browse the repository at this point in the history

  9. Update SGD to not consume lattice for new serializer

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    4731e40 View commit details
    Browse the repository at this point in the history

  10. Update the lattice reader sample size

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    5a01480 View commit details
    Browse the repository at this point in the history

  11. Add SequenceWithLatticeNode

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    1b2a256 View commit details
    Browse the repository at this point in the history

  12. Ensure lattice consumption in MB

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    4a96f60 View commit details
    Browse the repository at this point in the history

  13. Add parameters for the lattice symlit

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    5d7e995 View commit details
    Browse the repository at this point in the history

  14. Rename lattice file parameters

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    b2edf0c View commit details
    Browse the repository at this point in the history

  15. Ensure that lattice metadata is passed inside the node

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    fe31b52 View commit details
    Browse the repository at this point in the history

  16. Refactor lattice methods

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    b209dd0 View commit details
    Browse the repository at this point in the history

  17. Add symlistpath as a parameter to the SE node'

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    8aa9342 View commit details
    Browse the repository at this point in the history

  18. Adding binary packing

    @eldakms @vmazalov
    eldakms authored and vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    589aa30 View commit details
    Browse the repository at this point in the history

  19. Initial work on deserialization of the lattice

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    113cdf0 View commit details
    Browse the repository at this point in the history

  20. Make the new lattice node inherit from the legacy one

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    30bf710 View commit details
    Browse the repository at this point in the history

  21. Further work to enable lattice deserializer in Sequence Node

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    6e485ad View commit details
    Browse the repository at this point in the history

  22. Enabled more SE parameters

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    98594d5 View commit details
    Browse the repository at this point in the history

  23. Able to run up to lattice consupmtion in the SE node

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    71880a9 View commit details
    Browse the repository at this point in the history

  24. Able to run for single sequence

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    c36448c View commit details
    Browse the repository at this point in the history

  25. Initial support of multiple lattice sequences

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    8c8560b View commit details
    Browse the repository at this point in the history

  26. Testing parallel sequences

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    3dba770 View commit details
    Browse the repository at this point in the history

  27. Fix lattice index for last sequence

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    a01da0c View commit details
    Browse the repository at this point in the history

  28. Some cleanup of the SE with lattice node

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    f5edf14 View commit details
    Browse the repository at this point in the history

  29. update preferred device for lattice

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    cf28412 View commit details
    Browse the repository at this point in the history

  30. Fixing definesMbSize for HTK deserializer

    @eldakms @vmazalov
    eldakms authored and vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    815e3fc View commit details
    Browse the repository at this point in the history

  31. Fix indexing of lattice sequences

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    8ab9d94 View commit details
    Browse the repository at this point in the history

  32. Ensure lattice sequences matche the label sequences

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    6fc83ca View commit details
    Browse the repository at this point in the history

  33. Fixing non existing sequence

    @eldakms @vmazalov
    eldakms authored and vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    9041489 View commit details
    Browse the repository at this point in the history

  34. Some asserts and prints

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    36315e7 View commit details
    Browse the repository at this point in the history

  35. some debugging statements

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    0750688 View commit details
    Browse the repository at this point in the history

  36. Some optimization of the binary reader

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    60c2996 View commit details
    Browse the repository at this point in the history

  37. Add new lattice reader test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    d38c060 View commit details
    Browse the repository at this point in the history

  38. Remove tabs

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    9565deb View commit details
    Browse the repository at this point in the history

  39. Some gcc fixes

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    554610e View commit details
    Browse the repository at this point in the history

  40. Explicitly reference SequenceWithSoftmax node methods

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    4b975a0 View commit details
    Browse the repository at this point in the history

  41. Add lattice index builder to makefile

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    13206a4 View commit details
    Browse the repository at this point in the history

  42. Revert pre-compute break statement

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    646cf60 View commit details
    Browse the repository at this point in the history

  43. Update E2E test for lattice reader

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    c848303 View commit details
    Browse the repository at this point in the history

  44. Debug statement for the Linux test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    bd783aa View commit details
    Browse the repository at this point in the history

  45. Add check if lattice index file exists

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    4135324 View commit details
    Browse the repository at this point in the history

  46. Add some logs for the lattice reader'

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    fcb3a2f View commit details
    Browse the repository at this point in the history

  47. Some debugging of failed linux test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    8912cbf View commit details
    Browse the repository at this point in the history

  48. Some more debugging

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    a5b78ee View commit details
    Browse the repository at this point in the history

  49. Update e2e test baselines

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    0adc7be View commit details
    Browse the repository at this point in the history

  50. Return the SequenceTraining baselines

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    d4e71b8 View commit details
    Browse the repository at this point in the history

  51. Remove current dir logging

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    0ea1eaf View commit details
    Browse the repository at this point in the history

  52. Clean up boundary code

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    6e39b34 View commit details
    Browse the repository at this point in the history

  53. Fixing mbsize

    @eldakms @vmazalov
    eldakms authored and vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    733a4e7 View commit details
    Browse the repository at this point in the history

  54. Forcibly moving allocation of lattices to CPU

    @eldakms @vmazalov
    eldakms authored and vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    1d04199 View commit details
    Browse the repository at this point in the history

  55. Ensure the new reader output is identical to the old reader

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    8c60dca View commit details
    Browse the repository at this point in the history

  56. Some clean of the lattice reader change

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    e230130 View commit details
    Browse the repository at this point in the history

  57. Revert changes of the HTKMLFreader

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    6f9185f View commit details
    Browse the repository at this point in the history

  58. Ensure lattice sequence owns the buffer

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    75f3e43 View commit details
    Browse the repository at this point in the history

  59. Remove baselines for update

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    ce3ff8c View commit details
    Browse the repository at this point in the history

  60. Expose latticedeserializer in python

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    aaaa939 View commit details
    Browse the repository at this point in the history

  61. Update old lattice reader test to 1 utterances

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    9f1d5e2 View commit details
    Browse the repository at this point in the history

  62. Change the old lattice reader config back

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    5489a4e View commit details
    Browse the repository at this point in the history

  63. Further lattice reader cleanup

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    e57bac1 View commit details
    Browse the repository at this point in the history

  64. Ensure the clean up builds

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    73f2315 View commit details
    Browse the repository at this point in the history

  65. Update the SE with new reader test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    25d79f1 View commit details
    Browse the repository at this point in the history

  66. Disable parallel lattice constructoin

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    bf673b8 View commit details
    Browse the repository at this point in the history

  67. Ensure the e2e tests between old and new readers match

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    697793b View commit details
    Browse the repository at this point in the history

  68. Expose lattice reader to python

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    5bb82fa View commit details
    Browse the repository at this point in the history

  69. Python lattice reader test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    9c70a43 View commit details
    Browse the repository at this point in the history

  70. Update baselines to ensure they match

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    0a6a429 View commit details
    Browse the repository at this point in the history

  71. Include the python test and bump up the model version

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    eb88155 View commit details
    Browse the repository at this point in the history

  72. Update python test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    44f21e7 View commit details
    Browse the repository at this point in the history

  73. Update the python test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    2cac661 View commit details
    Browse the repository at this point in the history

  74. Remove main from the lattice test

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    f3e9993 View commit details
    Browse the repository at this point in the history

  75. Some fixes after rebase

    @vmazalov
    vmazalov committed Jan 12, 2018
    Configuration menu
    Copy the full SHA
    bcbf95b View commit details
    Browse the repository at this point in the history

Commits on Jan 13, 2018

  1. Change the docstring

    @vmazalov
    vmazalov committed Jan 13, 2018
    Configuration menu
    Copy the full SHA
    01c22ad View commit details
    Browse the repository at this point in the history

  2. add batch matmul

    Chengji Yao committed Jan 13, 2018
    Configuration menu
    Copy the full SHA
    15e705d View commit details
    Browse the repository at this point in the history

Commits on Jan 14, 2018

  1. Enable max seq size config in HTK reader

    @vmazalov
    vmazalov committed Jan 14, 2018
    Configuration menu
    Copy the full SHA
    7d67daf View commit details
    Browse the repository at this point in the history

  2. Introduce the maxSequenceSize config to the HTK feature reader

    @vmazalov
    vmazalov committed Jan 14, 2018
    Configuration menu
    Copy the full SHA
    5962f71 View commit details
    Browse the repository at this point in the history

  3. Restrict the size of lattice

    @vmazalov
    vmazalov committed Jan 14, 2018
    Configuration menu
    Copy the full SHA
    ea6834b View commit details
    Browse the repository at this point in the history

Commits on Jan 15, 2018

  1. Ensure the lattice serializer handles several chunks in a file

    @vmazalov
    vmazalov committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    f79324c View commit details
    Browse the repository at this point in the history

  2. Add some debugging statements

    @vmazalov
    vmazalov committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    1fe17f8 View commit details
    Browse the repository at this point in the history

  3. debug large lattice entries

    @vmazalov
    vmazalov committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    d26f547 View commit details
    Browse the repository at this point in the history

  4. Fix batch matmul test failures

    KeDengMS committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    1a81d41 View commit details
    Browse the repository at this point in the history

  5. Add logs for lattice index builder

    @vmazalov
    vmazalov committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    d929ecd View commit details
    Browse the repository at this point in the history

  6. Trim lattice TOC line entries

    @vmazalov
    vmazalov committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    77c95a4 View commit details
    Browse the repository at this point in the history

  7. Fixed typos in tutorials. 

    CNTK_206A_Basic_GAN.ipynb
CNTK_206B_DCGAN.ipynb
CNTK_206C_WGAN_LSGAN.ipynb
    Katsuya Iida committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    01c1509 View commit details
    Browse the repository at this point in the history

  8. Refactor lattice sequence initialization'

    @vmazalov
    vmazalov committed Jan 15, 2018
    Configuration menu
    Copy the full SHA
    e6dc1a4 View commit details
    Browse the repository at this point in the history

Commits on Jan 16, 2018

  1. Remove too much logging

    @vmazalov
    vmazalov committed Jan 16, 2018
    Configuration menu
    Copy the full SHA
    05cbd0a View commit details
    Browse the repository at this point in the history

  2. Update the max_sequence_length

    @vmazalov
    vmazalov committed Jan 16, 2018
    Configuration menu
    Copy the full SHA
    8b6af80 View commit details
    Browse the repository at this point in the history

  3. Integrate vadimma/LatticeSerializer into master

    Project Philly authored and CNTK Team committed Jan 16, 2018
    Configuration menu
    Copy the full SHA
    10d7130 View commit details
    Browse the repository at this point in the history

  4. Integrate yuqtang/TimesOnFreeAxes into master

    Project Philly authored and CNTK Team committed Jan 16, 2018
    Configuration menu
    Copy the full SHA
    5bdaed7 View commit details
    Browse the repository at this point in the history

Commits on Jan 17, 2018

  1. Enabling evaluation on GPU

    @eldakms
    eldakms committed Jan 17, 2018
    Configuration menu
    Copy the full SHA
    c3d01e4 View commit details
    Browse the repository at this point in the history

  2. Fixing BPTT for case when the minibatch size changes

    @eldakms
    eldakms committed Jan 17, 2018
    Configuration menu
    Copy the full SHA
    8066621 View commit details
    Browse the repository at this point in the history

  3. Adding find_by_uid to logging API.

    Spandan Tiwari committed Jan 17, 2018
    Configuration menu
    Copy the full SHA
    f07a98d View commit details
    Browse the repository at this point in the history

  4. Fixing bug in ONNX pooling node serialization.

    Spandan Tiwari committed Jan 17, 2018
    Configuration menu
    Copy the full SHA
    4836687 View commit details
    Browse the repository at this point in the history

Commits on Jan 19, 2018

  1. Integrate sptiwari/find_by_uid4 into master

    Project Philly authored and CNTK Team committed Jan 19, 2018
    Configuration menu
    Copy the full SHA
    f36c73c View commit details
    Browse the repository at this point in the history

  2. Replace CntkCsAssemblyVersion by CntkComponentVersion on VS projects

    Thiago Crepaldi committed Jan 19, 2018
    Configuration menu
    Copy the full SHA
    997d3a1 View commit details
    Browse the repository at this point in the history

Commits on Jan 20, 2018

  1. Support CNTK installation in VirtualEnv 

    CNTK installation using pip place dlls in the root folder of the python environment. Unlike Conda, VirtualEnv doesn't add this folder in library search path and hence 'import CNTK' is broken in VirtualEnv. Fix: Place CNTK libraries in site-packages/CNTK/lib and add the same to library search path.
    Manik Jindal committed Jan 20, 2018
    Configuration menu
    Copy the full SHA
    3765da9 View commit details
    Browse the repository at this point in the history

Commits on Jan 22, 2018

  1. Removing find_by_uid API.

    Spandan Tiwari committed Jan 22, 2018
    Configuration menu
    Copy the full SHA
    eeb0645 View commit details
    Browse the repository at this point in the history

Commits on Jan 23, 2018

  1. CNTK support for CUDA 9 

    CNTK now supports CUDA 9/cuDNN 7. This requires an update to build environment to Ubuntu 16/GCC 5 for Linux, and Visual Studio 2017/VCTools 14.11 for Windows. With CUDA 9, CNTK also added a preview for 16-bit floating point (a.k.a FP16) computation.

Please check out the example of FP16 in ResNet50 at /Examples/Image/Classification/ResNet/Python/TrainResNet_ImageNet_Distributed.py

Notes on FP16 preview:
* FP16 implementation on CPU is not optimized, and it's not supposed to be used in CPU inference directly. User needs to convert the model to 32-bit floating point before running on CPU.
* Loss/Criterion for FP16 training needs to be 32bit for accumulation without overflow, using cast function. Please check the example above.
* Readers do not have FP16 output unless using numpy to feed data, cast from FP32 to FP16 is needed. Please check the example above.
* FP16 gradient aggregation is currently only implemented on GPU using NCCL2. Distributed training with FP16 with MPI is not supported.
* FP16 math is a subset of current FP32 implementation. Some model may get Feature Not Implemented exception using FP16.
* FP16 is currently not supported in BrainScript. Please use Python for FP16.

To setup build and runtime environment on Windows:
* Install [Visual Studio 2017](https://www.visualstudio.com/downloads/) with following workloads and components. From command line (use Community version installer as example):
    vs_community.exe --add Microsoft.VisualStudio.Workload.NativeDesktop --add Microsoft.VisualStudio.Workload.ManagedDesktop --add Microsoft.VisualStudio.Workload.Universal --add Microsoft.Component.PythonTools --add Microsoft.VisualStudio.Component.VC.Tools.14.11
* Install [NVidia CUDA 9](https://developer.nvidia.com/cuda-90-download-archive?target_os=Windows&target_arch=x86_64)
* From PowerShell, run:
    /Tools/devInstall/Windows/DevInstall.ps1
* Start VCTools 14.11 command line, run:
    cmd /k "%VS2017INSTALLDIR%\VC\Auxiliary\Build\vcvarsall.bat" x64 --vcvars_ver=14.11
* Open /CNTK.sln from the VCTools 14.11 command line. Note that starting CNTK.sln other than VCTools 14.11 command line, would causes CUDA 9 [build error](https://developercommunity.visualstudio.com/content/problem/163758/vs-2017-155-doesnt-support-cuda-9.html).

To setup build and runtime environment on Linux using docker, please build Unbuntu 16.04 docker image using Dockerfiles /Tools/docker. For other Linux systems, please refer to the Dockerfiles to setup dependent libraries for CNTK.
    KeDengMS committed Jan 23, 2018
    Configuration menu
    Copy the full SHA
    3cf3af5 View commit details
    Browse the repository at this point in the history

  2. Remove redefinition of CNTK_COMPONENT_VERSION 

    Math.vcsproj inherits this macro from CNTK.Common.props already
    Thiago Crepaldi committed Jan 23, 2018
    Configuration menu
    Copy the full SHA
    e883520 View commit details
    Browse the repository at this point in the history

  3. Rename default Anaconda python environment 

    Depending on the CNTK setup steps followed, either cntk-py<version> or
<cntkdev-py<version> is installed. cntk-py<version> will be adopted as
the default from now on.
    Thiago Crepaldi committed Jan 23, 2018
    Configuration menu
    Copy the full SHA
    a40285c View commit details
    Browse the repository at this point in the history

  4. Integrate sptiwari/remove_find_by_uid into master

    Project Philly authored and CNTK Team committed Jan 23, 2018
    Configuration menu
    Copy the full SHA
    a087490 View commit details
    Browse the repository at this point in the history

  5. Fix Dockerfiles under Tools/docker

    KeDengMS committed Jan 23, 2018
    Configuration menu
    Copy the full SHA
    9f6cced View commit details
    Browse the repository at this point in the history

Commits on Jan 24, 2018

  1. Merge branch 'kedeng/fixDocker2'

    KeDengMS committed Jan 24, 2018
    Configuration menu
    Copy the full SHA
    387725d View commit details
    Browse the repository at this point in the history

  2. Fix linux binary drop script for nccl

    KeDengMS committed Jan 24, 2018
    Configuration menu
    Copy the full SHA
    851ea5d View commit details
    Browse the repository at this point in the history

Commits on Jan 25, 2018

  1. Adding halide based binary convolution operators and its dependancies

    @jaliyae
    jaliyae committed Jan 25, 2018
    Configuration menu
    Copy the full SHA
    a7a52d7 View commit details
    Browse the repository at this point in the history

  2. Fix TensorBoard on_write_test_summary method (#2878)

    @rnrneverdies
    rnrneverdies committed Jan 25, 2018
    Configuration menu
    Copy the full SHA
    af26e4f View commit details
    Browse the repository at this point in the history

  3. Update CNTK_204_Sequence_To_Sequence.ipynb 

    remove redundant code.
create_reader() is correctly defined in block 10.
    @zingdle
    zingdle committed Jan 25, 2018
    Configuration menu
    Copy the full SHA
    32c4e04 View commit details
    Browse the repository at this point in the history

  4. Integrate jaliyaek/halide_squash2 into master

    Project Philly authored and CNTK Team committed Jan 25, 2018
    Configuration menu
    Copy the full SHA
    ffc7507 View commit details
    Browse the repository at this point in the history

Commits on Jan 26, 2018

  1. Remove Python 3.4 support

    Manik Jindal committed Jan 26, 2018
    Configuration menu
    Copy the full SHA
    343f383 View commit details
    Browse the repository at this point in the history

Commits on Jan 27, 2018

  1. Making Halide default in Jenkin builds

    @jaliyae
    jaliyae committed Jan 27, 2018
    Configuration menu
    Copy the full SHA
    dd47e21 View commit details
    Browse the repository at this point in the history

Commits on Jan 28, 2018

  1. Integrate jaliyaek/config into master

    Project Philly authored and CNTK Team committed Jan 28, 2018
    Configuration menu
    Copy the full SHA
    950ac47 View commit details
    Browse the repository at this point in the history

  2. Fix build if optional MKLDNN is not present

    Manik Jindal committed Jan 28, 2018
    Configuration menu
    Copy the full SHA
    ba9c2e7 View commit details
    Browse the repository at this point in the history

Commits on Jan 29, 2018

  1. Update post-build.cmd

    @mhamilton723
    mhamilton723 committed Jan 29, 2018
    Configuration menu
    Copy the full SHA
    be931c8 View commit details
    Browse the repository at this point in the history

  2. Edit makefile

    @mhamilton723
    mhamilton723 committed Jan 29, 2018
    Configuration menu
    Copy the full SHA
    08cc45c View commit details
    Browse the repository at this point in the history

  3. Add image writer SOs and zlib dlls

    @mhamilton723
    mhamilton723 committed Jan 29, 2018
    Configuration menu
    Copy the full SHA
    6db8b6e View commit details
    Browse the repository at this point in the history

  4. add debug dll support

    @mhamilton723
    mhamilton723 committed Jan 29, 2018
    Configuration menu
    Copy the full SHA
    166c8bc View commit details
    Browse the repository at this point in the history

  5. fix variable

    @mhamilton723
    mhamilton723 committed Jan 29, 2018
    Configuration menu
    Copy the full SHA
    8569868 View commit details
    Browse the repository at this point in the history

  6. fix makefile

    @mhamilton723
    mhamilton723 committed Jan 29, 2018
    Configuration menu
    Copy the full SHA
    6e846aa View commit details
    Browse the repository at this point in the history

Commits on Jan 30, 2018

  1. Break Nuget packages 

    Separate out dependencies from Nuget packages to reduce size of each
individual package because Nuget service do not allow to upload a
package of size more than 333MB.

Current list of packages:
CNTK.Deps.Cuda - Cuda dlls for GPU package
CNTK.Deps.MKL - MKL dlls
CNTK.Deps.OpenCV.Zip - OpenCV and Zip dlls
CNTK.CPUOnly - CNTK CPU only Nuget package. Dependencies: MKL, OpenCV.Zip
CNTK.GPU - CNTK GPU Nuget package. Dependencies: Cuda, MKL, OpenCV.Zip
CNTK.UWP.CPUOnly - CNTK CPU Nuget package for UWP.
Microsoft.Research.CNTK.CpuEval-mkl - Legacy CNTK eval Nuget package. Dependencies: MKL
    Manik Jindal committed Jan 30, 2018
    Configuration menu
    Copy the full SHA
    6ea2c67 View commit details
    Browse the repository at this point in the history

  2. Merge pull request #2887 from Microsoft/marhamil-fix-jar 

    Update post-build.cmd
    @mhamilton723
    mhamilton723 committed Jan 30, 2018
    Configuration menu
    Copy the full SHA
    00158ff View commit details
    Browse the repository at this point in the history

  3. add ImageScaler, fix ConvTranspose

    @liqunfu
    liqunfu committed Jan 30, 2018
    Configuration menu
    Copy the full SHA
    2a764a7 View commit details
    Browse the repository at this point in the history

  4. Updating release notes with ONNX and other feature work (e.g. group c… 

    …onvolution).
    @spandantiwari
    spandantiwari committed Jan 30, 2018
    Configuration menu
    Copy the full SHA
    f18bcda View commit details
    Browse the repository at this point in the history

  5. Pooling missing code fix

    @liqunfu
    liqunfu committed Jan 30, 2018
    Configuration menu
    Copy the full SHA
    e91df39 View commit details
    Browse the repository at this point in the history

Commits on Jan 31, 2018

  1. broadcast of scalar

    @liqunfu
    liqunfu committed Jan 31, 2018
    Configuration menu
    Copy the full SHA
    22cea7c View commit details
    Browse the repository at this point in the history

  2. Making axis attribute optional in Concat.

    Spandan Tiwari committed Jan 31, 2018
    Configuration menu
    Copy the full SHA
    4bb3031 View commit details
    Browse the repository at this point in the history

  3. Merge branch 'master' into liqun/onnx17Stage2

    Manik Jindal committed Jan 31, 2018
    Configuration menu
    Copy the full SHA
    b09ff2b View commit details
    Browse the repository at this point in the history

  4. Update current_iteration.md

    @jaliyae
    jaliyae committed Jan 31, 2018
    Configuration menu
    Copy the full SHA
    688b96e View commit details
    Browse the repository at this point in the history

  5. Update current_iteration.md

    @jaliyae
    jaliyae committed Jan 31, 2018
    Configuration menu
    Copy the full SHA
    abfc3a8 View commit details
    Browse the repository at this point in the history

Commits on Feb 1, 2018

  1. add hierarchical softmax

    Chengji Yao committed Feb 1, 2018
    Configuration menu
    Copy the full SHA
    7877e3f View commit details
    Browse the repository at this point in the history

  2. MakeBinaryDrop: Change Cuda libs to 9.0

    Manik Jindal committed Feb 1, 2018
    Configuration menu
    Copy the full SHA
    e129bcf View commit details
    Browse the repository at this point in the history

  3. Integrate yacheo/h-softmax into master

    Project Philly authored and CNTK Team committed Feb 1, 2018
    Configuration menu
    Copy the full SHA
    42a47a3 View commit details
    Browse the repository at this point in the history

  4. Merge release/2.4 

    Squashed commit of the following:

commit 82e025b
Merge: 0702c09 ed5451c
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 21:23:23 2018 -0800

    Merge branch 'release/2.4' into manikj/merge-release-2.4

commit 0702c09
Merge: 32f6844 1afe404
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 21:23:11 2018 -0800

    Merge branch 'release/2.4' (early part) into manikj/merge-release-2.4

commit 32f6844
Merge: 42a47a3 1f15d46
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 21:21:55 2018 -0800

    Merge branch 'release/2.4' (early part) into manikj/merge-release-2.4

commit ed5451c
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 19:32:04 2018 -0800

    MakeBinaryDrop: Fix lib names

commit c528a56
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 12:35:40 2018 -0800

    Update README.md

commit 490a58e
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 12:08:57 2018 -0800

    Add Halide to Release notes

commit 2196299
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 11:48:45 2018 -0800

    MakeBinaryDrop: Change Cuda libs to 9.0

commit 3df48dd
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 06:06:29 2018 -0800

    Add release notes

commit 1afe404
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 06:05:05 2018 -0800

    Replace 2.4+ with 2.4

commit 1f15d46
Author: Manik Jindal <manikj@microsoft.com>
Date:   Wed Jan 31 06:04:24 2018 -0800

    Bump version to 2.4
    Manik Jindal committed Feb 1, 2018
    Configuration menu
    Copy the full SHA
    fbbbaa9 View commit details
    Browse the repository at this point in the history

  5. Adding find_by_uid to logging API.

    Spandan Tiwari committed Feb 1, 2018
    Configuration menu
    Copy the full SHA
    3a19d2f View commit details
    Browse the repository at this point in the history

Commits on Feb 2, 2018

  1. Node timing and profile details format in chrome://tracing. 

    Working example in ./Examples/Image/Classification/MLP/Python/SimpleMNIST.py

Note that node timing would be added to profiler details when profiler is enabled, i.e.

    import cntk as C
    C.debugging.debug.set_node_timing(True)
    C.debugging.start_profiler()
    C.debugging.enable_profiler()
    trainer|evaluator|function executions
    trainer|evaluator|function.print_node_timing()
    C.debugging.stop_profiler()
    KeDengMS committed Feb 2, 2018
    Configuration menu
    Copy the full SHA
    3660b7a View commit details
    Browse the repository at this point in the history

Commits on Feb 5, 2018

  1. Backward compat bug fix in LeakyReLu reverting alph to type double.

    Spandan Tiwari committed Feb 5, 2018
    Configuration menu
    Copy the full SHA
    977e402 View commit details
    Browse the repository at this point in the history

  2. Passin device ID in DistComm for NCCL initilization

    @junjieqian
    junjieqian committed Feb 5, 2018
    Configuration menu
    Copy the full SHA
    6556e1f View commit details
    Browse the repository at this point in the history

Commits on Feb 6, 2018

  1. Integrate sptiwari/onnx_leakyrelu_fix into master

    Project Philly authored and CNTK Team committed Feb 6, 2018
    Configuration menu
    Copy the full SHA
    d615602 View commit details
    Browse the repository at this point in the history

Commits on Feb 7, 2018

  1. Fix typo

    @k-ujihara
    k-ujihara committed Feb 7, 2018
    Configuration menu
    Copy the full SHA
    56e1f3b View commit details
    Browse the repository at this point in the history

  2. Add KERA_BACKGROUND=cntk into CNTK Dockerfiles

    Thiago Crepaldi committed Feb 7, 2018
    Configuration menu
    Copy the full SHA
    c5abe7f View commit details
    Browse the repository at this point in the history

Commits on Feb 8, 2018

  1. Bug fix in ONNX broadcasting for scalar.

    Spandan Tiwari committed Feb 8, 2018
    Configuration menu
    Copy the full SHA
    aa1aaf8 View commit details
    Browse the repository at this point in the history

  2. Integrate sptiwari/onnx_scalar_constant into master

    Project Philly authored and CNTK Team committed Feb 8, 2018
    Configuration menu
    Copy the full SHA
    1142563 View commit details
    Browse the repository at this point in the history

Commits on Feb 9, 2018

  1. Update README.md

    @mx-iao
    mx-iao committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    08ba214 View commit details
    Browse the repository at this point in the history

  2. Merge pull request #2822 from tsbertalan/master 

    Grammar fix.
    @mx-iao
    mx-iao committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    062ce4d View commit details
    Browse the repository at this point in the history

  3. Merge pull request #2940 from kazuyaujihara/patch/csexampletypo 

    Fix typo in C# examples
    @mx-iao
    mx-iao committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    06ee7b8 View commit details
    Browse the repository at this point in the history

  4. Merge pull request #2880 from rnrneverdies/master 

    Fix TensorBoard on_write_test_summary method
    @mx-iao
    mx-iao committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    2b94e7d View commit details
    Browse the repository at this point in the history

  5. Merge pull request #2881 from zingdle/patch-2 

    Remove redundant code in CNTK_204_Sequence_To_Sequence.ipynb
    @mx-iao
    mx-iao committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    8b71ff0 View commit details
    Browse the repository at this point in the history

  6. Merge pull request #2843 from kaiidams/kaiidams/typotut206 

    Fixed typos in tutorials.
    @mx-iao
    mx-iao committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    bd121f6 View commit details
    Browse the repository at this point in the history

  7. Merge pull request #2690 from MingyuJoo/CNTK-2-PretrainedModels 

    Adding ResNet101_ImageNet_CNTK/ResNet152_ImageNet_CNTK in download_model.py.
    @mx-iao
    mx-iao committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    08bcbdc View commit details
    Browse the repository at this point in the history

  8. Fast tensor ops using MKL 

    - Accelerates some common tensor ops in Intel CPU inference for float32, especially for fully connected networks
- Can be turned on/off by cntk.cntk_py.enable_cpueval_optimization()/cntk.cntk_py.disable_cpueval_optimization()
    KeDengMS committed Feb 9, 2018
    Configuration menu
    Copy the full SHA
    19719a6 View commit details
    Browse the repository at this point in the history

Commits on Feb 13, 2018

  1. First round of changes for ONNX FreeDimension support.

    Spandan Tiwari committed Feb 13, 2018
    Configuration menu
    Copy the full SHA
    aa7447e View commit details
    Browse the repository at this point in the history

Commits on Feb 14, 2018

  1. Fix for ONNX ConvTranspose node loading issue.

    Spandan Tiwari committed Feb 14, 2018
    Configuration menu
    Copy the full SHA
    6da3408 View commit details
    Browse the repository at this point in the history

  2. Fix Tutorial 201B for convergence issue.

    KeDengMS committed Feb 14, 2018
    Configuration menu
    Copy the full SHA
    206db8c View commit details
    Browse the repository at this point in the history

  3. Make the lattice deserialization parallel 

    (cherry picked from commit 4961226)
    @vmazalov
    vmazalov committed Feb 14, 2018
    Configuration menu
    Copy the full SHA
    1a1e08c View commit details
    Browse the repository at this point in the history

  4. Introduce lattice reader verbosity 

    (cherry picked from commit d465167)
    @vmazalov
    vmazalov committed Feb 14, 2018
    Configuration menu
    Copy the full SHA
    c3b6f1f View commit details
    Browse the repository at this point in the history

  5. Some refactoring 

    (cherry picked from commit 9d2cda7)
    @vmazalov
    vmazalov committed Feb 14, 2018
    Configuration menu
    Copy the full SHA
    838a433 View commit details
    Browse the repository at this point in the history

  6. Integrate kedeng/fix201B2 into master

    Project Philly authored and CNTK Team committed Feb 14, 2018
    Configuration menu
    Copy the full SHA
    54ba085 View commit details
    Browse the repository at this point in the history

  7. Add support for FreeDimension in Pooling/Unpooling

    KeDengMS committed Feb 14, 2018
    Configuration menu
    Copy the full SHA
    2237dd0 View commit details
    Browse the repository at this point in the history

Commits on Feb 15, 2018

  1. Introduce latticeConfigPath to SE node 

    (cherry picked from commit dcdad16)
    @vmazalov
    vmazalov committed Feb 15, 2018
    Configuration menu
    Copy the full SHA
    e971570 View commit details
    Browse the repository at this point in the history

  2. Integrate vadimma/latpar into master

    Project Philly authored and CNTK Team committed Feb 15, 2018
    Configuration menu
    Copy the full SHA
    e186ddd View commit details
    Browse the repository at this point in the history

  3. Add nightly build badges

    Manik Jindal committed Feb 15, 2018
    Configuration menu
    Copy the full SHA
    5f1d710 View commit details
    Browse the repository at this point in the history

Commits on Feb 16, 2018

  1. Adding mean_variance_normalization CNTK and ONNX op, and LayerNormali… 

    …zation ONNX suort.
    Spandan Tiwari committed Feb 16, 2018
    Configuration menu
    Copy the full SHA
    4017a16 View commit details
    Browse the repository at this point in the history

Commits on Feb 20, 2018

  1. Move hard-coded CNTK version to a common place 

    This will help to build CNTK for nighlty builds with required CNTK
version. If environment variable 'BUILD_CNTK_VERSION' is set, then CNTK
will be build for that version as a public release('+' won't be appended
to CNTK version). Otherwise hard-coded CNTK version will be used as a
private build (2.4+).
    Manik Jindal committed Feb 20, 2018
    Configuration menu
    Copy the full SHA
    461b82d View commit details
    Browse the repository at this point in the history

Commits on Feb 21, 2018

  1. Disable MKL-DNN for now before we pick up the fix for AMD cache size … 

    …fix.

In MKL-DNN 0.12 release, there's a cache size detection issue causing crash on AMD Ryzen CPUs.
This change disables MKL-DNN to unblock AMD Ryzen users for now.
    KeDengMS committed Feb 21, 2018
    Configuration menu
    Copy the full SHA
    9032174 View commit details
    Browse the repository at this point in the history

  2. Handle malformed lattice

    @vmazalov
    vmazalov committed Feb 21, 2018
    Configuration menu
    Copy the full SHA
    1c78107 View commit details
    Browse the repository at this point in the history

  3. Minor refactoring

    @vmazalov
    vmazalov committed Feb 21, 2018
    Configuration menu
    Copy the full SHA
    523af50 View commit details
    Browse the repository at this point in the history

Commits on Feb 23, 2018

  1. Fix crash in CBF when crossing sweep boundary 

    This fixes issue #2479 #2964
    KeDengMS committed Feb 23, 2018
    Configuration menu
    Copy the full SHA
    695bdf7 View commit details
    Browse the repository at this point in the history