Skip to content
[Jenkins]Jenkins
log in
Success

#692 (May 22, 2020, 11:47:00 AM)

Started 5 yr 0 mo ago
Took 1 hr 47 min on powerci-docker1
Build Artifacts
tf_nightly-2.3.0.dev20200522-cp36-cp36m-linux_ppc64le.whl297.47 MiB view

Started by timer

Revision: f8c0e68a8aa5d575a19129ec67c9ed6262652082
Repository: https://github.com/tensorflow/tensorflow.git
  • refs/remotes/origin/master
Changes
  1. Enabe BF16 SoftmaxGrad(Sum), and fix accuracy by accum type. (details)
  2. Fixed eager mode gradient checkpointing by eliminating unecessary persistence of intermediate activations in memory (details)
  3. Removed API change and added tests. (details)
  4. Fix issue in tf.image.extract_glimpse (details)
  5. Add test cases for tf.image.extract_glimpse (details)
  6. Fix test failure caused by API changes in tests (details)
  7. Fix test failure due to changes of the fix for centered=False and normalized=False (details)
  8. Fix incorrect doc test (details)
  9. Fix incorrect doc example with centered=False (details)
  10. Use ExtractGlimpseV2 and ExtractGlimpse to make sure C++ kernel is backward compatible (details)
  11. Expand test case to cover both old kernel (ExtractGlimpse) and new kernel (ExtractGlimpseV2) (details)
  12. Add api_def_ExtractGlimpseV2.pbtxt (details)
  13. Update API golden (details)
  14. Reroute tf.compat.v1.extract_glimpse to use gen_image_ops.extract_glimpse (old API) (details)
  15. Update RELEADE.md to capture the breaking change of `tf.image.extract_glimpse` (details)
  16. Update doc example of v1 to keep old behavior with usage of tf.compat.v1.image.extract_glimpse (details)
  17. Modified recompute_grad to handle fwd mode diff (details)
  18. Addressed PR comments (details)
  19. Addressed PR comments (details)
  20. Reorganized tests for recompute grad (details)
  21. Support BF16 Softmax and add UT. (details)
  22. Fixed tests (details)
  23. Added grad_wrapper to accomodate graph mode (details)
  24. Fixed test (details)
  25. Addressed PR comments (details)
  26. Fixed sanity test issue (details)
  27. Fixed test (details)
  28. Allow rounding loss in Reduce UT. (details)
  29. Fixed test (details)
  30. Fixed test (details)
  31. Update TF Lite roadmap (details)
  32. [TF:TRT] Delay the computation of GraphProperties used by ConvertAfterShapes. (details)
  33. Add tf.Yield and tf.IfRegion op to model non-functional conditionals (details)
  34. Prevent extra downloads for the xtensa-xpg target. (details)
  35. Limit FillOp custom folder to int and float types to avoid crash (details)
  36. Internal change (details)
  37. Remove illegal BroadcastTo op compiler tests (details)
  38. [TF] Add support for more than one outer batch dimension to tf.nn.convolution. (details)
  39. Enable skipped test as Variable.assign(name=xxx) now works with CentralStorage (details)
  40. Support ShardedVariable in `tf.keras.layers.Embedding`. (details)
  41. Fix makefile benchmark script to build on mac os. (details)
  42. Move the feature_column.LinearModel to estimator which is the only caller for it. (details)
  43. cupti tracer allow device synchronization before gputracer stop. (details)
  44. [tf.data service] Perform tf.data service compression within tf.data. (details)
  45. Implement a pass that converts readonly reference variables to the corresponding (details)
  46. Fix for hello_world memory error (details)
  47. Switches keras.backend.placeholder + keras.backend.function to build a keras model when running eagerly (instead of trying to directly lift ops out of a graph into a concretefunction). Allows us to strip most of EagerDefinedExecutionFunction from the keras backend. (details)
  48. Internal change (details)
  49. Take device locality into account during prioritization. (details)
  50. Nit: Fix some typos. (details)
  51. [NFC] Fix typos and adopt Google style variable names (details)
  52. Enable gradient tests for tf.linalg.cholesky in eager mode. (details)
  53. Exporting CategoryEncoding layer. (details)
  54. Support the pybind11 `_dtypes.DType` in `tf.as_dtype`. (details)
  55. [TF] Add support for more than one outer batch dimension to tf.nn.convolution. (details)
  56. [tf.data service] Apply dataset options to tf.data service side datasets. (details)
  57. Disable two tests on windows due to image issue. (details)
  58. Disable flaky test. (details)
  59. switch capture_tpu_profile to new api of profiler_client. because some options is dropped, therefore two flags are deprecated. (details)
  60. Add memory cost breakdown per operation type and memory space to OpMetrics. (details)
  61. PR #39548: [INTEL MKL] Fix conv_ops_test and remapper_test (details)
  62. Adding uint32 support for more variable related operations. (details)
  63. Update ops-related pbtxt files. (details)
  64. Reapplying #39042 to avoid breaking some internal users. (details)
  65. Fix SideEffecting semantics of xla_hlo ops without regions. (details)
  66. Go: Update generated wrapper functions for TensorFlow ops. (details)
  67. Fix infinite loop in GetMatchingPaths in GCS (details)
  68. Handle case when input is complex64 type. (details)
  69. Simplify and address missing features in TPU Extract Head Outside Compilation pass. (details)
  70. Enable 1st order gradient tests for tf.linalg.svd in eager mode. (details)
  71. Enable tests for tf.linalg.matrix_logarithm in eager mode. (details)
  72. Don't disable all TPU tests just the ones that fail. (details)
  73. Fix issues where index_lookup was improperly handling hard vocab caps. Add tests. (details)
  74. Add bfloat16 support for SparseSegmentMean*/SparseSegmentSqrtN* (details)
  75. Improve diagnostic when a mutable global tensor is found (details)
  76. Reduce Layer.__call__ overhead by ~5%. (details)
  77. Test hello world example binary can run (details)
  78. Temporarily blacklist keras model_coverage_lib_tests from TF2 testing (details)
  79. Add vlogging for compression data size. (details)
  80. Test that micro speech example binary can run (details)
  81. FloorDiv and FloorMod dynamic shape support (details)
  82. Improve side-effecting semantics of xla_hlo ops with regions. (details)
  83. Pfor: support vectorizing tensorlists captured by while_loop. (details)
  84. Make tf_saved_model-related passes check their input invariants. (details)
  85. Update ops-related pbtxt files. (details)
  86. Integrate LLVM at https://github.com/llvm/llvm-project/commit/1108f5c737db (details)
  87. Remove superfluous tracemes in compression_utils. (details)
  88. Bump open source llvm revision to 1108f5c737dbdab0277874a7e5b237491839c43a (details)
  89. Internal TF change. (details)
  90. [Grappler] Do not add data inputs after control inputs in ImplementationSelector (details)
  91. Add bfloat16 support for SparseSegmentMean*/SparseSegmentSqrtN* (details)
  92. Internal change (details)
  93. [XLA] Fixup the bug in tautological compare simplifier, as spotted by Sanjoy (details)
  94. [TF/XLA] Support F64 conversion for tf.cumsum (details)
  95. Fix TFLiteConverter2 API Documentation to read frozen_graphs. (details)
  96. [tfdbg2] Document the need to call set_soft_device_placement(True) on TPUs (details)
  97. Add tf.function test for device placement logging (details)
  98. Update sparse input documentation. (details)
  99. Rearrange the binary __operator__ code in TensorFlow to be more dispatch-friendly. (details)
  100. Disable multi-threaded Conv optimizations w/ non-const filters (details)
  101. [XLA] Introduce asynchronous collective-permute (CollectivePermuteStart and CollectivePermuteDone) HLO opcodes. (details)
  102. Disable multi-threaded Conv optimizations w/ non-const filters (details)
  103. [XLA] Add AllGather to the HLO matchers. (details)
  104. [XLA] CopyStart/CopyDone times should use exclusive indices, not inclusive. (details)
  105. Reduce Layer.__call__ overhead by ~5-10%. (details)
  106. [XLA] Use all-gather in SPMD to replicate a tiled tensor. (details)
  107. compat: Update forward compatibility horizon to 2020-05-22 (details)
  108. Update GraphDef version to 409. (details)