tree: ef8dadcfc6c64194ebb8aba9f139d3008dfcb49e [path history] [tgz]
  1. all_reduce/
  2. android/
  3. batching/
  4. bayesflow/
  5. benchmark_tools/
  6. boosted_trees/
  7. cloud/
  8. cluster_resolver/
  9. cmake/
  10. coder/
  11. compiler/
  12. copy_graph/
  13. crf/
  14. cudnn_rnn/
  15. data/
  16. decision_trees/
  17. deprecated/
  18. distributions/
  19. eager/
  20. estimator/
  21. factorization/
  22. feature_column/
  23. ffmpeg/
  24. framework/
  25. fused_conv/
  26. gan/
  27. gdr/
  28. graph_editor/
  29. grid_rnn/
  30. hooks/
  31. hvx/
  32. image/
  33. input_pipeline/
  34. integrate/
  35. kafka/
  36. keras/
  37. kernel_methods/
  38. kfac/
  39. labeled_tensor/
  40. layers/
  41. learn/
  42. legacy_seq2seq/
  43. libsvm/
  44. linalg/
  45. linear_optimizer/
  46. lite/
  47. lookup/
  48. losses/
  49. makefile/
  50. memory_stats/
  51. meta_graph_transform/
  52. metrics/
  53. model_pruning/
  54. mpi/
  55. mpi_collectives/
  56. nccl/
  57. nearest_neighbor/
  58. nn/
  59. opt/
  60. periodic_resample/
  61. pi_examples/
  62. predictor/
  63. py2tf/
  64. quantization/
  65. quantize/
  66. receptive_field/
  67. reduce_slice_ops/
  68. remote_fused_graph/
  69. resampler/
  70. rnn/
  71. saved_model/
  72. seq2seq/
  73. session_bundle/
  74. signal/
  75. slim/
  76. solvers/
  77. sparsemax/
  78. specs/
  79. staging/
  80. stat_summarizer/
  81. stateless/
  82. summary/
  83. tensor_forest/
  84. tensorboard/
  85. tensorrt/
  86. testing/
  87. text/
  88. tfprof/
  89. timeseries/
  90. tpu/
  91. training/
  92. util/
  93. verbs/
  94. __init__.py
  95. BUILD
  96. README.md
tensorflow/contrib/README.md

TensorFlow contrib

Any code in this directory is not officially supported, and may change or be removed at any time without notice.

The contrib directory contains project directories, each of which has designated owners. It is meant to contain features and contributions that eventually should get merged into core TensorFlow, but whose interfaces may still change, or which require some testing to see whether they can find broader acceptance. We are trying to keep duplication within contrib to a minimum, so you may be asked to refactor code in contrib to use some feature inside core or in another project in contrib rather than reimplementing the feature.

When adding a project, please stick to the following directory structure: Create a project directory in contrib/, and mirror the portions of the TensorFlow tree that your project requires underneath contrib/my_project/.

For example, let's say you create foo ops in two files: foo_ops.py and foo_ops_test.py. If you were to merge those files directly into TensorFlow, they would live in tensorflow/python/ops/foo_ops.py and tensorflow/python/kernel_tests/foo_ops_test.py. In contrib/, they are part of project foo, and their full paths are contrib/foo/python/ops/foo_ops.py and contrib/foo/python/kernel_tests/foo_ops_test.py.