diff options
author | A. Unique TensorFlower <nobody@tensorflow.org> | 2016-05-24 09:47:50 -0800 |
---|---|---|
committer | TensorFlower Gardener <gardener@tensorflow.org> | 2016-05-24 10:58:11 -0700 |
commit | ad4d5f5a4391eb89146dd6f3ed40625c807d28c0 (patch) | |
tree | 1a1a093abbed5a6e727ebad33e7ed71d68015825 | |
parent | 3b31c151b48bd75da0486354015ea0f1846deb82 (diff) |
Update ops-related pbtxt files.
Change: 123122170
-rw-r--r-- | tensorflow/core/ops/compat/ops_history.v0.pbtxt | 45 | ||||
-rw-r--r-- | tensorflow/core/ops/ops.pbtxt | 54 |
2 files changed, 98 insertions, 1 deletions
diff --git a/tensorflow/core/ops/compat/ops_history.v0.pbtxt b/tensorflow/core/ops/compat/ops_history.v0.pbtxt index 41f05a0e96..87f036fe64 100644 --- a/tensorflow/core/ops/compat/ops_history.v0.pbtxt +++ b/tensorflow/core/ops/compat/ops_history.v0.pbtxt @@ -19643,6 +19643,51 @@ op { } } op { + name: "SparseDenseCwiseAdd" + input_arg { + name: "sp_indices" + type: DT_INT64 + } + input_arg { + name: "sp_values" + type_attr: "T" + } + input_arg { + name: "sp_shape" + type: DT_INT64 + } + input_arg { + name: "dense" + type_attr: "T" + } + output_arg { + name: "output" + type_attr: "T" + } + attr { + name: "T" + type: "type" + allowed_values { + list { + type: DT_FLOAT + type: DT_DOUBLE + type: DT_INT64 + type: DT_INT32 + type: DT_UINT8 + type: DT_UINT16 + type: DT_INT16 + type: DT_INT8 + type: DT_COMPLEX64 + type: DT_COMPLEX128 + type: DT_QINT8 + type: DT_QUINT8 + type: DT_QINT32 + type: DT_HALF + } + } + } +} +op { name: "SparseDenseCwiseDiv" input_arg { name: "sp_indices" diff --git a/tensorflow/core/ops/ops.pbtxt b/tensorflow/core/ops/ops.pbtxt index 4291a5d6cd..f56843a45b 100644 --- a/tensorflow/core/ops/ops.pbtxt +++ b/tensorflow/core/ops/ops.pbtxt @@ -11179,6 +11179,58 @@ op { description: "Concatenation is with respect to the dense versions of these sparse tensors.\nIt is assumed that each input is a `SparseTensor` whose elements are ordered\nalong increasing dimension number.\n\nAll inputs\' shapes must match, except for the concat dimension. The\n`indices`, `values`, and `shapes` lists must have the same length.\n\nThe output shape is identical to the inputs\', except along the concat\ndimension, where it is the sum of the inputs\' sizes along that dimension.\n\nThe output elements will be resorted to preserve the sort order along\nincreasing dimension number.\n\nThis op runs in `O(M log M)` time, where `M` is the total number of non-empty\nvalues across all inputs. This is due to the need for an internal sort in\norder to concatenate efficiently across an arbitrary dimension.\n\nFor example, if `concat_dim = 1` and the inputs are\n\n sp_inputs[0]: shape = [2, 3]\n [0, 2]: \"a\"\n [1, 0]: \"b\"\n [1, 1]: \"c\"\n\n sp_inputs[1]: shape = [2, 4]\n [0, 1]: \"d\"\n [0, 2]: \"e\"\n\nthen the output will be\n\n shape = [2, 7]\n [0, 2]: \"a\"\n [0, 4]: \"d\"\n [0, 5]: \"e\"\n [1, 0]: \"b\"\n [1, 1]: \"c\"\n\nGraphically this is equivalent to doing\n\n [ a] concat [ d e ] = [ a d e ]\n [b c ] [ ] [b c ]" } op { + name: "SparseDenseCwiseAdd" + input_arg { + name: "sp_indices" + description: "2-D. `N x R` matrix with the indices of non-empty values in a\nSparseTensor, possibly not in canonical ordering." + type: DT_INT64 + } + input_arg { + name: "sp_values" + description: "1-D. `N` non-empty values corresponding to `sp_indices`." + type_attr: "T" + } + input_arg { + name: "sp_shape" + description: "1-D. Shape of the input SparseTensor." + type: DT_INT64 + } + input_arg { + name: "dense" + description: "`R`-D. The dense Tensor operand." + type_attr: "T" + } + output_arg { + name: "output" + description: "1-D. The `N` values that are operated on." + type_attr: "T" + } + attr { + name: "T" + type: "type" + allowed_values { + list { + type: DT_FLOAT + type: DT_DOUBLE + type: DT_INT64 + type: DT_INT32 + type: DT_UINT8 + type: DT_UINT16 + type: DT_INT16 + type: DT_INT8 + type: DT_COMPLEX64 + type: DT_COMPLEX128 + type: DT_QINT8 + type: DT_QUINT8 + type: DT_QINT32 + type: DT_HALF + } + } + } + summary: "Adds up a SparseTensor and a dense Tensor, using these special rules:" + description: "(1) Broadcasts the dense side to have the same shape as the sparse side, if\n eligible;\n(2) Then, only the dense values pointed to by the indices of the SparseTensor\n participate in the cwise addition.\n\nBy these rules, the result is a logical SparseTensor with exactly the same\nindices and shape, but possibly with different non-zero values. The output of\nthis Op is the resultant non-zero values." +} +op { name: "SparseDenseCwiseDiv" input_arg { name: "sp_indices" @@ -11280,7 +11332,7 @@ op { } } summary: "Component-wise multiplies a SparseTensor by a dense Tensor." - description: "*Limitation*: this Op only broadcasts the dense side to the sparse side, but not\nthe other direction." + description: "The output locations corresponding to the implicitly zero elements in the sparse\ntensor will be zero (i.e., will not take up storage space), regardless of the\ncontents of the dense tensor (even if it\'s +/-INF and that INF*0 == NaN).\n\n*Limitation*: this Op only broadcasts the dense side to the sparse side, but not\nthe other direction." } op { name: "SparseMatMul" |