diff options
author | Eric Kunze <eric.kunze@arm.com> | 2020-11-04 08:20:09 -0800 |
---|---|---|
committer | Eric Kunze <eric.kunze@arm.com> | 2020-11-05 10:47:42 -0800 |
commit | d77db2a476454eb549512a7097813cb4a5b352b1 (patch) | |
tree | 11e3dc5f94c284cb6b22dbc825701eb2ce0d13b6 /chapters/data_layout.adoc | |
parent | 57e79c0cf2b37d37c4895ee08fd7c2534ae44cad (diff) | |
download | specification-d77db2a476454eb549512a7097813cb4a5b352b1.tar.gz |
Clean input argument rank limitations
Comparison ops, similar to the other elementwise ops, can operate
on tensors of any rank.
For those operators with rank limitations, make them consistent and
remove "dims".
Signed-off-by: Eric Kunze <eric.kunze@arm.com>
Change-Id: I33eb00122cec4c6625f080f9ed640f1d874fae87
Diffstat (limited to 'chapters/data_layout.adoc')
-rw-r--r-- | chapters/data_layout.adoc | 10 |
1 files changed, 5 insertions, 5 deletions
diff --git a/chapters/data_layout.adoc b/chapters/data_layout.adoc index bec551b..4e6d24d 100644 --- a/chapters/data_layout.adoc +++ b/chapters/data_layout.adoc @@ -17,7 +17,7 @@ Concatenate two tensors along a given axis. No data conversion happens during a |=== |Argument|Type|Name|Shape|Description -|Input|in_t*|input1|shape1|Input tensor from 1 to 4 dims +|Input|in_t*|input1|shape1|Input tensor with rank from 1 to 4 |Input|in_t*|input2|shape2|Input tensor with rank matching input1 |Attribute|int|axis|-|Axis along which concatenation is to occur. |Output|out_t*|output|shape|Output tensor of same type, size as the input tensor @@ -106,7 +106,7 @@ Returns a tensor with the same type/values as the input, with a new shape specif |=== |Argument|Type|Name|Shape|Description -|Input|in_t*|input1|shape1|Input tensor from 1 to 4 dims +|Input|in_t*|input1|shape1|Input tensor |Attribute|int|new_shape|[rank(output)]|List of values, with each element giving the size of the result tensor for the given dimension. At most one dimension may be given as-1 to automatically calculate the dimension size. |Output|out_t*|output|shape|Output tensor of same type, size as the input tensor |=== @@ -180,7 +180,7 @@ Extracts a slice of the input tensor 0 on the given axis, beginning at the start |=== |Argument|Type|Name|Shape|Description -|Input|in_t*|input1|shape1|Input tensor from 1 to 4 dims +|Input|in_t*|input1|shape1|Input tensor with rank from 1 to 4 |Attribute|int|start|[rank(input1)]|List of integer coordinates, of length equal to the rank of input 0. Start coordinate for slicing. |Attribute|int|size|[rank(input1)]|List of integer size values, of length equal to the rank of input 0. Size of the input to be used. |Output|out_t*|output|shape|Output tensor of same type as the input tensor @@ -221,7 +221,7 @@ Replicates input 0 multiplies times along each dimension. |=== |Argument|Type|Name|Shape|Description -|Input|in_t*|input1|shape1|Input tensor from 1 to 4 dims +|Input|in_t*|input1|shape1|Input tensor with rank from 1 to 4 |Attribute|int|multiplies|[rank(shape1)]|Number of times to replicate input1 in each dimension |Output|out_t*|output|shape|Output tensor of same type, rank as the input tensor |=== @@ -262,7 +262,7 @@ Permutes the dimensions based on perm. |=== |Argument|Type|Name|Shape|Description -|Input|in_t*|input1|shape1|Input tensor from 1 to 4 dims +|Input|in_t*|input1|shape1|Input tensor with rank from 1 to 4 |Attribute|int|perms|[rank(input1)]|List of integers of length equal to the rank of input1. |Output|out_t*|output|shape|Output tensor of same type, rank as the input tensor |=== |