Add type info to MatMul spec (#4855)
This commit is contained in:
parent
a868d88d49
commit
3866896ba6
@ -57,14 +57,17 @@ Two attributes, `transpose_a` and `transpose_b` specify embedded transposition f
|
||||
|
||||
**Inputs**:
|
||||
|
||||
* **1**: Input batch of matrices A. Rank >= 1. Required.
|
||||
* **1**: Tensor of type *T* with matrices A. Rank >= 1. Required.
|
||||
|
||||
* **2**: Input batch of matrices B. Rank >= 1. Required.
|
||||
* **2**: Tensor of type *T* with matrices B. Rank >= 1. Required.
|
||||
|
||||
**Outputs**
|
||||
|
||||
* **1**: Tensor with results of the multiplication.
|
||||
* **1**: Tensor of type *T* with results of the multiplication.
|
||||
|
||||
**Types**:
|
||||
|
||||
* *T*: any supported floating point or integer type.
|
||||
|
||||
**Example**
|
||||
|
||||
|
Loading…
Reference in New Issue
Block a user