MatMulBiasAddFusionPass
Description
Fuses the matmul/matmulv2/batchmatmul/batchmatmulv2 and biasadd/add operators into the matmul/matmulv2batchmatmul/batchmatmulv2 operator.
Before:
After: 
Restrictions
- One of the two inputs of the add operator must have a dimension of 1.
- The bias dimension is the same as the n dimensions of the matmul or matmulv2 output.
- The output dimension of matmul/matmulv2 is 2.
Parent topic: Graph Fusion Patterns