Riccardo Alberghi

Results 3 comments of Riccardo Alberghi

Inconsistency in conv+bn fusion and addition of useless reshapes in tf2onnx==1.16.1

I conducted a little more testing. Downgrading to onnx==1.14 and tf2onnx==1.15.1/37820d solves the issue. With tf2onnx==1.16.0 the issue reappears, I dont know if this helps for identifying the problem.

Inconsistency in conv+bn fusion and addition of useless reshapes in tf2onnx==1.16.1

@fatcat-z any news?

[Performance] CoreML not being used to it's fullest capacity - custom transformer

Any news? I am facing a similar performance issue. LayerNorm and MultiHeadAttention seems not to be implemented as operators in CoreML. Any plans to support them?