Added tf dtype generics support in eager mode. #19588
Conversation
This is done by extending IRGen to use the new AnyTensorHandle based compiler rt entry points.
|
@swift-ci please test tensorflow |
|
This is great, it's awesome to see it all come together!! |
| @@ -43,6 +43,46 @@ void *swift_tfc_CreateScalarFloatTensor(int32_t val) { | |||
| return tensor; | |||
| } | |||
|
|
|||
| void *swift_tfc_CreateScalarIntTensor(int64_t val, int32_t dtype, | |||
There was a problem hiding this comment.
I'm not sure why ints and floats have to be in different code paths. Can you explain a bit more, and also add some doc comments at the declaration site?
There was a problem hiding this comment.
The float version should take as input a float or double value, instead of int64_t val, so there should be different C APIs for them.
BTW, currently we only have integer related data types like IGM.Int32Ty, but there is no IGM.FloatTy. That can be added though.
There was a problem hiding this comment.
It would be useful to add doc documents, because these arguments' types do not directly reflect the underlying data type, and would require the call and the function implementation to cast things back and forth.
…API TF_MakeInternalErrorStatus().
|
@swift-ci please clean test tensorflow |
|
Thanks for the review! CI passed, so I'm merging. |
This is done by extending IRGen to use the new AnyTensorHandle based compiler rt
entry points.