Add Convert operator (a.k.a. Transmitter)

assigned to @olivierbichler

added 1 commit

• ce1cbb42 - Updated

Compare with previous version

added 1 commit

• 39cf8659 - Added automatic GPU reallocation

Compare with previous version

added 1 commit

• dd2529ec - Minor adaptations

Compare with previous version

added 1 commit

• 8b9a6658 - Added safeguards

Compare with previous version

added 1 commit

• adb33744 - Added CUDA Convert test

Compare with previous version

added 1 commit

• 1ae93537 - Added Producer impl

Compare with previous version

added 1 commit

• f8eb193f - Adapted Conv impl to use cast/from features

Compare with previous version

added 1 commit

• 4e00cee0 - Fixed unsupported CuDNN type

Compare with previous version

added 1 commit

• 1f0339d8 - Added missing float16 type

Compare with previous version

added 1 commit

• 10f6c6be - Fixed typo

Compare with previous version

added 1 commit

• 1c8e1637 - Added thrust_copy specializations for half float

Compare with previous version

added 1 commit

• 1ccc5bae - Added half support in CudaContext.hpp

Compare with previous version

added 1 commit

• 8536d095 - Fixed types

Compare with previous version

added 1 commit

• dcdee743 - Missing backend

Compare with previous version

added 1 commit

• 2b674142 - Fixed threshold to make test pass

Compare with previous version

added 1 commit

• 594e9898 - Renamed Convert to CastMove

Compare with previous version

added 1 commit

• d30c27b1 - Adaptations to core changes

Compare with previous version

added 1 commit

• 15242a20 - Fixed reviewed issues

Compare with previous version

marked this merge request as ready