[mlgo][nfc] Make `LoggedFeatureSpec` an implementation detail
authorMircea Trofin <mtrofin@google.com>
Wed, 30 Nov 2022 18:03:45 +0000 (10:03 -0800)
committerMircea Trofin <mtrofin@google.com>
Wed, 30 Nov 2022 23:57:58 +0000 (15:57 -0800)
commit1ee3bb17c39579de21ea0bd526e79bb932b8b1c3
tree6b369368ec7a6dc775f52af9f8446e7ac172b65c
parent459ad97221fcdfa41d647a11b0ff02b5133e27c6
[mlgo][nfc] Make `LoggedFeatureSpec` an implementation detail

It's an artifact very specific to using TFAgents during training, so it
belongs with ModelUnderTrainingRunner.

Differential Revision: https://reviews.llvm.org/D139031
13 files changed:
llvm/include/llvm/Analysis/ModelUnderTrainingRunner.h
llvm/include/llvm/Analysis/TensorSpec.h
llvm/include/llvm/Analysis/Utils/TFUtils.h
llvm/include/llvm/Analysis/Utils/TrainingLogger.h
llvm/lib/Analysis/DevelopmentModeInlineAdvisor.cpp
llvm/lib/Analysis/ModelUnderTrainingRunner.cpp
llvm/lib/Analysis/TFLiteUtils.cpp
llvm/lib/Analysis/TensorSpec.cpp
llvm/lib/Analysis/TrainingLogger.cpp
llvm/lib/CodeGen/MLRegallocEvictAdvisor.cpp
llvm/lib/CodeGen/MLRegallocPriorityAdvisor.cpp
llvm/unittests/Analysis/TFUtilsTest.cpp
llvm/unittests/Analysis/TrainingLoggerTest.cpp