saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-07-11 17:48:34 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

disable computation reduction optimization for non-gpu build (#8251)

Browse source 
* disable computation reduction optimization for non-gpu build

* fix comments in pr

* add cpu execution provider

* apply the core provider list to computation reduction optimizer

* try macro

Co-authored-by: Cheng Tang <chenta@microsoft.com>
This commit is contained in:
Tang, Cheng 2021-07-01 16:43:51 -07:00 committed by GitHub
parent 9cfe642b34
commit a9a2394fa5
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 2 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
orttraining/orttraining/core/optimizer/graph_transformer_utils.cc
View file

@ -111,8 +111,9 @@ std::vector<std::unique_ptr<GraphTransformer>> GeneratePreTrainingTransformers(
execution_provider, false /*skip_dequantize_linear*/, compatible_eps, weights_to_train));
transformers.emplace_back(std::make_unique<ReshapeFusion>(compatible_eps));
transformers.emplace_back(std::make_unique<ConcatSliceElimination>(compatible_eps));
#if defined(USE_CUDA) || defined(USE_ROCM)
transformers.emplace_back(std::make_unique<ComputationReductionTransformer>(compatible_eps));
#endif
if (config.gelu_recompute) {
transformers.emplace_back(std::make_unique<GeluRecompute>());
}