thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 4

Added initial implementation for algorithm and opMult

Browse source
This commit is contained in:
Alejandro Saucedo 2020-08-21 19:15:07 +01:00
parent 0d18dc50e6
commit d59dc41ffc
11 changed files with 385 additions and 28 deletions
Download patch file Download diff file Expand all files Collapse all files

2
src/OpCreateTensor.hpp
View file

@ -31,6 +31,8 @@ class OpCreateTensor : public OpBase
void record() override;
void postSubmit() override;
private:
std::shared_ptr<Tensor> mPrimaryTensor;
std::shared_ptr<Tensor> mStagingTensor;