Tuning Assistant for Floating point to Fixed point Optimization
-
Updated
Mar 26, 2022 - C++
Tuning Assistant for Floating point to Fixed point Optimization
Rigorous Floating-Point Mixed-Precision Tuner
FPLearner is a DL-based tool to predict performance and accuracy of mixed-precision programs that can be used in dynamic precision tuners to save time.
POPiX: Floating to Fixed-Point Code Generator
Recurrence-aware static value range analysis for TAFFO, developed with LLVM.
Reinforcement learning for precision tuning
RLTuner is a tool to identify optimal mixed-precision configurations using Deep Q-Network(DQN).
To associate your repository with the precision-tuning topic, visit your repo's landing page and select "manage topics."