RL Quant

DNN mixed-precision quantization with reinforcement learning, derived from HAQ.

ReRAM Accelerator Mixed-Precision Quantization

Running RL Search

Use run_reram.sh to run RL search.

./run_reram.sh

This will call reram_search.py and start RL search. reram_search.py script calls comp_reram/comp_reram.py script to evaluate the performance a certain quantization scheme.

Evaluating Quantization Configuration

To evaluate the loss or accuracy of a specific quantization configuration, use comp_reram/comp_reram.py script.

python3 comp_reram/comp_reram.py

Testing Environment

• Ubuntu 18.04.4 LTS
• Python 3.6.9
• PyTorch 1.4.0
• GPU: NVIDIA Tesla P100