This folder contains the official implementation of paper It's All In the Teacher: Zero-Shot Quantization Brought Closer to the Teacher on GDFQ, Qimera, AutoReCon framework.
We recommend using Python virtual environment to run this code.
You can install requirements with the command below.
pip install -r requirements.txt
ait_code
├── figs
├── AutoReCon_AIT
│ ├── main.py
│ ├── optimizer.py # GI implementation
│ ├── option.py
│ ├── trainer.py
│ ├── {DATASET}_{NETWORK}.hocon # Setting files
│ ├── run_{DATASET}_{NETWORK}_{BITWIDTH}bit.sh # Train scripts
│ ├── trainer.py
│ └── ... # Utils
├── GDFQ_AIT
│ └── ... # Similar to above
├── Qimera_AIT
│ └── ... # Similar to above
├── LICENSE.md
├── README.md
└── requirements.txt
For Imagenet training, change the path of the validation set in .hocon file. To train the model described in the paper, run one of this command:
./run_cifar10_4bit.sh
./run_cifar100_4bit.sh
./run_imgnet_resnet18_4bit.sh
./run_imgnet_resnet50_4bit.sh
./run_imgnet_mobilenet_v2_4bit.sh
The script name is same for all experiment framework.
This project is licensed under the terms of the GNU General Public License v3.0