Skip to content

Commit

Permalink
initial commit for submission
Browse files Browse the repository at this point in the history
  • Loading branch information
HUJH511 committed Apr 12, 2024
1 parent bcba7ba commit 5c86801
Show file tree
Hide file tree
Showing 31 changed files with 9,262 additions and 0 deletions.
18 changes: 18 additions & 0 deletions VLSI24/submitted_notebooks/LearnAFE/.gitignore
Original file line number Diff line number Diff line change
@@ -0,0 +1,18 @@
# ---------- User-specified files to ignore ----------
# Edit the following if necessary
.conda/*
AFE_Config/SpiceAC/*
!AFE_Config/SpiceAC/README.md
AFE_Config/SpiceTrans/*
!AFE_Config/SpiceTrans/README.md
SpeechCommands/*
!SpeechCommands/README.md
ckpts/*
!ckpts/README.md
logs/*
!logs/README.md
*.pyc
*.pt
*.pth
*.tar.gz
# ---------- End of user-specified list of files to ignore ----------
Original file line number Diff line number Diff line change
@@ -0,0 +1,7 @@
C1_W 22
C2_W 60
I1 300
I2 100
scale 1.298
C_spk 5
Vth 1
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
### Directory to place initial design file
10 changes: 10 additions & 0 deletions VLSI24/submitted_notebooks/LearnAFE/AFE_Config/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
# Directory to place configurations of analog front-end

```
./LearnAFE_KWS/AFE_Config
|
├───Design
│ └───DSF16.txt
├───SpiceAC
└───SpiceTrans
```
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
### Directory to place spice files
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
### Directory to place spice files
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
21 changes: 21 additions & 0 deletions VLSI24/submitted_notebooks/LearnAFE/LICENSE
Original file line number Diff line number Diff line change
@@ -0,0 +1,21 @@
MIT License

Copyright (c) 2024 HU JINHAI

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
7,471 changes: 7,471 additions & 0 deletions VLSI24/submitted_notebooks/LearnAFE/LearnAFE.ipynb

Large diffs are not rendered by default.

126 changes: 126 additions & 0 deletions VLSI24/submitted_notebooks/LearnAFE/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,126 @@
# LearnAFE: A Learnable Analog Front-End for Keyword Spotting
This repository contains code for keyword spotting with learnable analog front-end.


## Table of Content
1. [Structure](#structure)
2. [Getting Started](#getting-started)
3. [Customization](#customization)
4. [Architecture](#architecture)
5. [Results](#results)
6. [License](#license)
7. [Acknowledgement](#acknowledgement)


## Structure
While it is possible to have a stand-alone notebook, we have decided to split up the code into modules to improve readability and documentation. The following outline the directory structure.
```
./LearnAFE_KWS/
|
├───AFE_Config
│ ├───Design
│ ├───SpiceAC
│ └───SpiceTrans
├───ckpts
├───Figures
├───logs
├───SpeechCommands
│ └───speech_commands_v0.02
├───src
│ ├───data
│ ├───models
│ ├───utils
│ ├───train.py
│ └───test.py
├───.gitignore
├───environment.yaml
├───README.md
└───LearnAFE_KWS.ipynb
```


## Getting Started

### Environment Setup
```
./home/
|
├───EDA_Tools
│ ├───magic
│ └───open_pdks
└───LearnAFE_KWS
```
Suggest to setup following environment under `EDA_Tools`
```
%%install ngspice and other possible configuration files
cd /home/EDA_Tools
sudo pip3 install flake8
sudo pip3 install setuptools-scm
sudo apt update
sudo apt install build-essential tcl8.6 tcl8.6-dev tk8.6 tk8.6-dev libx11-dev libcairo2-dev libxpm-dev
sudo apt update
sudo apt install flex
sudo apt install bison
sudo apt install libcairo2-dev libjpeg-dev
sudo apt install ngspice
sudo apt install gwave
sudo apt-get install xterm
sudo apt-get install vim-gtk
sudo apt-get install build-essential tcl-dev tk-dev libcairo2-dev libncurses-dev libx11-dev
git clone https://github.com/RTimothyEdwards/magic.git
cd magic
./configure
make
sudo make install
```
Open PDK Installation
```
%%download and install open pdk
cd /home/EDA_Tools
git clone https://github.com/RTimothyEdwards/open_pdks
cd open_pdks
./configure --enable-sky130-pdk --enable-sram-sky130
make
sudo make install
make veryclean
```

%% Python Setup
The `environment.yml` can be used as a reference for the packages required. Before running the notebook, ensure that the circuit-design tools and the required packages have been installed successfully. Additionally, create a textfile in SpeechCommands -> speech_commands_v0.02 -> noise_list.txt with the following content. This is used to add background noise to the data for a more robust result.
```
%% noise_list.txt
_background_noise_/doing_the_dishes.wav
_background_noise_/dude_miaowing.wav
_background_noise_/exercise_bike.wav
_background_noise_/pink_noise.wav
_background_noise_/running_tap.wav
_background_noise_/white_noise.wav
```


## Customization
- Code-a-Chip
| Data Augmentation | Filter Version | SR | Hop length | Overlap | Max samples | Model | Freeze |
|----------------------|----------------|-----|------------|---------|-------------|----------------|--------|
| Resample + Noise | DSF16 | 20k | 640 | False | 30 | LearnAFE_v2 | False |


## Architecture
- Learnable filter-bank on time domain

Waveform --> `START TRAIN` --> { Time domain filter } --> Filtered waveform --> { Spiking + Framing } --> Feature Map --> { DSCNN } --> Classification Result


## Results
| Data Augmentation | Model | AFE Config | Freeze | SR | Hop_len | Overlap | Train Acc | Valid Acc | Test Acc | 20dB | 5dB |
|-------------------|---------------|----------------|--------|------|---------|---------|-----------|-----------|----------|------|------|
| Resample + Noise | LearnAFE_v2 | Design/DSF16 | True | 20k | 640 | False | 94.0 | 91.3 | 90.2 | 89.1 | 82.7 |
| Resample + Noise | LearnAFE_v2 | Design/DSF16 | False | 20k | 640 | False | 95.6 | 93.8 | 92.7 | 92.4 | 89.4 |


## License
This project is licensed under the MIT License. See LICENSE for more details.


## Acknowledgement
This work was supported by the Agency for Science, Technology and Research (A*STAR), Singapore under the Nanosystems at the Edge programme, grant No. A18A1b0055. We thank Professor Zhengya Zhang for his insightful comments to strengthen this work.
12 changes: 12 additions & 0 deletions VLSI24/submitted_notebooks/LearnAFE/SpeechCommands/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,12 @@
# Directory to place Speech Commands Dataset

A noise list file is added under `SpeechCommands/speech_commands_v0.02`
```
%% noise_list.txt
_background_noise_/doing_the_dishes.wav
_background_noise_/dude_miaowing.wav
_background_noise_/exercise_bike.wav
_background_noise_/pink_noise.wav
_background_noise_/running_tap.wav
_background_noise_/white_noise.wav
```
1 change: 1 addition & 0 deletions VLSI24/submitted_notebooks/LearnAFE/ckpts/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
# Directory to place temporary models/checkpoints
181 changes: 181 additions & 0 deletions VLSI24/submitted_notebooks/LearnAFE/environment.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,181 @@
name: /home/LearnAFE_KWS/.conda
channels:
- conda-forge
- defaults
dependencies:
- _libgcc_mutex=0.1
- _openmp_mutex=4.5
- asttokens=2.4.1
- blas=1.0
- brotli=1.0.9
- brotli-bin=1.0.9
- bzip2=1.0.8
- ca-certificates=2024.3.11
- comm=0.2.2
- contourpy=1.2.0
- cycler=0.11.0
- cyrus-sasl=2.1.28
- dbus=1.13.18
- debugpy=1.6.7
- decorator=5.1.1
- exceptiongroup=1.2.0
- executing=2.0.1
- expat=2.5.0
- fontconfig=2.14.1
- fonttools=4.25.0
- freetype=2.12.1
- glib=2.78.4
- glib-tools=2.78.4
- gst-plugins-base=1.14.1
- gstreamer=1.14.1
- icu=73.1
- importlib-metadata=7.1.0
- importlib_metadata=7.1.0
- ipykernel=6.29.3
- ipython=8.22.2
- jedi=0.19.1
- jpeg=9e
- jupyter_client=8.6.1
- jupyter_core=5.7.2
- kiwisolver=1.4.4
- krb5=1.20.1
- lcms2=2.12
- ld_impl_linux-64=2.38
- lerc=3.0
- libbrotlicommon=1.0.9
- libbrotlidec=1.0.9
- libbrotlienc=1.0.9
- libclang=14.0.6
- libclang13=14.0.6
- libcups=2.4.2
- libdeflate=1.17
- libedit=3.1.20230828
- libffi=3.4.4
- libgcc-ng=13.2.0
- libgfortran-ng=11.2.0
- libgfortran5=11.2.0
- libglib=2.78.4
- libgomp=13.2.0
- libiconv=1.16
- libllvm14=14.0.6
- libopenblas=0.3.21
- libpng=1.6.39
- libpq=12.17
- libsodium=1.0.18
- libstdcxx-ng=11.2.0
- libtiff=4.5.1
- libuuid=1.41.5
- libwebp-base=1.3.2
- libxcb=1.15
- libxkbcommon=1.0.1
- libxml2=2.10.4
- lz4-c=1.9.4
- matplotlib=3.8.0
- matplotlib-base=3.8.0
- matplotlib-inline=0.1.6
- munkres=1.1.4
- mysql=5.7.24
- ncurses=6.4
- nest-asyncio=1.6.0
- numpy=1.26.4
- numpy-base=1.26.4
- openjpeg=2.4.0
- openssl=3.2.1
- packaging=24.0
- parso=0.8.3
- pcre2=10.42
- pexpect=4.9.0
- pickleshare=0.7.5
- pillow=10.2.0
- pip=23.3.1
- platformdirs=4.2.0
- ply=3.11
- prompt-toolkit=3.0.42
- psutil=5.9.8
- ptyprocess=0.7.0
- pure_eval=0.2.2
- pygments=2.17.2
- pyparsing=3.0.9
- pyqt=5.15.10
- pyqt5-sip=12.13.0
- python=3.11.8
- python-dateutil=2.9.0
- python_abi=3.11
- pyzmq=25.1.2
- qt-main=5.15.2
- readline=8.2
- setuptools=68.2.2
- sip=6.7.12
- six=1.16.0
- sqlite=3.41.2
- stack_data=0.6.2
- tk=8.6.12
- tornado=6.4
- traitlets=5.14.2
- typing_extensions=4.10.0
- wcwidth=0.2.13
- wheel=0.41.2
- xz=5.4.6
- zeromq=4.3.5
- zipp=3.17.0
- zlib=1.2.13
- zstd=1.5.5
- pip:
- ax-platform==0.3.7
- botorch==0.10.0
- cffi==1.16.0
- cmake==3.29.0.1
- filelock==3.13.3
- fsspec==2024.3.1
- gpytorch==1.11
- ipywidgets==8.1.2
- jaxtyping==0.2.28
- jinja2==3.1.3
- joblib==1.3.2
- jupyterlab-widgets==3.0.10
- linear-operator==0.5.1
- lit==18.1.2
- markupsafe==2.1.5
- mpmath==1.3.0
- multipledispatch==1.0.0
- mypy-extensions==1.0.0
- networkx==3.2.1
- nvidia-cublas-cu12==12.1.3.1
- nvidia-cuda-cupti-cu12==12.1.105
- nvidia-cuda-nvrtc-cu12==12.1.105
- nvidia-cuda-runtime-cu12==12.1.105
- nvidia-cudnn-cu12==8.9.2.26
- nvidia-cufft-cu12==11.0.2.54
- nvidia-curand-cu12==10.3.2.106
- nvidia-cusolver-cu12==11.4.5.107
- nvidia-cusparse-cu12==12.1.0.106
- nvidia-nccl-cu12==2.19.3
- nvidia-nvjitlink-cu12==12.4.99
- nvidia-nvtx-cu12==12.1.105
- opt-einsum==3.3.0
- pandas==2.2.1
- plotly==5.20.0
- pycparser==2.21
- pyre-extensions==0.0.30
- pyro-api==0.1.2
- pyro-ppl==1.9.0
- pysoundfile==0.9.0.post1
- pytz==2024.1
- scikit-learn==1.4.1.post1
- scipy==1.12.0
- soundfile==0.12.1
- sympy==1.12
- tabulate==0.9.0
- tenacity==8.2.3
- threadpoolctl==3.4.0
- torch==2.2.1
- torchaudio==2.2.1
- torchinfo==1.8.0
- torchvision==0.17.1
- tqdm==4.66.2
- triton==2.2.0
- typeguard==2.13.3
- typing-inspect==0.9.0
- tzdata==2024.1
- widgetsnbextension==4.0.10
prefix: /home/LearnAFE_KWS/.conda
Loading

0 comments on commit 5c86801

Please sign in to comment.