Merge tutorials, add tritony-sample

Author: kimdwkimdw

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.5.0
+    rev: v4.6.0
     hooks:
       - id: check-json
       - id: end-of-file-fixer
@@ -11,7 +11,7 @@ repos:
       - id: check-added-large-files
         args: [--maxkb=4096]
   - repo: https://github.com/psf/black
-    rev: 24.2.0
+    rev: 24.3.0
     hooks:
       - id: black
         args: [--line-length=120]
@@ -27,3 +27,9 @@ repos:
       - id: flake8
         types: [python]
         args: ["--max-line-length", "88", "--ignore", "E203,E501,W503,W504"]
+  - repo: https://github.com/shellcheck-py/shellcheck-py
+    rev: v0.10.0.1
+    hooks:
+      - id: shellcheck
+        exclude: test
+        args: ["--exclude", "SC1091,SC1073,SC2086,SC2128"]

README.md

@@ -1,17 +1,26 @@
-# RTZR Streaming STT API(gRPC) with Mic. Interface
+# Python Audio Processing Tutorials
 
-## Requirements
-This repository need `PyAudio` library. First, you need to read installation instruction `PyAudio` [this link](https://pypi.org/project/PyAudio/). 
+Welcome to the Python Audio Processing Tutorials repository. This repository contains two main projects focusing on real-time audio processing using different technologies and frameworks.
 
-And Then,
+## Projects
+
+### 1. RTZR Streaming STT API(gRPC) with Mic. Interface
+
+This project demonstrates how to set up a real-time streaming Speech-to-Text (STT) API using gRPC with microphone interface capabilities. It requires the installation of the `PyAudio` library and additional setup for gRPC communication.
+
+- **Requirements**: PyAudio (see [installation instructions](https://pypi.org/project/PyAudio/)).
+- **Setup**: Download necessary `.proto` files and generate gRPC client code. For details, check the project's [README](./python-stt-sample/).
+
+### 2. Triton, Tritony VAD Client Sample
+
+This project showcases the use of the Triton Inference Server and Tritony Voice Activity Detection (VAD) to process and analyze audio streams effectively. It involves setting up a Docker container and running a Triton server.
+
+- **Requirements**: pydub, tritonclient, tritony
+- **Setup**: Build and run a Docker image for the Triton Inference Server. Testing is facilitated through predefined scripts and pytest. For more information, visit the project's [README](./tritony-sample/).
+
+## General Installation
+
+Most project dependencies can be installed via pip:
 
 ```bash
 pip install -r requirements.txt
-```
- Download `.proto` file from this [Link](https://github.com/vito-ai/openapi-grpc/blob/main/protos/vito-stt-client.proto). And save `.proto` file in the project directory.
- 
- And run this command to generate grpc file.
- ```bash
- python -m grpc_tools.protoc -I. --python_out=./src --grpc_python_out=./src ./vito-stt-client.proto
- ```
- 

python-stt-sample/README.md

@@ -0,0 +1,17 @@
+# RTZR Streaming STT API(gRPC) with Mic. Interface
+
+## Requirements
+This repository need `PyAudio` library. First, you need to read installation instruction `PyAudio` [this link](https://pypi.org/project/PyAudio/). 
+
+And Then,
+
+```bash
+pip install -r requirements.txt
+```
+ Download `.proto` file from this [Link](https://github.com/vito-ai/openapi-grpc/blob/main/protos/vito-stt-client.proto). And save `.proto` file in the project directory.
+ 
+ And run this command to generate grpc file.
+ ```bash
+ python -m grpc_tools.protoc -I. --python_out=./src --grpc_python_out=./src ./vito-stt-client.proto
+ ```
+ 

config.ini python-stt-sample/config.iniconfig.ini renamed to python-stt-sample/config.ini

@@ -28,10 +28,9 @@
 
 import grpc
 import soundfile as sf
-from requests import Session
-
 import vito_stt_client_pb2 as pb
 import vito_stt_client_pb2_grpc as pb_grpc
+from requests import Session
 
 API_BASE = "https://openapi.vito.ai"
 GRPC_SERVER_URL = "grpc-openapi.vito.ai:443"

requirements.txt python-stt-sample/requirements.txtrequirements.txt renamed to python-stt-sample/requirements.txt

@@ -7,10 +7,9 @@
 
 import grpc
 import pyaudio
-from requests import Session
-
 import vito_stt_client_pb2 as pb
 import vito_stt_client_pb2_grpc as pb_grpc
+from requests import Session
 
 API_BASE = "https://openapi.vito.ai"
 GRPC_SERVER_URL = "grpc-openapi.vito.ai:443"

src/sample_stt.py python-stt-sample/src/sample_stt.pysrc/sample_stt.py renamed to python-stt-sample/src/sample_stt.py

@@ -0,0 +1,10 @@
+FROM nvcr.io/nvidia/tritonserver:23.08-py3
+
+RUN apt update && apt install -y --no-install-recommends \
+    libsndfile1 \
+    && apt clean \
+    && rm -rf /var/lib/apt/lists/*;
+
+RUN pip install --upgrade pip && \
+    pip install torch torchaudio==2.1.2 && \
+    pip install pyannote.audio

src/stream_stt.py python-stt-sample/src/stream_stt.pysrc/stream_stt.py renamed to python-stt-sample/src/stream_stt.py

@@ -0,0 +1,31 @@
+# Triton, Tritony VAD Client sample
+
+## Requirements
+This repository requires `pydub`, `tritonclient`, `tritony` python libraries. So you need to run following command.
+
+```bash
+pip install -r requirements.txt
+```
+
+Next, you can build Docker image using shell script.
+
+```bash
+/bin/bash ./build_triton_server.sh
+```
+or
+
+For manually building Docker image, you must assign your docker image tag `triton-vad-server:latest`.
+
+## Run Triton Inference Server
+To start Triton Inference Server, run the below command.
+You will need to set the `HF_TOKEN` environment variable to your Hugging Face API token. ( https://huggingface.co/docs/hub/security-tokens )
+
+```bash
+export HF_TOKEN=YOUR_HF_TOKEN
+/bin/bash ./bin/run_triton_server.sh
+```
+## Test
+First, you must run Triton Inference Server.
+```bash
+pytest -s
+```

tritony-sample/Dockerfile

@@ -0,0 +1,18 @@
+#!/bin/bash
+
+HERE=$(dirname "$(readlink -f $0)")
+PARENT_DIR=$(dirname "$HERE")
+
+docker run -it --rm --name triton_tritony \
+    -p8100:8000   \
+    -p8101:8001   \
+    -p8102:8002    \
+    -v "${PARENT_DIR}"/model_repository:/model_repository:ro \
+    -e OMP_NUM_THREADS=4 \
+    -e OPENBLAS_NUM_THREADS=4 \
+    -e HF_TOKEN=$HF_TOKEN \
+    --shm-size=4g  \
+    triton-vad-server:latest \
+    tritonserver --model-repository=/model_repository \
+    --exit-timeout-secs 15 \
+    --min-supported-compute-capability 7.0 \