Switched to a pip installable package with automated releases

Author: cVarun26

.github/workflows/release.yml

@@ -0,0 +1,33 @@
+name: Build and Release
+
+on:
+  push:
+    tags:
+      - 'v*'  # Triggers on version tags like v1.0.0
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.9'
+
+      - name: Install build tools
+        run: |
+          python -m pip install --upgrade pip
+          pip install build
+
+      - name: Build package
+        run: python -m build
+
+      - name: Upload wheel to GitHub Release
+        uses: softprops/action-gh-release@v1
+        with:
+          files: dist/*.whl
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }} 

README.md

@@ -1,23 +1,55 @@
-# Benchmarking Tool for NLP Models
+# BenchTool
 
-A lightweight tool to benchmark the performance of NLP models based on medical text datasets. It supports HuggingFace, ONNX, and ORT models and works with input data from CSV files or a PostgreSQL database.
-
----
+A lightweight tool to benchmark the performance of NLP models based on medical text datasets. Supports HuggingFace, ONNX, and ORT models, and works with input data from CSV files or a PostgreSQL database.
 
+## Features
 - Compare multiple NLP models
 - Supports HuggingFace, ONNX, and ORT model types
 - Input from CSV or PostgreSQL database
 - Outputs performance metrics like inference time and throughput
 - Configurable via a single YAML file
 
----
+## Installation
+
+Download the latest release wheel from the [GitHub Releases page](https://github.com/Holmusk/BenchMark/releases) and install with pip:
+
+```sh
+pip install https://github.com/Holmusk/BenchMark/releases/download/v1.0.0/benchtool-1.0.0-py3-none-any.whl
+```
+
+## Usage
+
+Prepare your `config.yml` and input CSV in your working directory. Then run:
+
+```sh
+benchtool
+```
+
+## Example `config.yml`
+
+```yaml
+model:
+  name: "dbmdz/bert-large-cased-finetuned-conll03-english"
+  task: ""
 
-##  Installation
+input:
+  mode: "csv"
+  csv:
+    path: "100notes.csv"
+    column: "note_text"
 
-```git clone https://github.com/Holmusk/BenchMark.git```
-```cd your-directory```
+benchmark:
+  batch_size: 16
+  runs: 2
+  warmup_runs: 1
+  save_output: true
+  output_path: "./benchmark_results.csv"
 
-create a virtual environment: ```python3 -m venv Enviornment```\
-activate:  ```source Enviornment/bin/activate```
+system:
+  device: "cpu"
+  num_threads: 8
+```
 
-To install dependenices: ```pip install -r requirements.txt```
+## Notes
+- Make sure your `config.yml` and input CSV are in the directory where you run `benchtool`.
+- For more details, see the code and comments in the repository.

bench

@@ -0,0 +1,24 @@
+from setuptools import setup, find_packages
+
+setup(
+    name='benchtool',
+    version='1.0.0',
+    description='A lightweight tool to benchmark the performance of NLP models based on medical text datasets.',
+    author='Holmusk',
+    author_email='varun.c@holmusk.com',
+    packages=find_packages(include=['utils', 'utils.*']),
+    py_modules=['main', 'benchmark'],
+    install_requires=[
+        r.strip() for r in open('requirements.txt').readlines() if r.strip() and not r.startswith('#')
+    ],
+    entry_points={
+        'console_scripts': [
+            'benchtool = main:main',
+        ],
+    },
+    include_package_data=True,
+    package_data={
+        '': ['config.yml']
+    },
+    python_requires='>=3.7',
+)