llm-batch-processor 0.1.4

A package to process CSV text data in batches using OpenAI API

LLM Batch Processor 🚀

A Python package to process CSV text data in batches using OpenAI's GPT API.

---

📌 Features

✅ Processes large CSV files (50K+ rows) in batches of 10 rows
✅ Uses OpenAI GPT-4 API (for now) to generate responses for each row
✅ Saves the results in a new CSV column
✅ CLI support (llm-process) for easy execution
✅ Modular and scalable package design

---

📂 Project Structure

llm_batch_processor/
│── llm_batch_processor/  # The package
│   │── __init__.py
│   │── config.py         # Configuration settings
│   │── api_client.py     # Handles OpenAI API requests
│   │── processor.py      # Batch processing logic
│── scripts/
│   │── main.py           # CLI entry point
│── data/                 # Place CSV files here
│── setup.py              # Package setup
│── setup.cfg
│── pyproject.toml
│── requirements.txt
│── README.md
│── .gitignore
│── LICENSE

---

🛠 Installation with repo

1️⃣ Clone the repository

git clone https://github.com/karthikRavichandran/llm_batch_processor.git
cd llm_batch_processor

Or

2️⃣ Install the package

pip install llm-batch-process

3️⃣ Set up OpenAI API Key

Create a .env file in the root directory and add:

OPENAI_API_KEY=your_api_key_here
Project_Id=your_project_id #optional

or

export OPENAI_API_KEY="your_api_key_here"
export Project_Id="your_project_id" #Optional 

---

🚀 Usage

1️⃣ Prepare your CSV file

• Place your input CSV file in the data/ folder or mention the location using --input_csv
• Ensure it has a column containing text (default: "text") or mention the column name using --text_column
• Place your prompt in prompt.txt or mention it using --prompt_file

2️⃣ Run the batch processor

llm-process 

CLI args

  --api_key API_KEY     OpenAI API Key (default: from environment variable)
  --project_id PROJECT_ID
                        Project ID (default: from environment variable)
  --batch_size BATCH_SIZE
                        Batch size for processing (default: 10)
  --input_csv INPUT_CSV
                        Path to input CSV file (default: data/input.csv)
  --output_csv OUTPUT_CSV
                        Path to output CSV file (default: data/output.csv)
  --text_column TEXT_COLUMN
                        Column name containing input text (default: text)
  --response_column RESPONSE_COLUMN
                        Column name for storing responses (default: response)
  --prompt_file PROMPT_FILE
                        Path to the text file containing the prompt

This will:

• Read data from data/input.csv
• Process text in batches of 10 rows
• Send each batch to OpenAI GPT-4
• Save responses in data/output.csv under a new column "response"

---

⚙️ Configuration

Modify config.py for:

BATCH_SIZE = 10  # Number of rows per batch
INPUT_CSV = "data/input.csv"
OUTPUT_CSV = "data/output.csv"
TEXT_COLUMN = "text"
RESPONSE_COLUMN = "response"

---

🛠 Development

Install dependencies

pip install -r requirements.txt

Run tests

(TODO: Add unit tests)

pytest

---

📦 Publish to PyPI

1️⃣ Build the package

python setup.py sdist bdist_wheel

2️⃣ Upload to PyPI

twine upload dist/*

3️⃣ Install from PyPI

pip install llm_batch_processor

---

📜 License

This project is licensed under the MIT License.

---

💡 Contributing

1. Fork the repo
2. Create a feature branch (git checkout -b feature-branch)
3. Commit changes (git commit -m "Added new feature")
4. Push to branch (git push origin feature-branch)
5. Create a Pull Request 🚀

---

📧 Contact

For support or collaboration, reach out via Karthik Ravichandran.

---