bert_conda

Perform Sentiment Analysis with BERT and Qwak

Overview

This project demonstrates how to label sentiment in a text prompt using a pre-trained BERT model with JFrog ML.

It showcases how to:

• Define the QwakModel class
• Initialize the pre-trained BERT model
• Predict phrase sentiment using JFrog ML's API

The code is designed to work seamlessly with JFrog ML and serves as a practical example.

How to Test Locally

1. Clone the Repository: Clone this GitHub repository to your local machine.

2. Install Dependencies: Make sure you have the required dependencies installed, as specified in the conda.yml file.

   conda env create -f ./main/conda.yml
   conda activate bert_sentiment_analysis

3. Install and Configure the JFrog ML SDK: Use your account JFrog ML API Key to set up your SDK locally.

   pip install qwak-sdk
   qwak configure

4. Run the Model Locally: Execute the following command to test the model locally:

   python test_model_locally.py

How to Run Remotely on JFrog ML

1. Build on JFrog ML:

   Create a new model on Qwak using the command:

   qwak models create "BERT Sentiment Analysis" --project "Sample Project"

   Initiate a model build with:

   qwak models build --model-id <your-model-id> ./bert_conda

2. Deploy the Model on JFrog ML with a Real-Time Endpoint:

   To deploy your model via the CLI, use the following command:

   qwak models deploy realtime --model-id <your-model-id> --build-id <your-build-id>

3. Test the Live Model with a Sample Request:

   Install the Qwak Inference SDK:

   pip install qwak-inference

   Call the Real-Time endpoint using your Model ID from JFrog ML:

   python test_live_mode.py <your-qwak-model-id>

Project Structure

.
├── main                   # Main directory containing core code
│   ├── __init__.py        # An empty file that indicates this directory is a Python package
│   ├── model.py           # Defines the Credit Risk Model
│   └── conda.yaml         # Conda environment configurationdata
|
├── test_model_locally.py  # Script to test the model locally
├── test_live_model.py     # Script to test the live model with a sample REST request
└── README.md              # Documentation