https://github.com/furk4neg3/ibm_hf-model-loading-inference

A practical guide on loading and performing inference with Hugging Face models. Demonstrates both manual and pipeline-based approaches for text classification and generation, completed as part of IBM’s Generative AI Engineering with LLMs Specialization course.
https://github.com/furk4neg3/ibm_hf-model-loading-inference

Last synced: about 2 months ago
JSON representation

A practical guide on loading and performing inference with Hugging Face models. Demonstrates both manual and pipeline-based approaches for text classification and generation, completed as part of IBM’s Generative AI Engineering with LLMs Specialization course.

• Host: GitHub
• URL: https://github.com/furk4neg3/ibm_hf-model-loading-inference
• Owner: furk4neg3
• Created: 2024-11-10T14:38:38.000Z (about 2 months ago)
• Default Branch: main
• Last Pushed: 2024-11-10T14:41:50.000Z (about 2 months ago)
• Last Synced: 2024-11-10T15:32:31.548Z (about 2 months ago)
• Language: Jupyter Notebook
• Size: 0 Bytes
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Loading Models and Inference with Hugging Face

This project explores how to load and perform inference with pretrained models using the Hugging Face `transformers` library. Using models like DistilBERT and GPT-2, we carry out text classification and text generation tasks. This project also highlights the simplicity of the Hugging Face `pipeline()` function for efficient NLP model inference.

## Overview

This lab project demonstrates:

- Loading and tokenizing input text for model inference

- Performing text classification and text generation using Hugging Face models

- Comparing traditional inference with the Hugging Face `pipeline()` function for streamlined deployment

## Table of Contents

1. [Introduction to Model Loading](#introduction-to-model-loading)

2. [Text Classification with DistilBERT](#text-classification-with-distilbert)

3. [Text Generation with GPT-2](#text-generation-with-gpt-2)

4. [Using the Hugging Face Pipeline](#using-the-hugging-face-pipeline)

5. [Requirements](#requirements)

6. [References](#references)

## Introduction to Model Loading

In this section, we explore loading pretrained models and tokenizing text inputs for NLP tasks. This approach enables us to use models like DistilBERT and GPT-2 without relying solely on the pipeline method.

## Text Classification with DistilBERT

Using DistilBERT, we classify text into categories. This part demonstrates setting up the model, tokenizing inputs, and processing classification outputs.

## Text Generation with GPT-2

In this section, we use the GPT-2 model to generate text. We cover input preparation, running inference, and processing generated sequences.

## Using the Hugging Face Pipeline

The `pipeline()` function in Hugging Face simplifies model loading and inference by providing an easy-to-use API for common NLP tasks like classification and generation.

## Requirements

- Python 3.7+

- Hugging Face Transformers Library

- PyTorch

  ## References

- [IBM AI Engineering Professional Certificate](https://www.coursera.org/professional-certificates/ai-engineer?)

- [Generative AI Engineering with LLMs Specialization](https://www.coursera.org/specializations/generative-ai-engineering-with-llms)

- [Generative AI Engineering and Fine-Tuning Transformers](https://www.coursera.org/learn/generative-ai-engineering-and-fine-tuning-transformers?specialization=generative-ai-engineering-with-llms)