AI — Hugging Face

What is Hugging Face?

Platform for sharing and using ML models.

Install

pip install transformers datasets

Use Pre-trained Model

from transformers import pipeline

# Sentiment analysis
classifier = pipeline("sentiment-analysis")
result = classifier("I love this!")

# Text generation
generator = pipeline("text-generation", model="gpt2")
text = generator("Once upon a time")

Fine-tuning

from transformers import AutoModelForSequenceClassification, Trainer

model = AutoModelForSequenceClassification.from_pretrained("bert-base")
trainer = Trainer(model=model, train_dataset=dataset)
trainer.train()

Datasets

from datasets import load_dataset

dataset = load_dataset("imdb")

Mini Practice

1. Install Hugging Face
2. Use pre-trained models
3. Fine-tune a model
4. Explore datasets

Up Next

Continue with Model Training — training models.