We’ve used the Hugging Face pipeline to understand text (sentiment-analysis) and answer questions (question-answering). Now, let’s use it for its most famous task: Text Generation. In this guide, we’ll focus on Hugging Face Text Generation and show you how to get started.
We will use gpt-2, one of the original famous models, to have our Python script complete our thoughts.
Step 1: Installation
pip install transformers # You also need the 'torch' (PyTorch) backend pip install torch
Step 2: The Code
The pipeline makes this incredibly simple.
from transformers import pipeline
# 1. Load the text-generation pipeline
# This downloads the gpt-2 model (it might take a minute!)
generator = pipeline("text-generation", model="gpt-2")
# 2. Create your prompt
prompt = "Python is the best programming language for"
# 3. Generate the text
# max_length is the total length (prompt + new text)
generated_text = generator(prompt, max_length=50, num_return_sequences=1)
# 4. Print the result
print(generated_text[0]['generated_text'])Step 3: The Result
The model will take your prompt and continue it, often in creative ways. You might get something like:
"Python is the best programming language for data analysis, machine learning, and web development. It is also a very popular language for beginners..."
You can play with this by changing the prompt and increasing the max_length. You are now running a powerful generative AI model right in your Python script!
