How to beginner · 3 min read

How to use sentiment analysis pipeline Hugging Face

Q: How to use sentiment analysis pipeline Hugging Face

Use Hugging Face's pipeline function with the task set to sentiment-analysis to analyze text sentiment easily. Install transformers and run the pipeline on your input text to get sentiment labels and confidence scores.

Quick answer

Use Hugging Face's pipeline function with the task set to sentiment-analysis to analyze text sentiment easily. Install transformers and run the pipeline on your input text to get sentiment labels and confidence scores.

PREREQUISITES

Python 3.8+
pip install transformers>=4.0.0
pip install torch (or tensorflow, depending on backend)

Setup

Install the transformers library from Hugging Face and a deep learning backend like torch or tensorflow. Set up your Python environment accordingly.

bash

pip install transformers torch

Step by step

Import the pipeline from transformers, create a sentiment analysis pipeline, and pass your text to get sentiment predictions.

python

from transformers import pipeline

# Create sentiment analysis pipeline
sentiment_analyzer = pipeline('sentiment-analysis')

# Analyze sentiment of a sample text
result = sentiment_analyzer('I love using Hugging Face transformers!')
print(result)

output

[{'label': 'POSITIVE', 'score': 0.9998}]

Common variations

Specify a different model by passing model='distilbert-base-uncased-finetuned-sst-2-english' to pipeline.
Use batch inputs by passing a list of texts.
Run asynchronously with asyncio and custom wrappers.

python

from transformers import pipeline

# Use a specific model
sentiment_analyzer = pipeline('sentiment-analysis', model='distilbert-base-uncased-finetuned-sst-2-english')

texts = ["I love this!", "This is terrible."]
results = sentiment_analyzer(texts)
print(results)

output

[{'label': 'POSITIVE', 'score': 0.9998}, {'label': 'NEGATIVE', 'score': 0.9996}]

Troubleshooting

If you get an import error, ensure transformers and torch are installed.
For slow performance, check if a GPU is available and configured.
If you see model download errors, verify your internet connection or use a local cache.

✅

Key Takeaways

Use Hugging Face's pipeline with task sentiment-analysis for quick sentiment detection.
You can specify different pretrained models to customize accuracy and speed.
Batch processing multiple texts improves throughput and efficiency.

Verified 2026-04 · distilbert-base-uncased-finetuned-sst-2-english

Verify ↗