Streamline MLOps with ZenML: Seamless Integration with Hugging Face

from zenml import pipeline, step
from zenml.integrations.huggingface import deploy_to_hub

@step
def preprocess_data(raw_data):
    # Preprocess the raw data
    preprocessed_data = ...
    return preprocessed_data

@step
def train_model(preprocessed_data):
    # Train the model using Hugging Face's pre-trained models
    model = ...
    return model

@step
def evaluate_model(model, test_data):
    # Evaluate the model performance
    metrics = ...
    return metrics

@pipeline
def ml_pipeline(raw_data, test_data):
    preprocessed_data = preprocess_data(raw_data)
    model = train_model(preprocessed_data)
    metrics = evaluate_model(model, test_data)
    deploy_to_hub(model)  # Deploy the model to Hugging Face Hub

# Run the pipeline
ml_pipeline(raw_data, test_data)

from transformers import AutoModelForSequenceClassification, AutoTokenizer

# Load a pre-trained model from Hugging Face Hub
model_name = "distilbert-base-uncased-finetuned-sst-2-english"
model = AutoModelForSequenceClassification.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)

# Use the model for inference
text = "This movie was fantastic!"
inputs = tokenizer(text, return_tensors="pt")
outputs = model(**inputs)
sentiment = outputs.logits.argmax().item()

print("Sentiment:", "Positive" if sentiment == 1 else "Negative")