Subiendo app

app.py

@@ -0,0 +1,64 @@
+# Import the necessary libraries
+import gradio as gr  # Gradio is a library to quickly build and share demos for ML models
+import joblib        # joblib is used here to load the trained model from a file
+import numpy as np   # NumPy for numerical operations (if needed for array manipulation)
+from huggingface_hub import hf_hub_download
+
+HF_TOKEN = 'hf_your_token_here'  # Replace with your actual Hugging Face token
+
+# Replace with your actual Hugging Face model repo ID and file names
+# For example, repo_id="username/iris-decision-tree"
+# Use repo_type="model" if it's a model repository
+model_path = hf_hub_download(
+    repo_id="brjapon/iris-dt",
+    filename="iris_dt.joblib",      # The model file stored in the HF repo
+    repo_type="model"               # Could also be 'dataset' if you're storing it that way
+)
+
+# Load the trained model
+pipeline = joblib.load(model_path)
+
+# Define a function that takes the four iris measurements as input
+# and returns the predicted iris species label.
+def predict_iris(sepal_length, sepal_width, petal_length, petal_width):
+    # Convert the input parameters into a 2D list/array because
+    # scikit-learn's predict() expects a 2D array of shape (n_samples, n_features)
+    input = np.array([[sepal_length, sepal_width, petal_length, petal_width]])
+    prediction = pipeline.predict(input)
+    
+    # Convert the prediction to the string label
+    if prediction == 0:
+        return 'iris-setosa'
+    elif prediction == 1:
+        return 'Iris-versicolor'
+    elif prediction == 2:
+        return 'Iris-virginica'
+    else:
+        return "Invalid prediction"
+
+# Create a Gradio Interface:
+# - fn: the function to call for inference
+# - inputs: a list of component types to collect user input (in this case, four numeric values)
+# - outputs: how the prediction is displayed (in this case, as text)
+# - live: whether to update the output in real-time as the user types
+interface = gr.Interface(
+    fn=predict_iris,
+    inputs=["number", "number", "number", "number"],
+    outputs="text",
+    live=True,
+    title="Iris Species Identifier",
+    description="Enter the four measurements to predict the Iris species."
+)
+
+# Run the interface when this script is executed directly.
+# This will launch a local Gradio server and open a user interface in the browser.
+if __name__ == "__main__":
+    # To create a public link, set the parameter share=True
+    interface.launch()
+    
+'''
+# The Flag button allows users (or testers) to mark or “flag”
+# a particular input-output interaction for later review.
+# When someone clicks Flag, Gradio saves the input values (and often the output) to a log.csv file
+# letting you keep track of interesting or potentially problematic cases for debugging or analysis later on
+'''

requirements.txt

@@ -0,0 +1,5 @@
+scikit-learn
+huggingface_hub
+datasets
+gradio
+joblib