| import gradio as gr |
| import hopsworks |
| import joblib |
| import pandas as pd |
| import numpy as np |
| project = hopsworks.login() |
| fs = project.get_feature_store() |
|
|
| mr = project.get_model_registry() |
| model = mr.get_model("iris_model", version=1) |
| model_dir = model.download() |
| model = joblib.load(model_dir + "/iris_model.pkl") |
| feature_view = fs.get_feature_view(name="iris", version=1) |
| batch_data = feature_view.get_batch_data() |
|
|
| def greet(sep_length, sep_width, pet_length, pet_width): |
| df = pd.DataFrame({ "sepal_length": sep_length, "sepal_width": sep_width, "petal_length": pet_length, "petal_width": pet_width}, index=[0]) |
| prediction = model.predict(df)[0] |
| return prediction, f"Images\{prediction}.jpg" |
|
|
| predict = gr.Interface( |
| fn=greet, |
| inputs=[ |
| gr.Textbox(placeholder="Sepal length here", label = "Sepal Length"), |
| gr.Textbox(placeholder="Sepal width here", label = "Sepal Width"), |
| gr.Textbox(placeholder="Petal length here", label = "Petal Length"), |
| gr.Textbox(placeholder="Petal width here", label = "Petal Width")], |
| outputs=[gr.Textbox(label = "Prediction"), gr.Image(type="pil", label="Iris")], |
| ) |
| predict.launch() |
