app.py

import gradio as gr
from huggingface_hub import HfApi
from datetime import datetime, timedelta
import pandas as pd

# Initialize the Hugging Face API
api = HfApi()

def get_recent_models(min_likes, days_ago, filter_string, search_string):
    # Get the current date and date from `days_ago` days ago
    today = datetime.utcnow().replace(tzinfo=None)
    start_date = (today - timedelta(days=days_ago)).replace(tzinfo=None)
    
    # Initialize an empty list to store the filtered models
    recent_models = []
    
    filter_substrings = [sub.strip().lower() for sub in filter_string.split(';')]
    search_substrings = [term.strip().lower() for term in search_string.split(';')]
    
    # Use a generator to fetch models in batches, sorted by likes in descending order
    for model in api.list_models(sort="likes", direction=-1):
        if model.likes >= min_likes:
            if hasattr(model, "created_at") and model.created_at:
                # Ensure created_at is offset-naive
                created_at_date = model.created_at.replace(tzinfo=None)
                if search_substrings != []:
                    if any(term in model.modelId.lower() for term in search_substrings):
                        if filter_substrings != []:
                            if not any(sub in model.modelId.lower() for sub in filter_substrings):
                                if created_at_date >= start_date:
                                    task = model.pipeline_tag if hasattr(model, "pipeline_tag") else "N/A"
                                    recent_models.append({
                                        "Model ID": f'<a href="https://huggingface.co/{model.modelId}" target="_blank">{model.modelId}</a>',
                                        "Likes": model.likes,
                                        "Creation Date": model.created_at.strftime("%Y-%m-%d %H:%M"),
                                        "Task": task
                                    })
                        else:
                            if created_at_date >= start_date:
                                task = model.pipeline_tag if hasattr(model, "pipeline_tag") else "N/A"
                                recent_models.append({
                                    "Model ID": f'<a href="https://huggingface.co/{model.modelId}" target="_blank">{model.modelId}</a>',
                                    "Likes": model.likes,
                                    "Creation Date": model.created_at.strftime("%Y-%m-%d %H:%M"),
                                    "Task": task
                                })
                else:
                    if filter_substrings != []:
                        if not any(sub in model.modelId.lower() for sub in filter_substrings):
                            if created_at_date >= start_date:
                                task = model.pipeline_tag if hasattr(model, "pipeline_tag") else "N/A"
                                recent_models.append({
                                    "Model ID": f'<a href="https://huggingface.co/{model.modelId}" target="_blank">{model.modelId}</a>',
                                    "Likes": model.likes,
                                    "Creation Date": model.created_at.strftime("%Y-%m-%d %H:%M"),
                                    "Task": task
                                })
                    else:
                        if created_at_date >= start_date:
                            task = model.pipeline_tag if hasattr(model, "pipeline_tag") else "N/A"
                            recent_models.append({
                                "Model ID": f'<a href="https://huggingface.co/{model.modelId}" target="_blank">{model.modelId}</a>',
                                "Likes": model.likes,
                                "Creation Date": model.created_at.strftime("%Y-%m-%d %H:%M"),
                                "Task": task
                            })
        else:
            # Since the models are sorted by likes in descending order,
            # we can stop once we hit a model with 10 or fewer likes
            break
    
    # Convert the list of dictionaries to a pandas DataFrame
    df = pd.DataFrame(recent_models)
    
    return df

# Define the Gradio interface
def launch_interface():
    iface = gr.Interface(
        fn=get_recent_models,
        inputs=[
            gr.Slider(minimum=1, maximum=100, step=1, value=5, label="Minimum Likes"),
            gr.Slider(minimum=1, maximum=30, step=1, value=3, label="Days Ago"),
            gr.Text(label="Filter", max_lines=1),
            gr.Text(label="Search", max_lines=1)
        ],
        outputs=gr.DataFrame(
            headers=["Model ID", "Likes", "Creation Date", "Task"],
            wrap=True,
            datatype=["html", "number", "str"],
        ),
        title="Model Drops Tracker 🚀",
        description="Overwhelmed by the rapid pace of model releases? 😅 You're not alone! That's exactly why I built this tool. Easily filter recent models from the Hub by setting a minimum number of likes and the number of days since their release. Click on a model to see its card. Use `;` to split filter and search"
    )
    iface.launch()

if __name__ == "__main__":
    launch_interface()