Prbhvv/mllm-recommender-api

1# Deployment Guide 🚀2 3This guide explains how to deploy your **Next-Gen Recommender** for free using **Hugging Face Spaces** (Backend) and **Vercel** (Frontend).4 5## Part 1: Backend (Hugging Face Spaces)6 7We use Hugging Face Spaces because it provides free CPU resources (16GB RAM) suitable for running your PyTorch model.8 91.  **Create a Space**:10    *   Go to [huggingface.co/spaces](https://huggingface.co/spaces) and create a new Space.11    *   **Name**: `mllm-recommender-api` (or similar).12    *   **SDK**: Select **Docker**.13    *   **Hardware**: Free CPU (Basic).14 152.  **Upload Files**:16    *   Clone your Space locally:17        ```bash18        git clone https://huggingface.co/spaces/YOUR_USERNAME/mllm-recommender-api19        ```20    *   Copy **ALL** files from your project folder (`MLLM-MSR`) into this cloned directory.21    *   **Important**: Ensure `Dockerfile`, `requirements.txt`, `api/`, `gnn_fusion/`, `models/`, `data/`, and `embeddings/` are included.22    *   *Note: Large files (>10MB) require `git lfs`.*23        ```bash24        git lfs install25        git lfs track "*.pt" "*.npz" "*.csv"26        ```27    *   Push the files:28        ```bash29        git add .30        git commit -m "Initial deploy"31        git push32        ```33 343.  **Get Backend URL**:35    *   Once built, your API will be available at: `https://YOUR_USERNAME-mllm-recommender-api.hf.space`.36    *   Test it: `https://YOUR_USERNAME-mllm-recommender-api.hf.space/health`37 38---39 40## Part 2: Frontend (Vercel)41 421.  **Push Code to GitHub**:43    *   Create a new repository on GitHub.44    *   Push your `web` folder code to it.45 462.  **Deploy on Vercel**:47    *   Go to [vercel.com](https://vercel.com) and "Add New Project".48    *   Import your GitHub repository.49    *   **Root Directory**: Select `web` (since your frontend is in the `web` subfolder).50    *   **Environment Variables**:51        *   Add `VITE_API_URL` = `https://YOUR_USERNAME-mllm-recommender-api.hf.space` (The URL from Part 1).52    *   Click **Deploy**.53 54## 🎉 Done!55 56Your app is now live!57- **Frontend**: `https://your-project.vercel.app`58- **Backend**: `https://YOUR_USERNAME-mllm-recommender-api.hf.space`59