DLRM examples to create prediction using TorchRec library. This shows basic example of how to use TorchRec library quickly locally. (#3043)

Summary:

Adding end to end example of using TorchRec library with DLRM to create recommendation. You can use this example in local machine to run and test how TorchRec is used for simple use case.

Differential Revision: D75989524

Author: yupadhyay

examples/prediction/README.md

@@ -0,0 +1,95 @@
+# DLRM Prediction Example
+
+This example demonstrates how to use a Deep Learning Recommendation Model (DLRM) for making predictions. The code includes:
+
+1. A custom DLRM implementation
+2. Training with random data
+3. Evaluation
+4. Making sample predictions
+
+## Cross-Platform Compatibility
+
+This implementation has been specifically designed to work on all platforms, including:
+- Linux
+- macOS
+- Windows
+
+Unlike the original torchrec implementation, this version uses a custom SimpleDLRM class that doesn't depend on torchrec or fbgemm_gpu, avoiding compatibility issues on macOS and other platforms.
+
+## Dependencies
+
+Install the required dependencies:
+
+```bash
+# Install PyTorch
+pip install torch torchvision
+
+# Install NumPy
+pip install numpy
+```
+
+**Important**: torchrec is NOT required or used in this implementation. The code has been completely rewritten to avoid any dependencies on torchrec or fbgemm_gpu.
+
+## Running the Example Locally
+
+1. Download the `predict_using_torchrec.py` file to your local machine.
+
+2. Run the example:
+
+```bash
+python3 predict_using_torchrec.py
+```
+
+3. If you're using a different Python environment:
+
+```bash
+# For conda environments
+conda activate your_environment_name
+python predict_using_torchrec.py
+
+# For virtual environments
+source your_venv/bin/activate
+python predict_using_torchrec.py
+```
+
+## What to Expect
+
+When you run the example, you'll see:
+
+1. Training progress for 10 epochs with loss and learning rate information
+2. Evaluation results showing MSE and RMSE metrics
+3. Sample predictions for a specific user on multiple items
+
+## Implementation Details
+
+This example uses a custom SimpleDLRM implementation that:
+
+- Takes dense features and categorical features as input
+- Processes dense features through a bottom MLP
+- Processes categorical features through embedding tables
+- Computes feature interactions using dot products
+- Processes the interactions through a top MLP
+- Outputs rating predictions on a 0-5 scale
+
+The implementation is designed to be simple and easy to understand, while still capturing the key components of a DLRM model.
+
+## Troubleshooting
+
+If you encounter any issues:
+
+1. **Python version**: This code has been tested with Python 3.8+. Make sure you're using a compatible version.
+
+2. **PyTorch installation**: If you have issues with PyTorch, try installing a specific version:
+   ```bash
+   pip install torch==2.0.0 torchvision==0.15.0
+   ```
+
+3. **Memory issues**: If you run out of memory, try reducing the batch size by modifying this line in the code:
+   ```python
+   batch_size = 256  # Try a smaller value like 64 or 32
+   ```
+
+4. **CPU vs GPU**: The code automatically uses CUDA if available. To force CPU usage, modify:
+   ```python
+   device = torch.device("cpu")
+   ```

examples/prediction/__init__.py

@@ -0,0 +1,16 @@
+#!/usr/bin/env python3
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+# pyre-strict
+
+
+def main() -> None:
+    """DOC_STRING"""
+
+
+if __name__ == "__main__":
+    main()