|
|
--- |
|
|
title: VaibeSync User Matching Model |
|
|
emoji: 💕 |
|
|
colorFrom: pink |
|
|
colorTo: purple |
|
|
sdk: pytorch |
|
|
tags: |
|
|
- dating |
|
|
- user-matching |
|
|
- recommendation-system |
|
|
- pytorch-lightning |
|
|
--- |
|
|
|
|
|
# pluto User Matching Model |
|
|
|
|
|
This model powers the pluto dating app's intelligent user matching system. |
|
|
|
|
|
## Model Details |
|
|
|
|
|
- **Framework**: PyTorch Lightning |
|
|
- **Task**: User compatibility prediction |
|
|
- **Architecture**: Two-tower neural network |
|
|
- **Size**: 1.5 MB |
|
|
|
|
|
## Usage |
|
|
|
|
|
This model is automatically downloaded and used by the pluto ML API deployed on Railway. |
|
|
|
|
|
```python |
|
|
from models.user_match_two_tower import UserMatchingModel |
|
|
|
|
|
model = UserMatchingModel.load_from_checkpoint("last.ckpt") |
|
|
model.eval() |
|
|
``` |
|
|
|
|
|
## Railway Deployment |
|
|
|
|
|
Set these environment variables in your Railway deployment: |
|
|
|
|
|
``` |
|
|
MODEL_STORAGE_TYPE=huggingface |
|
|
HF_MODEL_REPO=ssourav15/vaibesync-user-matching-model |
|
|
``` |
|
|
|
|
|
The model will be automatically downloaded on first startup. |
|
|
|
