model: graph_conv-dot_product_h128_l3_edge_prediction | (graph_conv-dot_product_h128_l3) | WandB: nvog9n27

README.md

@@ -1,5 +1,5 @@
 ---
-tags: ['graph-neural-networks', 'biological-networks', 'napistu', 'pytorch', 'graph_conv', 'dot_product', 'edge_prediction']
+tags: ['napistu', 'napistu-torch', 'graph-neural-networks', 'biological-networks', 'pytorch', 'graph_conv', 'dot_product', 'edge_prediction']
 library_name: napistu-torch
 license: mit
 metrics:
@@ -7,7 +7,7 @@ metrics:
 - average_precision
 ---
 
-# 128-dim GraphConv model trained for edge prediction with a dot-product head
+# graph_conv-dot_product_h128_l3_edge_prediction
 
 This model was trained using [Napistu-Torch](https://www.shackett.org/napistu_torch/), a PyTorch framework for training graph neural networks on biological pathway networks.
 
@@ -39,7 +39,7 @@ optionally incorporating relation types for relation-aware prediction.
   - Type: `dot_product`
   - Relation-Aware: ✗
 
-**Training Date**: 2025-12-04
+**Training Date**: 2025-12-21
 
 For detailed experiment and training settings see this repository's `config.json` file.
 
@@ -47,15 +47,16 @@ For detailed experiment and training settings see this repository's `config.json
 
 | Metric | Value |
 |--------|-------|
-| Validation AUC | 0.7957 |
-| Test AUC | 0.7964 |
-| Validation AP | 0.7938 |
-| Test AP | 0.7947 |
+| Validation AUC | 0.8176 |
+| Test AUC | 0.8186 |
+| Validation AP | 0.8190 |
+| Test AP | 0.8194 |
 
 
 ## Links
 
-- 📊 [W&B Run](https://wandb.ai/napistu/napistu-experiments/runs/s6xbw29d)
+- 📊 [W&B Run](https://wandb.ai/napistu/napistu-experiments/runs/nvog9n27)
+- 🌐 [Napistu](https://napistu.com)
 - 💻 [GitHub Repository](https://github.com/napistu/Napistu-Torch)
 - 📖 [Read the Docs](https://napistu-torch.readthedocs.io/en/latest)
 - 📚 [Napistu Wiki](https://github.com/napistu/napistu/wiki)
@@ -69,8 +70,8 @@ To reproduce the environment used for training, run the following commands:
 ```bash
 pip install torch==2.8.0
 pip install torch-scatter torch-sparse -f https://data.pyg.org/whl/2.8.0+cpu.html
-pip install 'napistu==0.8.2'
-pip install 'napistu-torch[pyg,lightning]==0.2.13'
+pip install 'napistu==0.8.5'
+pip install 'napistu-torch[pyg,lightning]==0.3.2'
 ```
 
 ### 2. Setup Data Store
@@ -91,10 +92,10 @@ napistu_data_store = gcs_model_to_store(
 
 ### 3. Load Pretrained Model from HuggingFace Hub
 ```python
-from napistu_torch.ml.hugging_face import HuggingFaceLoader
+from napistu_torch.ml.hugging_face import HFModelLoader
 
 # Load checkpoint
-loader = HuggingFaceLoader("seanhacks/edge_prediction_dotprod_128e")
+loader = HFModelLoader("seanhacks/edge_prediction_dotprod_128e")
 checkpoint = loader.load_checkpoint()
 
 # Load config to reproduce experiment