https://github.com/feedzai/truncation-gap
Official code for the paper "Mind the truncation gap: challenges of learning on dynamic graphs"
https://github.com/feedzai/truncation-gap
Last synced: 4 months ago
JSON representation
Official code for the paper "Mind the truncation gap: challenges of learning on dynamic graphs"
- Host: GitHub
- URL: https://github.com/feedzai/truncation-gap
- Owner: feedzai
- Created: 2024-03-01T11:36:38.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-12-18T11:25:03.000Z (over 1 year ago)
- Last Synced: 2025-10-17T07:41:32.205Z (8 months ago)
- Language: Python
- Homepage: https://arxiv.org/abs/2412.21046
- Size: 36.1 KB
- Stars: 1
- Watchers: 5
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Mind the truncation gap: challenges of learning on dynamic graphs
Official code for the paper [Mind the truncation gap: challenges of learning on dynamic graphs with recurrent architectures](https://openreview.net/forum?id=QezxDgd5hf).
Includes our toy task implementation, together with the python scripts required to train and evaluate a GRNN model on this task using both full backpropagation and truncated backpropagation.
Also includes python scripts to reproduce the experiments on public dynamic graph datasets, comparing full and truncated backpropagation.
# Using the synthetic task
Our proposed synthetic link regression task can be used as a simple benchmark to test an algorithm's ability to learn longer term dependencies.
The following example instantiates the synthetic task with 100 nodes and a memory of 3 steps and uses it to generate edges:
```python
import jax
from jax import lax
from tgap.data.buffer_task import get_sampler_link_regression
init_data, step_data = get_sampler_link_regression(num_nodes=100, delay=3)
# initialize state
rng = jax.random.PRNGKey(123)
initial_data_state = init_data(rng)
# run 1 step
new_data_state, (edge_src, edge_dst, edge_feat, edge_target) = step_data(initial_data_state)
# run 1000 steps
new_data_state, (edges_src, edges_dst, edges_feat, edges_target) = lax.scan(step_data, initial_data_state, None, 1000)
```
# Running the synthetic task experiments
Install dependencies (with python >= 3.9):
```{bash}
pip install -r requirements.txt
```
Run full or truncated backprop:
```{bash}
# Full Backprop
python run_toy_task.py --method FBPTT
# Truncated Backprop
python run_toy_task.py --method TBPTT
```