cugraph extensions for DGL
Project description
cugraph_dgl
Description
RAPIDS cugraph_dgl provides a duck-typed version of the DGLGraph class, which uses cugraph for storing graph structure and node/edge feature data. Using cugraph as the backend allows DGL users to access a collection of GPU accelerated algorithms for graph analytics, such as centrality computation and community detection.
Conda
Install and update cugraph-dgl and the required dependencies using the command:
# CUDA 11
conda install -c rapidsai -c pytorch -c conda-forge -c nvidia -c dglteam/label/th23_cu118 cugraph-dgl
# CUDA 12
conda install -c rapidsai -c pytorch -c conda-forge -c nvidia -c dglteam/label/th23_cu121 cugraph-dgl
Build from Source
Create the conda development environment
mamba env create -n cugraph_dgl_dev --file conda/cugraph_dgl_dev_11.6.yml
Install in editable mode
pip install -e .
Run tests
pytest tests/*
Usage
+from cugraph_dgl.convert import cugraph_storage_from_heterograph
+cugraph_g = cugraph_storage_from_heterograph(dgl_g)
sampler = dgl.dataloading.NeighborSampler(
[15, 10, 5], prefetch_node_feats=['feat'], prefetch_labels=['label'])
train_dataloader = dgl.dataloading.DataLoader(
- dgl_g,
+ cugraph_g,
train_idx,
sampler,
device=device,
batch_size=1024,
shuffle=True,
drop_last=False,
num_workers=0)
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
File details
Details for the file cugraph_dgl_cu11-24.10.0.tar.gz
.
File metadata
- Download URL: cugraph_dgl_cu11-24.10.0.tar.gz
- Upload date:
- Size: 1.4 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.0.0 CPython/3.10.12
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 13b34c7076047960b9cf5092c47b8d8c91a33cab40739d47d6a1d413ba1bdeca |
|
MD5 | cad992121bbc82bd2adafbc44ade6113 |
|
BLAKE2b-256 | 65172139d1c4a1a87fa885be2a598c3bb0b078ace61f68d0ff97746f056458ca |