Serving Graph Compression for Graph Neural Networks

Authors: Si Si, Felix Yu, Ankit Singh Rawat, Cho-Jui Hsieh, Sanjiv Kumar

ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on semi-supervised node classification demonstrate that the proposed method can significantly reduce the serving space requirement for GNN inference. (Abstract) 4 EXPERIMENTAL RESULTS (Section title)
Researcher Affiliation Collaboration 1Google Research 2University of California, Los Angeles {sisidaisy, felixyu, ankitsrawat, sanjivk}@google.com {chohsieh}@ucla.cs.edu
Pseudocode Yes Algorithm 1 The Virtual Node Graph (VNG) algorithm
Open Source Code No The paper mentions using a third-party open-source implementation for GCN model training ("Cluster GCN s tensorflow implementation"), but does not state that the code for their proposed VNG method is open-source or provided.
Open Datasets Yes All above datasets are publicly available and are commonly used for benchmarking the performance of GNNs on node classification tasks. (Section 4) Arxiv: ... We use the same dataset and partition as in (Hu et al., 2020). (Section 4) Reddit: ... We use the same dataset and partition as in (Chiang et al., 2019). (Section 4) Product: ... based on a different preprocessing and split by Hu et al. (2020). (Section 4) Amazon2M: ... We use the same dataset and partition as in (Chiang et al., 2019). (Section 4)
Dataset Splits Yes Table 2: The statistics of Arxiv, Reddit, Product, and Amazon2M datasets. #Training Nodes #Validate Nodes #Labels #Features Serving size Arxiv: ... We use the same dataset and partition as in (Hu et al., 2020). Reddit: ... We use the same dataset and partition as in (Chiang et al., 2019). Product: ... based on a different preprocessing and split by Hu et al. (2020). Amazon2M: ... We use the same dataset and partition as in (Chiang et al., 2019).
Hardware Specification No The paper does not explicitly state the specific hardware, such as GPU or CPU models, used for running the experiments.
Software Dependencies No The paper mentions 'Cluster GCN s tensorflow implementation' but does not specify a version number for TensorFlow or any other software dependencies.
Experiment Setup Yes As for architecture, on all datasets, we consider a 4-layer GCN model with hidden dimensions 512, 256, 512, and 400 for Product, Arxiv, Reddit, and Amazon2M, respectively, and the mean aggregator from Hamilton et al. (2017).