Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / williamleif/GraphSAGE issues and pull requests

#36 - question about tf.stop_gradient in deepwalk test part

Issue - State: open - Opened by wierdoki over 6 years ago - 1 comment

#35 - Why is the <train_prefix>-class_map.json needed for unsupervised training?

Issue - State: open - Opened by steffenthoma over 6 years ago - 3 comments

#34 - how to gain embedding of new nodes

Issue - State: open - Opened by guott15 over 6 years ago - 1 comment

#33 - Reddit Source data query on Google BigQuery

Issue - State: closed - Opened by ayushbits over 6 years ago - 3 comments

#32 - Replicating the results in the paper

Issue - State: open - Opened by Jianbo-Lab over 6 years ago - 3 comments

#31 - A question about the example file

Issue - State: open - Opened by ShanYiran over 6 years ago - 2 comments

#30 - Make networkx version check Python3 friendly

Pull Request - State: closed - Opened by gokceneraslan over 6 years ago

#29 - Some problem of evaluating on new test data

Issue - State: open - Opened by pectinid16 over 6 years ago - 3 comments

#28 - Generating graph from data & Questions

Issue - State: open - Opened by amit-pande-2018 over 6 years ago - 5 comments

#27 - Input files for simple homogeneous network

Issue - State: closed - Opened by JunhengH almost 7 years ago - 2 comments

#26 - Embeddings for the supervised example

Issue - State: closed - Opened by annitrolla almost 7 years ago - 4 comments

#25 - Question about the deepwalk implementation

Issue - State: closed - Opened by jermainewang almost 7 years ago - 4 comments

#24 - Details in simple graph training and embedding

Issue - State: open - Opened by JunhengH almost 7 years ago - 2 comments

#23 - make adjacency matrix a placeholder to make saved graphdef smaller

Pull Request - State: closed - Opened by laurimi almost 7 years ago - 1 comment

#22 - How can I use trained model to embed new node in graph? Please help!

Issue - State: open - Opened by doannam020293 almost 7 years ago - 10 comments
Labels: help wanted

#21 - Weight decay was not applied

Issue - State: closed - Opened by dawenl almost 7 years ago - 1 comment

#20 - Why did LSTM aggregator obtain the highest f1 score?

Issue - State: open - Opened by xxxzhi almost 7 years ago - 4 comments
Labels: question

#19 - Saving GraphSAGE model -- ValueError: GraphDef cannot be larger than 2GB

Issue - State: closed - Opened by laurimi almost 7 years ago - 8 comments
Labels: enhancement

#18 - End of mini batch iteration

Issue - State: closed - Opened by Songweiping almost 7 years ago - 3 comments

#17 - Across graphs learning example

Issue - State: closed - Opened by kwonoh almost 7 years ago - 2 comments

#16 - Is f1_score evaluation wrong in ppi_eval.py?

Issue - State: closed - Opened by happylicio almost 7 years ago - 1 comment

#15 - input-format description wrong

Issue - State: closed - Opened by weidong8405347 about 7 years ago - 1 comment

#14 - Unsupervised XENT loss

Issue - State: closed - Opened by bkj about 7 years ago - 3 comments

#13 - Number nodes at each layers

Issue - State: closed - Opened by tamthuc1995 about 7 years ago - 4 comments

#12 - Double division by batch size

Issue - State: closed - Opened by bkj about 7 years ago - 2 comments

#11 - Add networkx version checking to utils.py

Pull Request - State: closed - Opened by bkj about 7 years ago - 2 comments

#10 - Expected behavior of `example_unsupervised.sh`

Issue - State: closed - Opened by bkj about 7 years ago - 4 comments

#9 - Parameters for LSTM model on Reddit dataset

Issue - State: closed - Opened by bkj about 7 years ago - 5 comments

#8 - adding Dockerfiles

Pull Request - State: closed - Opened by aksakalli about 7 years ago - 2 comments

#7 - Fix python 3 support

Issue - State: closed - Opened by williamleif about 7 years ago
Labels: enhancement

#6 - Add a requirements.txt

Issue - State: closed - Opened by williamleif about 7 years ago
Labels: enhancement

#5 - Question about sampling methods in minibatch

Issue - State: closed - Opened by cxfneo about 7 years ago - 3 comments
Labels: question

#4 - Questions about the algorithm

Issue - State: closed - Opened by DanqingZ about 7 years ago - 4 comments
Labels: question

#3 - Add easy support for featureless training / identity features.

Issue - State: closed - Opened by williamleif about 7 years ago
Labels: enhancement

#2 - how to feed data when the number of nodes is big?

Issue - State: closed - Opened by chenwgen about 7 years ago - 1 comment

#1 - how to save well-trained model?

Issue - State: closed - Opened by chenwgen about 7 years ago - 2 comments