Frozenmad
fbd420e8e5
PR [ #32 ] change_model_func_name -> dev
change model function name decode to lp_decode in response to PR #31
5 years ago
lihy96
1159328106
change model function name decode to lp_decode
5 years ago
Frozenmad
4b00022b10
PR [ #28 ] lp branch
* add link prediction trainer, solver, and model adaptation
5 years ago
Frozenmad
2b251671bf
move lp solver
5 years ago
Frozenmad
9815e39d82
fix tuple bugs, add lp to solver __init__
5 years ago
Frozenmad
34a452e25b
black format
5 years ago
Frozenmad
c28d1c3a3b
fix ensemble bug
5 years ago
Frozenmad
a08843dc3b
fix performance bugs of gcn on link prediction
5 years ago
Frozenmad
4878a0abe5
adjust metrics input
5 years ago
Frozenmad
5c88a3e57f
add comparison code
5 years ago
Frozenmad
e2321300df
update lp solver
5 years ago
Frozenmad
5602bf794e
adjust for new model
5 years ago
Frozenmad
7d831bb94e
Merge branch 'dev' into lp_solver
5 years ago
Frozenmad
3e9a11fa36
change the default num_workers to 0
5 years ago
lihy96
9cbd8af51a
add encode and decode in gat and graphsage
5 years ago
lihy96
ee6e58e802
roc auc (gcn on cora) ~0.911
5 years ago
Frozenmad
07d2329476
black style
5 years ago
lihy96
5b4c918975
link prediction
5 years ago
Frozenmad
13106d41d1
add stratify support, fix bugs
5 years ago
Frozenmad
7106622ec1
change url of paper
5 years ago
Frozenmad
25a404b7df
merge 'main' to 'dev'
5 years ago
null
3d8e9d1847
Implement partial GraphSAINT Methodolgy and refactor BaseModel
Reproduce partial implementation of GraphSAINT, a representative Subgraph-wise sampling method.
Add experimental BaseModel for future major version.
TODO: Migrate the base class for all the concrete models to ClassificationModel (ClassificationApproach)
5 years ago
Frozenmad
a062f6c535
Merge pull request #22 from Frozenmad/update_paper
update paper
5 years ago
Frozenmad
afddf27745
update paper
5 years ago
cluster32
ad4ef59294
add progress bar
5 years ago
null
ba81889c90
Refactor module.train and fix a bug in example for graph cls
5 years ago
Frozenmad
fe3f676e3b
[ #19 ] fix num_workers bug
fixed num_workers, set_start_method(fork) when num_workers>0
5 years ago
Frozenmad
de9c1ce59e
Merge branch 'dev' into multiprocess
5 years ago
Frozenmad
127332df22
Merge pull request #21 from THUMNLab/examples
Update examples & fix bugs in fe
5 years ago
Frozenmad
8ae5156cdd
update examples & fix bugs in fe
5 years ago
SwiftieH
b177e104e0
fixed num_workers, set_start_method(fork) when num_workers>0
5 years ago
null
93159cd363
Implement Node-wise(Neighbor) Sampling with GraphSAGE model
5 years ago
Frozenmad
38ab2b06b9
rebase trainer
5 years ago
Frozenmad
3badd90bb9
change es logic
5 years ago
Frozenmad
e6d42bbc4f
black format
5 years ago
Frozenmad
683f4db8e6
Merge pull request #18 from Frozenmad/cora_gat
Fix bugs in trainer
5 years ago
Frozenmad
d94923fb39
fix es bug and duplicate_from_hyperparameter
5 years ago
Frozenmad
deb49073b5
Merge pull request #17 from THUMNLab/fix_train_branch
Customized trainer support & performance debugging
5 years ago
Frozenmad
3f948eea7c
Merge pull request #16 from Frozenmad/add_trainer_to_solver
Add default argument support of trainer & model in solver
5 years ago
Frozenmad
614c9c0b84
fix bug in graph trainer
5 years ago
Frozenmad
7d5f51c4ff
update configs
5 years ago
Frozenmad
9031fd4618
construct trainer when building from configs
5 years ago
Frozenmad
d19bc12a22
customize init process
5 years ago
lihy96
4f20b5d956
rm model hp space from trainer hp space
5 years ago
Frozenmad
dbaa4dbd0f
fix hp space setting
5 years ago
Frozenmad
c2a589615e
fix device inconsistency
5 years ago
Frozenmad
0e0c5528b3
add trainer when loading configs
5 years ago
Frozenmad
49924154fe
add trainer to init func
5 years ago
lihy96
031ed5710f
add lr_scheduler setting
5 years ago
Frozenmad
e63c95da45
black format
5 years ago