-
Notifications
You must be signed in to change notification settings - Fork 0
/
makefile
42 lines (40 loc) · 2.79 KB
/
makefile
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
train = cora_lc
embedding = given_lipchitz##coefficents, Lipschitz_rw, Lipschitz_sp,Lipschitz_rw_only,spectral, Lipschitz_rw_node_weights
anchors = 35# default 20 for 500, 3 for 30
epochs = 40
need_training=true # true or false
model_name=graphsage ## gcn,gat,graphsage,mlp
num_layers=2
trained_on=gap
norm=percentile # MinMax, Standard, Normalizer, percentile,None
modelScore = ModelLinkPred ## ModelBasic, ModelAtten, ModelLinkPred, ModelLocalLinkPred
iter_update_params = 2
isAssigned_feature = false
gpu=0 ##0,1,2,3
wandb=false ## wandb on of true or false
num_run=1 ## for running multiple times in val and test
initial_type=kmeans_Linf ## random,kmeans,kmeans_Linf,kmeansSpectral
scoring_func=mlp # mlp, l1, l2, cosine, dot
initial_order=random ## random, degree,closeness,betweenness,core_value,cluster,kcore
num_perturbation=2
node_select=true
node_select_heuristic=diff_max_balanced # diff, diff_max, diff_max_scaled, diff_max_balanced
pool=mean ## mean, max, sum
update_reward=last ## last,best,last_non_nan,teacher_force_nan
hops=2 # 2,3,..
num_perturbation_inference=1000
cuttype=normalised ##normalised,kmin,sparsest,sparsest_weight
finetune_epochs=0
gamma=0.99
# threshold=0.5
# pos_weight=0.3
# num_times_samples=50
# num_edges_samples=50
## Add below line to incorporate above variables
# pos_weight=$(pos_weight) threshold=$(threshold) num_edges_samples=$(num_edges_samples) num_times_samples=$(num_times_samples)
main:
$(MAKE) -C src cuttype=$(cuttype) update_reward=$(update_reward) pool=$(pool) wandb=$(wandb) num_perturbation=$(num_perturbation) num_run=$(num_run) initial_order=$(initial_order) initial_type=$(initial_type) train=$(train) gpu=$(gpu) embedding=$(embedding) anchors=$(anchors) epochs=$(epochs) model_name=$(model_name) num_layers=$(num_layers) need_training=$(need_training) norm=$(norm) modelScore=$(modelScore) iter_update_params=$(iter_update_params) isAssigned_feature=$(isAssigned_feature) scoring_func=$(scoring_func) node_select=$(node_select) node_select_heuristic=$(node_select_heuristic) hops=$(hops) gamma=$(gamma)
finetune:
$(MAKE) -C src cuttype=$(cuttype) update_reward=$(update_reward) pool=$(pool) wandb=$(wandb) num_perturbation=$(num_perturbation) num_run=$(num_run) initial_order=$(initial_order) initial_type=$(initial_type) train=$(train) gpu=$(gpu) embedding=$(embedding) anchors=$(anchors) epochs=$(epochs) model_name=$(model_name) num_layers=$(num_layers) need_training=$(need_training) norm=$(norm) modelScore=$(modelScore) iter_update_params=$(iter_update_params) isAssigned_feature=$(isAssigned_feature) scoring_func=$(scoring_func) node_select=$(node_select) node_select_heuristic=$(node_select_heuristic) hops=$(hops) gamma=$(gamma) finetune_epochs=$(finetune_epochs) finetune
clean_pkl:
find ./data -name \*.pkl -type f -delete