============ Initialized logger ============
add_traspose_rels: True
aggregate: pna
aug: False
batch_size: 16
constrained_neg_prob: 0.0
dataset: fb237_v1
dropout: 0.1
early_stop: 50
emb_dim: 32
enclosing_sub_graph: False
exp_dir: GNN/experiments/fb237_v1_add
experiment_name: fb237_v1_add
gpu: 3
hop: 3
l2: 0.0001
load_model: False
loss: bce
lr: 0.0005
main_dir: GNN
margin: 10
max_links: 1000000
message: add
num_epochs: 10
num_gcn_layers: 6
num_neg_samples_per_link: 1
num_workers: 8
optimizer: Adam
residual: False
train_file: train
un_hop: 1
update: lstm
using_jk: False
valid_file: valid
============================================
No existing model found. Initializing new model..
Device: cuda
Input dim : 32, # Relations : 360
Total number of parameters: 189697
Starting training ...
====================================================================================================
Epoch 1 Training Performance:{'auc': 0.7266327876903611, 'auc_pr': 0.7177891832148854, 'acc': 0.6374558303886926} in 554.5582120418549 s 
Epoch 1 Validation Performance:{'auc': 0.778714542010112, 'auc_pr': 0.7507718751738568, 'acc': 0.6779141104294478} in 36.16291165351868 s 
Epoch 1 Better models found w.r.t AUC-PR. Saved it!
Epoch 1 with loss: 0.6054077690705321, best validation AUC-PR: 0.7507718751738568, weight_norm: 11.72094440460205 in 590.753271818161 s 
====================================================================================================
Epoch 2 Training Performance:{'auc': 0.8375488380288039, 'auc_pr': 0.8286491864664343, 'acc': 0.7371024734982332} in 557.2659213542938 s 
Epoch 2 Validation Performance:{'auc': 0.8524638153905345, 'auc_pr': 0.8302210641261085, 'acc': 0.7525562372188139} in 36.21615934371948 s 
Epoch 2 Better models found w.r.t AUC-PR. Saved it!
Epoch 2 with loss: 0.49374900556596596, best validation AUC-PR: 0.8302210641261085, weight_norm: 11.17192268371582 in 593.5122842788696 s 
====================================================================================================
Epoch 3 Training Performance:{'auc': 0.889970185945913, 'auc_pr': 0.8788896031213927, 'acc': 0.7895170789163722} in 548.2360923290253 s 
Epoch 3 Validation Performance:{'auc': 0.880869936141117, 'auc_pr': 0.8538295499287463, 'acc': 0.7617586912065439} in 36.61060881614685 s 
Epoch 3 Better models found w.r.t AUC-PR. Saved it!
Epoch 3 with loss: 0.41444880224036096, best validation AUC-PR: 0.8538295499287463, weight_norm: 10.707462310791016 in 584.8779771327972 s 
====================================================================================================
Epoch 4 Training Performance:{'auc': 0.9198191456449144, 'auc_pr': 0.9115848512978866, 'acc': 0.8280329799764429} in 546.7104346752167 s 
Epoch 4 Validation Performance:{'auc': 0.8780408245198037, 'auc_pr': 0.8445433862186886, 'acc': 0.7955010224948875} in 37.193673610687256 s 
Epoch 4 with loss: 0.3566387630718991, best validation AUC-PR: 0.8538295499287463, weight_norm: 10.313469886779785 in 583.92169880867 s 
====================================================================================================
Epoch 5 Training Performance:{'auc': 0.9359199279690233, 'auc_pr': 0.9281543276271895, 'acc': 0.8512367491166077} in 551.8700098991394 s 
Epoch 5 Validation Performance:{'auc': 0.8965126442261449, 'auc_pr': 0.8752348502123626, 'acc': 0.8241308793456033} in 35.34925413131714 s 
Epoch 5 Better models found w.r.t AUC-PR. Saved it!
Epoch 5 with loss: 0.31784677880823164, best validation AUC-PR: 0.8752348502123626, weight_norm: 9.966522216796875 in 587.251576423645 s 
====================================================================================================
Epoch 6 Training Performance:{'auc': 0.9494894707415777, 'auc_pr': 0.9437661214651087, 'acc': 0.871849234393404} in 548.9732737541199 s 
Epoch 6 Validation Performance:{'auc': 0.8860284123937253, 'auc_pr': 0.8611366391225779, 'acc': 0.7658486707566462} in 35.72619986534119 s 
Epoch 6 with loss: 0.28290178111397235, best validation AUC-PR: 0.8752348502123626, weight_norm: 9.657879829406738 in 584.7232611179352 s 
====================================================================================================
Epoch 7 Training Performance:{'auc': 0.95692530948209, 'auc_pr': 0.9508280568558947, 'acc': 0.8843345111896349} in 554.1366822719574 s 
Epoch 7 Validation Performance:{'auc': 0.9094935200170626, 'auc_pr': 0.8935647041258389, 'acc': 0.8292433537832311} in 37.164169788360596 s 
Epoch 7 Better models found w.r.t AUC-PR. Saved it!
Epoch 7 with loss: 0.26030473324253145, best validation AUC-PR: 0.8935647041258389, weight_norm: 9.379698753356934 in 591.3312313556671 s 
====================================================================================================
Epoch 8 Training Performance:{'auc': 0.9625092639993563, 'auc_pr': 0.9574478763911481, 'acc': 0.8959952885747938} in 570.8864712715149 s 
Epoch 8 Validation Performance:{'auc': 0.9095646137311236, 'auc_pr': 0.8823712927754117, 'acc': 0.8231083844580777} in 36.43942379951477 s 
Epoch 8 with loss: 0.2420457560933174, best validation AUC-PR: 0.8935647041258389, weight_norm: 9.125726699829102 in 607.3563058376312 s 
====================================================================================================
Epoch 9 Training Performance:{'auc': 0.9685275686354486, 'auc_pr': 0.964027676369466, 'acc': 0.905889281507656} in 551.2972538471222 s 
Epoch 9 Validation Performance:{'auc': 0.9090314108756655, 'auc_pr': 0.8945570910569464, 'acc': 0.8302658486707567} in 37.094566822052 s 
Epoch 9 Better models found w.r.t AUC-PR. Saved it!
Epoch 9 with loss: 0.21944265912069863, best validation AUC-PR: 0.8945570910569464, weight_norm: 8.891315460205078 in 588.427084684372 s 
====================================================================================================
Epoch 10 Training Performance:{'auc': 0.9717600280798722, 'auc_pr': 0.9676628672604485, 'acc': 0.9088339222614841} in 536.3847017288208 s 
Epoch 10 Validation Performance:{'auc': 0.9072812509148088, 'auc_pr': 0.8806036117781968, 'acc': 0.7924335378323109} in 34.66845202445984 s 
Epoch 10 with loss: 0.20846926049798503, best validation AUC-PR: 0.8945570910569464, weight_norm: 8.674671173095703 in 571.0695676803589 s 
====================================================================================================
