update cluster+bucket

bowang-lab · Jan 11, 2024 · 27eb9cc · 27eb9cc
2 parents 3501f94 + 815cf5a
commit 27eb9cc
Show file tree

Hide file tree

Showing 5 changed files with 13 additions and 31 deletions.
diff --git a/configs/GPS/cocosuperpixels-GPS.yaml b/configs/GPS/cocosuperpixels-GPS.yaml
@@ -36,7 +36,11 @@ model:
   type: GPSModel
   loss_fun: weighted_cross_entropy
 gt:
+<<<<<<< HEAD
   layer_type: CustomGatedGCN+Mamba_Hybrid_Degree_Bucket #Transformer  #Performer
+=======
+  layer_type: CustomGatedGCN+Mamba_Bucket #Transformer  #Performer
+>>>>>>> 815cf5a26a1097601438fcae49ae7f288ca469fd
   layers: 4
   n_heads: 8
   dim_hidden: 96  # `gt.dim_hidden` must match `gnn.dim_inner`
@@ -54,6 +58,7 @@ gnn:
   dropout: 0.0
   agg: mean
   normalize_adj: False
+<<<<<<< HEAD
 # optim:
 #   clip_grad_norm: True
 #   optimizer: adamW
@@ -70,7 +75,6 @@ optim:
     max_epoch: 300
     scheduler: cosine_with_warmup
     num_warmup_epochs: 10
-
 #optim:
 #  optimizer: adamW
 #  weight_decay: 0.0

diff --git a/configs/GPS/peptides-func-GPS.yaml b/configs/GPS/peptides-func-GPS.yaml
@@ -26,13 +26,6 @@ posenc_LapPE:
   dim_pe: 16
   layers: 2
   raw_norm_type: none
-# posenc_RWSE:
-#   enable: True
-#   kernel:
-#     times_func: range(1,17)
-#   model: Linear
-#   dim_pe: 20
-#   raw_norm_type: BatchNorm
 train:
   mode: custom
   batch_size: 128
@@ -43,7 +36,7 @@ model:
   loss_fun: cross_entropy
   graph_pooling: mean
 gt:
-  layer_type: CustomGatedGCN+Mamba_Cluster_Bucket #Mamba_Hybrid_Degree
+  layer_type: CustomGatedGCN+Mamba_Cluster #Mamba_Hybrid_Degree
   n_heads: 4
   dim_hidden: 96  # `gt.dim_hidden` must match `gnn.dim_inner`
   dropout: 0.0
@@ -58,19 +51,11 @@ gnn:
   batchnorm: True
   act: relu
   dropout: 0.0
-# optim:
-#   clip_grad_norm: True
-#   optimizer: adamW
-#   weight_decay: 0.0
-#   base_lr: 0.0003
-#   max_epoch: 200
-#   scheduler: cosine_with_warmup
-#   num_warmup_epochs: 10
 optim:
   clip_grad_norm: True
   optimizer: adamW
-  weight_decay: 0.01
-  base_lr: 0.001
+  weight_decay: 0.0
+  base_lr: 0.0003
   max_epoch: 200
   scheduler: cosine_with_warmup
   num_warmup_epochs: 10

diff --git a/configs/GPS/peptides-struct-GPS.yaml b/configs/GPS/peptides-struct-GPS.yaml
@@ -53,19 +53,11 @@ gnn:
   batchnorm: True
   act: relu
   dropout: 0.0
-#optim:
-#  clip_grad_norm: True
-#  optimizer: adamW
-#  weight_decay: 0.0
-#  base_lr: 0.0003
-#  max_epoch: 200
-#  scheduler: cosine_with_warmup
-#  num_warmup_epochs: 10
 optim:
   clip_grad_norm: True
   optimizer: adamW
-  weight_decay: 0.01
-  base_lr: 0.001
+  weight_decay: 0.0
+  base_lr: 0.0003
   max_epoch: 200
   scheduler: cosine_with_warmup
   num_warmup_epochs: 10

diff --git a/configs/GPS/vocsuperpixels-GPS.yaml b/configs/GPS/vocsuperpixels-GPS.yaml
@@ -1,7 +1,6 @@
 out_dir: results
 metric_best: f1
 wandb:
-  use: True
   project: PascalVOC-SP
   entity: tf-map
 dataset:

diff --git a/graphgps/layer/gps_layer.py b/graphgps/layer/gps_layer.py
@@ -339,6 +339,7 @@ def forward(self, batch):
                         h_attn = self.self_attn(h_dense)[mask][h_ind_perm_reverse]
                         mamba_arr.append(h_attn)
                     h_attn = sum(mamba_arr) / 5
+
             elif 'Mamba_Hybrid_Degree' in self.global_model_type:
                 if batch.split == 'train':
                     h_ind_perm = permute_within_batch(batch.batch)
@@ -379,6 +380,7 @@ def forward(self, batch):
                         #h_attn = self.self_attn(h_dense)[mask][h_ind_perm_reverse]
                         mamba_arr.append(h_attn)
                     h_attn = sum(mamba_arr) / 5
+
             elif self.global_model_type == 'Mamba_Eigen':
                 deg = degree(batch.edge_index[0], batch.x.shape[0]).to(torch.long)
                 centrality = batch.EigCentrality
@@ -420,7 +422,7 @@ def forward(self, batch):
                     unique_cluster_n = len(torch.unique(batch.LouvainCluster))
                     permuted_louvain = torch.zeros(batch.LouvainCluster.shape).long().to(batch.LouvainCluster.device)
                     random_permute = torch.randperm(unique_cluster_n+1).long().to(batch.LouvainCluster.device)
-                    for i in range(len(torch.unique(batch.LouvainCluster))):
+                    for i in range(unique_cluster_n):
                         indices = torch.nonzero(batch.LouvainCluster == i).squeeze()
                         permuted_louvain[indices] = random_permute[i]
                     #h_ind_perm_1 = lexsort([deg[h_ind_perm], permuted_louvain[h_ind_perm], batch.batch[h_ind_perm]])