docs: ✏️ update configs

GestaltCogTeam · Sep 12, 2024 · c9d53cf · c9d53cf
1 parent a6d4276
commit c9d53cf
Show file tree

Hide file tree

Showing 11 changed files with 22 additions and 21 deletions.
diff --git a/baselines/DeepAR/ExchangeRate.py b/baselines/DeepAR/ExchangeRate.py
@@ -25,15 +25,14 @@
 NULL_VAL = regular_settings['NULL_VAL'] # Null value in the data
 # Model architecture and parameters
 MODEL_ARCH = DeepAR
-NUM_NODES = 8
 MODEL_PARAM = {
     'cov_feat_size' : 2,
     'embedding_size' : 32,
     'hidden_size' : 64,
     'num_layers': 3,
     'use_ts_id'   : True,
     'id_feat_size': 32,
-    'num_nodes': 7
+    'num_nodes': 8
     }
 NUM_EPOCHS = 100
 

diff --git a/baselines/STEP/README.md b/baselines/STEP/README.md
@@ -1 +1 @@
-STEP requires `timm` package. You can install it by `pip install timm`.
+STEP requires `timm` package. You can install it by `pip install timm==0.6.7`.
diff --git a/baselines/STEP/STEP_METR-LA.py b/baselines/STEP/STEP_METR-LA.py
@@ -28,7 +28,7 @@
 MODEL_ARCH = STEP
 MODEL_PARAM = {
     "dataset_name": DATA_NAME,
-    "pre_trained_tsformer_path": "checkpoints/TSFormer/METR-LA_100_2016_12/de9f10ca8535dbe99fb71072aab848ce/TSFormer_best_val_MAE.pt",
+    "pre_trained_tsformer_path": "checkpoints/TSFormer/METR-LA_100_2016_12/cd176b70ebb4620da5a289ad76355c75/TSFormer_best_val_MAE.pt",
     "short_term_len": INPUT_LEN_SHORT,
     "long_term_len": INPUT_LEN,
     "tsformer_args": {

diff --git a/baselines/STEP/TSFormer_METR-LA.py b/baselines/STEP/TSFormer_METR-LA.py
@@ -1,6 +1,5 @@
 import os
 import sys
-import torch
 from easydict import EasyDict
 sys.path.append(os.path.abspath(__file__ + '/../../..'))
 

diff --git a/baselines/STEP/arch/tsformer/transformer_layers.py b/baselines/STEP/arch/tsformer/transformer_layers.py
@@ -7,14 +7,17 @@ class TransformerLayers(nn.Module):
     def __init__(self, hidden_dim, nlayers, mlp_ratio, num_heads=4, dropout=0.1):
         super().__init__()
         self.d_model = hidden_dim
-        encoder_layers = TransformerEncoderLayer(hidden_dim, num_heads, hidden_dim*mlp_ratio, dropout)
+        # encoder_layers = TransformerEncoderLayer(hidden_dim, num_heads, hidden_dim*mlp_ratio, dropout)
+        encoder_layers = TransformerEncoderLayer(hidden_dim, num_heads, hidden_dim*mlp_ratio, dropout, batch_first=True)
         self.transformer_encoder = TransformerEncoder(encoder_layers, nlayers)
 
     def forward(self, src):
         B, N, L, D = src.shape
         src = src * math.sqrt(self.d_model)
         src = src.view(B*N, L, D)
-        src = src.transpose(0, 1)
+        # src = src.transpose(0, 1)
+        # output = self.transformer_encoder(src, mask=None)
+        # output = output.transpose(0, 1).view(B, N, L, D)
         output = self.transformer_encoder(src, mask=None)
-        output = output.transpose(0, 1).view(B, N, L, D)
+        output = output.view(B, N, L, D)
         return output
diff --git a/baselines/STEP/arch/tsformer/tsformer.py b/baselines/STEP/arch/tsformer/tsformer.py
@@ -1,6 +1,5 @@
 import torch
 from torch import nn
-from timm.models.vision_transformer import trunc_normal_
 
 from .patch import PatchEmbedding
 from .mask import MaskGenerator
@@ -66,6 +65,8 @@ def initialize_weights(self):
         # positional encoding
         nn.init.uniform_(self.positional_encoding.position_embedding, -.02, .02)
         # mask token
+        # import here to fix bugs related to set visible device
+        from timm.models.vision_transformer import trunc_normal_
         trunc_normal_(self.mask_token, std=.02)
 
     def encoding(self, long_term_history, mask=True):

diff --git a/baselines/STID/SD.py b/baselines/STID/SD.py
@@ -123,10 +123,8 @@
 }
 # Train data loader settings
 CFG.TRAIN.DATA = EasyDict()
-CFG.TRAIN.DATA.BATCH_SIZE = 64
+CFG.TRAIN.DATA.BATCH_SIZE = 32
 CFG.TRAIN.DATA.SHUFFLE = True
-# Early stopping
-CFG.TRAIN.EARLY_STOPPING_PATIENCE = 10 # Early stopping patience. Default: None. If not specified, the early stopping will not be used.
 
 ############################## Validation Configuration ##############################
 CFG.VAL = EasyDict()

diff --git a/baselines/STWave/PEMS07.py b/baselines/STWave/PEMS07.py
@@ -159,20 +159,20 @@ def loadGraph(adj_mx, hs, ls):
 }
 # Train data loader settings
 CFG.TRAIN.DATA = EasyDict()
-CFG.TRAIN.DATA.BATCH_SIZE = 64
+CFG.TRAIN.DATA.BATCH_SIZE = 16
 CFG.TRAIN.DATA.SHUFFLE = True
 
 ############################## Validation Configuration ##############################
 CFG.VAL = EasyDict()
 CFG.VAL.INTERVAL = 1
 CFG.VAL.DATA = EasyDict()
-CFG.VAL.DATA.BATCH_SIZE = 64
+CFG.VAL.DATA.BATCH_SIZE = 16
 
 ############################## Test Configuration ##############################
 CFG.TEST = EasyDict()
 CFG.TEST.INTERVAL = 1
 CFG.TEST.DATA = EasyDict()
-CFG.TEST.DATA.BATCH_SIZE = 64
+CFG.TEST.DATA.BATCH_SIZE = 16
 
 ############################## Evaluation Configuration ##############################
 

diff --git a/baselines/Triformer/Electricity.py b/baselines/Triformer/Electricity.py
@@ -116,20 +116,20 @@
 }
 # Train data loader settings
 CFG.TRAIN.DATA = EasyDict()
-CFG.TRAIN.DATA.BATCH_SIZE = 64
+CFG.TRAIN.DATA.BATCH_SIZE = 16
 CFG.TRAIN.DATA.SHUFFLE = True
 
 ############################## Validation Configuration ##############################
 CFG.VAL = EasyDict()
 CFG.VAL.INTERVAL = 1
 CFG.VAL.DATA = EasyDict()
-CFG.VAL.DATA.BATCH_SIZE = 64
+CFG.VAL.DATA.BATCH_SIZE = 16
 
 ############################## Test Configuration ##############################
 CFG.TEST = EasyDict()
 CFG.TEST.INTERVAL = 1
 CFG.TEST.DATA = EasyDict()
-CFG.TEST.DATA.BATCH_SIZE = 64
+CFG.TEST.DATA.BATCH_SIZE = 16
 
 ############################## Evaluation Configuration ##############################
 

diff --git a/baselines/Triformer/Weather.py b/baselines/Triformer/Weather.py
@@ -23,7 +23,7 @@
 NULL_VAL = regular_settings['NULL_VAL'] # Null value in the data
 # Model architecture and parameters
 MODEL_ARCH = Triformer
-NUM_NODES = 21
+NUM_NODES = 8
 MODEL_PARAM = {
     "num_nodes": NUM_NODES,
     "lag": INPUT_LEN,

diff --git a/experiments/train.py b/experiments/train.py
@@ -15,8 +15,9 @@
 
 def parse_args():
     parser = ArgumentParser(description="Run time series forecasting model in BasicTS framework!")
-    # parser.add_argument("-c", "--cfg", default="baselines/STID/METR-LA.py", help="training config")
-    parser.add_argument("-c", "--cfg", default="baselines/STEP/STEP_METR-LA2.py", help="training config")
+    parser.add_argument("-c", "--cfg", default="baselines/STID/METR-LA.py", help="training config")
+    # parser.add_argument("-c", "--cfg", default="baselines/STEP/TSFormer_METR-LA.py", help="training config")
+    # parser.add_argument("-c", "--cfg", default="baselines/STEP/STEP_METR-LA.py", help="training config")
     # parser.add_argument("-c", "--cfg", default="baselines/DGCRN/PEMS-BAY.py", help="training config")
     # parser.add_argument("-c", "--cfg", default="baselines/DGCRN/example.py", help="training config")
     # parser.add_argument("-c", "--cfg", default="examples/complete_config.py", help="training config")