update to aws link

snap-stanford · Apr 7, 2021 · 53389b3 · 53389b3
1 parent 2229cb4
commit 53389b3
Show file tree

Hide file tree

Showing 5 changed files with 25 additions and 8 deletions.
diff --git a/ogb/lsc/mag240m.py b/ogb/lsc/mag240m.py
@@ -12,7 +12,10 @@
 
 class MAG240MDataset(object):
     version = 1
-    url = 'http://ogb-data.stanford.edu/data/lsc/mag240m_kddcup2021.zip'
+    # Old url hosted at Stanford
+    # url = 'http://ogb-data.stanford.edu/data/lsc/mag240m_kddcup2021.zip'
+    # New url hosted by DGL team at AWS--much faster to download
+    url = 'https://dgl-data.s3-accelerate.amazonaws.com/dataset/OGB-LSC/mag240m_kddcup2021.zip'
 
     __rels__ = {
         ('author', 'paper'): 'writes',

diff --git a/ogb/lsc/pcqm4m.py b/ogb/lsc/pcqm4m.py
@@ -23,8 +23,11 @@ def __init__(self, root = 'dataset', smiles2graph = smiles2graph, only_smiles=Fa
         self.only_smiles = only_smiles
         self.folder = osp.join(root, 'pcqm4m_kddcup2021')
         self.version = 1
-        self.url = f'http://ogb-data.stanford.edu/data/lsc/pcqm4m_kddcup2021.zip'
-        # self._use_smiles = False
+
+        # Old url hosted at Stanford
+        # self.url = f'http://ogb-data.stanford.edu/data/lsc/pcqm4m_kddcup2021.zip'
+        # New url hosted by DGL team at AWS--much faster to download
+        self.url = 'https://dgl-data.s3-accelerate.amazonaws.com/dataset/OGB-LSC/pcqm4m_kddcup2021.zip'
 
         # check version and update if necessary
         if osp.isdir(self.folder) and (not osp.exists(osp.join(self.folder, f'RELEASE_v{self.version}.txt'))):

diff --git a/ogb/lsc/pcqm4m_dgl.py b/ogb/lsc/pcqm4m_dgl.py
@@ -24,8 +24,11 @@ def __init__(self, root = 'dataset', smiles2graph = smiles2graph):
         self.smiles2graph = smiles2graph
         self.folder = osp.join(root, 'pcqm4m_kddcup2021')
         self.version = 1
-        self.url = f'http://ogb-data.stanford.edu/data/lsc/pcqm4m_kddcup2021.zip'
-        # self._use_smiles = False
+
+        # Old url hosted at Stanford
+        # self.url = f'http://ogb-data.stanford.edu/data/lsc/pcqm4m_kddcup2021.zip'
+        # New url hosted by DGL team at AWS--much faster to download
+        self.url = 'https://dgl-data.s3-accelerate.amazonaws.com/dataset/OGB-LSC/pcqm4m_kddcup2021.zip'
 
         # check version and update if necessary
         if osp.isdir(self.folder) and (not osp.exists(osp.join(self.folder, f'RELEASE_v{self.version}.txt'))):

diff --git a/ogb/lsc/pcqm4m_pyg.py b/ogb/lsc/pcqm4m_pyg.py
@@ -25,8 +25,11 @@ def __init__(self, root = 'dataset', smiles2graph = smiles2graph, transform=None
         self.smiles2graph = smiles2graph
         self.folder = osp.join(root, 'pcqm4m_kddcup2021')
         self.version = 1
-        self.url = f'http://ogb-data.stanford.edu/data/lsc/pcqm4m_kddcup2021.zip'
-        # self._use_smiles = False
+
+        # Old url hosted at Stanford
+        # self.url = f'http://ogb-data.stanford.edu/data/lsc/pcqm4m_kddcup2021.zip'
+        # New url hosted by DGL team at AWS--much faster to download
+        self.url = 'https://dgl-data.s3-accelerate.amazonaws.com/dataset/OGB-LSC/pcqm4m_kddcup2021.zip'
 
         # check version and update if necessary
         if osp.isdir(self.folder) and (not osp.exists(osp.join(self.folder, f'RELEASE_v{self.version}.txt'))):

diff --git a/ogb/lsc/wikikg90m.py b/ogb/lsc/wikikg90m.py
@@ -17,7 +17,12 @@ def __init__(self, root: str = 'dataset'):
 
         self.folder = osp.join(root, 'wikikg90m_kddcup2021')
         self.version = 1
-        self.url = f'http://ogb-data.stanford.edu/data/lsc/wikikg90m_kddcup2021.zip'
+
+        # Old url hosted at Stanford
+        # self.url = 'http://ogb-data.stanford.edu/data/lsc/wikikg90m_kddcup2021.zip'
+        # New url hosted by DGL team at AWS--much faster to download
+        self.url = 'https://dgl-data.s3-accelerate.amazonaws.com/dataset/OGB-LSC/wikikg90m_kddcup2021.zip'
+
         self.processed_dir = osp.join(self.folder, 'processed')
 
         if osp.isdir(self.folder) and (not osp.exists(osp.join(self.folder, f'RELEASE_v{self.version}.txt'))):