Update README.md

README.md

@@ -91,26 +91,27 @@ then, from within python load the datasets library
 and load the `MolData` datasets, e.g.,
 
     >>> MolData = datasets.load_dataset("maomlab/MolData")
-    Downloading readme: 100%|██████████| 5.23k/5.23k [00:00<00:00, 35.1kkB/s]
-    Downloading data: 100%|██████████| 34.5k//34.5k/ [00:00<00:00, 155kB/s]
-    Downloading data: 100%|██████████| 97.1k/97.1k [00:00<00:00, 587kB/s]
-    Generating test split: 100%|██████████| 594/594 [00:00<00:00, 12705.92 examples/s]
-    Generating train split: 100%|██████████| 1788/1788 [00:00<00:00, 43895.91 examples/s]
+        Generating train split: 100%|███████████████████████████████████████████████████████████████████████████████████████████| 138547273/138547273 [02:07<00:00, 1088043.12 examples/s]
+        Generating test split: 100%|██████████████████████████████████████████████████████████████████████████████████████████████| 17069726/17069726 [00:16<00:00, 1037407.67 examples/s]
+        Generating validation split: 100%|████████████████████████████████████████████████████████████████████████████████████████| 12728449/12728449 [00:11<00:00, 1093675.24 examples/s]
 
 and inspecting the loaded dataset
 
     >>> MolData
-    MolData
     DatasetDict({
-      test: Dataset({
-         features: ['SMILES', 'Y'],
-         num_rows: 594
-      })
-      train: Dataset({
-          features: ['SMILES', 'Y'],
-          num_rows: 1788
-      })    
+    train: Dataset({
+        features: ['SMILES', 'PUBCHEM_CID', 'split', 'AID', 'Y'],
+        num_rows: 138547273
     })
+    test: Dataset({
+        features: ['SMILES', 'PUBCHEM_CID', 'split', 'AID', 'Y'],
+        num_rows: 17069726
+    })
+    validation: Dataset({
+        features: ['SMILES', 'PUBCHEM_CID', 'split', 'AID', 'Y'],
+        num_rows: 12728449
+    })
+})
 
 ### Use a dataset to train a model
 One way to use the dataset is through the [MolFlux](https://exscientia.github.io/molflux/) package developed by Exscientia.