Fix of dotnet#3822 to address some inadvertently omitted fixes on VectorDataViewType documentation. (dotnet#3296)

Author: TomFinley

src/Microsoft.ML.DataView/VectorType.cs

@@ -26,10 +26,10 @@ public sealed class VectorDataViewType : StructuredDataViewType
         /// In the case where this is a multi-dimensional type, that is, a situation where <see cref="Dimensions"/>
         /// has length greater than one, since <see cref="VBuffer{T}"/> itself is a single dimensional structure,
         /// we must clarify what we mean. The indices represent a "flattened" view of the coordinates implicit in the
-        /// dimensions. We consider that that last dimension is the most "minor" index. In the case where <see cref="Dimensions"/>
+        /// dimensions. We consider that the last dimension is the most "minor" index. In the case where <see cref="Dimensions"/>
         /// has length <c>2</c>, this is commonly referred to as row-major order. So, if you hypothetically had
         /// dimensions of <c>{ 5, 2 }</c>, then the <see cref="VBuffer{T}"/> values would be all of length <c>10</c>,
-        /// and the flattened indices in the  of <c>0, 1, 2, 3, 4, ...</c> would correspond to "coordinates" of
+        /// and the flattened indices <c>0, 1, 2, 3, 4, ...</c> would correspond to "coordinates" of
         /// <c>(0, 0), (0, 1), (1, 0), (1, 1), (2, 0), ...</c>, respectively.
         /// </remarks>
         public ImmutableArray<int> Dimensions { get; }