Fix Relay docs formatting and grammar (apache#2471)

docs/langref/relay_expr.rst

@@ -11,7 +11,7 @@ Dataflow and Control Fragments
 For the purposes of comparing Relay to traditional computational graph-based IRs, it
 can be useful to consider Relay exrpessions in terms of dataflow and control fragments.
 Each portion of a Relay program containing expressions that only affect the dataflow can
-be viewed as a traditional comptuation graph when writing and expressing transformations.
+be viewed as a traditional computation graph when writing and expressing transformations.
 
 The dataflow fragment covers the set of Relay expressions that do not involve
 control flow. That is, any portion of a program containing only the following
@@ -31,8 +31,8 @@ fragment in Relay includes the following constructs:
 - Recursive Calls in Functions
 
 From the point of view of a computation graph, a function is a subgraph and a function call inlines the subgraph, substituting its arguments for the free variables in the subgraph with corresponding names.
-Thus if a function's body uses only dataflow constructs
-, a call to that function is in the dataflow fragment; conversely, if the
+Thus, if a function's body uses only dataflow constructs,
+a call to that function is in the dataflow fragment; conversely, if the
 function's body contains control flow, a call to that function is not part of the dataflow fragment.
 
 Variables
@@ -205,6 +205,7 @@ For example, one can define a polymorphic identity function for
 any Relay type as follows:
 
 .. code-block:: python
+
     fn<t : Type>(%x : t) -> t {
         %x
     }
@@ -213,13 +214,14 @@ The below definition is also polymorphic, but restricts its
 arguments to tensor types:
 
 .. code-block:: python
+
     fn<s : Shape, bt : BaseType>(%x : Tensor[s, bt]) {
         %x
     }
 
 Notice that the return type is omitted and will be inferred.
 
-*Note: :code:`where` syntax is not yet supported in the text format.*
+*Note: "where" syntax is not yet supported in the text format.*
 
 A function may also be subject to one or more type relations, such as in
 the following:

docs/langref/relay_op.rst

@@ -2,9 +2,9 @@ Relay Core Tensor Operators
 ===========================
 
 This page contains the list of core tensor operator primitives pre-defined in tvm.relay.
-The core tensor operator primitives covers typical workloads in deep learning.
-They can represent workloads in front-end frameworks, and provide basic building blocks for optimization.
-Since deep learning is a fast evolving field and it is that possible to have operators that are not in here.
+The core tensor operator primitives cover typical workloads in deep learning.
+They can represent workloads in front-end frameworks and provide basic building blocks for optimization.
+Since deep learning is a fast evolving field, it is possible to have operators that are not in here.
 
 
 .. note::

docs/langref/relay_type.rst

@@ -2,8 +2,8 @@
 Relay's Type System
 ===================
 
-We briefly introduced types while detailing Relay's expression language
-, but have not yet described its type system. Relay is
+We briefly introduced types while detailing Relay's expression language,
+but have not yet described its type system. Relay is
 a statically typed and type-inferred language, allowing programs to
 be fully typed while requiring just a few explicit type annotations.
 

python/tvm/relay/op/nn/nn.py

@@ -101,7 +101,7 @@ def conv2d_transpose(data,
                      kernel_layout="OIHW",
                      output_padding=(0, 0),
                      out_dtype=""):
-    """Two dimensional trnasposed convolution operator.
+    """Two dimensional transposed convolution operator.
 
     Parameters
     ----------

python/tvm/relay/op/transform.py

@@ -231,7 +231,7 @@ def full(fill_value, shape=(), dtype=""):
 
 
 def full_like(data, fill_value):
-    """Return an scalar value array with the same shape and type as the input array.
+    """Return a scalar value array with the same shape and type as the input array.
 
     Parameters
     ----------
@@ -288,7 +288,7 @@ def where(condition, x, y):
     return _make.where(condition, x, y)
 
 def broadcast_to(data, shape):
-    """Return an scalar value array with the same type, broadcast to
+    """Return a scalar value array with the same type, broadcast to
     the provided shape.
 
     Parameters
@@ -307,7 +307,7 @@ def broadcast_to(data, shape):
     return _make.broadcast_to(data, shape)
 
 def broadcast_to_like(data, broadcast_type):
-    """Return an scalar value array with the same shape and type as the input array.
+    """Return a scalar value array with the same shape and type as the input array.
 
     Parameters
     ----------
@@ -326,7 +326,7 @@ def broadcast_to_like(data, broadcast_type):
 
 
 def collapse_sum_like(data, collapse_type):
-    """Return an scalar value array with the same shape and type as the input array.
+    """Return a scalar value array with the same shape and type as the input array.
 
     Parameters
     ----------
@@ -377,7 +377,7 @@ def split(data, indices_or_sections, axis=0):
 
 
 def strided_slice(data, begin, end, strides=None):
-    """Strided slice of an array..
+    """Strided slice of an array.
 
     Parameters
     ----------