[IR] Improve documentation 1/n (#2227)

![image](https://github.com/user-attachments/assets/3877ec1b-3f4d-45b1-b8da-8e363ce711ff)

Author: justinchuby

.gitignore

@@ -100,6 +100,7 @@ dmypy.json
 *.onnxlib
 **/onnx_backend_test_code/**
 docs/auto_examples/*
+docs/intermediate_representation/generated/*
 tests/export/*
 tests/models/testoutputs/*
 tests/mylib.onnxlib

docs/_templates/classtemplate.rst

@@ -0,0 +1,14 @@
+.. role:: hidden
+    :class: hidden-section
+.. currentmodule:: {{ module }}
+
+
+{{ name | underline}}
+
+.. autoclass:: {{ name }}
+    :members:
+
+
+..
+  autogenerated from docs/_templates/classtemplate.rst
+  note it does not have :inherited-members:

docs/intermediate_representation/index.md

@@ -1,9 +1,24 @@
 # ONNX IR
 
+An in-memory IR that supports the full ONNX spec, designed for graph construction, analysis and transformation.
+
+## Features ✨
+
+- Full ONNX spec support: all valid models representable by ONNX protobuf, and a subset of invalid models (so you can load and fix them).
+- Low memory footprint: mmap'ed external tensors; unified interface for ONNX TensorProto, Numpy arrays and PyTorch Tensors etc. No tensor size limitation. Zero copies.
+- Straightforward access patterns: Access value information and traverse the graph topology at ease.
+- Robust mutation: Create as many iterators as you like on the graph while mutating it.
+- Speed: Performant graph manipulation, serialization/deserialization to Protobuf.
+- Pythonic and familiar APIs: Classes define Pythonic apis and still map to ONNX protobuf concepts in an intuitive way.
+- No protobuf dependency: The IR does not require protobuf once the model is converted to the IR representation, decoupling from the serialization format.
+
+## Get started
+
 ```{toctree}
 :maxdepth: 1
 
 getting_started
 tensors
 ir_api
+generated
 ```

docs/intermediate_representation/ir_api.md

@@ -1,9 +1,46 @@
 # onnxscript.ir
 
-<!-- TODO: Organize the orders and add tutorial -->
+```{eval-rst}
+.. automodule::onnxscript.ir
+```
+
+## IR objects
 
 ```{eval-rst}
-.. automodule:: onnxscript.ir
-    :members:
-    :undoc-members:
+.. currentmodule:: onnxscript
+.. autosummary::
+    :toctree: generated
+    :nosignatures:
+    :template: classtemplate.rst
+
+    ir.Model
+    ir.Graph
+    ir.GraphView
+    ir.Function
+    ir.Node
+    ir.Value
+    ir.Attr
+    ir.RefAttr
+    ir.Shape
+    ir.SymbolicDim
+    ir.TypeAndShape
+    ir.TensorType
+    ir.SparseTensorType
+    ir.SequenceType
+    ir.OptionalType
+    ir.Tensor
+    ir.ExternalTensor
+    ir.StringTensor
+```
+
+## Enums
+
+```{eval-rst}
+.. autosummary::
+    :toctree: generated
+    :nosignatures:
+    :template: classtemplate.rst
+
+    ir.DataType
+    ir.AttributeType
 ```

docs/intermediate_representation/tensors.md

@@ -167,7 +167,7 @@ The following example shows how to create a `FLOAT8E4M3FN` tensor, transform its
     print("tensor.numpy():", tensor.numpy())  # [0.00195312 0.00585938]
 
     # Compute
-    times_100 = tensor.numpy() * 100
+    times_100 = tensor.numpy() * np.array(100, dtype=tensor.numpy().dtype)
     print("times_100:", times_100)
 
     # Create a new tensor out of the new value; dtype must be specified