# "Dynamic Shapes" sample

This sample shows how to

1. Create a TensorFlow program that includes dynamic shapes in program inputs
   and outputs
2. Import that program into IREE's compiler
3. Compile that program to an IREE VM bytecode module
4. Load the compiled program using IREE's high level runtime C API
5. Call exported functions on the loaded program

Steps 1-2 are performed in Python via the
[`dynamic_shapes.ipynb`](./dynamic_shapes.ipynb)
[Colab](https://research.google.com/colaboratory/) notebook:

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/google/iree/blob/main/samples/dynamic_shapes/dynamic_shapes.ipynb)

Step 3 should be performed on your development host machine

Steps 4-5 are in [`main.c`](./main.c)

The program used to demonstrate includes functions with varying uses of
dynamic shapes:

```python
class DynamicShapesModule(tf.Module):
  # reduce_sum_1d (dynamic input size, static output size)
  #   e.g. [1, 2, 3] -> 6
  @tf.function(input_signature=[tf.TensorSpec([None], tf.int32)])
  def reduce_sum_1d(self, values):
    return tf.math.reduce_sum(values)

  # reduce_sum_2d (partially dynamic input size, static output size)
  #   e.g. [[1, 2, 3], [10, 20, 30]] -> [11, 22, 33]
  @tf.function(input_signature=[tf.TensorSpec([None, 3], tf.int32)])
  def reduce_sum_2d(self, values):
    return tf.math.reduce_sum(values, 0)

  # add_one (dynamic input size, dynamic output size)
  #   e.g. [1, 2, 3] -> [2, 3, 4]
  @tf.function(input_signature=[tf.TensorSpec([None], tf.int32)])
  def add_one(self, values):
    return tf.math.add(values, tf.constant(1, dtype=tf.int32))
```

## Background

Tensors are multi-dimensional arrays with a uniform type (e.g. int32, float32)
and a shape. Shapes consist of a rank and a list of dimensions and may be
static (i.e. fully known and fixed) or varying degrees of dynamic. See
TensorFlow's [Introduction to Tensors](https://www.tensorflow.org/guide/tensor)
for more information on how tensors are used in TensorFlow programs.

Dynamic shapes are useful for passing variable sized batches as input,
receiving variable length sentences of text as output, etc.

NOTE: as in other domains, providing more information to a compiler allows it
to generate more efficient code. As a general rule, the slowest varying
dimensions of program data like batch index or timestep are safer to treat as
dynamic than faster varying dimensions like image x/y/channel. See
[this paper](https://arxiv.org/pdf/2006.03031.pdf) for a discussion of the
challenges imposed by dynamic shapes and one project's approach to addressing
them.

## Instructions

1. Run the Colab notebook and download the `dynamic_shapes.mlir` file it
    generates

2. Build the `iree-compile` tool (see
    [here](https://google.github.io/iree/building-from-source/getting-started/)
    for general instructions on building using CMake)

    ```
    cmake -B ../iree-build/ -DCMAKE_BUILD_TYPE=RelWithDebInfo .
    cmake --build ../iree-build/ --target iree_tools_iree-compile
    ```

3. Compile the `dynamic_shapes.mlir` file using `iree-compile`. The
    [dylib-llvm-aot](https://google.github.io/iree/deployment-configurations/cpu-dylib/)
    configuration has the best support for dynamic shapes:

    ```
    ../iree-build/iree/tools/iree-compile \
        -iree-mlir-to-vm-bytecode-module \
        -iree-hal-target-backends=dylib-llvm-aot \
        -iree-input-type=mhlo \
        dynamic_shapes.mlir -o dynamic_shapes_dylib.vmfb
    ```

4. Build the `iree_samples_dynamic_shapes` CMake target

    ```
    cmake --build ../iree-build/ --target iree_samples_dynamic_shapes
    ```

5. Run the sample binary:

   ```
   ../iree-build/samples/dynamic_shapes/dynamic-shapes \
       /path/to/dynamic_shapes_dylib.vmfb dylib
   ```