“Dynamic Shapes” sample

This sample shows how to

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

Steps 1-2 are performed in Python via the dynamic_shapes.ipynb Colab notebook:

Step 3 should be performed on your development host machine

Steps 4-5 are in main.c

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

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 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 for a discussion of the challenges imposed by dynamic shapes and one project's approach to addressing them.

Instructions

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

Build the iree-compile tool (see here for general instructions on building using CMake)

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

Compile the dynamic_shapes.mlir file using iree-compile. The CPU configuration has the best support for dynamic shapes:

../iree-build/tools/iree-compile \
    --iree-hal-target-backends=llvm-cpu \
    --iree-input-type=mhlo_legacy \
    dynamic_shapes.mlir -o dynamic_shapes_cpu.vmfb

Build the iree_samples_dynamic_shapes CMake target
```
cmake --build ../iree-build/ --target iree_samples_dynamic_shapes
```
Alternatively if using a non-CMake build system the Makefile provided can be used as a reference of how to use the IREE runtime in an external project.

Run the sample binary:

../iree-build/samples/dynamic_shapes/dynamic-shapes \
    /path/to/dynamic_shapes_cpu.vmfb local-task