malmaud / functorch

functorch is a prototype of JAX-like composable function transforms for PyTorch.

functorch

Why functorch? | Install guide | Transformations | Future Plans

This library is currently under heavy development - if you have suggestions on the API or use-cases you'd like to be covered, please open an github issue or reach out. We'd love to hear about how you're using the library.

functorch is a prototype of JAX-like composable FUNCtion transforms for pyTORCH.

It aims to provide composable vmap and grad transforms that work with PyTorch modules and PyTorch autograd with good eager-mode performance. Because this project requires some investment, we'd love to hear from and work with early adopters to shape the design. Please reach out on the issue tracker if you're interested in using this for your project.

In addition, there is experimental functionality to trace through these transformations using FX in order to capture the results of these transforms ahead of time. This would allow us to compile the results of vmap or grad to improve performance.

Why composable function transforms?

There are a number of use cases that are tricky to do in PyTorch today:

• computing per-sample-gradients (or other per-sample quantities)
• running ensembles of models on a single machine
• efficiently batching together tasks in the inner-loop of MAML
• efficiently computing Jacobians and Hessians
• efficiently computing batched Jacobians and Hessians

Composing vmap, grad, and vjp transforms allows us to express the above without designing a separate subsystem for each. This idea of composable function transforms comes from the JAX framework.

Install

Colab

Follow the instructions in this Colab notebook

Binaries

First, set up an environment. We will be installing a nightly PyTorch binary as well as functorch. If you're using conda, create a conda environment:

conda create --name functorch
conda activate functorch

If you wish to use venv instead:

python -m venv functorch-env
source functorch-env/bin/activate

Next, install one of the following following PyTorch nightly binaries.

# For CUDA 10.2
pip install --pre torch -f https://download.pytorch.org/whl/nightly/cu102/torch_nightly.html --upgrade
# For CUDA 11.1
pip install --pre torch -f https://download.pytorch.org/whl/nightly/cu111/torch_nightly.html --upgrade
# For CPU-only build
pip install --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html --upgrade

If you already have a nightly of PyTorch installed and wanted to upgrade it (recommended!), append --upgrade to one of those commands.

Install functorch:

pip install ninja  # Makes the build go faster
pip install --user "git+https://github.com/pytorch/functorch.git"

Run a quick sanity check in python:

>>> import torch
>>> from functorch import vmap
>>> x = torch.randn(3)
>>> y = vmap(torch.sin)(x)
>>> assert torch.allclose(y, x.sin())

From Source

functorch is a PyTorch C++ Extension module. To install,

• Install PyTorch from source. functorch usually runs on the latest development version of PyTorch.
• Run python setup.py install. You can use DEBUG=1 to compile in debug mode.

Then, try to run some tests to make sure all is OK:

pytest test/test_vmap.py -v
pytest test/test_eager_transforms.py -v

What are the transforms?

Right now, we support the following transforms:

• grad, vjp, jacrev
• vmap

Furthermore, we have some utilities for working with PyTorch modules.

• make_functional(model)
• make_functional_with_buffers(model)

vmap

Note: vmap imposes restrictions on the code that it can be used on. For more details, please read its docstring.

vmap(func)(*inputs) is a transform that adds a dimension to all Tensor operations in func. vmap(func) returns a few function that maps func over some dimension (default: 0) of each Tensor in inputs.

vmap is useful for hiding batch dimensions: one can write a function func that runs on examples and then lift it to a function that can take batches of examples with vmap(func), leading to a simpler modeling experience:

>>> from functorch import vmap
>>> batch_size, feature_size = 3, 5
>>> weights = torch.randn(feature_size, requires_grad=True)
>>>
>>> def model(feature_vec):
>>>     # Very simple linear model with activation
>>>     assert feature_vec.dim() == 1
>>>     return feature_vec.dot(weights).relu()
>>>
>>> examples = torch.randn(batch_size, feature_size)
>>> result = vmap(model)(examples)

grad

grad(func)(*inputs) assumes func returns a single-element Tensor. It compute the gradients of the output of func w.r.t. to inputs[0].

>>> from functorch import grad
>>> x = torch.randn([])
>>> cos_x = grad(lambda x: torch.sin(x))(x)
>>> assert torch.allclose(cos_x, x.cos())
>>>
>>> # Second-order gradients
>>> neg_sin_x = grad(grad(lambda x: torch.sin(x)))(x)
>>> assert torch.allclose(neg_sin_x, -x.sin())

When composed with vmap, grad can be used to compute per-sample-gradients:

>>> from functorch import vmap
>>> batch_size, feature_size = 3, 5
>>>
>>> def model(weights,feature_vec):
>>>     # Very simple linear model with activation
>>>     assert feature_vec.dim() == 1
>>>     return feature_vec.dot(weights).relu()
>>>
>>> def compute_loss(weights, example, target):
>>>     y = model(weights, example)
>>>     return ((y - target) ** 2).mean()  # MSELoss
>>>
>>> weights = torch.randn(feature_size, requires_grad=True)
>>> examples = torch.randn(batch_size, feature_size)
>>> targets = torch.randn(batch_size)
>>> inputs = (weights,examples, targets)
>>> grad_weight_per_example = vmap(grad(compute_loss), in_dims=(None, 0, 0))(*inputs)

vjp and jacrev

>>> from functorch import vjp
>>> outputs, vjp_fn = vjp(func, inputs); vjps = vjp_fn(*cotangents)

The vjp transform applies func to inputs and returns a new function that computes vjps given some cotangents Tensors.

>>> from functorch import jacrev
>>> x = torch.randn(5)
>>> jacobian = jacrev(torch.sin)(x)
>>> expected = torch.diag(torch.cos(x))
>>> assert torch.allclose(jacobian, expected)

Use jacrev to compute the jacobian. This can be composed with vmap to produce batched jacobians:

>>> x = torch.randn(64, 5)
>>> jacobian = vmap(jacrev(torch.sin))(x)
>>> assert jacobian.shape == (64, 5, 5)

jacrev can be composed with itself to produce hessians:

>>> def f(x):
>>>   return x.sin().sum()
>>>
>>> x = torch.randn(5)
>>> hessian = jacrev(jacrev(f))(x)

Tracing through the transformations

We can also trace through these transformations in order to capture the results as new code using make_fx. There is also experimental integration with the NNC compiler (only works on CPU for now!).

>>> from functorch import make_fx, grad
>>> def f(x):
>>>     return torch.sin(x).sum()
>>> x = torch.randn(100)
>>> grad_f = make_fx(grad(f))(x)
>>> print(grad_f.code)

def forward(self, x_1):
    sin = torch.ops.aten.sin(x_1)
    sum_1 = torch.ops.aten.sum(sin, None);  sin = None
    cos = torch.ops.aten.cos(x_1);  x_1 = None
    _tensor_constant0 = self._tensor_constant0
    mul = torch.ops.aten.mul(_tensor_constant0, cos);  _tensor_constant0 = cos = None
    return mul

We can also try compiling it with NNC (even more experimental)!.

>>> from functorch import nnc_jit
>>> jit_f = nnc_jit(grad(f))

Check examples/nnc for some example benchmarks.

Working with NN modules: make_functional and friends

Sometimes you may want to perform a transform with respect to the parameters and/or buffers of an nn.Module. This can happen for example in:

• model ensembling, where all of your weights and buffers have an additional dimension
• per-sample-gradient computation where you want to compute per-sample-grads of the loss with respect to the model parameters

Our solution to this right now is an API that, given an nn.Module, creates a stateless version of it that can be called like a function.

• make_functional(model) returns a functional version of model and the model.parameters()
• make_functional_with_buffers(model) returns a functional version of model and the model.parameters() and model.buffers().

Here's an example where we compute per-sample-gradients using an nn.Linear layer:

import torch
from functorch import make_functional, vmap, grad

model = torch.nn.Linear(3, 3)
data = torch.randn(64, 3)
targets = torch.randn(64, 3)

func_model, params = make_functional(model)

def compute_loss(params, data, targets):
    preds = func_model(params, data)
    return torch.mean((preds - targets) ** 2)

per_sample_grads = vmap(grad(compute_loss), (None, 0, 0))(params, data, targets)

If you're making an ensemble of models, you may find combine_state_for_ensemble useful.

Debugging

functorch._C.dump_tensor: Dumps dispatch keys on stack functorch._C._set_vmap_fallback_warning_enabled(False) if the vmap warning spam bothers you.

Future Plans

In the end state, we'd like to upstream this into PyTorch once we iron out the design details. To figure out the details, we need your help -- please send us your use cases by starting a conversation in the issue tracker or try out the prototype.

License

Functorch has a BSD-style license, as found in the LICENSE file.

Citing functorch

If you use functorch in your publication, please cite it by using the following BibTeX entry.

@Misc{functorch2021,
  author =       {Horace He, Richard Zou},
  title =        {functorch: JAX-like composable function transforms for PyTorch},
  howpublished = {\url{https://github.com/pytorch/functorch}},
  year =         {2021}
}