Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

https://github.com/cgarciae/treex

A Pytree Module system for Deep Learning in JAX
https://github.com/cgarciae/treex

deep-learning jax

Last synced: 4 months ago
JSON representation

A Pytree Module system for Deep Learning in JAX

Lists

README 

        
_Deprecation Notice_: This library was an experiment trying to get pytree Modules working with Flax-like colletions. I'd currently recommend the following alternatives:

* Just custom pytrees: [simple_pytree](https://github.com/cgarciae/simple-pytree)

* Pytree module system: [equinox](https://github.com/patrick-kidger/equinox)

* Production ready module system: [flax](https://github.com/google/flax)



# Treex



_A Pytree Module system for Deep Learning in JAX_



#### Main Features



* 💡 **Intuitive**: Modules contain their own parameters and respect Object Oriented semantics like in PyTorch and Keras.

* 🌳 **Pytree-based**:  Modules are Pytrees whose leaves are its parameters, meaning they are fully compatible with `jit`, `grad`, `vmap`, etc.



Treex is implemented on top of [Treeo](https://github.com/cgarciae/treeo) and reexports all of its API for convenience.



[Getting Started](#getting-started) | [User Guide](https://cgarciae.github.io/treex/user-guide/intro) | [Examples](#examples) |  [Documentation](https://cgarciae.github.io/treex)



## What is included?

* A base `Module` class.

* A `nn` module for with common layers implemented as wrappers over Flax layers.

* A `losses` module with common loss functions.

* A `metrics` module with common metrics.

* An `Optimizer` class that can wrap any optax optimizer.

## Why Treex?



Show



Despite all JAX benefits, current Module systems are not intuitive to new users and add additional complexity not present in frameworks like PyTorch or Keras. Treex takes inspiration from S4TF and delivers an intuitive experience using JAX Pytree infrastructure.



Current Alternative's Drawbacks and Solutions



Currently we have many alternatives like Flax, Haiku, Objax, that have one or more of the following drawbacks:



* Module structure and parameter structure are separate, and parameters have to be manipulated around by the end-user, which is not intuitive. In Treex, parameters are stored in the modules themselves and can be accessed directly.

* Monadic architecture adds complexity. Flax and Haiku use an `apply` method to call modules that set a context with parameters, rng, and different metadata, which adds additional overhead to the API and creates an asymmetry in how Modules are being used inside and outside a context. In Treex, modules can be called directly.

* Among different frameworks, parameter surgery requires special consideration and is challenging to implement. Consider a standard workflow such as transfer learning, transferring parameters and state from a  pre-trained module or submodule as part of a new module; in different frameworks, we have to know precisely how to extract their parameters and how to insert them into the new parameter structure/dictionaries such that it is in agreement with the new module structure. In Treex, just as in PyTorch / Keras, we enable to pass the (sub)module to the new module, and parameters are automatically added to the new structure.

* Multiple frameworks deviate from JAX semantics and require particular versions of `jit`, `grad`, `vmap`, etc., which makes it harder to integrate with other JAX libraries. Treex's Modules are plain old JAX PyTrees and are compatible with any JAX library that supports them.

* Other Pytree-based approaches like Parallax and Equinox do not have a total state management solution to handle complex states as encountered in Flax. Treex has the Filter and Update API, which is very expressive and can effectively handle systems with a complex state.



## Installation

Install using pip:

```bash

pip install treex

```



## Getting Started



This is a small appetizer to give you a feel for how using Treex looks like, be sure to checkout the [User Guide](https://cgarciae.github.io/treex/user-guide/intro) for a more in-depth explanation.

```python

import treex as tx

import numpy as np

import jax, optax



# create some data

x = np.random.uniform(size=(50, 1))

y = 1.3 * x ** 2 - 0.3 + np.random.normal(size=x.shape)



# initialize a Module, its simple

model = tx.MLP([64, 1]).init(key=42, inputs=x)

# define an optimizer, init with model params

optimizer = tx.Optimizer(optax.adam(4e-3)).init(model)



# define loss function, notice

# Modules are jit-abel and differentiable 🤯

@jax.jit

@jax.grad

def loss_fn(model: tx.MLP, x, y):

    # forward is a simple call

    preds = model(x)

    # MSE

    return ((preds - y) ** 2).mean()



# basic training loop

for step in range(500):



    # grads have the same type as model

    grads: tx.MLP = loss_fn(model, x, y)

    # apply the gradient updates

    model = optimizer.update(grads, model)



# Pytorch-like eval mode

model = model.eval()

preds = model(x)

```

#### Custom Modules



Show



Modules are Treeo `Tree`s, which are Pytrees. When creating core layers you often mark fields that will contain state that JAX should be aware as `nodes` by assigning class variables to the output of functions like `tx.Parameter.node()`:



```python

import treex as tx



class Linear(tx.Module):

    # use Treeo's API to define Parameter nodes

    w: jnp.ndarray = tx.Parameter.node()

    b: jnp.ndarray = tx.Parameter.node()



    def __init__(self, features_out: int):

        self.features_out = features_out



    def __call__(self, x: jnp.ndarray) -> jnp.ndarray:

        # init will call forward, we can know if we are inside it

        if self.initializing():

            # `next_key` only available during `init`

            key = tx.next_key() 

            # leverage shape inference

            self.w = jax.random.uniform(

                key, shape=[x.shape[-1], self.features_out]

            )

            self.b = jnp.zeros(shape=[self.features_out])



        # linear forward

        return jnp.dot(x, self.w) + self.b



model = Linear(10).init(key=42, inputs=x)

```

Node field types (e.g. `tx.Parameter`) are called Kinds and Treex exports a whole family of Kinds which serve for differente purposes such as holding non-differentiable state (`tx.BatchStats`), metric's state (`tx.MetricState`), logging, etc. Checkout the [kinds](https://cgarciae.github.io/treex/user-guide/kinds) section for more information.



#### Composite Modules



Show



Composite Modules usually hold and call other Modules within them, while they would be instantiate inside `__init__` and used later in `__call__` like in Pytorch / Keras, in Treex you usually leverage the `@tx.compact` decorator over the `__call__` method to define the submodules inline.

```python

class MLP(tx.Module):

    def __init__(self, features: Sequence[int]):

        self.features = features



    # compact lets you define submodules on the fly

    @tx.compact

    def __call__(self, x: jnp.ndarray) -> jnp.ndarray:

        for units in self.features[:-1]:

            x = Linear(units)(x)

            x = jax.nn.relu(x)



        return Linear(self.features[-1])(x)



model = MLP([32, 10]).init(key=42, inputs=x)

```

Under the hood all calls to submodule constructors (e.g. `Linear(...)`) inside `compact` are assigned to fields in the parent Module (`MLP`) so they are part of the same Pytree, their field names are available under the `._subtrees` attribute. `compact` must always define submodules in the same order.



## Status

Treex is in an early stage, things might break between versions but we will respect semanting versioning. Since Treex layers are numerically equivalent to Flax, it borrows some maturity and yields more confidence over its results. Feedback is much appreciated.



**Roadmap**:



- Wrap all Flax Linen Modules

- Implement more layers, losses, and metrics.

- Create applications and pretrained Modules.



Contributions are welcomed!



## Sponsors 💚

* [Quansight](https://www.quansight.com) - paid development time



## Examples

Checkout the [/examples](examples) directory for more detailed examples. Here are a few additional toy examples:



#### Linear Regression

This is a simple but realistic example of how Treex is used.



```python

from functools import partial

from typing import Union

import jax

import jax.numpy as jnp

import matplotlib.pyplot as plt

import numpy as np

import optax

import treex as tx



x = np.random.uniform(size=(500, 1))

y = 1.4 * x - 0.3 + np.random.normal(scale=0.1, size=(500, 1))



# differentiate only w.r.t. parameters

def loss_fn(params, model, x, y):

    # merge params into model

    model = model.merge(params)



    preds = model(x)

    loss = jnp.mean((preds - y) ** 2)



    # the model may contain state updates

    # so it should be returned

    return loss, model



grad_fn = jax.value_and_grad(loss_fn, has_aux=True)



# both model and optimizer are jit-able

@jax.jit

def train_step(model, x, y, optimizer):

    # select only the parameters

    params = model.parameters()



    (loss, model), grads = grad_fn(params, model, x, y)



    # update params and model

    params = optimizer.update(grads, params)

    model = model.merge(params)



    # return new model and optimizer

    return loss, model, optimizer



model = tx.Linear(1).init(42, x)

optimizer = tx.Optimizer(optax.adam(0.01)).init(model)



for step in range(300):

    loss, model, optimizer = train_step(model, x, y, optimizer)

    if step % 50 == 0:

        print(f"loss: {loss:.4f}")



# eval mode "turns off" layers like Dropout / BatchNorm

model = model.eval()



X_test = np.linspace(x.min(), x.max(), 100)[:, None]

preds = model(X_test)



plt.scatter(x, y, c="k", label="data")

plt.plot(X_test, preds, c="b", linewidth=2, label="prediction")

plt.legend()

plt.show()

```



#### A Stateful Module

Here is an example of creating a stateful module of a `RollingMean` metric and using them with `jax.jit`. For a real use cases use the metrics inside `treex.metrics`.



```python

class RollingMean(tx.Module):

    count: jnp.ndarray = tx.State.node()

    total: jnp.ndarray = tx.State.node()



    def __init__(self):

        self.count = jnp.array(0, dtype=jnp.int32)

        self.total = jnp.array(0.0, dtype=jnp.float32)



    def __call__(self, x: jnp.ndarray) -> jnp.ndarray:

        self.count += np.prod(x.shape)

        self.total += x.sum()



        return self.total / self.count



@jax.jit

def update(x: jnp.ndarray, metric: RollingMean) -> Tuple[jnp.ndarray, RollingMean]:

    mean = metric(x)

    return mean, metric # return mean value and updated metric



metric = RollingMean()



for i in range(10):

    x = np.random.uniform(-1, 1, size=(100, 1))

    mean, metric = update(x, metric)



print(mean)

```