You like pytorch? You like micrograd? You love tinygrad! ❤️
Go to file
George Hotz 1654008c1f conv stride support 2020-10-26 08:54:43 -07:00
.github/workflows remove install from workflow 2020-10-24 09:03:22 -05:00
test conv stride support 2020-10-26 08:54:43 -07:00
tinygrad conv stride support 2020-10-26 08:54:43 -07:00
.gitignore start tinygrad 2020-10-17 22:57:01 -07:00
LICENSE readme 2020-10-18 11:27:37 -07:00
README.md oops, set to None 2020-10-25 08:28:18 -07:00
requirements.txt everyone's a dev 2020-10-24 09:01:15 -05:00

README.md

tinygrad

Unit Tests

For something in between a pytorch and a karpathy/micrograd

This may not be the best deep learning framework, but it is a deep learning framework.

The Tensor class is a wrapper around a numpy array, except it does Tensor things.

Example

import numpy as np
from tinygrad.tensor import Tensor

x = Tensor(np.eye(3))
y = Tensor(np.array([[2.0,0,-2.0]]))
z = y.dot(x).sum()
z.backward()

print(x.grad)  # dz/dx
print(y.grad)  # dz/dy

Same example in torch

import torch

x = torch.eye(3, requires_grad=True)
y = torch.tensor([[2.0,0,-2.0]], requires_grad=True)
z = y.matmul(x).sum()
z.backward()

print(x.grad)  # dz/dx
print(y.grad)  # dz/dy

Neural networks?

It turns out, a decent autograd tensor library is 90% of what you need for neural networks. Add an optimizer (SGD, RMSprop, and Adam implemented) from tinygrad.optim, write some boilerplate minibatching code, and you have all you need.

Neural network example (from test/test_mnist.py)

from tinygrad.tensor import Tensor
import tinygrad.optim as optim
from tinygrad.utils import layer_init_uniform

class TinyBobNet:
  def __init__(self):
    self.l1 = Tensor(layer_init_uniform(784, 128))
    self.l2 = Tensor(layer_init_uniform(128, 10))

  def forward(self, x):
    return x.dot(self.l1).relu().dot(self.l2).logsoftmax()

model = TinyBobNet()
optim = optim.SGD([model.l1, model.l2], lr=0.001)

# ... and complete like pytorch, with (x,y) data

out = model.forward(x)
loss = out.mul(y).mean()
loss.backward()
optim.step()