+
Skip to content

Changing conv2d and conv2d_batched to methods of tensors #221

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Oct 7, 2022
Merged

Changing conv2d and conv2d_batched to methods of tensors #221

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions src/nn/conv.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,7 +98,7 @@ where
type Output = Tensor3D<O, { (H + 2 * P - K) / S + 1 }, { (W + 2 * P - K) / S + 1 }, T>;

fn forward(&self, x: Tensor3D<I, H, W, T>) -> Self::Output {
conv2d::<T, I, O, K, S, P, H, W>(x, &self.weight, &self.bias)
x.conv2d::<O, K, S, P>(&self.weight, &self.bias)
}
}

Expand All @@ -120,7 +120,7 @@ where
type Output = Tensor4D<B, O, { (H + 2 * P - K) / S + 1 }, { (W + 2 * P - K) / S + 1 }, T>;

fn forward(&self, x: Tensor4D<B, I, H, W, T>) -> Self::Output {
conv2d_batched::<T, B, I, O, K, S, P, H, W>(x, &self.weight, &self.bias)
x.conv2d::<O, K, S, P>(&self.weight, &self.bias)
}
}

Expand Down
148 changes: 62 additions & 86 deletions src/tensor_ops/conv.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,91 +2,67 @@ use crate::devices::{Cpu, DeviceConv2D};
use crate::gradients::Tape;
use crate::prelude::*;

/// **Requires Nightly** Perform a 2d convolution.
///
/// TODO docstring
pub fn conv2d<
TAPE: Tape,
const IN_CHAN: usize,
const OUT_CHAN: usize,
const KERNEL: usize,
const STRIDE: usize,
const PADDING: usize,
const IN_HEIGHT: usize,
const IN_WIDTH: usize,
>(
x: Tensor3D<IN_CHAN, IN_HEIGHT, IN_WIDTH, TAPE>,
filters: &Tensor4D<OUT_CHAN, IN_CHAN, KERNEL, KERNEL>,
bias: &Tensor1D<OUT_CHAN>,
) -> Tensor3D<
OUT_CHAN,
{ (IN_HEIGHT + 2 * PADDING - KERNEL) / STRIDE + 1 },
{ (IN_WIDTH + 2 * PADDING - KERNEL) / STRIDE + 1 },
TAPE,
> {
let mut result = Tensor3D::zeros();
<Cpu as DeviceConv2D<STRIDE, PADDING>>::conv_forward(
x.data(),
filters.data(),
bias.data(),
result.mut_data(),
);

let f = filters.clone();
let (x, mut tape) = x.split_tape();
let phf = filters.phantom();
let phb = bias.phantom();
let phr = result.phantom();
tape.add_backward_op(move |grads| {
let (fg, bg, ig, rg) = grads.muts_and_ref(&phf, &phb, &x, &phr);
<Cpu as DeviceConv2D<STRIDE, PADDING>>::conv_backward(x.data(), f.data(), rg, ig, fg, bg);
});
result.put_tape(tape)
}
impl<const C: usize, const H: usize, const W: usize, T: Tape> Tensor3D<C, H, W, T> {
/// **Requires Nightly** Perform a 2d convolution
///
/// TODO docstring
pub fn conv2d<const O: usize, const K: usize, const S: usize, const P: usize>(
self,
filters: &Tensor4D<O, C, K, K>,
bias: &Tensor1D<O>,
) -> Tensor3D<O, { (H + 2 * P - K) / S + 1 }, { (W + 2 * P - K) / S + 1 }, T> {
let mut result = Tensor3D::zeros();
<Cpu as DeviceConv2D<S, P>>::conv_forward(
self.data(),
filters.data(),
bias.data(),
result.mut_data(),
);

/// **Requires Nightly** Perform a batched 2d convolution
///
/// TODO docstring
pub fn conv2d_batched<
TAPE: Tape,
const BATCH_SIZE: usize,
const IN_CHAN: usize,
const OUT_CHAN: usize,
const KERNEL: usize,
const STRIDE: usize,
const PADDING: usize,
const IN_HEIGHT: usize,
const IN_WIDTH: usize,
>(
x: Tensor4D<BATCH_SIZE, IN_CHAN, IN_HEIGHT, IN_WIDTH, TAPE>,
filters: &Tensor4D<OUT_CHAN, IN_CHAN, KERNEL, KERNEL>,
bias: &Tensor1D<OUT_CHAN>,
) -> Tensor4D<
BATCH_SIZE,
OUT_CHAN,
{ (IN_HEIGHT + 2 * PADDING - KERNEL) / STRIDE + 1 },
{ (IN_WIDTH + 2 * PADDING - KERNEL) / STRIDE + 1 },
TAPE,
> {
let mut result = Tensor4D::zeros();
for (x_i, r_i) in x.data().iter().zip(result.mut_data().iter_mut()) {
<Cpu as DeviceConv2D<STRIDE, PADDING>>::conv_forward(x_i, filters.data(), bias.data(), r_i);
let f = filters.clone();
let (x, mut tape) = self.split_tape();
let phf = filters.phantom();
let phb = bias.phantom();
let phr = result.phantom();
tape.add_backward_op(move |grads| {
let (fg, bg, ig, rg) = grads.muts_and_ref(&phf, &phb, &x, &phr);
<Cpu as DeviceConv2D<S, P>>::conv_backward(x.data(), f.data(), rg, ig, fg, bg);
});
result.put_tape(tape)
}
}

let f = filters.clone();

let (x, mut tape) = x.split_tape();
let phf = filters.phantom();
let phb = bias.phantom();
let phr = result.phantom();
tape.add_backward_op(move |grads| {
let (fg, bg, ig, r_grad) = grads.muts_and_ref(&phf, &phb, &x, &phr);
let f = f.data();
for ((x_i, rg_i), ig_i) in x.data().iter().zip(r_grad.iter()).zip(ig.iter_mut()) {
<Cpu as DeviceConv2D<STRIDE, PADDING>>::conv_backward(x_i, f, rg_i, ig_i, fg, bg);
impl<const B: usize, const C: usize, const H: usize, const W: usize, T: Tape>
Tensor4D<B, C, H, W, T>
{
/// **Requires Nightly** Perform a batched 2d convolution
///
/// TODO docstring
pub fn conv2d<const O: usize, const K: usize, const S: usize, const P: usize>(
self,
filters: &Tensor4D<O, C, K, K>,
bias: &Tensor1D<O>,
) -> Tensor4D<B, O, { (H + 2 * P - K) / S + 1 }, { (W + 2 * P - K) / S + 1 }, T> {
let mut result = Tensor4D::zeros();
for (x_i, r_i) in self.data().iter().zip(result.mut_data().iter_mut()) {
<Cpu as DeviceConv2D<S, P>>::conv_forward(x_i, filters.data(), bias.data(), r_i);
}
});
result.put_tape(tape)

let f = filters.clone();

let (x, mut tape) = self.split_tape();
let phf = filters.phantom();
let phb = bias.phantom();
let phr = result.phantom();
tape.add_backward_op(move |grads| {
let (fg, bg, ig, r_grad) = grads.muts_and_ref(&phf, &phb, &x, &phr);
let f = f.data();
for ((x_i, rg_i), ig_i) in x.data().iter().zip(r_grad.iter()).zip(ig.iter_mut()) {
<Cpu as DeviceConv2D<S, P>>::conv_backward(x_i, f, rg_i, ig_i, fg, bg);
}
});
result.put_tape(tape)
}
}

#[cfg(test)]
Expand All @@ -111,7 +87,7 @@ mod tests {
[-0.86713916, 0.52773184, -0.95238322],
[-0.64531374, 0.77809018, -0.49099201],
]]);
let result = conv2d::<_, 1, 2, 2, 1, 0, 2, 3>(x.trace(), &weight, &bias);
let result = x.trace().conv2d::<2, 2, 1, 0>(&weight, &bias);
assert_close(
result.data(),
&[[[0.24369538, 0.71453357]], [[-0.69169492, -0.06172103]]],
Expand Down Expand Up @@ -154,7 +130,7 @@ mod tests {
[-0.31547278, 0.58071911, 0.86612970],
]]);

let result = conv2d::<OwnedTape, 1, 2, 2, 2, 0, 2, 3>(x.trace(), &weight, &bias);
let result = x.trace().conv2d::<2, 2, 2, 0>(&weight, &bias);
assert_close(result.data(), &[[[-0.29368058]], [[0.30018353]]]);

let g = backward(result.exp().mean());
Expand Down Expand Up @@ -188,7 +164,7 @@ mod tests {

let x = tensor([[[-0.32224107, -0.32800716]], [[-1.13570976, 0.93713200]]]);

let result = conv2d::<OwnedTape, 2, 3, 2, 1, 1, 1, 2>(x.trace(), &weight, &bias);
let result = x.trace().conv2d::<3, 2, 1, 1>(&weight, &bias);

#[rustfmt::skip]
assert_close(
Expand Down Expand Up @@ -236,7 +212,7 @@ mod tests {
#[rustfmt::skip]
let x = tensor([[[0.69103152, 0.25624934],[-0.38448590, 0.03110456],[0.83753252, 0.53786588],[1.15540242, -0.54148245]]]);

let result = conv2d::<OwnedTape, 1, 2, 3, 3, 4, 4, 2>(x.trace(), &weight, &bias);
let result = x.trace().conv2d::<2, 3, 3, 4>(&weight, &bias);

#[rustfmt::skip]
assert_close(
Expand Down Expand Up @@ -283,7 +259,7 @@ mod tests {
[[[-0.22305037, 0.63030297], [0.65323567, -0.68972057]],[[-0.50617385, -0.87281805], [0.30253950, -1.75082350]]],
[[[1.65487242, 0.44441956], [-0.45107457, 1.41857898]],[[1.00477660, -0.16381662], [0.40009478, -0.57880658]]],
]);
let result = conv2d_batched::<OwnedTape, 3, 2, 3, 1, 1, 0, 2, 2>(x.trace(), &weight, &bias);
let result = x.trace().conv2d::<3, 1, 1, 0>(&weight, &bias);

#[rustfmt::skip]
assert_close(
Expand Down
点击 这是indexloc提供的php浏览器服务，不要输入任何密码和下载