注意
前往底部 下载完整示例代码。
层归一化¶
在本教程中，你将编写一个高性能的层归一化 kernel，其运行速度比 PyTorch 实现更快。
通过本教程，你将学习到
在 Triton 中实现反向传播。
在 Triton 中实现并行归约。
动机¶
LayerNorm 算子首次在 [BA2016] 中提出，用于提高序列模型（例如 Transformers）或小批量神经网络的性能。它接受向量 \(x\) 作为输入，并生成形状相同的向量 \(y\) 作为输出。归一化是通过减去 \(x\) 的均值并除以标准差来实现的。归一化后，应用具有权重 \(w\) 和偏置 \(b\) 的可学习线性变换。前向传播可以表示为
\[y = \frac{ x - \text{E}[x] }{ \sqrt{\text{Var}(x) + \epsilon} } * w + b\]
其中 \(\epsilon\) 是添加到分母中的一个用于数值稳定性的小常数。我们先来看看前向传播的实现。
import torch
import triton
import triton.language as tl
try:
# This is https://github.com/NVIDIA/apex, NOT the apex on PyPi, so it
# should not be added to extras_require in setup.py.
import apex
HAS_APEX = True
except ModuleNotFoundError:
HAS_APEX = False
DEVICE = triton.runtime.driver.active.get_active_torch_device()
@triton.jit
def _layer_norm_fwd_fused(
X, # pointer to the input
Y, # pointer to the output
W, # pointer to the weights
B, # pointer to the biases
Mean, # pointer to the mean
Rstd, # pointer to the 1/std
stride, # how much to increase the pointer when moving by 1 row
N, # number of columns in X
eps, # epsilon to avoid division by zero
BLOCK_SIZE: tl.constexpr,
):
# Map the program id to the row of X and Y it should compute.
row = tl.program_id(0)
Y += row * stride
X += row * stride
# Compute mean
mean = 0
_mean = tl.zeros([BLOCK_SIZE], dtype=tl.float32)
for off in range(0, N, BLOCK_SIZE):
cols = off + tl.arange(0, BLOCK_SIZE)
a = tl.load(X + cols, mask=cols < N, other=0.).to(tl.float32)
_mean += a
mean = tl.sum(_mean, axis=0) / N
# Compute variance
_var = tl.zeros([BLOCK_SIZE], dtype=tl.float32)
for off in range(0, N, BLOCK_SIZE):
cols = off + tl.arange(0, BLOCK_SIZE)
x = tl.load(X + cols, mask=cols < N, other=0.).to(tl.float32)
x = tl.where(cols < N, x - mean, 0.)
_var += x * x
var = tl.sum(_var, axis=0) / N
rstd = 1 / tl.sqrt(var + eps)
# Write mean / rstd
tl.store(Mean + row, mean)
tl.store(Rstd + row, rstd)
# Normalize and apply linear transformation
for off in range(0, N, BLOCK_SIZE):
cols = off + tl.arange(0, BLOCK_SIZE)
mask = cols < N
w = tl.load(W + cols, mask=mask)
b = tl.load(B + cols, mask=mask)
x = tl.load(X + cols, mask=mask, other=0.).to(tl.float32)
x_hat = (x - mean) * rstd
y = x_hat * w + b
# Write output
tl.store(Y + cols, y, mask=mask)
反向传播¶
层归一化算子的反向传播比前向传播稍微复杂一些。设 \(\hat{x}\) 是线性变换前的归一化输入 \(\frac{ x - \text{E}[x] }{ \sqrt{\text{Var}(x) + \epsilon} }\)，则 \(x\) 的向量-雅可比积 (VJP) \(\nabla_{x}\) 由下式给出
\[\nabla_{x} = \frac{1}{\sigma}\Big( \nabla_{y} \odot w - \underbrace{ \big( \frac{1}{N} \hat{x} \cdot (\nabla_{y} \odot w) \big) }_{c_1} \odot \hat{x} - \underbrace{ \frac{1}{N} \nabla_{y} \cdot w }_{c_2} \Big)\]
其中 \(\odot\) 表示逐元素乘法，\(\cdot\) 表示点积，\(\sigma\) 是标准差。\(c_1\) 和 \(c_2\) 是为了提高后续实现的易读性而引入的中间常数。
对于权重 \(w\) 和偏置 \(b\)，VJP \(\nabla_{w}\) 和 \(\nabla_{b}\) 更直接
\[\nabla_{w} = \nabla_{y} \odot \hat{x} \quad \text{and} \quad \nabla_{b} = \nabla_{y}\]
由于同一批次中的所有行都使用相同的权重 \(w\) 和偏置 \(b\)，它们的梯度需要求和。为了高效地执行此步骤，我们使用并行归约策略：每个 kernel 实例将某些行的部分 \(\nabla_{w}\) 和 \(\nabla_{b}\) 累加到 \(\text{GROUP_SIZE_M}\) 个独立缓冲区中的一个。这些缓冲区保留在 L2 缓存中，然后由另一个函数进一步归约以计算实际的 \(\nabla_{w}\) 和 \(\nabla_{b}\)。
假设输入行数 \(M = 4\) 且 \(\text{GROUP_SIZE_M} = 2\)，下面是 \(\nabla_{w}\) 的并行归约策略图示（\(\nabla_{b}\) 为简洁起见省略）
在阶段 1 中，颜色相同的 X 行共享同一缓冲区，因此使用锁来确保同一时间只有一个 kernel 实例写入缓冲区。在阶段 2 中，缓冲区被进一步归约以计算最终的 \(\nabla_{w}\) 和 \(\nabla_{b}\)。在以下实现中，阶段 1 由函数 _layer_norm_bwd_dx_fused 实现，阶段 2 由函数 _layer_norm_bwd_dwdb 实现。
@triton.jit
def _layer_norm_bwd_dx_fused(DX, # pointer to the input gradient
DY, # pointer to the output gradient
DW, # pointer to the partial sum of weights gradient
DB, # pointer to the partial sum of biases gradient
X, # pointer to the input
W, # pointer to the weights
Mean, # pointer to the mean
Rstd, # pointer to the 1/std
Lock, # pointer to the lock
stride, # how much to increase the pointer when moving by 1 row
N, # number of columns in X
GROUP_SIZE_M: tl.constexpr, BLOCK_SIZE_N: tl.constexpr):
# Map the program id to the elements of X, DX, and DY it should compute.
row = tl.program_id(0)
cols = tl.arange(0, BLOCK_SIZE_N)
mask = cols < N
X += row * stride
DY += row * stride
DX += row * stride
# Offset locks and weights/biases gradient pointer for parallel reduction
lock_id = row % GROUP_SIZE_M
Lock += lock_id
Count = Lock + GROUP_SIZE_M
DW = DW + lock_id * N + cols
DB = DB + lock_id * N + cols
# Load data to SRAM
x = tl.load(X + cols, mask=mask, other=0).to(tl.float32)
dy = tl.load(DY + cols, mask=mask, other=0).to(tl.float32)
w = tl.load(W + cols, mask=mask).to(tl.float32)
mean = tl.load(Mean + row)
rstd = tl.load(Rstd + row)
# Compute dx
xhat = (x - mean) * rstd
wdy = w * dy
xhat = tl.where(mask, xhat, 0.)
wdy = tl.where(mask, wdy, 0.)
c1 = tl.sum(xhat * wdy, axis=0) / N
c2 = tl.sum(wdy, axis=0) / N
dx = (wdy - (xhat * c1 + c2)) * rstd
# Write dx
tl.store(DX + cols, dx, mask=mask)
# Accumulate partial sums for dw/db
partial_dw = (dy * xhat).to(w.dtype)
partial_db = (dy).to(w.dtype)
while tl.atomic_cas(Lock, 0, 1) == 1:
pass
count = tl.load(Count)
# First store doesn't accumulate
if count == 0:
tl.atomic_xchg(Count, 1)
else:
partial_dw += tl.load(DW, mask=mask)
partial_db += tl.load(DB, mask=mask)
tl.store(DW, partial_dw, mask=mask)
tl.store(DB, partial_db, mask=mask)
# need a barrier to ensure all threads finished before
# releasing the lock
tl.debug_barrier()
# Release the lock
tl.atomic_xchg(Lock, 0)
@triton.jit
def _layer_norm_bwd_dwdb(DW, # pointer to the partial sum of weights gradient
DB, # pointer to the partial sum of biases gradient
FINAL_DW, # pointer to the weights gradient
FINAL_DB, # pointer to the biases gradient
M, # GROUP_SIZE_M
N, # number of columns
BLOCK_SIZE_M: tl.constexpr, BLOCK_SIZE_N: tl.constexpr):
# Map the program id to the elements of DW and DB it should compute.
pid = tl.program_id(0)
cols = pid * BLOCK_SIZE_N + tl.arange(0, BLOCK_SIZE_N)
dw = tl.zeros((BLOCK_SIZE_M, BLOCK_SIZE_N), dtype=tl.float32)
db = tl.zeros((BLOCK_SIZE_M, BLOCK_SIZE_N), dtype=tl.float32)
# Iterate through the rows of DW and DB to sum the partial sums.
for i in range(0, M, BLOCK_SIZE_M):
rows = i + tl.arange(0, BLOCK_SIZE_M)
mask = (rows[:, None] < M) & (cols[None, :] < N)
offs = rows[:, None] * N + cols[None, :]
dw += tl.load(DW + offs, mask=mask, other=0.)
db += tl.load(DB + offs, mask=mask, other=0.)
# Write the final sum to the output.
sum_dw = tl.sum(dw, axis=0)
sum_db = tl.sum(db, axis=0)
tl.store(FINAL_DW + cols, sum_dw, mask=cols < N)
tl.store(FINAL_DB + cols, sum_db, mask=cols < N)
基准测试¶
现在我们可以比较我们的 kernel 与 PyTorch 的性能。这里我们关注每个 feature 小于 64KB 的输入。具体来说，可以将 'mode': 'backward' 设置为对反向传播进行基准测试。
class LayerNorm(torch.autograd.Function):
@staticmethod
def forward(ctx, x, normalized_shape, weight, bias, eps):
# allocate output
y = torch.empty_like(x)
# reshape input data into 2D tensor
x_arg = x.reshape(-1, x.shape[-1])
M, N = x_arg.shape
mean = torch.empty((M, ), dtype=torch.float32, device=x.device)
rstd = torch.empty((M, ), dtype=torch.float32, device=x.device)
# Less than 64KB per feature: enqueue fused kernel
MAX_FUSED_SIZE = 65536 // x.element_size()
BLOCK_SIZE = min(MAX_FUSED_SIZE, triton.next_power_of_2(N))
if N > BLOCK_SIZE:
raise RuntimeError("This layer norm doesn't support feature dim >= 64KB.")
# heuristics for number of warps
num_warps = min(max(BLOCK_SIZE // 256, 1), 8)
# enqueue kernel
_layer_norm_fwd_fused[(M, )]( #
x_arg, y, weight, bias, mean, rstd, #
x_arg.stride(0), N, eps, #
BLOCK_SIZE=BLOCK_SIZE, num_warps=num_warps, num_ctas=1)
ctx.save_for_backward(x, weight, bias, mean, rstd)
ctx.BLOCK_SIZE = BLOCK_SIZE
ctx.num_warps = num_warps
ctx.eps = eps
return y
@staticmethod
def backward(ctx, dy):
x, w, b, m, v = ctx.saved_tensors
# heuristics for amount of parallel reduction stream for DW/DB
N = w.shape[0]
GROUP_SIZE_M = 64
if N <= 8192: GROUP_SIZE_M = 96
if N <= 4096: GROUP_SIZE_M = 128
if N <= 1024: GROUP_SIZE_M = 256
# allocate output
locks = torch.zeros(2 * GROUP_SIZE_M, dtype=torch.int32, device=w.device)
_dw = torch.zeros((GROUP_SIZE_M, N), dtype=x.dtype, device=w.device)
_db = torch.zeros((GROUP_SIZE_M, N), dtype=x.dtype, device=w.device)
dw = torch.empty((N, ), dtype=w.dtype, device=w.device)
db = torch.empty((N, ), dtype=w.dtype, device=w.device)
dx = torch.empty_like(dy)
# enqueue kernel using forward pass heuristics
# also compute partial sums for DW and DB
x_arg = x.reshape(-1, x.shape[-1])
M, N = x_arg.shape
_layer_norm_bwd_dx_fused[(M, )]( #
dx, dy, _dw, _db, x, w, m, v, locks, #
x_arg.stride(0), N, #
BLOCK_SIZE_N=ctx.BLOCK_SIZE, #
GROUP_SIZE_M=GROUP_SIZE_M, #
num_warps=ctx.num_warps)
grid = lambda meta: (triton.cdiv(N, meta['BLOCK_SIZE_N']), )
# accumulate partial sums in separate kernel
_layer_norm_bwd_dwdb[grid](
_dw, _db, dw, db, min(GROUP_SIZE_M, M), N, #
BLOCK_SIZE_M=32, #
BLOCK_SIZE_N=128, num_ctas=1)
return dx, None, dw, db, None
layer_norm = LayerNorm.apply
def test_layer_norm(M, N, dtype, eps=1e-5, device=DEVICE):
# create data
x_shape = (M, N)
w_shape = (x_shape[-1], )
weight = torch.rand(w_shape, dtype=dtype, device=device, requires_grad=True)
bias = torch.rand(w_shape, dtype=dtype, device=device, requires_grad=True)
x = -2.3 + 0.5 * torch.randn(x_shape, dtype=dtype, device=device)
dy = .1 * torch.randn_like(x)
x.requires_grad_(True)
# forward pass
y_tri = layer_norm(x, w_shape, weight, bias, eps)
y_ref = torch.nn.functional.layer_norm(x, w_shape, weight, bias, eps).to(dtype)
# backward pass (triton)
y_tri.backward(dy, retain_graph=True)
dx_tri, dw_tri, db_tri = [_.grad.clone() for _ in [x, weight, bias]]
x.grad, weight.grad, bias.grad = None, None, None
# backward pass (torch)
y_ref.backward(dy, retain_graph=True)
dx_ref, dw_ref, db_ref = [_.grad.clone() for _ in [x, weight, bias]]
# compare
assert torch.allclose(y_tri, y_ref, atol=1e-2, rtol=0)
assert torch.allclose(dx_tri, dx_ref, atol=1e-2, rtol=0)
assert torch.allclose(db_tri, db_ref, atol=1e-2, rtol=0)
assert torch.allclose(dw_tri, dw_ref, atol=1e-2, rtol=0)
@triton.testing.perf_report(
triton.testing.Benchmark(
x_names=['N'],
x_vals=[512 * i for i in range(2, 32)],
line_arg='provider',
line_vals=['triton', 'torch'] + (['apex'] if HAS_APEX else []),
line_names=['Triton', 'Torch'] + (['Apex'] if HAS_APEX else []),
styles=[('blue', '-'), ('green', '-'), ('orange', '-')],
ylabel='GB/s',
plot_name='layer-norm-backward',
args={'M': 4096, 'dtype': torch.float16, 'mode': 'backward'},
))
def bench_layer_norm(M, N, dtype, provider, mode='backward', eps=1e-5, device=DEVICE):
# create data
x_shape = (M, N)
w_shape = (x_shape[-1], )
weight = torch.rand(w_shape, dtype=dtype, device=device, requires_grad=True)
bias = torch.rand(w_shape, dtype=dtype, device=device, requires_grad=True)
x = -2.3 + 0.5 * torch.randn(x_shape, dtype=dtype, device=device)
dy = .1 * torch.randn_like(x)
x.requires_grad_(True)
quantiles = [0.5, 0.2, 0.8]
def y_fwd():
if provider == "triton":
return layer_norm(x, w_shape, weight, bias, eps) # noqa: F811, E704
if provider == "torch":
return torch.nn.functional.layer_norm(x, w_shape, weight, bias, eps) # noqa: F811, E704
if provider == "apex":
apex_layer_norm = (apex.normalization.FusedLayerNorm(w_shape).to(x.device).to(x.dtype))
return apex_layer_norm(x) # noqa: F811, E704
# forward pass
if mode == 'forward':
gbps = lambda ms: 2 * x.numel() * x.element_size() * 1e-9 / (ms * 1e-3)
ms, min_ms, max_ms = triton.testing.do_bench(y_fwd, quantiles=quantiles, rep=500)
# backward pass
if mode == 'backward':
y = y_fwd()
gbps = lambda ms: 3 * x.numel() * x.element_size() * 1e-9 / (ms * 1e-3) # noqa: F811, E704
ms, min_ms, max_ms = triton.testing.do_bench(lambda: y.backward(dy, retain_graph=True), quantiles=quantiles,
grad_to_none=[x], rep=500)
return gbps(ms), gbps(max_ms), gbps(min_ms)
test_layer_norm(1151, 8192, torch.float16)
bench_layer_norm.run(save_path='.', print_data=True)
layer-norm-backward:
N Triton Torch
0 1024.0 117.028572 378.092307
1 1536.0 193.005236 438.857146
2 2048.0 243.326738 511.999982
3 2560.0 315.886895 568.888888
4 3072.0 366.805965 604.327881
5 3584.0 445.678757 547.872604
6 4096.0 491.520012 561.737163
7 4608.0 544.788169 567.138460
8 5120.0 590.769237 566.267298
9 5632.0 551.706116 563.200014
10 6144.0 589.823972 562.809189
11 6656.0 636.430303 566.468098
12 7168.0 807.661947 539.285244
13 7680.0 899.121934 546.943612
14 8192.0 959.063396 544.620479
15 8704.0 700.993291 560.042913
16 9216.0 737.279965 565.687967
17 9728.0 765.481982 569.443892
18 10240.0 768.000019 561.095906
19 10752.0 816.607619 552.565304
20 11264.0 842.168226 558.545467
21 11776.0 833.699137 566.380743
22 12288.0 852.346855 571.534916
23 12800.0 880.229204 577.443635
24 13312.0 926.052153 577.735965
25 13824.0 926.748629 577.001726
26 14336.0 945.230752 566.826996
27 14848.0 950.271992 568.344496
28 15360.0 959.999966 578.712698
29 15872.0 938.246290 578.917929
参考文献¶
[BA2016]
Jimmy Lei Ba and Jamie Ryan Kiros and Geoffrey E. Hinton, “Layer Normalization”, Arxiv 2016
脚本总运行时间： (0 分钟 29.005 秒)