stable-diffusion-webui/modules/sd_hijack_optimizations.py

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

669 lines
24 KiB
Python
Raw Normal View History

from __future__ import annotations
import math
import psutil
import torch
from torch import einsum
from ldm.util import default
from einops import rearrange
2023-05-18 21:03:27 +00:00
from modules import shared, errors, devices, sub_quadratic_attention
from modules.hypernetworks import hypernetwork
2022-10-11 08:09:51 +00:00
import ldm.modules.attention
import ldm.modules.diffusionmodules.model
2023-07-12 20:52:43 +00:00
import sgm.modules.attention
import sgm.modules.diffusionmodules.model
diffusionmodules_model_AttnBlock_forward = ldm.modules.diffusionmodules.model.AttnBlock.forward
2023-07-12 20:52:43 +00:00
sgm_diffusionmodules_model_AttnBlock_forward = sgm.modules.diffusionmodules.model.AttnBlock.forward
class SdOptimization:
name: str = None
label: str | None = None
cmd_opt: str | None = None
priority: int = 0
def title(self):
if self.label is None:
return self.name
return f"{self.name} - {self.label}"
def is_available(self):
return True
def apply(self):
pass
def undo(self):
ldm.modules.attention.CrossAttention.forward = hypernetwork.attention_CrossAttention_forward
ldm.modules.diffusionmodules.model.AttnBlock.forward = diffusionmodules_model_AttnBlock_forward
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = hypernetwork.attention_CrossAttention_forward
sgm.modules.diffusionmodules.model.AttnBlock.forward = sgm_diffusionmodules_model_AttnBlock_forward
class SdOptimizationXformers(SdOptimization):
name = "xformers"
cmd_opt = "xformers"
priority = 100
def is_available(self):
2023-06-04 08:33:51 +00:00
return shared.cmd_opts.force_enable_xformers or (shared.xformers_available and torch.cuda.is_available() and (6, 0) <= torch.cuda.get_device_capability(shared.device) <= (9, 0))
def apply(self):
ldm.modules.attention.CrossAttention.forward = xformers_attention_forward
ldm.modules.diffusionmodules.model.AttnBlock.forward = xformers_attnblock_forward
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = xformers_attention_forward
sgm.modules.diffusionmodules.model.AttnBlock.forward = xformers_attnblock_forward
class SdOptimizationSdpNoMem(SdOptimization):
name = "sdp-no-mem"
label = "scaled dot product without memory efficient attention"
cmd_opt = "opt_sdp_no_mem_attention"
priority = 80
def is_available(self):
return hasattr(torch.nn.functional, "scaled_dot_product_attention") and callable(torch.nn.functional.scaled_dot_product_attention)
def apply(self):
ldm.modules.attention.CrossAttention.forward = scaled_dot_product_no_mem_attention_forward
ldm.modules.diffusionmodules.model.AttnBlock.forward = sdp_no_mem_attnblock_forward
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = scaled_dot_product_no_mem_attention_forward
sgm.modules.diffusionmodules.model.AttnBlock.forward = sdp_no_mem_attnblock_forward
class SdOptimizationSdp(SdOptimizationSdpNoMem):
name = "sdp"
label = "scaled dot product"
cmd_opt = "opt_sdp_attention"
priority = 70
def apply(self):
ldm.modules.attention.CrossAttention.forward = scaled_dot_product_attention_forward
ldm.modules.diffusionmodules.model.AttnBlock.forward = sdp_attnblock_forward
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = scaled_dot_product_attention_forward
sgm.modules.diffusionmodules.model.AttnBlock.forward = sdp_attnblock_forward
class SdOptimizationSubQuad(SdOptimization):
name = "sub-quadratic"
cmd_opt = "opt_sub_quad_attention"
priority = 10
def apply(self):
ldm.modules.attention.CrossAttention.forward = sub_quad_attention_forward
ldm.modules.diffusionmodules.model.AttnBlock.forward = sub_quad_attnblock_forward
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = sub_quad_attention_forward
sgm.modules.diffusionmodules.model.AttnBlock.forward = sub_quad_attnblock_forward
class SdOptimizationV1(SdOptimization):
name = "V1"
label = "original v1"
cmd_opt = "opt_split_attention_v1"
priority = 10
def apply(self):
ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward_v1
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = split_cross_attention_forward_v1
class SdOptimizationInvokeAI(SdOptimization):
name = "InvokeAI"
cmd_opt = "opt_split_attention_invokeai"
@property
def priority(self):
return 1000 if not torch.cuda.is_available() else 10
def apply(self):
ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward_invokeAI
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = split_cross_attention_forward_invokeAI
class SdOptimizationDoggettx(SdOptimization):
name = "Doggettx"
cmd_opt = "opt_split_attention"
priority = 90
def apply(self):
ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward
ldm.modules.diffusionmodules.model.AttnBlock.forward = cross_attention_attnblock_forward
2023-07-12 20:52:43 +00:00
sgm.modules.attention.CrossAttention.forward = split_cross_attention_forward
sgm.modules.diffusionmodules.model.AttnBlock.forward = cross_attention_attnblock_forward
def list_optimizers(res):
res.extend([
SdOptimizationXformers(),
SdOptimizationSdpNoMem(),
SdOptimizationSdp(),
SdOptimizationSubQuad(),
SdOptimizationV1(),
SdOptimizationInvokeAI(),
SdOptimizationDoggettx(),
])
2022-10-07 07:17:52 +00:00
if shared.cmd_opts.xformers or shared.cmd_opts.force_enable_xformers:
try:
import xformers.ops
shared.xformers_available = True
except Exception:
errors.report("Cannot import xformers", exc_info=True)
def get_available_vram():
if shared.device.type == 'cuda':
stats = torch.cuda.memory_stats(shared.device)
mem_active = stats['active_bytes.all.current']
mem_reserved = stats['reserved_bytes.all.current']
mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device())
mem_free_torch = mem_reserved - mem_active
mem_free_total = mem_free_cuda + mem_free_torch
return mem_free_total
else:
return psutil.virtual_memory().available
# see https://github.com/basujindal/stable-diffusion/pull/117 for discussion
2023-07-13 06:30:33 +00:00
def split_cross_attention_forward_v1(self, x, context=None, mask=None, **kwargs):
h = self.heads
q_in = self.to_q(x)
context = default(context, x)
context_k, context_v = hypernetwork.apply_hypernetworks(shared.loaded_hypernetworks, context)
2022-10-11 08:09:51 +00:00
k_in = self.to_k(context_k)
v_in = self.to_v(context_v)
del context, context_k, context_v, x
2023-05-10 08:05:02 +00:00
q, k, v = (rearrange(t, 'b n (h d) -> (b h) n d', h=h) for t in (q_in, k_in, v_in))
del q_in, k_in, v_in
dtype = q.dtype
if shared.opts.upcast_attn:
q, k, v = q.float(), k.float(), v.float()
with devices.without_autocast(disable=not shared.opts.upcast_attn):
r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype)
for i in range(0, q.shape[0], 2):
end = i + 2
s1 = einsum('b i d, b j d -> b i j', q[i:end], k[i:end])
s1 *= self.scale
s2 = s1.softmax(dim=-1)
del s1
r1[i:end] = einsum('b i j, b j d -> b i d', s2, v[i:end])
del s2
del q, k, v
r1 = r1.to(dtype)
r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h)
del r1
return self.to_out(r2)
2022-10-11 08:09:51 +00:00
# taken from https://github.com/Doggettx/stable-diffusion and modified
2023-07-13 06:30:33 +00:00
def split_cross_attention_forward(self, x, context=None, mask=None, **kwargs):
h = self.heads
q_in = self.to_q(x)
context = default(context, x)
2022-10-07 07:17:52 +00:00
context_k, context_v = hypernetwork.apply_hypernetworks(shared.loaded_hypernetworks, context)
2022-10-11 08:09:51 +00:00
k_in = self.to_k(context_k)
v_in = self.to_v(context_v)
2022-10-07 07:17:52 +00:00
dtype = q_in.dtype
if shared.opts.upcast_attn:
q_in, k_in, v_in = q_in.float(), k_in.float(), v_in if v_in.device.type == 'mps' else v_in.float()
with devices.without_autocast(disable=not shared.opts.upcast_attn):
k_in = k_in * self.scale
del context, x
2023-05-10 08:05:02 +00:00
q, k, v = (rearrange(t, 'b n (h d) -> (b h) n d', h=h) for t in (q_in, k_in, v_in))
del q_in, k_in, v_in
r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype)
mem_free_total = get_available_vram()
gb = 1024 ** 3
tensor_size = q.shape[0] * q.shape[1] * k.shape[1] * q.element_size()
modifier = 3 if q.element_size() == 2 else 2.5
mem_required = tensor_size * modifier
steps = 1
if mem_required > mem_free_total:
steps = 2 ** (math.ceil(math.log(mem_required / mem_free_total, 2)))
# print(f"Expected tensor size:{tensor_size/gb:0.1f}GB, cuda free:{mem_free_cuda/gb:0.1f}GB "
# f"torch free:{mem_free_torch/gb:0.1f} total:{mem_free_total/gb:0.1f} steps:{steps}")
if steps > 64:
max_res = math.floor(math.sqrt(math.sqrt(mem_free_total / 2.5)) / 8) * 64
raise RuntimeError(f'Not enough memory, use lower resolution (max approx. {max_res}x{max_res}). '
f'Need: {mem_required / 64 / gb:0.1f}GB free, Have:{mem_free_total / gb:0.1f}GB free')
slice_size = q.shape[1] // steps
for i in range(0, q.shape[1], slice_size):
end = min(i + slice_size, q.shape[1])
s1 = einsum('b i d, b j d -> b i j', q[:, i:end], k)
s2 = s1.softmax(dim=-1, dtype=q.dtype)
del s1
r1[:, i:end] = einsum('b i j, b j d -> b i d', s2, v)
del s2
del q, k, v
r1 = r1.to(dtype)
r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h)
del r1
return self.to_out(r2)
# -- Taken from https://github.com/invoke-ai/InvokeAI and modified --
mem_total_gb = psutil.virtual_memory().total // (1 << 30)
2023-07-12 20:52:43 +00:00
def einsum_op_compvis(q, k, v):
s = einsum('b i d, b j d -> b i j', q, k)
s = s.softmax(dim=-1, dtype=s.dtype)
return einsum('b i j, b j d -> b i d', s, v)
2023-07-12 20:52:43 +00:00
def einsum_op_slice_0(q, k, v, slice_size):
r = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype)
for i in range(0, q.shape[0], slice_size):
end = i + slice_size
r[i:end] = einsum_op_compvis(q[i:end], k[i:end], v[i:end])
return r
2023-07-12 20:52:43 +00:00
def einsum_op_slice_1(q, k, v, slice_size):
r = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype)
for i in range(0, q.shape[1], slice_size):
end = i + slice_size
r[:, i:end] = einsum_op_compvis(q[:, i:end], k, v)
return r
2023-07-12 20:52:43 +00:00
def einsum_op_mps_v1(q, k, v):
if q.shape[0] * q.shape[1] <= 2**16: # (512x512) max q.shape[1]: 4096
return einsum_op_compvis(q, k, v)
else:
slice_size = math.floor(2**30 / (q.shape[0] * q.shape[1]))
if slice_size % 4096 == 0:
slice_size -= 1
return einsum_op_slice_1(q, k, v, slice_size)
2023-07-12 20:52:43 +00:00
def einsum_op_mps_v2(q, k, v):
if mem_total_gb > 8 and q.shape[0] * q.shape[1] <= 2**16:
return einsum_op_compvis(q, k, v)
else:
return einsum_op_slice_0(q, k, v, 1)
2023-07-12 20:52:43 +00:00
def einsum_op_tensor_mem(q, k, v, max_tensor_mb):
size_mb = q.shape[0] * q.shape[1] * k.shape[1] * q.element_size() // (1 << 20)
if size_mb <= max_tensor_mb:
return einsum_op_compvis(q, k, v)
div = 1 << int((size_mb - 1) / max_tensor_mb).bit_length()
if div <= q.shape[0]:
return einsum_op_slice_0(q, k, v, q.shape[0] // div)
return einsum_op_slice_1(q, k, v, max(q.shape[1] // div, 1))
2023-07-12 20:52:43 +00:00
def einsum_op_cuda(q, k, v):
stats = torch.cuda.memory_stats(q.device)
mem_active = stats['active_bytes.all.current']
mem_reserved = stats['reserved_bytes.all.current']
mem_free_cuda, _ = torch.cuda.mem_get_info(q.device)
mem_free_torch = mem_reserved - mem_active
mem_free_total = mem_free_cuda + mem_free_torch
# Divide factor of safety as there's copying and fragmentation
return einsum_op_tensor_mem(q, k, v, mem_free_total / 3.3 / (1 << 20))
2023-07-12 20:52:43 +00:00
def einsum_op(q, k, v):
if q.device.type == 'cuda':
return einsum_op_cuda(q, k, v)
if q.device.type == 'mps':
if mem_total_gb >= 32 and q.shape[0] % 32 != 0 and q.shape[0] * q.shape[1] < 2**18:
return einsum_op_mps_v1(q, k, v)
return einsum_op_mps_v2(q, k, v)
# Smaller slices are faster due to L2/L3/SLC caches.
# Tested on i7 with 8MB L3 cache.
return einsum_op_tensor_mem(q, k, v, 32)
2023-07-12 20:52:43 +00:00
2023-07-13 06:30:33 +00:00
def split_cross_attention_forward_invokeAI(self, x, context=None, mask=None, **kwargs):
h = self.heads
q = self.to_q(x)
context = default(context, x)
context_k, context_v = hypernetwork.apply_hypernetworks(shared.loaded_hypernetworks, context)
k = self.to_k(context_k)
2022-10-11 09:13:17 +00:00
v = self.to_v(context_v)
del context, context_k, context_v, x
dtype = q.dtype
if shared.opts.upcast_attn:
q, k, v = q.float(), k.float(), v if v.device.type == 'mps' else v.float()
with devices.without_autocast(disable=not shared.opts.upcast_attn):
k = k * self.scale
2023-05-10 08:05:02 +00:00
q, k, v = (rearrange(t, 'b n (h d) -> (b h) n d', h=h) for t in (q, k, v))
r = einsum_op(q, k, v)
r = r.to(dtype)
return self.to_out(rearrange(r, '(b h) n d -> b n (h d)', h=h))
2022-10-11 03:55:48 +00:00
# -- End of code from https://github.com/invoke-ai/InvokeAI --
# Based on Birch-san's modified implementation of sub-quadratic attention from https://github.com/Birch-san/diffusers/pull/1
2023-01-06 21:42:47 +00:00
# The sub_quad_attention_forward function is under the MIT License listed under Memory Efficient Attention in the Licenses section of the web UI interface
2023-07-13 06:30:33 +00:00
def sub_quad_attention_forward(self, x, context=None, mask=None, **kwargs):
assert mask is None, "attention-mask not currently implemented for SubQuadraticCrossAttnProcessor."
h = self.heads
q = self.to_q(x)
context = default(context, x)
context_k, context_v = hypernetwork.apply_hypernetworks(shared.loaded_hypernetworks, context)
k = self.to_k(context_k)
v = self.to_v(context_v)
del context, context_k, context_v, x
q = q.unflatten(-1, (h, -1)).transpose(1,2).flatten(end_dim=1)
k = k.unflatten(-1, (h, -1)).transpose(1,2).flatten(end_dim=1)
v = v.unflatten(-1, (h, -1)).transpose(1,2).flatten(end_dim=1)
2023-04-14 06:22:48 +00:00
if q.device.type == 'mps':
q, k, v = q.contiguous(), k.contiguous(), v.contiguous()
dtype = q.dtype
if shared.opts.upcast_attn:
q, k = q.float(), k.float()
x = sub_quad_attention(q, k, v, q_chunk_size=shared.cmd_opts.sub_quad_q_chunk_size, kv_chunk_size=shared.cmd_opts.sub_quad_kv_chunk_size, chunk_threshold=shared.cmd_opts.sub_quad_chunk_threshold, use_checkpoint=self.training)
x = x.to(dtype)
x = x.unflatten(0, (-1, h)).transpose(1,2).flatten(start_dim=2)
out_proj, dropout = self.to_out
x = out_proj(x)
x = dropout(x)
return x
2023-07-12 20:52:43 +00:00
def sub_quad_attention(q, k, v, q_chunk_size=1024, kv_chunk_size=None, kv_chunk_size_min=None, chunk_threshold=None, use_checkpoint=True):
bytes_per_token = torch.finfo(q.dtype).bits//8
batch_x_heads, q_tokens, _ = q.shape
_, k_tokens, _ = k.shape
qk_matmul_size_bytes = batch_x_heads * bytes_per_token * q_tokens * k_tokens
if chunk_threshold is None:
chunk_threshold_bytes = int(get_available_vram() * 0.9) if q.device.type == 'mps' else int(get_available_vram() * 0.7)
elif chunk_threshold == 0:
chunk_threshold_bytes = None
else:
chunk_threshold_bytes = int(0.01 * chunk_threshold * get_available_vram())
if kv_chunk_size_min is None and chunk_threshold_bytes is not None:
kv_chunk_size_min = chunk_threshold_bytes // (batch_x_heads * bytes_per_token * (k.shape[2] + v.shape[2]))
elif kv_chunk_size_min == 0:
kv_chunk_size_min = None
if chunk_threshold_bytes is not None and qk_matmul_size_bytes <= chunk_threshold_bytes:
# the big matmul fits into our memory limit; do everything in 1 chunk,
# i.e. send it down the unchunked fast-path
kv_chunk_size = k_tokens
with devices.without_autocast(disable=q.dtype == v.dtype):
return sub_quadratic_attention.efficient_dot_product_attention(
q,
k,
v,
query_chunk_size=q_chunk_size,
kv_chunk_size=kv_chunk_size,
kv_chunk_size_min = kv_chunk_size_min,
use_checkpoint=use_checkpoint,
)
def get_xformers_flash_attention_op(q, k, v):
if not shared.cmd_opts.xformers_flash_attention:
return None
try:
flash_attention_op = xformers.ops.MemoryEfficientAttentionFlashAttentionOp
fw, bw = flash_attention_op
if fw.supports(xformers.ops.fmha.Inputs(query=q, key=k, value=v, attn_bias=None)):
return flash_attention_op
except Exception as e:
errors.display_once(e, "enabling flash attention")
return None
2023-07-13 06:30:33 +00:00
def xformers_attention_forward(self, x, context=None, mask=None, **kwargs):
2022-10-07 02:21:49 +00:00
h = self.heads
q_in = self.to_q(x)
context = default(context, x)
2022-10-11 08:09:51 +00:00
context_k, context_v = hypernetwork.apply_hypernetworks(shared.loaded_hypernetworks, context)
2022-10-11 08:09:51 +00:00
k_in = self.to_k(context_k)
v_in = self.to_v(context_v)
2023-05-10 08:05:02 +00:00
q, k, v = (rearrange(t, 'b n (h d) -> b n h d', h=h) for t in (q_in, k_in, v_in))
2022-10-07 02:21:49 +00:00
del q_in, k_in, v_in
dtype = q.dtype
if shared.opts.upcast_attn:
2023-03-21 10:50:22 +00:00
q, k, v = q.float(), k.float(), v.float()
out = xformers.ops.memory_efficient_attention(q, k, v, attn_bias=None, op=get_xformers_flash_attention_op(q, k, v))
2022-10-07 02:21:49 +00:00
out = out.to(dtype)
out = rearrange(out, 'b n h d -> b n (h d)', h=h)
2022-10-07 02:21:49 +00:00
return self.to_out(out)
2023-07-12 20:52:43 +00:00
2023-03-06 19:33:13 +00:00
# Based on Diffusers usage of scaled dot product attention from https://github.com/huggingface/diffusers/blob/c7da8fd23359a22d0df2741688b5b4f33c26df21/src/diffusers/models/cross_attention.py
# The scaled_dot_product_attention_forward function contains parts of code under Apache-2.0 license listed under Scaled Dot Product Attention in the Licenses section of the web UI interface
2023-07-13 06:30:33 +00:00
def scaled_dot_product_attention_forward(self, x, context=None, mask=None, **kwargs):
2023-03-06 19:33:13 +00:00
batch_size, sequence_length, inner_dim = x.shape
if mask is not None:
mask = self.prepare_attention_mask(mask, sequence_length, batch_size)
mask = mask.view(batch_size, self.heads, -1, mask.shape[-1])
h = self.heads
q_in = self.to_q(x)
context = default(context, x)
context_k, context_v = hypernetwork.apply_hypernetworks(shared.loaded_hypernetworks, context)
k_in = self.to_k(context_k)
v_in = self.to_v(context_v)
head_dim = inner_dim // h
q = q_in.view(batch_size, -1, h, head_dim).transpose(1, 2)
k = k_in.view(batch_size, -1, h, head_dim).transpose(1, 2)
v = v_in.view(batch_size, -1, h, head_dim).transpose(1, 2)
2023-03-06 19:33:13 +00:00
del q_in, k_in, v_in
dtype = q.dtype
if shared.opts.upcast_attn:
2023-03-24 12:29:16 +00:00
q, k, v = q.float(), k.float(), v.float()
2023-03-06 19:33:13 +00:00
# the output of sdp = (batch, num_heads, seq_len, head_dim)
hidden_states = torch.nn.functional.scaled_dot_product_attention(
q, k, v, attn_mask=mask, dropout_p=0.0, is_causal=False
)
hidden_states = hidden_states.transpose(1, 2).reshape(batch_size, -1, h * head_dim)
hidden_states = hidden_states.to(dtype)
# linear proj
hidden_states = self.to_out[0](hidden_states)
# dropout
hidden_states = self.to_out[1](hidden_states)
return hidden_states
2023-07-12 20:52:43 +00:00
2023-07-13 06:30:33 +00:00
def scaled_dot_product_no_mem_attention_forward(self, x, context=None, mask=None, **kwargs):
with torch.backends.cuda.sdp_kernel(enable_flash=True, enable_math=True, enable_mem_efficient=False):
return scaled_dot_product_attention_forward(self, x, context, mask)
2023-07-12 20:52:43 +00:00
def cross_attention_attnblock_forward(self, x):
h_ = x
h_ = self.norm(h_)
q1 = self.q(h_)
k1 = self.k(h_)
v = self.v(h_)
# compute attention
b, c, h, w = q1.shape
q2 = q1.reshape(b, c, h*w)
del q1
q = q2.permute(0, 2, 1) # b,hw,c
del q2
k = k1.reshape(b, c, h*w) # b,c,hw
del k1
h_ = torch.zeros_like(k, device=q.device)
mem_free_total = get_available_vram()
tensor_size = q.shape[0] * q.shape[1] * k.shape[2] * q.element_size()
mem_required = tensor_size * 2.5
steps = 1
if mem_required > mem_free_total:
steps = 2**(math.ceil(math.log(mem_required / mem_free_total, 2)))
slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1]
for i in range(0, q.shape[1], slice_size):
end = i + slice_size
w1 = torch.bmm(q[:, i:end], k) # b,hw,hw w[b,i,j]=sum_c q[b,i,c]k[b,c,j]
w2 = w1 * (int(c)**(-0.5))
del w1
w3 = torch.nn.functional.softmax(w2, dim=2, dtype=q.dtype)
del w2
# attend to values
v1 = v.reshape(b, c, h*w)
w4 = w3.permute(0, 2, 1) # b,hw,hw (first hw of k, second of q)
del w3
h_[:, :, i:end] = torch.bmm(v1, w4) # b, c,hw (hw of q) h_[b,c,j] = sum_i v[b,c,i] w_[b,i,j]
del v1, w4
h2 = h_.reshape(b, c, h, w)
del h_
h3 = self.proj_out(h2)
del h2
h3 += x
return h3
2023-07-12 20:52:43 +00:00
2022-10-17 19:18:59 +00:00
def xformers_attnblock_forward(self, x):
try:
h_ = x
h_ = self.norm(h_)
q = self.q(h_)
k = self.k(h_)
v = self.v(h_)
b, c, h, w = q.shape
2023-05-10 08:05:02 +00:00
q, k, v = (rearrange(t, 'b c h w -> b (h w) c') for t in (q, k, v))
dtype = q.dtype
if shared.opts.upcast_attn:
q, k = q.float(), k.float()
2022-10-17 21:02:50 +00:00
q = q.contiguous()
k = k.contiguous()
v = v.contiguous()
out = xformers.ops.memory_efficient_attention(q, k, v, op=get_xformers_flash_attention_op(q, k, v))
out = out.to(dtype)
2022-10-17 19:18:59 +00:00
out = rearrange(out, 'b (h w) c -> b c h w', h=h)
out = self.proj_out(out)
return x + out
except NotImplementedError:
return cross_attention_attnblock_forward(self, x)
2023-07-12 20:52:43 +00:00
2023-03-10 17:48:41 +00:00
def sdp_attnblock_forward(self, x):
h_ = x
h_ = self.norm(h_)
q = self.q(h_)
k = self.k(h_)
v = self.v(h_)
b, c, h, w = q.shape
2023-05-10 08:05:02 +00:00
q, k, v = (rearrange(t, 'b c h w -> b (h w) c') for t in (q, k, v))
2023-03-10 17:48:41 +00:00
dtype = q.dtype
if shared.opts.upcast_attn:
q, k, v = q.float(), k.float(), v.float()
2023-03-10 17:48:41 +00:00
q = q.contiguous()
k = k.contiguous()
v = v.contiguous()
out = torch.nn.functional.scaled_dot_product_attention(q, k, v, dropout_p=0.0, is_causal=False)
out = out.to(dtype)
out = rearrange(out, 'b (h w) c -> b c h w', h=h)
out = self.proj_out(out)
return x + out
2023-07-12 20:52:43 +00:00
2023-03-10 17:48:41 +00:00
def sdp_no_mem_attnblock_forward(self, x):
with torch.backends.cuda.sdp_kernel(enable_flash=True, enable_math=True, enable_mem_efficient=False):
return sdp_attnblock_forward(self, x)
2023-07-12 20:52:43 +00:00
def sub_quad_attnblock_forward(self, x):
h_ = x
h_ = self.norm(h_)
q = self.q(h_)
k = self.k(h_)
v = self.v(h_)
b, c, h, w = q.shape
2023-05-10 08:05:02 +00:00
q, k, v = (rearrange(t, 'b c h w -> b (h w) c') for t in (q, k, v))
q = q.contiguous()
k = k.contiguous()
v = v.contiguous()
out = sub_quad_attention(q, k, v, q_chunk_size=shared.cmd_opts.sub_quad_q_chunk_size, kv_chunk_size=shared.cmd_opts.sub_quad_kv_chunk_size, chunk_threshold=shared.cmd_opts.sub_quad_chunk_threshold, use_checkpoint=self.training)
out = rearrange(out, 'b (h w) c -> b c h w', h=h)
out = self.proj_out(out)
return x + out