2023-02-08 21:51:19 +00:00
|
|
|
from .k_diffusion import sampling as k_diffusion_sampling
|
|
|
|
from .k_diffusion import external as k_diffusion_external
|
2023-01-03 06:53:32 +00:00
|
|
|
import torch
|
|
|
|
import contextlib
|
2023-02-08 19:05:31 +00:00
|
|
|
import model_management
|
2023-01-03 06:53:32 +00:00
|
|
|
|
|
|
|
class CFGDenoiser(torch.nn.Module):
|
|
|
|
def __init__(self, model):
|
|
|
|
super().__init__()
|
|
|
|
self.inner_model = model
|
|
|
|
|
|
|
|
def forward(self, x, sigma, uncond, cond, cond_scale):
|
2023-01-26 17:06:48 +00:00
|
|
|
if len(uncond[0]) == len(cond[0]) and x.shape[0] * x.shape[2] * x.shape[3] < (96 * 96): #TODO check memory instead
|
2023-01-03 06:53:32 +00:00
|
|
|
x_in = torch.cat([x] * 2)
|
|
|
|
sigma_in = torch.cat([sigma] * 2)
|
|
|
|
cond_in = torch.cat([uncond, cond])
|
|
|
|
uncond, cond = self.inner_model(x_in, sigma_in, cond=cond_in).chunk(2)
|
|
|
|
else:
|
|
|
|
cond = self.inner_model(x, sigma, cond=cond)
|
|
|
|
uncond = self.inner_model(x, sigma, cond=uncond)
|
|
|
|
return uncond + (cond - uncond) * cond_scale
|
|
|
|
|
2023-01-26 17:06:48 +00:00
|
|
|
class CFGDenoiserComplex(torch.nn.Module):
|
|
|
|
def __init__(self, model):
|
|
|
|
super().__init__()
|
|
|
|
self.inner_model = model
|
|
|
|
def forward(self, x, sigma, uncond, cond, cond_scale):
|
2023-02-08 19:05:31 +00:00
|
|
|
def get_area_and_mult(cond, x_in, sigma):
|
|
|
|
area = (x_in.shape[2], x_in.shape[3], 0, 0)
|
|
|
|
strength = 1.0
|
|
|
|
min_sigma = 0.0
|
|
|
|
max_sigma = 999.0
|
|
|
|
if 'area' in cond[1]:
|
|
|
|
area = cond[1]['area']
|
|
|
|
if 'strength' in cond[1]:
|
|
|
|
strength = cond[1]['strength']
|
|
|
|
if 'min_sigma' in cond[1]:
|
|
|
|
min_sigma = cond[1]['min_sigma']
|
|
|
|
if 'max_sigma' in cond[1]:
|
|
|
|
max_sigma = cond[1]['max_sigma']
|
|
|
|
if sigma < min_sigma or sigma > max_sigma:
|
|
|
|
return None
|
|
|
|
input_x = x_in[:,:,area[2]:area[0] + area[2],area[3]:area[1] + area[3]]
|
|
|
|
mult = torch.ones_like(input_x) * strength
|
|
|
|
|
|
|
|
rr = 8
|
|
|
|
if area[2] != 0:
|
|
|
|
for t in range(rr):
|
|
|
|
mult[:,:,area[2]+t:area[2]+1+t,:] *= ((1.0/rr) * (t + 1))
|
|
|
|
if (area[0] + area[2]) < x_in.shape[2]:
|
|
|
|
for t in range(rr):
|
|
|
|
mult[:,:,area[0] + area[2] - 1 - t:area[0] + area[2] - t,:] *= ((1.0/rr) * (t + 1))
|
|
|
|
if area[3] != 0:
|
|
|
|
for t in range(rr):
|
|
|
|
mult[:,:,:,area[3]+t:area[3]+1+t] *= ((1.0/rr) * (t + 1))
|
|
|
|
if (area[1] + area[3]) < x_in.shape[3]:
|
|
|
|
for t in range(rr):
|
|
|
|
mult[:,:,:,area[1] + area[3] - 1 - t:area[1] + area[3] - t] *= ((1.0/rr) * (t + 1))
|
|
|
|
return (input_x, mult, cond[0], area)
|
|
|
|
|
|
|
|
def calc_cond_uncond_batch(cond, uncond, x_in, sigma, max_total_area):
|
2023-01-26 17:06:48 +00:00
|
|
|
out_cond = torch.zeros_like(x_in)
|
|
|
|
out_count = torch.ones_like(x_in)/100000.0
|
2023-02-08 19:05:31 +00:00
|
|
|
|
|
|
|
out_uncond = torch.zeros_like(x_in)
|
|
|
|
out_uncond_count = torch.ones_like(x_in)/100000.0
|
|
|
|
|
2023-01-26 17:06:48 +00:00
|
|
|
sigma_cmp = sigma[0]
|
2023-02-08 19:05:31 +00:00
|
|
|
COND = 0
|
|
|
|
UNCOND = 1
|
2023-01-26 17:06:48 +00:00
|
|
|
|
2023-02-08 19:05:31 +00:00
|
|
|
to_run = []
|
2023-01-26 17:06:48 +00:00
|
|
|
for x in cond:
|
2023-02-08 19:05:31 +00:00
|
|
|
p = get_area_and_mult(x, x_in, sigma_cmp)
|
|
|
|
if p is None:
|
2023-01-26 17:06:48 +00:00
|
|
|
continue
|
2023-02-08 19:05:31 +00:00
|
|
|
|
|
|
|
to_run += [(p, COND)]
|
|
|
|
for x in uncond:
|
|
|
|
p = get_area_and_mult(x, x_in, sigma_cmp)
|
|
|
|
if p is None:
|
|
|
|
continue
|
|
|
|
|
|
|
|
to_run += [(p, UNCOND)]
|
|
|
|
|
|
|
|
while len(to_run) > 0:
|
|
|
|
first = to_run[0]
|
|
|
|
first_shape = first[0][0].shape
|
|
|
|
to_batch = []
|
|
|
|
for x in range(len(to_run)):
|
|
|
|
if to_run[x][0][0].shape == first_shape:
|
|
|
|
if to_run[x][0][2].shape == first[0][2].shape:
|
|
|
|
to_batch += [x]
|
|
|
|
if (len(to_batch) * first_shape[0] * first_shape[2] * first_shape[3] >= max_total_area):
|
|
|
|
break
|
|
|
|
|
|
|
|
to_batch.reverse()
|
|
|
|
input_x = []
|
|
|
|
mult = []
|
|
|
|
c = []
|
|
|
|
cond_or_uncond = []
|
|
|
|
area = []
|
|
|
|
for x in to_batch:
|
|
|
|
o = to_run.pop(x)
|
|
|
|
p = o[0]
|
|
|
|
input_x += [p[0]]
|
|
|
|
mult += [p[1]]
|
|
|
|
c += [p[2]]
|
|
|
|
area += [p[3]]
|
|
|
|
cond_or_uncond += [o[1]]
|
|
|
|
|
|
|
|
batch_chunks = len(cond_or_uncond)
|
|
|
|
input_x = torch.cat(input_x)
|
|
|
|
c = torch.cat(c)
|
|
|
|
sigma_ = torch.cat([sigma] * batch_chunks)
|
|
|
|
|
|
|
|
output = self.inner_model(input_x, sigma_, cond=c).chunk(batch_chunks)
|
2023-01-26 17:06:48 +00:00
|
|
|
del input_x
|
2023-02-08 19:05:31 +00:00
|
|
|
|
|
|
|
for o in range(batch_chunks):
|
|
|
|
if cond_or_uncond[o] == COND:
|
|
|
|
out_cond[:,:,area[o][2]:area[o][0] + area[o][2],area[o][3]:area[o][1] + area[o][3]] += output[o] * mult[o]
|
|
|
|
out_count[:,:,area[o][2]:area[o][0] + area[o][2],area[o][3]:area[o][1] + area[o][3]] += mult[o]
|
|
|
|
else:
|
|
|
|
out_uncond[:,:,area[o][2]:area[o][0] + area[o][2],area[o][3]:area[o][1] + area[o][3]] += output[o] * mult[o]
|
|
|
|
out_uncond_count[:,:,area[o][2]:area[o][0] + area[o][2],area[o][3]:area[o][1] + area[o][3]] += mult[o]
|
2023-01-26 17:06:48 +00:00
|
|
|
del mult
|
|
|
|
|
|
|
|
out_cond /= out_count
|
|
|
|
del out_count
|
2023-02-08 19:05:31 +00:00
|
|
|
out_uncond /= out_uncond_count
|
|
|
|
del out_uncond_count
|
|
|
|
|
|
|
|
return out_cond, out_uncond
|
2023-01-26 17:06:48 +00:00
|
|
|
|
|
|
|
|
2023-02-08 19:05:31 +00:00
|
|
|
max_total_area = model_management.maximum_batch_area()
|
|
|
|
cond, uncond = calc_cond_uncond_batch(cond, uncond, x, sigma, max_total_area)
|
2023-01-26 17:06:48 +00:00
|
|
|
return uncond + (cond - uncond) * cond_scale
|
2023-01-03 06:53:32 +00:00
|
|
|
|
|
|
|
def simple_scheduler(model, steps):
|
|
|
|
sigs = []
|
|
|
|
ss = len(model.sigmas) / steps
|
|
|
|
for x in range(steps):
|
|
|
|
sigs += [float(model.sigmas[-(1 + int(x * ss))])]
|
|
|
|
sigs += [0.0]
|
|
|
|
return torch.FloatTensor(sigs)
|
|
|
|
|
2023-01-26 17:06:48 +00:00
|
|
|
def create_cond_with_same_area_if_none(conds, c):
|
|
|
|
if 'area' not in c[1]:
|
|
|
|
return
|
|
|
|
|
|
|
|
c_area = c[1]['area']
|
|
|
|
smallest = None
|
|
|
|
for x in conds:
|
|
|
|
if 'area' in x[1]:
|
|
|
|
a = x[1]['area']
|
|
|
|
if c_area[2] >= a[2] and c_area[3] >= a[3]:
|
|
|
|
if a[0] + a[2] >= c_area[0] + c_area[2]:
|
|
|
|
if a[1] + a[3] >= c_area[1] + c_area[3]:
|
|
|
|
if smallest is None:
|
|
|
|
smallest = x
|
|
|
|
elif 'area' not in smallest[1]:
|
|
|
|
smallest = x
|
|
|
|
else:
|
|
|
|
if smallest[1]['area'][0] * smallest[1]['area'][1] > a[0] * a[1]:
|
|
|
|
smallest = x
|
|
|
|
else:
|
|
|
|
if smallest is None:
|
|
|
|
smallest = x
|
|
|
|
if smallest is None:
|
|
|
|
return
|
|
|
|
if 'area' in smallest[1]:
|
|
|
|
if smallest[1]['area'] == c_area:
|
|
|
|
return
|
|
|
|
n = c[1].copy()
|
|
|
|
conds += [[smallest[0], n]]
|
2023-01-03 06:53:32 +00:00
|
|
|
|
|
|
|
class KSampler:
|
|
|
|
SCHEDULERS = ["karras", "normal", "simple"]
|
|
|
|
SAMPLERS = ["sample_euler", "sample_euler_ancestral", "sample_heun", "sample_dpm_2", "sample_dpm_2_ancestral",
|
|
|
|
"sample_lms", "sample_dpm_fast", "sample_dpm_adaptive", "sample_dpmpp_2s_ancestral", "sample_dpmpp_sde",
|
|
|
|
"sample_dpmpp_2m"]
|
|
|
|
|
|
|
|
def __init__(self, model, steps, device, sampler=None, scheduler=None, denoise=None):
|
|
|
|
self.model = model
|
|
|
|
if self.model.parameterization == "v":
|
2023-02-08 21:51:19 +00:00
|
|
|
self.model_wrap = k_diffusion_external.CompVisVDenoiser(self.model, quantize=True)
|
2023-01-03 06:53:32 +00:00
|
|
|
else:
|
2023-02-08 21:51:19 +00:00
|
|
|
self.model_wrap = k_diffusion_external.CompVisDenoiser(self.model, quantize=True)
|
2023-01-26 17:06:48 +00:00
|
|
|
self.model_k = CFGDenoiserComplex(self.model_wrap)
|
2023-01-03 06:53:32 +00:00
|
|
|
self.device = device
|
|
|
|
if scheduler not in self.SCHEDULERS:
|
|
|
|
scheduler = self.SCHEDULERS[0]
|
|
|
|
if sampler not in self.SAMPLERS:
|
|
|
|
sampler = self.SAMPLERS[0]
|
|
|
|
self.scheduler = scheduler
|
|
|
|
self.sampler = sampler
|
|
|
|
self.sigma_min=float(self.model_wrap.sigmas[0])
|
|
|
|
self.sigma_max=float(self.model_wrap.sigmas[-1])
|
|
|
|
self.set_steps(steps, denoise)
|
|
|
|
|
|
|
|
def _calculate_sigmas(self, steps):
|
|
|
|
sigmas = None
|
|
|
|
|
|
|
|
discard_penultimate_sigma = False
|
|
|
|
if self.sampler in ['sample_dpm_2', 'sample_dpm_2_ancestral']:
|
|
|
|
steps += 1
|
|
|
|
discard_penultimate_sigma = True
|
|
|
|
|
|
|
|
if self.scheduler == "karras":
|
2023-02-08 21:51:19 +00:00
|
|
|
sigmas = k_diffusion_sampling.get_sigmas_karras(n=steps, sigma_min=self.sigma_min, sigma_max=self.sigma_max, device=self.device)
|
2023-01-03 06:53:32 +00:00
|
|
|
elif self.scheduler == "normal":
|
|
|
|
sigmas = self.model_wrap.get_sigmas(steps).to(self.device)
|
|
|
|
elif self.scheduler == "simple":
|
|
|
|
sigmas = simple_scheduler(self.model_wrap, steps).to(self.device)
|
|
|
|
else:
|
|
|
|
print("error invalid scheduler", self.scheduler)
|
|
|
|
|
|
|
|
if discard_penultimate_sigma:
|
|
|
|
sigmas = torch.cat([sigmas[:-2], sigmas[-1:]])
|
|
|
|
return sigmas
|
|
|
|
|
|
|
|
def set_steps(self, steps, denoise=None):
|
|
|
|
self.steps = steps
|
|
|
|
if denoise is None:
|
|
|
|
self.sigmas = self._calculate_sigmas(steps)
|
|
|
|
else:
|
|
|
|
new_steps = int(steps/denoise)
|
|
|
|
sigmas = self._calculate_sigmas(new_steps)
|
|
|
|
self.sigmas = sigmas[-(steps + 1):]
|
|
|
|
|
|
|
|
|
2023-01-31 08:09:38 +00:00
|
|
|
def sample(self, noise, positive, negative, cfg, latent_image=None, start_step=None, last_step=None, force_full_denoise=False):
|
2023-01-03 06:53:32 +00:00
|
|
|
sigmas = self.sigmas
|
|
|
|
sigma_min = self.sigma_min
|
|
|
|
|
2023-01-31 08:09:38 +00:00
|
|
|
if last_step is not None and last_step < (len(sigmas) - 1):
|
2023-01-03 06:53:32 +00:00
|
|
|
sigma_min = sigmas[last_step]
|
|
|
|
sigmas = sigmas[:last_step + 1]
|
2023-01-31 08:09:38 +00:00
|
|
|
if force_full_denoise:
|
|
|
|
sigmas[-1] = 0
|
|
|
|
|
2023-01-03 06:53:32 +00:00
|
|
|
if start_step is not None:
|
2023-01-31 08:09:38 +00:00
|
|
|
if start_step < (len(sigmas) - 1):
|
|
|
|
sigmas = sigmas[start_step:]
|
|
|
|
else:
|
|
|
|
if latent_image is not None:
|
|
|
|
return latent_image
|
|
|
|
else:
|
|
|
|
return torch.zeros_like(noise)
|
2023-01-03 06:53:32 +00:00
|
|
|
|
|
|
|
noise *= sigmas[0]
|
|
|
|
if latent_image is not None:
|
|
|
|
noise += latent_image
|
|
|
|
|
2023-01-26 17:06:48 +00:00
|
|
|
positive = positive[:]
|
|
|
|
negative = negative[:]
|
|
|
|
#make sure each cond area has an opposite one with the same area
|
|
|
|
for c in positive:
|
|
|
|
create_cond_with_same_area_if_none(negative, c)
|
|
|
|
for c in negative:
|
|
|
|
create_cond_with_same_area_if_none(positive, c)
|
|
|
|
|
2023-01-03 06:53:32 +00:00
|
|
|
if self.model.model.diffusion_model.dtype == torch.float16:
|
|
|
|
precision_scope = torch.autocast
|
|
|
|
else:
|
|
|
|
precision_scope = contextlib.nullcontext
|
|
|
|
|
|
|
|
with precision_scope(self.device):
|
|
|
|
if self.sampler == "sample_dpm_fast":
|
2023-02-08 21:51:19 +00:00
|
|
|
samples = k_diffusion_sampling.sample_dpm_fast(self.model_k, noise, sigma_min, sigmas[0], self.steps, extra_args={"cond":positive, "uncond":negative, "cond_scale": cfg})
|
2023-01-03 06:53:32 +00:00
|
|
|
elif self.sampler == "sample_dpm_adaptive":
|
2023-02-08 21:51:19 +00:00
|
|
|
samples = k_diffusion_sampling.sample_dpm_adaptive(self.model_k, noise, sigma_min, sigmas[0], extra_args={"cond":positive, "uncond":negative, "cond_scale": cfg})
|
2023-01-03 06:53:32 +00:00
|
|
|
else:
|
2023-02-08 21:51:19 +00:00
|
|
|
samples = getattr(k_diffusion_sampling, self.sampler)(self.model_k, noise, sigmas, extra_args={"cond":positive, "uncond":negative, "cond_scale": cfg})
|
2023-01-03 06:53:32 +00:00
|
|
|
return samples.to(torch.float32)
|