Commit 4df63d2d authored by AUTOMATIC's avatar AUTOMATIC

split samplers into one more files for k-diffusion

parent 27447410
from collections import deque from modules import sd_samplers_compvis, sd_samplers_kdiffusion, shared
import torch
import inspect
import k_diffusion.sampling
import ldm.models.diffusion.ddim
import ldm.models.diffusion.plms
from modules import prompt_parser, devices, sd_samplers_common, sd_samplers_compvis
from modules.shared import opts, state
import modules.shared as shared
from modules.script_callbacks import CFGDenoiserParams, cfg_denoiser_callback
# imports for functions that previously were here and are used by other modules # imports for functions that previously were here and are used by other modules
from modules.sd_samplers_common import samples_to_image_grid, sample_to_image from modules.sd_samplers_common import samples_to_image_grid, sample_to_image
samplers_k_diffusion = [
('Euler a', 'sample_euler_ancestral', ['k_euler_a', 'k_euler_ancestral'], {}),
('Euler', 'sample_euler', ['k_euler'], {}),
('LMS', 'sample_lms', ['k_lms'], {}),
('Heun', 'sample_heun', ['k_heun'], {}),
('DPM2', 'sample_dpm_2', ['k_dpm_2'], {'discard_next_to_last_sigma': True}),
('DPM2 a', 'sample_dpm_2_ancestral', ['k_dpm_2_a'], {'discard_next_to_last_sigma': True}),
('DPM++ 2S a', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a'], {}),
('DPM++ 2M', 'sample_dpmpp_2m', ['k_dpmpp_2m'], {}),
('DPM++ SDE', 'sample_dpmpp_sde', ['k_dpmpp_sde'], {}),
('DPM fast', 'sample_dpm_fast', ['k_dpm_fast'], {}),
('DPM adaptive', 'sample_dpm_adaptive', ['k_dpm_ad'], {}),
('LMS Karras', 'sample_lms', ['k_lms_ka'], {'scheduler': 'karras'}),
('DPM2 Karras', 'sample_dpm_2', ['k_dpm_2_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True}),
('DPM2 a Karras', 'sample_dpm_2_ancestral', ['k_dpm_2_a_ka'], {'scheduler': 'karras', 'discard_next_to_last_sigma': True}),
('DPM++ 2S a Karras', 'sample_dpmpp_2s_ancestral', ['k_dpmpp_2s_a_ka'], {'scheduler': 'karras'}),
('DPM++ 2M Karras', 'sample_dpmpp_2m', ['k_dpmpp_2m_ka'], {'scheduler': 'karras'}),
('DPM++ SDE Karras', 'sample_dpmpp_sde', ['k_dpmpp_sde_ka'], {'scheduler': 'karras'}),
]
samplers_data_k_diffusion = [
sd_samplers_common.SamplerData(label, lambda model, funcname=funcname: KDiffusionSampler(funcname, model), aliases, options)
for label, funcname, aliases, options in samplers_k_diffusion
if hasattr(k_diffusion.sampling, funcname)
]
all_samplers = [ all_samplers = [
*samplers_data_k_diffusion, *sd_samplers_kdiffusion.samplers_data_k_diffusion,
sd_samplers_common.SamplerData('DDIM', lambda model: sd_samplers_compvis.VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), [], {}), *sd_samplers_compvis.samplers_data_compvis,
sd_samplers_common.SamplerData('PLMS', lambda model: sd_samplers_compvis.VanillaStableDiffusionSampler(ldm.models.diffusion.plms.PLMSSampler, model), [], {}),
] ]
all_samplers_map = {x.name: x for x in all_samplers} all_samplers_map = {x.name: x for x in all_samplers}
...@@ -69,8 +31,8 @@ def create_sampler(name, model): ...@@ -69,8 +31,8 @@ def create_sampler(name, model):
def set_samplers(): def set_samplers():
global samplers, samplers_for_img2img global samplers, samplers_for_img2img
hidden = set(opts.hide_samplers) hidden = set(shared.opts.hide_samplers)
hidden_img2img = set(opts.hide_samplers + ['PLMS']) hidden_img2img = set(shared.opts.hide_samplers + ['PLMS'])
samplers = [x for x in all_samplers if x.name not in hidden] samplers = [x for x in all_samplers if x.name not in hidden]
samplers_for_img2img = [x for x in all_samplers if x.name not in hidden_img2img] samplers_for_img2img = [x for x in all_samplers if x.name not in hidden_img2img]
...@@ -83,257 +45,3 @@ def set_samplers(): ...@@ -83,257 +45,3 @@ def set_samplers():
set_samplers() set_samplers()
sampler_extra_params = {
'sample_euler': ['s_churn', 's_tmin', 's_tmax', 's_noise'],
'sample_heun': ['s_churn', 's_tmin', 's_tmax', 's_noise'],
'sample_dpm_2': ['s_churn', 's_tmin', 's_tmax', 's_noise'],
}
class CFGDenoiser(torch.nn.Module):
def __init__(self, model):
super().__init__()
self.inner_model = model
self.mask = None
self.nmask = None
self.init_latent = None
self.step = 0
def combine_denoised(self, x_out, conds_list, uncond, cond_scale):
denoised_uncond = x_out[-uncond.shape[0]:]
denoised = torch.clone(denoised_uncond)
for i, conds in enumerate(conds_list):
for cond_index, weight in conds:
denoised[i] += (x_out[cond_index] - denoised_uncond[i]) * (weight * cond_scale)
return denoised
def forward(self, x, sigma, uncond, cond, cond_scale, image_cond):
if state.interrupted or state.skipped:
raise sd_samplers_common.InterruptedException
conds_list, tensor = prompt_parser.reconstruct_multicond_batch(cond, self.step)
uncond = prompt_parser.reconstruct_cond_batch(uncond, self.step)
batch_size = len(conds_list)
repeats = [len(conds_list[i]) for i in range(batch_size)]
x_in = torch.cat([torch.stack([x[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [x])
image_cond_in = torch.cat([torch.stack([image_cond[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [image_cond])
sigma_in = torch.cat([torch.stack([sigma[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [sigma])
denoiser_params = CFGDenoiserParams(x_in, image_cond_in, sigma_in, state.sampling_step, state.sampling_steps)
cfg_denoiser_callback(denoiser_params)
x_in = denoiser_params.x
image_cond_in = denoiser_params.image_cond
sigma_in = denoiser_params.sigma
if tensor.shape[1] == uncond.shape[1]:
cond_in = torch.cat([tensor, uncond])
if shared.batch_cond_uncond:
x_out = self.inner_model(x_in, sigma_in, cond={"c_crossattn": [cond_in], "c_concat": [image_cond_in]})
else:
x_out = torch.zeros_like(x_in)
for batch_offset in range(0, x_out.shape[0], batch_size):
a = batch_offset
b = a + batch_size
x_out[a:b] = self.inner_model(x_in[a:b], sigma_in[a:b], cond={"c_crossattn": [cond_in[a:b]], "c_concat": [image_cond_in[a:b]]})
else:
x_out = torch.zeros_like(x_in)
batch_size = batch_size*2 if shared.batch_cond_uncond else batch_size
for batch_offset in range(0, tensor.shape[0], batch_size):
a = batch_offset
b = min(a + batch_size, tensor.shape[0])
x_out[a:b] = self.inner_model(x_in[a:b], sigma_in[a:b], cond={"c_crossattn": [tensor[a:b]], "c_concat": [image_cond_in[a:b]]})
x_out[-uncond.shape[0]:] = self.inner_model(x_in[-uncond.shape[0]:], sigma_in[-uncond.shape[0]:], cond={"c_crossattn": [uncond], "c_concat": [image_cond_in[-uncond.shape[0]:]]})
devices.test_for_nans(x_out, "unet")
if opts.live_preview_content == "Prompt":
sd_samplers_common.store_latent(x_out[0:uncond.shape[0]])
elif opts.live_preview_content == "Negative prompt":
sd_samplers_common.store_latent(x_out[-uncond.shape[0]:])
denoised = self.combine_denoised(x_out, conds_list, uncond, cond_scale)
if self.mask is not None:
denoised = self.init_latent * self.mask + self.nmask * denoised
self.step += 1
return denoised
class TorchHijack:
def __init__(self, sampler_noises):
# Using a deque to efficiently receive the sampler_noises in the same order as the previous index-based
# implementation.
self.sampler_noises = deque(sampler_noises)
def __getattr__(self, item):
if item == 'randn_like':
return self.randn_like
if hasattr(torch, item):
return getattr(torch, item)
raise AttributeError("'{}' object has no attribute '{}'".format(type(self).__name__, item))
def randn_like(self, x):
if self.sampler_noises:
noise = self.sampler_noises.popleft()
if noise.shape == x.shape:
return noise
if x.device.type == 'mps':
return torch.randn_like(x, device=devices.cpu).to(x.device)
else:
return torch.randn_like(x)
class KDiffusionSampler:
def __init__(self, funcname, sd_model):
denoiser = k_diffusion.external.CompVisVDenoiser if sd_model.parameterization == "v" else k_diffusion.external.CompVisDenoiser
self.model_wrap = denoiser(sd_model, quantize=shared.opts.enable_quantization)
self.funcname = funcname
self.func = getattr(k_diffusion.sampling, self.funcname)
self.extra_params = sampler_extra_params.get(funcname, [])
self.model_wrap_cfg = CFGDenoiser(self.model_wrap)
self.sampler_noises = None
self.stop_at = None
self.eta = None
self.default_eta = 1.0
self.config = None
self.last_latent = None
self.conditioning_key = sd_model.model.conditioning_key
def callback_state(self, d):
step = d['i']
latent = d["denoised"]
if opts.live_preview_content == "Combined":
sd_samplers_common.store_latent(latent)
self.last_latent = latent
if self.stop_at is not None and step > self.stop_at:
raise sd_samplers_common.InterruptedException
state.sampling_step = step
shared.total_tqdm.update()
def launch_sampling(self, steps, func):
state.sampling_steps = steps
state.sampling_step = 0
try:
return func()
except sd_samplers_common.InterruptedException:
return self.last_latent
def number_of_needed_noises(self, p):
return p.steps
def initialize(self, p):
self.model_wrap_cfg.mask = p.mask if hasattr(p, 'mask') else None
self.model_wrap_cfg.nmask = p.nmask if hasattr(p, 'nmask') else None
self.model_wrap_cfg.step = 0
self.eta = p.eta or opts.eta_ancestral
k_diffusion.sampling.torch = TorchHijack(self.sampler_noises if self.sampler_noises is not None else [])
extra_params_kwargs = {}
for param_name in self.extra_params:
if hasattr(p, param_name) and param_name in inspect.signature(self.func).parameters:
extra_params_kwargs[param_name] = getattr(p, param_name)
if 'eta' in inspect.signature(self.func).parameters:
extra_params_kwargs['eta'] = self.eta
return extra_params_kwargs
def get_sigmas(self, p, steps):
discard_next_to_last_sigma = self.config is not None and self.config.options.get('discard_next_to_last_sigma', False)
if opts.always_discard_next_to_last_sigma and not discard_next_to_last_sigma:
discard_next_to_last_sigma = True
p.extra_generation_params["Discard penultimate sigma"] = True
steps += 1 if discard_next_to_last_sigma else 0
if p.sampler_noise_scheduler_override:
sigmas = p.sampler_noise_scheduler_override(steps)
elif self.config is not None and self.config.options.get('scheduler', None) == 'karras':
sigma_min, sigma_max = (0.1, 10) if opts.use_old_karras_scheduler_sigmas else (self.model_wrap.sigmas[0].item(), self.model_wrap.sigmas[-1].item())
sigmas = k_diffusion.sampling.get_sigmas_karras(n=steps, sigma_min=sigma_min, sigma_max=sigma_max, device=shared.device)
else:
sigmas = self.model_wrap.get_sigmas(steps)
if discard_next_to_last_sigma:
sigmas = torch.cat([sigmas[:-2], sigmas[-1:]])
return sigmas
def sample_img2img(self, p, x, noise, conditioning, unconditional_conditioning, steps=None, image_conditioning=None):
steps, t_enc = sd_samplers_common.setup_img2img_steps(p, steps)
sigmas = self.get_sigmas(p, steps)
sigma_sched = sigmas[steps - t_enc - 1:]
xi = x + noise * sigma_sched[0]
extra_params_kwargs = self.initialize(p)
if 'sigma_min' in inspect.signature(self.func).parameters:
## last sigma is zero which isn't allowed by DPM Fast & Adaptive so taking value before last
extra_params_kwargs['sigma_min'] = sigma_sched[-2]
if 'sigma_max' in inspect.signature(self.func).parameters:
extra_params_kwargs['sigma_max'] = sigma_sched[0]
if 'n' in inspect.signature(self.func).parameters:
extra_params_kwargs['n'] = len(sigma_sched) - 1
if 'sigma_sched' in inspect.signature(self.func).parameters:
extra_params_kwargs['sigma_sched'] = sigma_sched
if 'sigmas' in inspect.signature(self.func).parameters:
extra_params_kwargs['sigmas'] = sigma_sched
self.model_wrap_cfg.init_latent = x
self.last_latent = x
samples = self.launch_sampling(t_enc + 1, lambda: self.func(self.model_wrap_cfg, xi, extra_args={
'cond': conditioning,
'image_cond': image_conditioning,
'uncond': unconditional_conditioning,
'cond_scale': p.cfg_scale
}, disable=False, callback=self.callback_state, **extra_params_kwargs))
return samples
def sample(self, p, x, conditioning, unconditional_conditioning, steps=None, image_conditioning = None):
steps = steps or p.steps
sigmas = self.get_sigmas(p, steps)
x = x * sigmas[0]
extra_params_kwargs = self.initialize(p)
if 'sigma_min' in inspect.signature(self.func).parameters:
extra_params_kwargs['sigma_min'] = self.model_wrap.sigmas[0].item()
extra_params_kwargs['sigma_max'] = self.model_wrap.sigmas[-1].item()
if 'n' in inspect.signature(self.func).parameters:
extra_params_kwargs['n'] = steps
else:
extra_params_kwargs['sigmas'] = sigmas
self.last_latent = x
samples = self.launch_sampling(steps, lambda: self.func(self.model_wrap_cfg, x, extra_args={
'cond': conditioning,
'image_cond': image_conditioning,
'uncond': unconditional_conditioning,
'cond_scale': p.cfg_scale
}, disable=False, callback=self.callback_state, **extra_params_kwargs))
return samples
from collections import namedtuple, deque from collections import namedtuple
import numpy as np import numpy as np
import torch import torch
from PIL import Image from PIL import Image
...@@ -64,6 +64,7 @@ class InterruptedException(BaseException): ...@@ -64,6 +64,7 @@ class InterruptedException(BaseException):
# MPS fix for randn in torchsde # MPS fix for randn in torchsde
# XXX move this to separate file for MPS
def torchsde_randn(size, dtype, device, seed): def torchsde_randn(size, dtype, device, seed):
if device.type == 'mps': if device.type == 'mps':
generator = torch.Generator(devices.cpu).manual_seed(int(seed)) generator = torch.Generator(devices.cpu).manual_seed(int(seed))
......
import math import math
import ldm.models.diffusion.ddim
import ldm.models.diffusion.plms
import numpy as np import numpy as np
import torch import torch
...@@ -7,6 +9,12 @@ from modules.shared import state ...@@ -7,6 +9,12 @@ from modules.shared import state
from modules import sd_samplers_common, prompt_parser, shared from modules import sd_samplers_common, prompt_parser, shared
samplers_data_compvis = [
sd_samplers_common.SamplerData('DDIM', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), [], {}),
sd_samplers_common.SamplerData('PLMS', lambda model: VanillaStableDiffusionSampler(ldm.models.diffusion.plms.PLMSSampler, model), [], {}),
]
class VanillaStableDiffusionSampler: class VanillaStableDiffusionSampler:
def __init__(self, constructor, sd_model): def __init__(self, constructor, sd_model):
self.sampler = constructor(sd_model) self.sampler = constructor(sd_model)
......
...@@ -2,18 +2,12 @@ from collections import deque ...@@ -2,18 +2,12 @@ from collections import deque
import torch import torch
import inspect import inspect
import k_diffusion.sampling import k_diffusion.sampling
import ldm.models.diffusion.ddim
import ldm.models.diffusion.plms
from modules import prompt_parser, devices, sd_samplers_common, sd_samplers_compvis from modules import prompt_parser, devices, sd_samplers_common, sd_samplers_compvis
from modules.shared import opts, state from modules.shared import opts, state
import modules.shared as shared import modules.shared as shared
from modules.script_callbacks import CFGDenoiserParams, cfg_denoiser_callback from modules.script_callbacks import CFGDenoiserParams, cfg_denoiser_callback
# imports for functions that previously were here and are used by other modules
from modules.sd_samplers_common import samples_to_image_grid, sample_to_image
samplers_k_diffusion = [ samplers_k_diffusion = [
('Euler a', 'sample_euler_ancestral', ['k_euler_a', 'k_euler_ancestral'], {}), ('Euler a', 'sample_euler_ancestral', ['k_euler_a', 'k_euler_ancestral'], {}),
('Euler', 'sample_euler', ['k_euler'], {}), ('Euler', 'sample_euler', ['k_euler'], {}),
...@@ -40,50 +34,6 @@ samplers_data_k_diffusion = [ ...@@ -40,50 +34,6 @@ samplers_data_k_diffusion = [
if hasattr(k_diffusion.sampling, funcname) if hasattr(k_diffusion.sampling, funcname)
] ]
all_samplers = [
*samplers_data_k_diffusion,
sd_samplers_common.SamplerData('DDIM', lambda model: sd_samplers_compvis.VanillaStableDiffusionSampler(ldm.models.diffusion.ddim.DDIMSampler, model), [], {}),
sd_samplers_common.SamplerData('PLMS', lambda model: sd_samplers_compvis.VanillaStableDiffusionSampler(ldm.models.diffusion.plms.PLMSSampler, model), [], {}),
]
all_samplers_map = {x.name: x for x in all_samplers}
samplers = []
samplers_for_img2img = []
samplers_map = {}
def create_sampler(name, model):
if name is not None:
config = all_samplers_map.get(name, None)
else:
config = all_samplers[0]
assert config is not None, f'bad sampler name: {name}'
sampler = config.constructor(model)
sampler.config = config
return sampler
def set_samplers():
global samplers, samplers_for_img2img
hidden = set(opts.hide_samplers)
hidden_img2img = set(opts.hide_samplers + ['PLMS'])
samplers = [x for x in all_samplers if x.name not in hidden]
samplers_for_img2img = [x for x in all_samplers if x.name not in hidden_img2img]
samplers_map.clear()
for sampler in all_samplers:
samplers_map[sampler.name.lower()] = sampler.name
for alias in sampler.aliases:
samplers_map[alias.lower()] = sampler.name
set_samplers()
sampler_extra_params = { sampler_extra_params = {
'sample_euler': ['s_churn', 's_tmin', 's_tmax', 's_noise'], 'sample_euler': ['s_churn', 's_tmin', 's_tmax', 's_noise'],
'sample_heun': ['s_churn', 's_tmin', 's_tmax', 's_noise'], 'sample_heun': ['s_churn', 's_tmin', 's_tmax', 's_noise'],
...@@ -92,6 +42,13 @@ sampler_extra_params = { ...@@ -92,6 +42,13 @@ sampler_extra_params = {
class CFGDenoiser(torch.nn.Module): class CFGDenoiser(torch.nn.Module):
"""
Classifier free guidance denoiser. A wrapper for stable diffusion model (specifically for unet)
that can take a noisy picture and produce a noise-free picture using two guidances (prompts)
instead of one. Originally, the second prompt is just an empty string, but we use non-empty
negative prompt.
"""
def __init__(self, model): def __init__(self, model):
super().__init__() super().__init__()
self.inner_model = model self.inner_model = model
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment