mirror of https://github.com/vladmandic/automatic
301 lines
12 KiB
Python
301 lines
12 KiB
Python
|
|
import numpy as np
|
|
import torch
|
|
from diffusers.configuration_utils import ConfigMixin, register_to_config
|
|
from diffusers.schedulers.scheduling_utils import SchedulerMixin, SchedulerOutput
|
|
|
|
|
|
class RungeKutta67Scheduler(SchedulerMixin, ConfigMixin):
|
|
"""
|
|
RK6_7S: 6th-order Runge-Kutta scheduler with 7 stages.
|
|
Adapted from the RES4LYF repository.
|
|
(Note: Defined as 5th order in some contexts, but follows the 7-stage tableau).
|
|
"""
|
|
|
|
order = 1
|
|
|
|
@register_to_config
|
|
def __init__(
|
|
self,
|
|
num_train_timesteps: int = 1000,
|
|
beta_start: float = 0.00085,
|
|
beta_end: float = 0.012,
|
|
beta_schedule: str = "linear",
|
|
trained_betas: np.ndarray | list[float] | None = None,
|
|
prediction_type: str = "epsilon",
|
|
use_karras_sigmas: bool = False,
|
|
use_exponential_sigmas: bool = False,
|
|
use_beta_sigmas: bool = False,
|
|
use_flow_sigmas: bool = False,
|
|
sigma_min: float | None = None,
|
|
sigma_max: float | None = None,
|
|
rho: float = 7.0,
|
|
shift: float | None = None,
|
|
base_shift: float = 0.5,
|
|
max_shift: float = 1.15,
|
|
use_dynamic_shifting: bool = False,
|
|
timestep_spacing: str = "linspace",
|
|
clip_sample: bool = False,
|
|
sample_max_value: float = 1.0,
|
|
set_alpha_to_one: bool = False,
|
|
skip_prk_steps: bool = False,
|
|
interpolation_type: str = "linear",
|
|
steps_offset: int = 0,
|
|
timestep_type: str = "discrete",
|
|
rescale_betas_zero_snr: bool = False,
|
|
final_sigmas_type: str = "zero",
|
|
):
|
|
if trained_betas is not None:
|
|
self.betas = torch.tensor(trained_betas, dtype=torch.float32)
|
|
elif beta_schedule == "linear":
|
|
self.betas = torch.linspace(beta_start, beta_end, num_train_timesteps, dtype=torch.float32)
|
|
elif beta_schedule == "scaled_linear":
|
|
self.betas = torch.linspace(beta_start**0.5, beta_end**0.5, num_train_timesteps, dtype=torch.float32) ** 2
|
|
else:
|
|
raise NotImplementedError(f"{beta_schedule} is not implemented")
|
|
|
|
self.alphas = 1.0 - self.betas
|
|
self.alphas_cumprod = torch.cumprod(self.alphas, dim=0)
|
|
self.init_noise_sigma = 1.0
|
|
|
|
# internal state
|
|
self.num_inference_steps = None
|
|
self.timesteps = torch.from_numpy(np.arange(0, num_train_timesteps)[::-1].copy())
|
|
self.sigmas = None
|
|
self.model_outputs = []
|
|
self.sample_at_start_of_step = None
|
|
self._sigmas_cpu = None
|
|
self._timesteps_cpu = None
|
|
self._step_index = None
|
|
|
|
def set_timesteps(
|
|
self,
|
|
num_inference_steps: int,
|
|
device: str | torch.device = None,
|
|
mu: float | None = None, dtype: torch.dtype = torch.float32):
|
|
self.num_inference_steps = num_inference_steps
|
|
|
|
# 1. Spacing
|
|
if self.config.timestep_spacing == "linspace":
|
|
timesteps = np.linspace(self.config.num_train_timesteps - 1, 0, num_inference_steps, dtype=float).copy()
|
|
elif self.config.timestep_spacing == "leading":
|
|
step_ratio = self.config.num_train_timesteps // num_inference_steps
|
|
timesteps = (np.arange(0, num_inference_steps) * step_ratio).round()[::-1].copy().astype(float)
|
|
elif self.config.timestep_spacing == "trailing":
|
|
step_ratio = self.config.num_train_timesteps / num_inference_steps
|
|
timesteps = np.arange(self.config.num_train_timesteps, 0, -step_ratio).round().copy().astype(float)
|
|
timesteps -= step_ratio
|
|
else:
|
|
raise ValueError(f"timestep_spacing must be one of 'linspace', 'leading', or 'trailing', got {self.config.timestep_spacing}")
|
|
|
|
# Ensure trailing ends at 0
|
|
if self.config.timestep_spacing == "trailing":
|
|
timesteps = np.maximum(timesteps, 0)
|
|
|
|
sigmas = np.array(((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5)
|
|
if self.config.interpolation_type == "linear":
|
|
sigmas = np.interp(timesteps, np.arange(len(sigmas)), sigmas)
|
|
elif self.config.interpolation_type == "log_linear":
|
|
sigmas = np.exp(np.interp(timesteps, np.arange(len(sigmas)), np.log(sigmas)))
|
|
else:
|
|
raise ValueError(f"interpolation_type must be one of 'linear' or 'log_linear', got {self.config.interpolation_type}")
|
|
|
|
if self.config.use_karras_sigmas:
|
|
sigma_min = self.config.sigma_min if self.config.sigma_min is not None else sigmas[-1]
|
|
sigma_max = self.config.sigma_max if self.config.sigma_max is not None else sigmas[0]
|
|
rho = self.config.rho
|
|
ramp = np.linspace(0, 1, num_inference_steps)
|
|
sigmas = (sigma_max ** (1 / rho) + ramp * (sigma_min ** (1 / rho) - sigma_max ** (1 / rho))) ** rho
|
|
elif self.config.use_exponential_sigmas:
|
|
sigma_min = self.config.sigma_min if self.config.sigma_min is not None else sigmas[-1]
|
|
sigma_max = self.config.sigma_max if self.config.sigma_max is not None else sigmas[0]
|
|
sigmas = np.exp(np.linspace(np.log(sigma_max), np.log(sigma_min), num_inference_steps))
|
|
elif self.config.use_beta_sigmas:
|
|
sigma_min = self.config.sigma_min if self.config.sigma_min is not None else sigmas[-1]
|
|
sigma_max = self.config.sigma_max if self.config.sigma_max is not None else sigmas[0]
|
|
alpha, beta = 0.6, 0.6
|
|
ramp = np.linspace(0, 1, num_inference_steps)
|
|
try:
|
|
import torch.distributions as dist
|
|
|
|
b = dist.Beta(alpha, beta)
|
|
ramp = b.sample((num_inference_steps,)).sort().values.numpy()
|
|
except Exception:
|
|
pass
|
|
sigmas = sigma_max * (1 - ramp) + sigma_min * ramp
|
|
elif self.config.use_flow_sigmas:
|
|
sigmas = np.linspace(1.0, 1 / 1000, num_inference_steps)
|
|
|
|
# 3. Shifting
|
|
if self.config.use_dynamic_shifting and mu is not None:
|
|
sigmas = mu * sigmas / (1 + (mu - 1) * sigmas)
|
|
elif self.config.shift is not None:
|
|
sigmas = self.config.shift * sigmas / (1 + (self.config.shift - 1) * sigmas)
|
|
|
|
# RK6_7s c values: [0, 1/3, 2/3, 1/3, 1/2, 1/2, 1]
|
|
c_values = [0, 1 / 3, 2 / 3, 1 / 3, 1 / 2, 1 / 2, 1]
|
|
|
|
sigmas_expanded = []
|
|
for i in range(len(sigmas) - 1):
|
|
s_curr = sigmas[i]
|
|
s_next = sigmas[i + 1]
|
|
for c in c_values:
|
|
sigmas_expanded.append(s_curr + c * (s_next - s_curr))
|
|
sigmas_expanded.append(0.0)
|
|
|
|
sigmas_interpolated = np.array(sigmas_expanded)
|
|
# Linear remapping for Flow Matching
|
|
timesteps_expanded = sigmas_interpolated * self.config.num_train_timesteps
|
|
self.sigmas = torch.from_numpy(sigmas_interpolated).to(device=device, dtype=dtype)
|
|
self.timesteps = torch.from_numpy(timesteps_expanded + self.config.steps_offset).to(device=device, dtype=dtype)
|
|
|
|
self.init_noise_sigma = self.sigmas.max().item() if self.sigmas.numel() > 0 else 1.0
|
|
self._sigmas_cpu = self.sigmas.detach().cpu().numpy()
|
|
self._timesteps_cpu = self.timesteps.detach().cpu().numpy()
|
|
self._step_index = None
|
|
|
|
self.model_outputs = []
|
|
self.sample_at_start_of_step = None
|
|
|
|
@property
|
|
def step_index(self):
|
|
"""
|
|
The index counter for the current timestep. It will increase 1 after each scheduler step.
|
|
"""
|
|
return self._step_index
|
|
|
|
def index_for_timestep(self, timestep, schedule_timesteps=None):
|
|
from .scheduler_utils import index_for_timestep
|
|
if schedule_timesteps is None:
|
|
schedule_timesteps = self.timesteps
|
|
return index_for_timestep(timestep, schedule_timesteps)
|
|
|
|
def _init_step_index(self, timestep):
|
|
if self._step_index is None:
|
|
if isinstance(timestep, torch.Tensor):
|
|
timestep = timestep.to(self.timesteps.device)
|
|
self._step_index = self.index_for_timestep(timestep)
|
|
|
|
def scale_model_input(self, sample: torch.Tensor, timestep: float | torch.Tensor) -> torch.Tensor:
|
|
if self._step_index is None:
|
|
self._init_step_index(timestep)
|
|
if self.config.prediction_type == "flow_prediction":
|
|
return sample
|
|
sigma = self._sigmas_cpu[self._step_index]
|
|
return sample / ((sigma**2 + 1) ** 0.5)
|
|
|
|
def step(
|
|
self,
|
|
model_output: torch.Tensor,
|
|
timestep: float | torch.Tensor,
|
|
sample: torch.Tensor,
|
|
return_dict: bool = True,
|
|
) -> SchedulerOutput | tuple:
|
|
if self._step_index is None:
|
|
self._init_step_index(timestep)
|
|
|
|
step_index = self._step_index
|
|
stage_index = step_index % 7
|
|
|
|
base_step_index = (step_index // 7) * 7
|
|
sigma_curr = self._sigmas_cpu[base_step_index]
|
|
sigma_next_idx = min(base_step_index + 7, len(self._sigmas_cpu) - 1)
|
|
sigma_next = self._sigmas_cpu[sigma_next_idx]
|
|
h = sigma_next - sigma_curr
|
|
|
|
sigma_t = self._sigmas_cpu[step_index]
|
|
alpha_t = 1 / (sigma_t**2 + 1) ** 0.5
|
|
sigma_actual = sigma_t * alpha_t
|
|
|
|
prediction_type = getattr(self.config, "prediction_type", "epsilon")
|
|
if prediction_type == "epsilon":
|
|
denoised = sample - sigma_t * model_output
|
|
elif prediction_type == "v_prediction":
|
|
alpha_t = 1 / (sigma_t**2 + 1) ** 0.5
|
|
sigma_actual = sigma_t * alpha_t
|
|
denoised = alpha_t * sample - sigma_actual * model_output
|
|
elif prediction_type == "flow_prediction":
|
|
denoised = sample - sigma_t * model_output
|
|
elif prediction_type == "sample":
|
|
denoised = model_output
|
|
else:
|
|
raise ValueError(f"prediction_type error: {prediction_type}")
|
|
|
|
if self.config.clip_sample:
|
|
denoised = denoised.clamp(-self.config.sample_max_value, self.config.sample_max_value)
|
|
|
|
# derivative = (x - x0) / sigma
|
|
derivative = (sample - denoised) / sigma_t if sigma_t > 1e-6 else torch.zeros_like(sample)
|
|
|
|
if self.sample_at_start_of_step is None:
|
|
if stage_index > 0:
|
|
# Mid-step fallback for Img2Img/Inpainting
|
|
sigma_next_t = self._sigmas_cpu[self._step_index + 1]
|
|
dt = sigma_next_t - sigma_t
|
|
prev_sample = sample + dt * derivative
|
|
self._step_index += 1
|
|
if not return_dict:
|
|
return (prev_sample,)
|
|
return SchedulerOutput(prev_sample=prev_sample)
|
|
|
|
self.sample_at_start_of_step = sample
|
|
self.model_outputs = [derivative] * stage_index
|
|
|
|
# Butcher Tableau A matrix for rk6_7s
|
|
a = [
|
|
[],
|
|
[1 / 3],
|
|
[0, 2 / 3],
|
|
[1 / 12, 1 / 3, -1 / 12],
|
|
[-1 / 16, 9 / 8, -3 / 16, -3 / 8],
|
|
[0, 9 / 8, -3 / 8, -3 / 4, 1 / 2],
|
|
[9 / 44, -9 / 11, 63 / 44, 18 / 11, 0, -16 / 11],
|
|
]
|
|
|
|
# Butcher Tableau B weights for rk6_7s
|
|
b = [11 / 120, 0, 27 / 40, 27 / 40, -4 / 15, -4 / 15, 11 / 120]
|
|
|
|
if stage_index == 0:
|
|
self.model_outputs = [derivative]
|
|
self.sample_at_start_of_step = sample
|
|
else:
|
|
self.model_outputs.append(derivative)
|
|
|
|
if stage_index < 6:
|
|
# Predict next stage sample: y_next_stage = y_start + h * sum(a[stage_index+1][j] * k[j])
|
|
next_a_row = a[stage_index + 1]
|
|
sum_ak = torch.zeros_like(derivative)
|
|
for j, weight in enumerate(next_a_row):
|
|
sum_ak += weight * self.model_outputs[j]
|
|
|
|
prev_sample = self.sample_at_start_of_step + h * sum_ak
|
|
else:
|
|
# Final 7th stage complete, calculate final step
|
|
sum_bk = torch.zeros_like(derivative)
|
|
for j, weight in enumerate(b):
|
|
sum_bk += weight * self.model_outputs[j]
|
|
|
|
prev_sample = self.sample_at_start_of_step + h * sum_bk
|
|
|
|
# Clear state
|
|
self.model_outputs = []
|
|
self.sample_at_start_of_step = None
|
|
|
|
self._step_index += 1
|
|
|
|
if not return_dict:
|
|
return (prev_sample,)
|
|
return SchedulerOutput(prev_sample=prev_sample)
|
|
|
|
def add_noise(
|
|
self,
|
|
original_samples: torch.Tensor,
|
|
noise: torch.Tensor,
|
|
timesteps: torch.Tensor,
|
|
) -> torch.Tensor:
|
|
from .scheduler_utils import add_noise_to_sample
|
|
return add_noise_to_sample(original_samples, noise, self.sigmas, timesteps, self.timesteps)
|
|
|
|
def __len__(self):
|
|
return self.config.num_train_timesteps
|