298977887/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2025-04-18 01:53:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'comfyanonymous:master' into multigpu_support

Browse Source
This commit is contained in:
Jedrzej Kosinski 2025-01-07 20:11:27 -06:00 committed by GitHub
commit 66838ebd39
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 15 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
app/app_settings.py
View File

@ -1,6 +1,7 @@
import os import os
import json import json
from aiohttp import web from aiohttp import web
import logging
class AppSettings(): class AppSettings():
@ -11,8 +12,12 @@ class AppSettings():
file = self.user_manager.get_request_user_filepath( file = self.user_manager.get_request_user_filepath(
request, "comfy.settings.json") request, "comfy.settings.json")
if os.path.isfile(file): if os.path.isfile(file):
with open(file) as f: try:
return json.load(f) with open(file) as f:
return json.load(f)
except:
logging.error(f"The user settings file is corrupted: {file}")
return {}
else: else:
return {} return {}

10
comfy/k_diffusion/sampling.py
View File

@ -475,7 +475,7 @@ class DPMSolver(nn.Module):
return x_3, eps_cache return x_3, eps_cache
def dpm_solver_fast(self, x, t_start, t_end, nfe, eta=0., s_noise=1., noise_sampler=None): def dpm_solver_fast(self, x, t_start, t_end, nfe, eta=0., s_noise=1., noise_sampler=None):
noise_sampler = default_noise_sampler(x) if noise_sampler is None else noise_sampler noise_sampler = default_noise_sampler(x, seed=self.extra_args.get("seed", None)) if noise_sampler is None else noise_sampler
if not t_end > t_start and eta: if not t_end > t_start and eta:
raise ValueError('eta must be 0 for reverse sampling') raise ValueError('eta must be 0 for reverse sampling')
@ -514,7 +514,7 @@ class DPMSolver(nn.Module):
return x return x
def dpm_solver_adaptive(self, x, t_start, t_end, order=3, rtol=0.05, atol=0.0078, h_init=0.05, pcoeff=0., icoeff=1., dcoeff=0., accept_safety=0.81, eta=0., s_noise=1., noise_sampler=None): def dpm_solver_adaptive(self, x, t_start, t_end, order=3, rtol=0.05, atol=0.0078, h_init=0.05, pcoeff=0., icoeff=1., dcoeff=0., accept_safety=0.81, eta=0., s_noise=1., noise_sampler=None):
noise_sampler = default_noise_sampler(x) if noise_sampler is None else noise_sampler noise_sampler = default_noise_sampler(x, seed=self.extra_args.get("seed", None)) if noise_sampler is None else noise_sampler
if order not in {2, 3}: if order not in {2, 3}:
raise ValueError('order should be 2 or 3') raise ValueError('order should be 2 or 3')
forward = t_end > t_start forward = t_end > t_start
@ -894,7 +894,8 @@ def DDPMSampler_step(x, sigma, sigma_prev, noise, noise_sampler):
def generic_step_sampler(model, x, sigmas, extra_args=None, callback=None, disable=None, noise_sampler=None, step_function=None): def generic_step_sampler(model, x, sigmas, extra_args=None, callback=None, disable=None, noise_sampler=None, step_function=None):
extra_args = {} if extra_args is None else extra_args extra_args = {} if extra_args is None else extra_args
noise_sampler = default_noise_sampler(x) if noise_sampler is None else noise_sampler seed = extra_args.get("seed", None)
noise_sampler = default_noise_sampler(x, seed=seed) if noise_sampler is None else noise_sampler
s_in = x.new_ones([x.shape[0]]) s_in = x.new_ones([x.shape[0]])
for i in trange(len(sigmas) - 1, disable=disable): for i in trange(len(sigmas) - 1, disable=disable):
@ -914,7 +915,8 @@ def sample_ddpm(model, x, sigmas, extra_args=None, callback=None, disable=None,
@torch.no_grad() @torch.no_grad()
def sample_lcm(model, x, sigmas, extra_args=None, callback=None, disable=None, noise_sampler=None): def sample_lcm(model, x, sigmas, extra_args=None, callback=None, disable=None, noise_sampler=None):
extra_args = {} if extra_args is None else extra_args extra_args = {} if extra_args is None else extra_args
noise_sampler = default_noise_sampler(x) if noise_sampler is None else noise_sampler seed = extra_args.get("seed", None)
noise_sampler = default_noise_sampler(x, seed=seed) if noise_sampler is None else noise_sampler
s_in = x.new_ones([x.shape[0]]) s_in = x.new_ones([x.shape[0]])
for i in trange(len(sigmas) - 1, disable=disable): for i in trange(len(sigmas) - 1, disable=disable):
denoised = model(x, sigmas[i] * s_in, **extra_args) denoised = model(x, sigmas[i] * s_in, **extra_args)

3
comfy/text_encoders/t5.py
View File

@ -227,8 +227,9 @@ class T5(torch.nn.Module):
super().__init__() super().__init__()
self.num_layers = config_dict["num_layers"] self.num_layers = config_dict["num_layers"]
model_dim = config_dict["d_model"] model_dim = config_dict["d_model"]
inner_dim = config_dict["d_kv"] * config_dict["num_heads"]
self.encoder = T5Stack(self.num_layers, model_dim, model_dim, config_dict["d_ff"], config_dict["dense_act_fn"], config_dict["is_gated_act"], config_dict["num_heads"], config_dict["model_type"] != "umt5", dtype, device, operations) self.encoder = T5Stack(self.num_layers, model_dim, inner_dim, config_dict["d_ff"], config_dict["dense_act_fn"], config_dict["is_gated_act"], config_dict["num_heads"], config_dict["model_type"] != "umt5", dtype, device, operations)
self.dtype = dtype self.dtype = dtype
self.shared = operations.Embedding(config_dict["vocab_size"], model_dim, device=device, dtype=dtype) self.shared = operations.Embedding(config_dict["vocab_size"], model_dim, device=device, dtype=dtype)