mirror of
https://git.datalinker.icu/comfyanonymous/ComfyUI
synced 2026-11-20 16:07:12 +08:00
Fix some resolutions not working.
This commit is contained in:
parent
d19dea22e6
commit
18e958d6cb
@ -7,15 +7,14 @@ import torch.nn.functional as F
|
|||||||
from comfy.ldm.modules.diffusionmodules.mmdit import Mlp, TimestepEmbedder, PatchEmbed, RMSNorm
|
from comfy.ldm.modules.diffusionmodules.mmdit import Mlp, TimestepEmbedder, PatchEmbed, RMSNorm
|
||||||
from comfy.ldm.modules.diffusionmodules.util import timestep_embedding
|
from comfy.ldm.modules.diffusionmodules.util import timestep_embedding
|
||||||
from torch.utils import checkpoint
|
from torch.utils import checkpoint
|
||||||
from tqdm import tqdm
|
|
||||||
|
|
||||||
from .attn_layers import Attention, CrossAttention
|
from .attn_layers import Attention, CrossAttention
|
||||||
from .poolers import AttentionPool
|
from .poolers import AttentionPool
|
||||||
from .posemb_layers import get_2d_rotary_pos_embed, get_fill_resize_and_crop
|
from .posemb_layers import get_2d_rotary_pos_embed, get_fill_resize_and_crop
|
||||||
|
|
||||||
def calc_rope(x, patch_size, head_size):
|
def calc_rope(x, patch_size, head_size):
|
||||||
th = x.shape[2] // patch_size
|
th = (x.shape[2] + (patch_size // 2)) // patch_size
|
||||||
tw = x.shape[3] // patch_size
|
tw = (x.shape[3] + (patch_size // 2)) // patch_size
|
||||||
base_size = 512 // 8 // patch_size
|
base_size = 512 // 8 // patch_size
|
||||||
start, stop = get_fill_resize_and_crop((th, tw), base_size)
|
start, stop = get_fill_resize_and_crop((th, tw), base_size)
|
||||||
sub_args = [start, stop, (th, tw)]
|
sub_args = [start, stop, (th, tw)]
|
||||||
@ -327,7 +326,7 @@ class HunYuanDiT(nn.Module):
|
|||||||
text_states = torch.where(clip_t5_mask.unsqueeze(2), text_states, self.text_embedding_padding.to(text_states))
|
text_states = torch.where(clip_t5_mask.unsqueeze(2), text_states, self.text_embedding_padding.to(text_states))
|
||||||
|
|
||||||
_, _, oh, ow = x.shape
|
_, _, oh, ow = x.shape
|
||||||
th, tw = oh // self.patch_size, ow // self.patch_size
|
th, tw = (oh + (self.patch_size // 2)) // self.patch_size, (ow + (self.patch_size // 2)) // self.patch_size
|
||||||
|
|
||||||
|
|
||||||
# Get image RoPE embedding according to `reso`lution.
|
# Get image RoPE embedding according to `reso`lution.
|
||||||
@ -382,8 +381,8 @@ class HunYuanDiT(nn.Module):
|
|||||||
if return_dict:
|
if return_dict:
|
||||||
return {'x': x}
|
return {'x': x}
|
||||||
if self.learn_sigma:
|
if self.learn_sigma:
|
||||||
return x[:,:self.out_channels // 2]
|
return x[:,:self.out_channels // 2,:oh,:ow]
|
||||||
return x
|
return x[:,:,:oh,:ow]
|
||||||
|
|
||||||
def unpatchify(self, x, h, w):
|
def unpatchify(self, x, h, w):
|
||||||
"""
|
"""
|
||||||
|
|||||||
Loading…
x
Reference in New Issue
Block a user