mirrors/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2025-09-14 13:35:05 +00:00
Code Issues Packages Projects Releases Wiki Activity

P2 of qwen edit model. (#9412)

Browse Source 
* P2 of qwen edit model.

* Typo.

* Fix normal qwen.

* Fix.

* Make the TextEncodeQwenImageEdit also set the ref latent.

If you don't want it to set the ref latent and want to use the
ReferenceLatent node with your custom latent instead just disconnect the
VAE.
This commit is contained in:
comfyanonymous
2025-08-18 19:38:34 -07:00
committed by GitHub
parent bd2ab73976
commit 4977f203fa
10 changed files with 565 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
comfy/text_encoders/qwen_image.py
View File

@@ -15,13 +15,27 @@ class QwenImageTokenizer(sd1_clip.SD1Tokenizer):
def __init__(self, embedding_directory=None, tokenizer_data={}):
super().__init__(embedding_directory=embedding_directory, tokenizer_data=tokenizer_data, name="qwen25_7b", tokenizer=Qwen25_7BVLITokenizer)
self.llama_template = "<|im_start|>system\nDescribe the image by detailing the color, shape, size, texture, quantity, text, spatial relationships of the objects and background:<|im_end|>\n<|im_start|>user\n{}<|im_end|>\n<|im_start|>assistant\n"
self.llama_template_images = "<|im_start|>system\nDescribe the key features of the input image \\(color, shape, size, texture, objects, background\\), then explain how the user's text instruction should alter or modify the image. Generate a new image that meets the user's requirements while maintaining consistency with the original input where appropriate.<|im_end|>\n<|im_start|>user\n<|vision_start|><|image_pad|><|vision_end|>{}<|im_end|>\n<|im_start|>assistant\n"
def tokenize_with_weights(self, text, return_word_ids=False, llama_template=None,**kwargs):
def tokenize_with_weights(self, text, return_word_ids=False, llama_template=None, images=[], **kwargs):
if llama_template is None:
llama_text = self.llama_template.format(text)
if len(images) > 0:
llama_text = self.llama_template_images.format(text)
else:
llama_text = self.llama_template.format(text)
else:
llama_text = llama_template.format(text)
return super().tokenize_with_weights(llama_text, return_word_ids=return_word_ids, **kwargs)
tokens = super().tokenize_with_weights(llama_text, return_word_ids=return_word_ids, **kwargs)
key_name = next(iter(tokens))
embed_count = 0
qwen_tokens = tokens[key_name]
for r in qwen_tokens:
for i in range(len(r)):
if r[i][0] == 151655:
if len(images) > embed_count:
r[i] = ({"type": "image", "data": images[embed_count], "original_type": "image"},) + r[i][1:]
embed_count += 1
return tokens
class Qwen25_7BVLIModel(sd1_clip.SDClipModel):