I don't think so,

@sweetmax797 if you're not sure, I can try and let you know,

is it some error unrelated to the one I mentioned in the description?

@sweetmax797 I dont think so because it recognise the text encoder:
"Attenpting to find mmproj file for text encoder...

Using mmproj 'QwenImageEdit2.5-VL-7B-Instruct-Q3_K_M-mmproj-F16.gguf' for text encoder 'QwenImageEdit2.5-VL-7B-Instruct-Q3_K_M-mmproj-F16.gguf'."
What I am getting is:
"clip missing: ['visual.patch_embed.proj.weight', 'visual.blocks.0.norm1.weight', 'visual.blocks.0.norm2.weight', 'visual.blocks.0.attn.qkv.weight', 'visual.blocks.0.attn.proj.weight', 'visual.blocks.0.mlp.gate_proj.weight', 'visual.blocks.0.mlp.up_proj.weight', 'visual.blocks.0.mlp.down_proj.weight', 'visual.blocks.1.norm1.weight', 'visual.blocks.1.norm2.weight', 'visual.blocks.1.attn.qkv.weight', 'visual.blocks.1.attn.proj.weight', 'visual.blocks.1.mlp.gate_proj.weight', 'visual.blocks.1.mlp.up_proj.weight', 'visual.blocks.1.mlp.down_proj.weight', 'visual.blocks.2.norm1.weight', 'visual.blocks.2.norm2.weight', 'visual.blocks.2.attn.qkv.weight', 'visual.blocks.2.attn.proj.weight', 'visual.blocks.2.mlp.gate_proj.weight', 'visual.blocks.2.mlp.up_proj.weight', 'visual.blocks.2.mlp.down_proj.weight', 'visual.blocks.3.norm1.weight', 'visual.blocks.3.norm2.weight', 'visual.blocks.3.attn.qkv.weight', 'visual.blocks.3.attn.proj.weight', 'visual.blocks.3.mlp.gate_proj.weight', 'visual.blocks.3.mlp.up_proj.weight', 'visual.blocks.3.mlp.down_proj.weight', 'visual.blocks.4.norm1.weight', 'visual.blocks.4.norm2.weight', 'visual.blocks.4.attn.qkv.weight', 'visual.blocks.4.attn.proj.weight', 'visual.blocks.4.mlp.gate_proj.weight', 'visual.blocks.4.mlp.up_proj.weight', 'visual.blocks.4.mlp.down_proj.weight', 'visual.blocks.5.norm1.weight', 'visual.blocks.5.norm2.weight', 'visual.blocks.5.attn.qkv.weight', 'visual.blocks.5.attn.proj.weight', 'visual.blocks.5.mlp.gate_proj.weight', 'visual.blocks.5.mlp.up_proj.weight', 'visual.blocks.5.mlp.down_proj.weight', 'visual.blocks.6.norm1.weight', 'visual.blocks.6.norm2.weight', 'visual.blocks.6.attn.qkv.weight', 'visual.blocks.6.attn.proj.weight', 'visual.blocks.6.mlp.gate_proj.weight', 'visual.blocks.6.mlp.up_proj.weight', 'visual.blocks.6.mlp.down_proj.weight', 'visual.blocks.7.norm1.weight', 'visual.blocks.7.norm2.weight', 'visual.blocks.7.attn.qkv.weight', 'visual.blocks.7.attn.proj.weight', 'visual.blocks.7.mlp.gate_proj.weight', 'visual.blocks.7.mlp.up_proj.weight', 'visual.blocks.7.mlp.down_proj.weight', 'visual.blocks.8.norm1.weight', 'visual.blocks.8.norm2.weight', 'visual.blocks.8.attn.qkv.weight', 'visual.blocks.8.attn.proj.weight', 'visual.blocks.8.mlp.gate_proj.weight', 'visual.blocks.8.mlp.up_proj.weight', 'visual.blocks.8.mlp.down_proj.weight', 'visual.blocks.9.norm1.weight', 'visual.blocks.9.norm2.weight', 'visual.blocks.9.attn.qkv.weight', 'visual.blocks.9.attn.proj.weight', 'visual.blocks.9.mlp.gate_proj.weight', 'visual.blocks.9.mlp.up_proj.weight', 'visual.blocks.9.mlp.down_proj.weight', 'visual.blocks.10.norm1.weight', 'visual.blocks.10.norm2.weight', 'visual.blocks.10.attn.qkv.weight', 'visual.blocks.10.attn.proj.weight', 'visual.blocks.10.mlp.gate_proj.weight', 'visual.blocks.10.mlp.up_proj.weight', 'visual.blocks.10.mlp.down_proj.weight', 'visual.blocks.11.norm1.weight', 'visual.blocks.11.norm2.weight', 'visual.blocks.11.attn.qkv.weight', 'visual.blocks.11.attn.proj.weight', 'visual.blocks.11.mlp.gate_proj.weight', 'visual.blocks.11.mlp.up_proj.weight', 'visual.blocks.11.mlp.down_proj.weight', 'visual.blocks.12.norm1.weight', 'visual.blocks.12.norm2.weight', 'visual.blocks.12.attn.qkv.weight', 'visual.blocks.12.attn.proj.weight', 'visual.blocks.12.mlp.gate_proj.weight', 'visual.blocks.12.mlp.up_proj.weight', 'visual.blocks.12.mlp.down_proj.weight', 'visual.blocks.13.norm1.weight', 'visual.blocks.13.norm2.weight', 'visual.blocks.13.attn.qkv.weight', 'visual.blocks.13.attn.proj.weight', 'visual.blocks.13.mlp.gate_proj.weight', 'visual.blocks.13.mlp.up_proj.weight', 'visual.blocks.13.mlp.down_proj.weight', 'visual.blocks.14.norm1.weight', 'visual.blocks.14.norm2.weight', 'visual.blocks.14.attn.qkv.weight', 'visual.blocks.14.attn.proj.weight', 'visual.blocks.14.mlp.gate_proj.weight', 'visual.blocks.14.mlp.up_proj.weight', 'visual.blocks.14.mlp.down_proj.weight', 'visual.blocks.15.norm1.weight', 'visual.blocks.15.norm2.weight', 'visual.blocks.15.attn.qkv.weight', 'visual.blocks.15.attn.proj.weight', 'visual.blocks.15.mlp.gate_proj.weight', 'visual.blocks.15.mlp.up_proj.weight', 'visual.blocks.15.mlp.down_proj.weight', 'visual.blocks.16.norm1.weight', 'visual.blocks.16.norm2.weight', 'visual.blocks.16.attn.qkv.weight', 'visual.blocks.16.attn.proj.weight', 'visual.blocks.16.mlp.gate_proj.weight', 'visual.blocks.16.mlp.up_proj.weight', 'visual.blocks.16.mlp.down_proj.weight', 'visual.blocks.17.norm1.weight', 'visual.blocks.17.norm2.weight', 'visual.blocks.17.attn.qkv.weight', 'visual.blocks.17.attn.proj.weight', 'visual.blocks.17.mlp.gate_proj.weight', 'visual.blocks.17.mlp.up_proj.weight', 'visual.blocks.17.mlp.down_proj.weight', 'visual.blocks.18.norm1.weight', 'visual.blocks.18.norm2.weight', 'visual.blocks.18.attn.qkv.weight', 'visual.blocks.18.attn.proj.weight', 'visual.blocks.18.mlp.gate_proj.weight', 'visual.blocks.18.mlp.up_proj.weight', 'visual.blocks.18.mlp.down_proj.weight', 'visual.blocks.19.norm1.weight', 'visual.blocks.19.norm2.weight', 'visual.blocks.19.attn.qkv.weight', 'visual.blocks.19.attn.proj.weight', 'visual.blocks.19.mlp.gate_proj.weight', 'visual.blocks.19.mlp.up_proj.weight', 'visual.blocks.19.mlp.down_proj.weight', 'visual.blocks.20.norm1.weight', 'visual.blocks.20.norm2.weight', 'visual.blocks.20.attn.qkv.weight', 'visual.blocks.20.attn.proj.weight', 'visual.blocks.20.mlp.gate_proj.weight', 'visual.blocks.20.mlp.up_proj.weight', 'visual.blocks.20.mlp.down_proj.weight', 'visual.blocks.21.norm1.weight', 'visual.blocks.21.norm2.weight', 'visual.blocks.21.attn.qkv.weight', 'visual.blocks.21.attn.proj.weight', 'visual.blocks.21.mlp.gate_proj.weight', 'visual.blocks.21.mlp.up_proj.weight', 'visual.blocks.21.mlp.down_proj.weight', 'visual.blocks.22.norm1.weight', 'visual.blocks.22.norm2.weight', 'visual.blocks.22.attn.qkv.weight', 'visual.blocks.22.attn.proj.weight', 'visual.blocks.22.mlp.gate_proj.weight', 'visual.blocks.22.mlp.up_proj.weight', 'visual.blocks.22.mlp.down_proj.weight', 'visual.blocks.23.norm1.weight', 'visual.blocks.23.norm2.weight', 'visual.blocks.23.attn.qkv.weight', 'visual.blocks.23.attn.proj.weight', 'visual.blocks.23.mlp.gate_proj.weight', 'visual.blocks.23.mlp.up_proj.weight', 'visual.blocks.23.mlp.down_proj.weight', 'visual.blocks.24.norm1.weight', 'visual.blocks.24.norm2.weight', 'visual.blocks.24.attn.qkv.weight', 'visual.blocks.24.attn.proj.weight', 'visual.blocks.24.mlp.gate_proj.weight', 'visual.blocks.24.mlp.up_proj.weight', 'visual.blocks.24.mlp.down_proj.weight', 'visual.blocks.25.norm1.weight', 'visual.blocks.25.norm2.weight', 'visual.blocks.25.attn.qkv.weight', 'visual.blocks.25.attn.proj.weight', 'visual.blocks.25.mlp.gate_proj.weight', 'visual.blocks.25.mlp.up_proj.weight', 'visual.blocks.25.mlp.down_proj.weight', 'visual.blocks.26.norm1.weight', 'visual.blocks.26.norm2.weight', 'visual.blocks.26.attn.qkv.weight', 'visual.blocks.26.attn.proj.weight', 'visual.blocks.26.mlp.gate_proj.weight', 'visual.blocks.26.mlp.up_proj.weight', 'visual.blocks.26.mlp.down_proj.weight', 'visual.blocks.27.norm1.weight', 'visual.blocks.27.norm2.weight', 'visual.blocks.27.attn.qkv.weight', 'visual.blocks.27.attn.proj.weight', 'visual.blocks.27.mlp.gate_proj.weight', 'visual.blocks.27.mlp.up_proj.weight', 'visual.blocks.27.mlp.down_proj.weight', 'visual.blocks.28.norm1.weight', 'visual.blocks.28.norm2.weight', 'visual.blocks.28.attn.qkv.weight', 'visual.blocks.28.attn.proj.weight', 'visual.blocks.28.mlp.gate_proj.weight', 'visual.blocks.28.mlp.up_proj.weight', 'visual.blocks.28.mlp.down_proj.weight', 'visual.blocks.29.norm1.weight', 'visual.blocks.29.norm2.weight', 'visual.blocks.29.attn.qkv.weight', 'visual.blocks.29.attn.proj.weight', 'visual.blocks.29.mlp.gate_proj.weight', 'visual.blocks.29.mlp.up_proj.weight', 'visual.blocks.29.mlp.down_proj.weight', 'visual.blocks.30.norm1.weight', 'visual.blocks.30.norm2.weight', 'visual.blocks.30.attn.qkv.weight', 'visual.blocks.30.attn.proj.weight', 'visual.blocks.30.mlp.gate_proj.weight', 'visual.blocks.30.mlp.up_proj.weight', 'visual.blocks.30.mlp.down_proj.weight', 'visual.blocks.31.norm1.weight', 'visual.blocks.31.norm2.weight', 'visual.blocks.31.attn.qkv.weight', 'visual.blocks.31.attn.proj.weight', 'visual.blocks.31.mlp.gate_proj.weight', 'visual.blocks.31.mlp.up_proj.weight', 'visual.blocks.31.mlp.down_proj.weight', 'visual.merger.ln_q.weight', 'visual.merger.mlp.0.weight', 'visual.merger.mlp.2.weight']"

Could it be the lightining lora? I was using the one for qwen image, since I have read some people saying it was performing better.

@simiyet looks like mismatch, did u download all files here or form other sources? i checked mine works fine,,, start with gguf loader node, see if it is updated,, if it is try to red-ownload them form : https://huggingface.co/unsloth/Qwen2.5-VL-7B-Instruct-GGUF/tree/main both text encoder mmproj file (at the bottom of the repo) see if it helps
alternatively try this repo too if 'unsloth' doesn't solve the issue:
https://huggingface.co/bartowski/Qwen2-VL-7B-Instruct-GGUF/tree/main

@simiyet lightning lora generally slows done the generation combined with gguf, (if i remember currently) or maybe it was something specific to my pc but try to run it without it and monitor the s/it in terminal

@sweetmax797 I downloaded both, from here (txt-enc-7b-Intruct-Q3) and from unsloth HF (Qwen2.5-VL-7B-Instruct-Q3_K_M.gguf), and got the same error. I'll try bartowski.
The lora actually halfen the generation time. But lora change didnt solve the issue.

Wait.. Do I need txt-enc-7b-Intruct-Q3 or mmprof-F16 verison? Sorry, first time trying QWEN models.

@simiyet yes, put them in text_encoders folder and in the GGUF clip loader, make sure u select the text-encoder model not the mmproj file

@sweetmax797 Oh so I need both.. I only downloaded the text encoder.. Hope this will solve the issue.

@simiyet Yes, it has a bit of a learning curve, but suddenly everything clicks, and the next thing you know, you're developing a custom node for it! :)

@sweetmax797 That solved the issue! Thank you so much!

@simiyet you're welcome!

there is a 7GB Q2 model QuantStack/Qwen-Image-Edit-GGUF at main

are you serious right now? u can run fp8 with 8gb vram, if u can't, there is wrong setup, not the vram

@sweetmax797 thx idk about that,cause ive seen on youtube the size matters to vram, like 10gb file require 10gbvram

@Seii1 oh no, they're wrong. the size of the model has nothing to do with the vram. ask ai. I run full wan2.2 fp16 on 10gb vram 64gb ram i7 , even full qwen-image-edit bf16 40gb in size with ease. not to mension all other applications and 3 big monitors running at 120ghz. u have 12gb vram, u can do much more

@sweetmax797 thx for the answer

its took like 20minutes just to load things xD , maybe i should download the GGuf version

@Seii1 what type of CPU do u have? how much ram. ddr4 or ddr5, and what kind of hard disk , windows or linux?

@sweetmax797 kinda old, 5600x, 32gb ram,NVME , if my ram 64 thats not a problem i think

@Seii1 64GB RAM + NVME great. do u have i5 or i7, anything above 10th generation is more than enough. Im guessing you have internal GPU too, if it is disabled, enable it via BIOS so u offload more of the OS to internal and let your nvidia be used only for AI.
----
You need to install sage attention and Triton. after that you need to use advanced diffiusion loader and use these settings:
{'compute_dtype': torch.bfloat16, 'sage_attention': 'sageattn_qk_int8_pv_fp16_triton', 'use_fp16_accumulation': True}

in comfyui manager search for this node pack: https://github.com/kijai/ComfyUI-KJNodes
It has this node and choose those settings, if u run fp8 change the weight type to default
https://imgur.com/a/fmxxtDF

@sweetmax797 im using 4060ti 16gb, i gues i just need more ram to load faster, thx for the tips

@Seii1 yes more ram in your situation is very useful, Ti is more for energy officency and have less cuda cores compare to regular ones. your ssd is great, ram is cheap, but first fix your workflow, and windows 11 is just not good, to much overhead, it eats your resources, Ubuntu or LinuxMint best for AI.

i have 16gb vram + 32ram before,many workflow need 20min, than i up to 64ram and every change