Patch fails to install

[DrJimmyBrungus]

I am trying to install for the first time today, and getting the following error when attempting to run the patch.
The readme says I can try to apply the hunks by hand if the install fails, but I'm completely clueless here and have no idea what that means. Are you able to assist?

Running the latest version of Forge Neo as of today - 2.28

E:\Stable Diffusion\sd-webui-forge-neo>git apply --verbose krea2-features-backend.patch
Checking patch backend/diffusion_engine/krea.py...
error: while searching for:
import torch
from huggingface_guess import model_list

from backend import memory_management
from backend.diffusion_engine.base import ForgeDiffusionEngine, ForgeObjects
from backend.modules.k_prediction import PredictionDiscreteFlow
from backend.patcher.clip import CLIP
from backend.patcher.unet import UnetPatcher
from backend.patcher.vae import VAE
from backend.text_processing.qwen3vl_engine import Qwen3VLTextProcessingEngine


class Krea2(ForgeDiffusionEngine):

error: patch failed: backend/diffusion_engine/krea.py:1
error: backend/diffusion_engine/krea.py: patch does not apply
Checking patch backend/nn/krea.py...
error: while searching for:
# https://github.com/Comfy-Org/ComfyUI/blob/v0.26.1/comfy/ldm/krea2/model.py

from typing import Optional

import torch

error: patch failed: backend/nn/krea.py:1
error: backend/nn/krea.py: patch does not apply
Checking patch backend/nn/llm/llama.py...
Hunk #3 succeeded at 447 (offset -2 lines).
Hunk #4 succeeded at 464 (offset -2 lines).
Hunk #5 succeeded at 512 (offset -2 lines).
Hunk #6 succeeded at 701 (offset -2 lines).
Hunk #7 succeeded at 725 (offset -2 lines).
Hunk #8 succeeded at 742 (offset -2 lines).
Checking patch backend/text_processing/qwen3vl_engine.py...
error: while searching for:
        self.vision_block = "<|vision_start|><|image_pad|><|vision_end|>"

    def tokenize(self, texts, images=[]):
        llama_texts = [((self.image_template.replace(self.vision_block, self.vision_block * len(images), 1) if images else self.llama_template).format(text)) if text else " " for text in texts]
        return self.tokenizer(llama_texts)["input_ids"]

    def tokenize_line(self, line: str, images=[]):

error: patch failed: backend/text_processing/qwen3vl_engine.py:37
error: backend/text_processing/qwen3vl_engine.py: patch does not apply