[Bug]: [Bug]: Z-Image-Turbo Text to Image Template throws OOM

### ComfyUI and Frontend Versions

### Description

Workflow throws OOM after any image generation on my 16Gb 5060TI

### Steps to Reproduce

Generate Picture, after that try to generate next picture

### Debug Logs

```powershell
`got prompt
Requested to load Lumina2
loaded completely; 12030.40 MB usable, 11739.55 MB loaded, full load: True
0%| | 0/9 [00:02
k_callback = lambda x: callback(x["i"], x["denoised"], x["x"], total_steps)
File "F:\ComfyUI\ComfyUI.torch2.9\latent_preview.py", line 105, in callback
preview_bytes = previewer.decode_latent_to_preview_image(preview_format, x0)
File "F:\ComfyUI\ComfyUI.torch2.9\latent_preview.py", line 27, in decode_latent_to_preview_image
preview_image = self.decode_latent_to_preview(x0)
File "F:\ComfyUI\ComfyUI.torch2.9\latent_preview.py", line 35, in decode_latent_to_preview
x_sample = self.taesd.decode(x0[:1])[0].movedim(0, 2)
File "F:\ComfyUI\ComfyUI.torch2.9\comfy\taesd\taesd.py", line 74, in decode
x_sample = self.taesd_decoder((x - self.vae_shift) * self.vae_scale)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1777, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1788, in _call_impl
return forward_call(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\container.py", line 250, in forward
input = module(input)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1777, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1788, in _call_impl
return forward_call(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\comfy\taesd\taesd.py", line 26, in forward
return self.fuse(self.conv(x) + self.skip(x))
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1777, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1788, in _call_impl
return forward_call(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\container.py", line 250, in forward
input = module(input)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1777, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\module.py", line 1788, in _call_impl
return forward_call(*args, **kwargs)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\modules\activation.py", line 144, in forward
return F.relu(input, inplace=self.inplace)
File "F:\ComfyUI\ComfyUI.torch2.9\venv\lib\site-packages\torch\nn\functional.py", line 1697, in relu
result = torch.relu(input)
torch.OutOfMemoryError: Allocation on device

Got an OOM, unloading all loaded models.
Prompt executed in 36.75 seconds`
```

### Browser Logs

--

### Setting JSON

--

### What browsers do you use to access the UI ?

Google Chrome