[Template Bug]: Ovis text-to-image generating same style/content images

### Confirmation

- [x] I have confirmed this is a workflow template mistake

### Template Name

ovis_text_to_image

### Problem Description

In subgraph mode this template only generates kids with balloons. Like the prompt is not passed to it. In full workflow mode it works fine, however. Running ComfyUI under Ubuntu 24.04. See attachment, please.

### ComfyUI Mode

Nodes 2.0 Mode

### Error Message / Logs

```text
got prompt
Using pytorch attention in VAE
Using pytorch attention in VAE
VAE load device: cuda:0, offload device: cpu, dtype: torch.bfloat16
CLIP/text encoder model load device: cuda:0, offload device: cpu, current: cpu, dtype: torch.float16
Requested to load OvisTEModel_
loaded completely; 22440.30 MB usable, 3281.74 MB loaded, full load: True
model weight dtype torch.bfloat16, manual cast: None
model_type FLOW
Requested to load Flux
loaded completely; 18961.96 MB usable, 14058.02 MB loaded, full load: True
100%|██████████| 20/20 [00:31<00:00, 1.57s/it]
Requested to load AutoencodingEngine
loaded completely; 1563.74 MB usable, 159.87 MB loaded, full load: True
Prompt executed in 39.02 seconds
got prompt
100%|██████████| 20/20 [00:31<00:00, 1.57s/it]
Prompt executed in 33.62 seconds
got prompt
100%|██████████| 20/20 [00:32<00:00, 1.63s/it]
Prompt executed in 34.88 seconds
got prompt
100%|██████████| 20/20 [00:31<00:00, 1.58s/it]
Prompt executed in 33.75 seconds
```