BUG:AI tagging works inconsistently — some images are not tagged

### Is there an existing issue for this?

- [x] I have searched the existing issues

### What happened?

## 🐞 Bug: AI tagging works inconsistently — some images are not tagged

### Summary
When using the **AI tagging** feature, tagging works only for some images.
For example, after adding **6 images**, only **3 of them receive tags**, while the remaining images are left untagged without any error message or feedback.

I’ve attached a short video demonstrating this behavior.

---

### Environment
- **App type:** Tauri + React
- **OS:** Windows 10 )
- **PictoPy v1.1.0**
-

---

### Steps to Reproduce
1. Add multiple images to the app (e.g. 5–6 images).
2. Enable or trigger the **AI tagging** feature.
3. Wait for the tagging process to complete.
4. Observe that only some images receive tags, while others remain untagged.

---

### Current Behavior
- AI tagging runs successfully for a subset of images.
- Some images are skipped or left untagged.
- No error, warning, or status message is shown for the failed images.
- From a user perspective, it’s unclear whether:
- tagging failed,
- tagging is still in progress,
- or those images are unsupported.

---

### Expected Behavior
- Either **all images** should be processed and tagged,
**or**
- The app should clearly indicate:
- which images failed to tag
- and why (e.g. unsupported format, model error, timeout, etc.).

---

### Possible Causes (speculation)
This could be related to:
- unsupported image formats or sizes
- timeouts or batching issues in the AI tagging pipeline
- silent errors from the AI service or model
- missing progress/error handling in the UI

---

### Why this matters
AI tagging is a key feature, and inconsistent behavior without feedback can confuse users and reduce trust in the system.
Clear success/failure handling would greatly improve the UX.

---

### Request to Maintainers

Could the team please confirm whether this behavior is expected?

If this is a confirmed issue, I’d love to be **assigned to it** so I can help investigate and work on a fix.

Thanks for the great project — happy to contribute! 🙌

### Record

- [x] I agree to follow this project's Code of Conduct