The LTX2.0 FP4 quantized model presents compatibility issues with LoRA implementations

### Custom Node Testing

- [ ] I have tried disabling custom nodes and the issue persists (see [how to disable custom nodes](https://docs.comfy.org/troubleshooting/custom-node-issues#step-1%3A-test-with-all-custom-nodes-disabled) if you need help)

### Expected Behavior

he FP4 model demonstrates limited compatibility with LoRA adaptations, whereas the FP8 model effectively executes LoRA functionality with precision. However, character generations employing LoRA within the FP4 framework reveal outputs that completely deviate from the trained subject matter.

### Actual Behavior

none

### Steps to Reproduce

none

### Debug Logs

```powershell
none
```

### Other

_No response_