[BugFix][5997203] Update Sqrt casts to FP16 (#1084)

### What does this PR do?

Type of change: Bug fix

Change cast nodes before Sqrt to FP16

### Testing
```
python torch_quant_to_onnx.py --quantize_mode=mxfp8 --timm_model_name=vit_base_patch16_224  --onnx_save_path=vit_base_patch16_224.mxfp8.onnx --calibration_data_size=512

python evaluate.py --onnx_path=vit_base_patch16_224.mxfp8.onnx --model_name=vit_base_patch16_224 --eval_data_size=100
```

### Before your PR is "*Ready for review*"

Make sure you read and follow [Contributor
guidelines](https://github.com/NVIDIA/Model-Optimizer/blob/main/CONTRIBUTING.md)
and your commits are signed (`git commit -s -S`).

Make sure you read and follow the [Security Best
Practices](https://github.com/NVIDIA/Model-Optimizer/blob/main/SECURITY.md#security-coding-practices-for-contributors)
(e.g. avoiding hardcoded `trust_remote_code=True`, `torch.load(...,
weights_only=False)`, `pickle`, etc.).

- Is this change backward compatible?:  ❌ 
- Casts before Sqrt are now FP16 instead of FP32
- If you copied code from any other sources or added a new PIP
dependency, did you follow guidance in `CONTRIBUTING.md`: ✅
- Did you write any new necessary tests?:N/A
- Did you update
[Changelog](https://github.com/NVIDIA/Model-Optimizer/blob/main/CHANGELOG.rst)?:
N/A


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Improved ONNX model export for quantized models with reduced precision
(fp16/bf16) by enhancing type casting handling during the export
process.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->

Signed-off-by: ajrasane <131806219+ajrasane@users.noreply.github.com>

Author: ajrasane

modelopt/torch/_deploy/utils/torch_onnx.py

@@ -608,7 +608,7 @@ def get_onnx_bytes_and_metadata(
                 op_block_list=["QuantizeLinear", "DequantizeLinear", "Div"],
             )
             # Change FP32 cast nodes feeding into Concat/Add to FP16
-            onnx_opt_graph = change_casts_to_fp16(onnx_opt_graph, ["Concat", "Add"])
+            onnx_opt_graph = change_casts_to_fp16(onnx_opt_graph, ["Concat", "Add", "Sqrt"])
         else:
             onnx_opt_graph = convert_to_f16(
                 onnx_opt_graph, low_precision_type=weights_dtype, keep_io_types=False