Increase tolerance for test_normalize_data_basic test

[MagnusBuehler]

Motivation and Context

The test "test_normalize_data_basic" randomly failed due to float16 precision. This fix increases the tolerance if the dtype is torch.float16 to 1e-2 instead of the default 1e-3. If this further fails randomly, increase tolerance further.

Public API Changes

• No Public API changes
• Yes, Public API changes (Details below)

---

How Has This Been Tested?

Needs to be tested through future CI runs.

Checklist

• The changes have been tested locally.
• Documentation has been updated (if the public API or usage changes).
• A entry has been added to CHANGELOG.md (if relevant for users).
• The code follows the project's style guidelines.
• I have considered the impact of these changes on the public API.

---

[gemini-code-assist]

Code Review

This pull request addresses a randomly failing test, test_normalize_data_basic, by increasing the absolute tolerance for torch.float16 precision. The change is logical and directly solves the issue of floating-point inaccuracies with half-precision floats. My review includes a suggestion to define the new tolerance values as named constants to improve code readability and make future adjustments easier, as hinted in the pull request description.