Add GPT-5.5 and GPT-5.6 Tiktoken support

[Copilot]

Adds Tiktoken model resolution for GPT-5.5 and GPT-5.6 using the o200k_base encoding.

Changes

• Support exact model names: gpt-5.5, gpt-5.6
• Support prefixed variants such as gpt-5.5-mini and gpt-5.6-nano
• Extend model creation coverage for both families

TiktokenTokenizer.CreateForModel("gpt-5.6");

[Copilot]

Pull request overview

Adds GPT-5.5 and GPT-5.6 model-name resolution to the ML.NET TiktokenTokenizer so CreateForModel(...) maps these models (and their -<variant> suffix forms) to the o200k_base encoding, with test coverage updates.

Changes:

• Extend TiktokenTokenizer model-to-encoding mappings to include gpt-5.5 / gpt-5.6 and gpt-5.5-* / gpt-5.6-* prefixes.
• Add test cases validating creation and supported-name handling for gpt-5.5 / gpt-5.6 and representative suffix variants.

Show a summary per file

File	Description
src/Microsoft.ML.Tokenizers/Model/TiktokenTokenizer.cs	Adds exact and prefix model mappings for GPT-5.5/GPT-5.6 to O200kBase.
test/Microsoft.ML.Tokenizers.Tests/TiktokenTests.cs	Extends supported-model test matrices to include GPT-5.5/GPT-5.6 (and variant examples).

Copilot's findings

• Files reviewed: 2/2 changed files
• Comments generated: 0

[tarekgh]

/ba-g opened the issue #7662 to track unrelated failures.

[tarekgh]

Thanks @stephentoub!