Add MiniMax M2.7 tutorial notebook and litellm integration

[octo-patch]

Summary

• Add minimax-m2.7.ipynb: hands-on tutorial notebook for MiniMax M2.7 (204K context) covering text generation, streaming, JSON structured output, Pydantic validation, tool/function calling, summarization, and data labelling via OpenAI-compatible API
• Add MiniMax M2.7 as a third provider in notebook 26 (multiple LLM providers with litellm), alongside GPT-4.1-mini and Gemini 2.5 Flash
• Update README.md with Model Explorations section
• Add 28 unit tests and 6 integration tests (34 total)

Why MiniMax?

MiniMax M2.7 is a powerful LLM with a 204K context window and an OpenAI-compatible API (https://api.minimax.io/v1). It supports:

• Chat completions with system prompts
• Streaming responses
• JSON structured output
• Function/tool calling
• Both standard (MiniMax-M2.7) and highspeed (MiniMax-M2.7-highspeed) variants

This makes it a great addition to the bootcamp's multi-provider tutorial, showing learners how easy it is to switch between different LLM providers.

Files Changed

File	Change
minimax-m2.7.ipynb	New tutorial notebook (32 cells)
26.multiple-llm-providers-with-litellm.ipynb	Added MiniMax M2.7 via litellm
README.md	Added Model Explorations section
tests/test_minimax.py	28 unit tests
tests/test_minimax_integration.py	6 integration tests
tests/__init__.py	Test package init

Test Plan

• 28 unit tests pass (config, format, structured output, tools, notebooks)
• 6 integration tests pass (completion, system prompt, JSON, streaming, highspeed, usage)
• Verify notebooks run correctly in Google Colab with MINIMAX_API_KEY set