docs: add Bedrock inference example notebook (llamastack#4586)

## Summary

Add an interactive Jupyter notebook demonstrating AWS Bedrock provider
inference via Llama Stack's OpenAI-compatible API.

**Notebook covers:**
- Simple chat completion (non-streaming)
- Streaming responses with SSE
- Multi-turn conversations (client-side context)
- System messages for role-based prompting
- Sampling parameters (temperature, top_p, max_tokens)
- Error handling patterns
- Batch processing multiple prompts
- Equivalent cURL commands for reference

**Also includes:**
- Prerequisites and server setup instructions
- API support matrix (what works/doesn't with Bedrock)
- Troubleshooting guide for common errors
- Bearer token refresh instructions

## Test plan
- [ ] Verified notebook renders correctly on GitHub
- [ ] Ran notebook cells with a live Bedrock server
- [ ] Confirmed all examples execute without errors

---------

Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Author: skamenan7