feat: add streaming tool use to llama-cpp-python

[lsorber]

Changes:

1. ✨ Enable streaming tool use for llama-cpp-python models (see below for details). The result is a more simple rag and async_rag implementation that opens the door to Agentic RAG and user-defined tools.
2. ✨ Update the query parameter description to require that it is a single-faceted question (i.e., a non-compound question) to encourage parallel function calling for compound questions.
3. ✅ Add fairly thorough tests for both rag and the improved chatml-function-calling chat handler.

Changes to llama-cpp-python's chatml-function-calling chat handler:

1. General:
   a. ✨ If no system message is supplied, add an empty system message to hold the tool metadata.
   b. ✨ Add function descriptions to the system message so that tool use is better informed (fixes chatml-function-callling not adding tool description to the prompt. abetlen/llama-cpp-python#1869).
   c. ✨ Replace print statements relating to JSON grammars with RuntimeWarning warnings.
   d. ✅ Add tests with fairly broad coverage of the different scenarios.
2. Case "Tool choice by user":
   a. ✨ Add support for more than one function call by making this a special case of "Automatic tool choice" with a single tool (subsumes Support parallel function calls with tool_choice abetlen/llama-cpp-python#1503).
3. Case "Automatic tool choice -> respond with a message":
   a. ✨ Use user-defined stop and max_tokens.
   b. 🐛 Replace incorrect use of follow-up grammar with user-defined grammar.
4. Case "Automatic tool choice -> one or more function calls":
   a. ✨ Add support for streaming the function calls (fixes Feature request: add support for streaming tool use abetlen/llama-cpp-python#1883).
   b. ✨ Make tool calling more robust by giving the LLM an explicit way to terminate the tool calls by wrapping them in a <function_calls></function_calls> block.
   c. 🐛 Add missing ":" stop token to determine whether to continue with another tool call, which prevented parallel function calling (fixes chatml-function-calling chat format fails to generate multi calls to the same tool abetlen/llama-cpp-python#1756).
   d. ✨ Set temperature=0 to determine whether to continue with another tool call, similar to the initial decision on whether to call a tool.

[lsorber]

Upstream PR to bring these improvements to llama-cpp-python: abetlen/llama-cpp-python#1884