Simplifying imports for llm_client and cross_encoder

[t41372]

Why?

• It didn’t make sense to expose only OpenAIClient while omitting OpenAIGenericClient, AnthropicClient, GeminiClient, and GroqClient. This led to confusion for users.
• A similar issue existed in cross_encoder, so I addressed that as well.
• I noticed related confusion mentioned in #331.

What’s fixed:

• Added the missing clients to __init__.py to ensure they’re properly exposed.

Note:

I didn’t open a separate issue to discuss this because I ran into the problem myself while using Graphiti, and it was quicker to submit a PR. Please feel free to close this if there are other considerations I missed.

---

[!IMPORTANT] Added missing client imports to __init__.py in llm_client and cross_encoder to ensure proper exposure.

• Imports:
  • Added AnthropicClient, GeminiClient, GroqClient, and OpenAIGenericClient to __init__.py in llm_client.
  • Added BGERerankerClient to __init__.py in cross_encoder.
• Exports:
  • Updated __all__ in llm_client/__init__.py to include all clients.
  • Updated __all__ in cross_encoder/__init__.py to include BGERerankerClient.

^{This description was created by for 003bae5fe89a421e46674b6769210327a95df796. You can customize this summary. It will automatically update as commits are pushed.}

[danielchalef]

All contributors have signed the CLA ✍️ ✅
_{Posted by the CLA Assistant Lite bot.}

[t41372]

I have read the CLA Document and I hereby sign the CLA

[danielchalef]

This has been superseded by #675