Bedrock and Azure support

[augustintsang]

Hi everyone,

I had a few questions and proposals regarding provider support and client configuration in the gollm implementation:

Amazon Bedrock support – Is anyone currently working on an implementation for Bedrock as a provider (similar to openai, ollama, etc)? If not, I’d be happy to take the lead on contributing this.

Azure OpenAI support – Is the streaming functionality being worked on currently?

Making model configuration less hardcoded – Currently, the model is selected using a mix of env vars and local getOpenAIModel() logic. Would it be acceptable to pass more of these LLM config parameters (api_key, base_url, model) through ClientOptions when instantiating a client, and override env-based defaults there?

Let me know if there's already work in progress here or if you'd like me to open an issue/PR!

Thanks,
Augustin