Rate limitation error message

[danbeibei]

Hi,
I rencently got an error when calling the github_copilot/claude-3.7-sonnet-thought model.

$ llm --model github_copilot/claude-3.7-sonnet-thought "Hi"
Error: Attempted to access streaming response content, without having called `read()`.

The GitHub Copilot VS Code extension indicates that I'm actually being rate-limited.
Could the LLM error message be updated to indicate this as well?

[cavanaug]

This issue was likely fixed with the changes I made for streaming awhile back. Try 0.3 and see if you encounter the same issue.