Inference and page functions should return errors instead of swallowing them #217
Description
Is your feature request related to a problem? Please describe.
When running page actions, such as observe, when an error occurs, it gets swallowed, not allowing us to handle them properly.
inference.py#observe catches the error and returns an empty response, page.py#observe returns an ObserveResult
When we have an error that is retryable, such as an LLM rate limit error, we would like to retry, but this behaviour makes it very hard to handle.
Describe the solution you'd like
The result should return a result or an error that can be handled accordingly.
Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.
Are you willing to contribute to implementing this feature or fix?
- Yes, I can submit a PR
- Yes, but I need guidance
- No, I cannot contribute at this time