-
Notifications
You must be signed in to change notification settings - Fork 1.3k
[Feature]: Provide a config array to configure fallback retry modells when using openrouter #2867
Copy link
Copy link
Open
Labels
type: featureBrand new functionality, features, pages, workflows, endpoints, etc.Brand new functionality, features, pages, workflows, endpoints, etc.
Description
1. Why Do You NEED This Feature?
when using models like:
- qwen/qwen3.6-plus:free
- nvidia/nemotron-3-super-120b-a12b:free
- qwen/qwen3-coder:free
it would be nice to not only be able to configure retry mechanisms to cope with times of huge usages
2. What Is NOT Possible Right Now?
currently:
[retry]
# Backoff multiplication factor for each retry attempt
backoff_factor = 2
# Initial backoff delay in milliseconds for retry operations
initial_backoff_ms = 200
# Maximum number of retry attempts
max_attempts = 8
# Minimum delay in milliseconds between retry attempts
min_delay_ms = 1000
# HTTP status codes that should trigger retries
status_codes = [429, 500, 502, 503, 504, 408, 522, 520, 529]
# Whether to suppress retry error logging and events
suppress_errors = falsedesired additionally:
fallback_model_after_max_attempts = true
fallback_models = [ qwen/qwen3.6-plus:free, qwen/qwen3-coder:free ]3. What WILL Be Possible With This Feature?
Allowing the model to be switched easily during an unsupervised run when developer is afk for a longer period of time and somehow the model failed!
Proposed Solution (User Experience)
- improvement for user experience with configurable automatic handling
- unsupervised working without potential time loss!
Alternatives Considered
No response
Feature Category
Configuration
Priority/Impact
High - Would significantly improve my workflow
Examples from Other Tools
No response
Additional Context
No response
Pre-submission Checklist
- I have searched existing issues and confirmed this is not a duplicate
- I am willing to submit a PR to implement this feature
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
type: featureBrand new functionality, features, pages, workflows, endpoints, etc.Brand new functionality, features, pages, workflows, endpoints, etc.