Closed jonathonbarton closed 1 week ago
Closing this as the default OpenRouter model has been switched to a 8k token model
Art - I submitted the PR even after seeing that Mantella is getting the Context Window from the API from OR. It's still useful for users who are running a local LLM...
Even in the current version this is still an issue that this PR addresses.
Updated wordsmithing for logging - generally overall, but specifically to address all the frequently asked questions in #issues on the Discord regarding the 4096 token low token count "warning" that is not actually any kind of error, but is just educational and requires no user action.