Motivation
Current multi-turn implementation is limited to one copy of dataset. Doesn't scale beyond concurrency=number of conversations.
Proposed Solution
Allow n_samples_to_issue > number of conversations.
Alternatives Considered
No response
Additional Context
No response
Motivation
Current multi-turn implementation is limited to one copy of dataset. Doesn't scale beyond concurrency=number of conversations.
Proposed Solution
Allow n_samples_to_issue > number of conversations.
Alternatives Considered
No response
Additional Context
No response