File size: 760 Bytes
3712700 3797ce7 3712700 3797ce7 3712700 3797ce7 3712700 3797ce7 3712700 3797ce7 3712700 3797ce7 3712700 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 | # Verifier RM data — conversation format
This revision contains the verifier reward-model examples formatted for language-model conversation training.
Each row has one column named `messages` with two turns:
- user: instruction, current completion, candidate text, and optional evidence
- assistant: the original `target_label` (`YES` or `NO`)
The user content uses this order: `<instruction>`, `<current_completion>`, and `<text_to_score>`. Non-empty `reference_solution` and `privileged_context` values are appended independently. The requested `privelaged_context` spelling is retained in the tag.
Split counts are train 48,607, validation 2,697, and test 2,651.
Source revision used for this conversion: `556b271958cd5c6afb2093b5333b64dba5c42d6a`.
|