6 ms·
The benchmark results seem unrealistically good, but I'm not sure from which angles I should challenge them.
by deyiao 2y ago
The benchmark results seem unrealistically good, but I'm not sure from which angles I should challenge them.
- ai-christianson 2y agoI think they're real. The model is performing better than claude-3-5-sonnet-20241022 on the claude leaderboard: https://aider.chat/docs/leaderboards/ https://aider.chat/docs/leaderboards/