Mahou-1.2a-mistral-7B: benchmark results
A DPO fine-tune released by flammen.ai of flammenai/flammen27-mistral-7B, using flammenai/FlameMix-DPO-v1, flammenai/Grill-preprod-v1_chatML, flammenai/Grill-preprod-v2_chatML. The release targets casual conversation and character roleplay. Its card specifies ChatML prompting. Two evaluator revisions share this card; they are not assumed to have identical weights. Provider: flammen.ai. Access: Open.
Aggregate Rating 1437 ± 31, rank #4301 of 10916 rated models, from 12 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - HellaSwag | 88.14 | Normalized accuracy (%) (10-shot) | 89.7 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 68.71 | MC2 (%) (0-shot) | 87.9 |
| Open LLM Leaderboard v1 - ARC Challenge | 68.77 | Normalized accuracy (%) (25-shot) | 81.3 |
| Open LLM Leaderboard v1 - GSM8K | 63.61 | Accuracy (%) (5-shot) | 77.2 |
| Open LLM Leaderboard v1 - MMLU | 64.73 | Accuracy (%) (5-shot) | 75.6 |
| Open LLM Leaderboard v1 - WinoGrande | 80.11 | Accuracy (%) (5-shot) | 72.4 |
| Open LLM Leaderboard - BBH | 51.18 | Score | 54.7 |
| Open LLM Leaderboard - IFEval | 45.52 | Score | 50.5 |
| Open LLM Leaderboard - MMLU-Pro | 31.63 | Score | 42.6 |
| Open LLM Leaderboard - MATH Level 5 | 6.87 | Score | 36.9 |
| Open LLM Leaderboard - MuSR | 38.96 | Score | 34.9 |
| Open LLM Leaderboard - GPQA | 27.18 | Score | 26.6 |
Interactive version: theaggregate.ai/model?slug=mahou-1-2a-mistral-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.