Mahou-1.2a-llama3-8B: benchmark results
A DPO fine-tune released by flammen.ai of nbeerbower/llama-3-Stheno-Mahou-8B, 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. Provider: flammen.ai. Access: Open.
Aggregate Rating 1452 ± 30, rank #3698 of 10916 rated models, from 12 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 71.11 | Accuracy (%) (5-shot) | 95.2 |
| Open LLM Leaderboard v1 - MMLU | 68.52 | Accuracy (%) (5-shot) | 91.6 |
| Open LLM Leaderboard v1 - ARC Challenge | 68.86 | Normalized accuracy (%) (25-shot) | 81.6 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 58.76 | MC2 (%) (0-shot) | 72.6 |
| Open LLM Leaderboard - MMLU-Pro | 38.17 | Score | 66.3 |
| Open LLM Leaderboard v1 - HellaSwag | 83.97 | Normalized accuracy (%) (10-shot) | 62.8 |
| Open LLM Leaderboard - IFEval | 50.93 | Score | 60.8 |
| Open LLM Leaderboard v1 - WinoGrande | 77.58 | Accuracy (%) (5-shot) | 57 |
| Open LLM Leaderboard - BBH | 50.94 | Score | 53.3 |
| Open LLM Leaderboard - GPQA | 28.86 | Score | 43.2 |
| Open LLM Leaderboard - MATH Level 5 | 8.38 | Score | 42.1 |
| Open LLM Leaderboard - MuSR | 38.47 | Score | 32.2 |
Interactive version: theaggregate.ai/model?slug=mahou-1-2a-llama3-8b · How It Works · Data refreshed daily, snapshot 2026-11-08.