Mahou-1.3-llama3-8B: benchmark results
A ORPO fine-tune released by flammen.ai of nbeerbower/llama-3-Daredevil-Mahou-8B, using flammenai/MahouMix-v1. The release targets casual conversation and character roleplay. Its card specifies ChatML prompting. Provider: flammen.ai. Access: Open.
Aggregate Rating 1455 ± 34, rank #3522 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 72.93 | Accuracy (%) (5-shot) | 98.3 |
| Open LLM Leaderboard v1 - MMLU | 68.68 | Accuracy (%) (5-shot) | 91.8 |
| Open LLM Leaderboard v1 - ARC Challenge | 69.2 | Normalized accuracy (%) (25-shot) | 82.5 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 61.05 | MC2 (%) (0-shot) | 77 |
| Open LLM Leaderboard v1 - HellaSwag | 84.85 | Normalized accuracy (%) (10-shot) | 69.3 |
| Open LLM Leaderboard v1 - WinoGrande | 78.85 | Accuracy (%) (5-shot) | 65.9 |
Interactive version: theaggregate.ai/model?slug=mahou-1-3-llama3-8b · How It Works · Data refreshed daily, snapshot 2026-11-08.