M7-7B: benchmark results
liminerity's M7 7B, a merge of their merge4 and merge2 models built with mergekit. Provider: liminerity. Released 2024-03-07. Access: Open.
Aggregate Rating 1468 ± 24, rank #885 of 1635 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - HellaSwag | 89.15 | Normalized accuracy (%) (10-shot) | 97.9 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 77.93 | MC2 (%) (0-shot) | 97.7 |
| Open LLM Leaderboard v1 - GSM8K | 71.72 | Accuracy (%) (5-shot) | 96.5 |
| Open Chinese LLM - TruthfulQA MC | 63.84 | Accuracy (%) | 95.1 |
| Open LLM Leaderboard v1 - WinoGrande | 84.77 | Accuracy (%) (5-shot) | 95 |
| Open LLM Leaderboard v1 - ARC Challenge | 72.87 | Normalized accuracy (%) (25-shot) | 94 |
| Open Portuguese LLM - FaQuAD NLI | 77.86 | Macro F1 (%) | 81.7 |
| Open Portuguese LLM - ASSIN2 RTE | 92.2 | Macro F1 (%) | 75.9 |
| Open LLM Leaderboard v1 - MMLU | 64.5 | Accuracy (%) (5-shot) | 71.5 |
| Open Chinese LLM - WinoGrande | 63.77 | Accuracy (%) | 60.7 |
| Open Portuguese LLM - BLUEX | 54.52 | Accuracy (%) | 60.4 |
| Open Chinese LLM - ARC Challenge | 53.58 | Accuracy (%) | 57.7 |
Interactive version: theaggregate.ai/model?slug=m7-7b · How It Works · Data refreshed daily, snapshot 2026-11-05.