m.star.7B: benchmark results
m.star.7b is liminerity's Mistral-family fine-tune of liminerity/M7-7b, produced using Unsloth and Hugging Face TRL according to its release card. It remains distinct from its merge parent. Provider: liminerity. Access: Open.
Aggregate Rating 1387 ± 34, rank #6636 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 54.06 | Accuracy (%) (5-shot) | 65.7 |
| Open LLM Leaderboard v1 - WinoGrande | 78.53 | Accuracy (%) (5-shot) | 63.8 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 53.93 | MC2 (%) (0-shot) | 61.2 |
| Open LLM Leaderboard v1 - ARC Challenge | 60.15 | Normalized accuracy (%) (25-shot) | 46.6 |
| Open LLM Leaderboard v1 - HellaSwag | 80.96 | Normalized accuracy (%) (10-shot) | 43 |
| Open LLM Leaderboard v1 - MMLU | 58.28 | Accuracy (%) (5-shot) | 43 |
Interactive version: theaggregate.ai/model?slug=m-star-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.