WikiHow-Mistral-Instruct-7B: benchmark results
A full fine-tune by ajibawa-2023 of Mistral-7B-Instruct-v0.2 on its WikiHow dataset for generating detailed tutorials, using ChatML prompts. Provider: ajibawa-2023. Access: Open.
Aggregate Rating 1379 ± 33, rank #6871 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 62.16 | MC2 (%) (0-shot) | 78.8 |
| Open LLM Leaderboard v1 - ARC Challenge | 60.92 | Normalized accuracy (%) (25-shot) | 50.3 |
| Open LLM Leaderboard v1 - GSM8K | 30.02 | Accuracy (%) (5-shot) | 47.1 |
| Open LLM Leaderboard v1 - MMLU | 58.57 | Accuracy (%) (5-shot) | 43.7 |
| Open LLM Leaderboard v1 - HellaSwag | 80.99 | Normalized accuracy (%) (10-shot) | 43.2 |
| Open LLM Leaderboard v1 - WinoGrande | 74.82 | Accuracy (%) (5-shot) | 38.9 |
Interactive version: theaggregate.ai/model?slug=wikihow-mistral-instruct-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.