Code-Mistral-7B [4bit] [rev: ed3b9ad58391]: benchmark results
A Mistral fine-tune by ajibawa-2023 using QLoRA and coding, Orca math and OpenHermes datasets, with ChatML prompts. Training method, stored weight dtype and evaluated runtime quantization are separate properties. Provider: ajibawa-2023. Access: Open.
Aggregate Rating 1421 ± 34, rank #5122 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 66.72 | Accuracy (%) (5-shot) | 82.3 |
| Open LLM Leaderboard v1 - WinoGrande | 81.22 | Accuracy (%) (5-shot) | 77.8 |
| Open LLM Leaderboard v1 - ARC Challenge | 63.57 | Normalized accuracy (%) (25-shot) | 63.3 |
| Open LLM Leaderboard v1 - HellaSwag | 83.71 | Normalized accuracy (%) (10-shot) | 60.9 |
| Open LLM Leaderboard v1 - MMLU | 63.38 | Accuracy (%) (5-shot) | 60.3 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 51.81 | MC2 (%) (0-shot) | 53.3 |
Interactive version: theaggregate.ai/model?slug=code-mistral-7b-4bit-rev-ed3b9ad58391 · How It Works · Data refreshed daily, snapshot 2026-11-08.