ChimeraLlama-3-8B-v3: benchmark results
mlabonne's DARE TIES merge of seven Llama 3 8B fine-tunes (Instruct, Dolphin 2.9, OrpoLlama and others) on the Meta-Llama-3-8B base. Provider: mlabonne. Released 2024-06-01. Access: Open.
Aggregate Rating 1439 ± 19, rank #1055 of 1656 rated models, from 37 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 71.87 | Accuracy (%) (5-shot) | 96.8 |
| Open LLM Leaderboard v1 - MMLU | 68.5 | Accuracy (%) (5-shot) | 91.5 |
| Open Medical LLM - MMLU Medical Genetics | 83 | Accuracy (%) | 84.1 |
| Open Medical LLM - PubMedQA | 76.4 | Accuracy (%) | 83.9 |
| Open Medical LLM - MedQA (USMLE) | 61.27 | Accuracy (%) | 82.8 |
| Open Medical LLM - MedMCQA | 57.35 | Accuracy (%) | 80.2 |
| Open Medical LLM - MMLU Professional Medicine | 74.63 | Accuracy (%) | 79.2 |
| Open Portuguese LLM - OAB Exams | 50.57 | Accuracy (%) | 77.6 |
| Open Portuguese LLM - BLUEX | 59.11 | Accuracy (%) | 75.8 |
| Open Portuguese LLM - ENEM | 69.84 | Accuracy (%) | 74.2 |
| Open Medical-LLM Leaderboard | 70.05 | Medical average (%) | 73.9 |
| Open Medical LLM | 70.05 | Average Accuracy (%) | 73.4 |
Interactive version: theaggregate.ai/model?slug=chimerallama-3-8b-v3 · How It Works · Data refreshed daily, snapshot 2026-10-20.