llama-3-youko-70B-instruct: benchmark results
Rinna Youko 70B instruction release using supervised fine-tuning and a 0.5-scaled Llama 3 chat-vector addition that skips embeddings. It adopts the Llama 3 chat format. Provider: rinna. Released 2024-07-25. Access: Open.
Aggregate Rating 1552 ± 27, rank #1681 of 10916 rated models, from 58 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM - Wikicorpus J TO E Comet Wmt22 | 77.93 | Score (%) | 99.3 |
| Open Japanese LLM - Wikicorpus J TO E Bert Score EN F1 | 91.63 | Score (%) | 96.6 |
| Open Japanese LLM - MT | 85.75 | Score (%) | 96.4 |
| Open Japanese LLM - ALT J TO E Comet Wmt22 | 89.05 | Score (%) | 94.7 |
| Open Japanese LLM - Wiki NER SET F1 | 14.16 | Score (%) | 93.1 |
| Open Japanese LLM - Wikicorpus E TO J Comet Wmt22 | 85.05 | Score (%) | 92.8 |
| Open Japanese LLM - Wikicorpus J TO E Bleu EN | 12.55 | Score (%) | 92.4 |
| Open Japanese LLM - Jemhopqa Char F1 | 63.48 | Score (%) | 92.3 |
| Open Japanese LLM - ALT J TO E Bert Score EN F1 | 95.84 | Score (%) | 91.6 |
| Open Japanese LLM - ALT J TO E Bleu EN | 18.56 | Score (%) | 91.1 |
| Open Japanese LLM - Wikicorpus E TO J Bert Score JA F1 | 84.71 | Score (%) | 90.9 |
| Open Japanese LLM - EL | 58.76 | Score (%) | 90 |
Interactive version: theaggregate.ai/model?slug=llama-3-youko-70b-instruct · How It Works · Data refreshed daily, snapshot 2026-11-08.