Llama3-TenyxChat-70B: benchmark results
Tenyx's DPO fine-tune of Llama 3 70B Instruct trained on UltraFeedback to improve chat quality while limiting catastrophic forgetting. Provider: Tenyx. Released 2024-04-26. Access: Open.
Aggregate Rating 1515 ± 22, rank #735 of 1656 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 86.28 | Accuracy (%) (5-shot) | 99.9 |
| Open LLM Leaderboard v1 - MMLU | 80.04 | Accuracy (%) (5-shot) | 99.6 |
| Open Chinese LLM - GSM8K | 72.71 | Accuracy (%) | 99.1 |
| Open LLM Leaderboard - IFEval | 80.87 | Score | 98.3 |
| Open Chinese LLM - CMMLU | 72.02 | Accuracy (%) | 94.1 |
| Open LLM Leaderboard - BBH | 65.11 | Score | 91.5 |
| Open LLM Leaderboard v1 - ARC Challenge | 72.1 | Normalized accuracy (%) (25-shot) | 91 |
| Open LLM Leaderboard - MMLU-Pro | 52.1 | Score | 90.5 |
| Open Chinese LLM Leaderboard | 67.86 | Official Nonzero-Task Mean (%) | 89.3 |
| Open Chinese LLM - ARC Challenge | 60.24 | Accuracy (%) | 87.7 |
| Open LLM Leaderboard v1 - WinoGrande | 82.95 | Accuracy (%) (5-shot) | 86.4 |
| Open Chinese LLM - C-Eval Semantic | 86.83 | Accuracy (%) | 82.5 |
Interactive version: theaggregate.ai/model?slug=llama3-tenyxchat-70b · How It Works · Data refreshed daily, snapshot 2026-10-20.