blossom-v5.1-34B: benchmark results
Azure99's Blossom v5.1 bilingual Chinese-English chat fine-tune of Yi-1.5-34B, trained in two stages on Wizard, Orca, Math and Blossom dialogue data (May 2024). Provider: Azure99. Released 2024-05-19. Access: Open.
Aggregate Rating 1499 ± 23, rank #786 of 1656 rated models, from 25 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 78.05 | Accuracy (%) (5-shot) | 99.1 |
| Open Portuguese LLM - FaQuAD NLI | 81.88 | Macro F1 (%) | 93.3 |
| Open Chinese LLM - GSM8K | 67.85 | Accuracy (%) | 92 |
| Open Chinese LLM - C-Eval Semantic | 89.74 | Accuracy (%) | 91.4 |
| Open Chinese LLM Leaderboard | 68.68 | Official Nonzero-Task Mean (%) | 90.8 |
| Open Chinese LLM - ARC Challenge | 61.18 | Accuracy (%) | 90.1 |
| Open Portuguese LLM - BLUEX | 67.18 | Accuracy (%) | 89.6 |
| Open LLM Leaderboard v1 - WinoGrande | 83.43 | Accuracy (%) (5-shot) | 88.8 |
| Open Chinese LLM - HellaSwag | 68.61 | Accuracy (%) | 88.6 |
| Open Chinese LLM - WinoGrande | 68.51 | Accuracy (%) | 87 |
| Open Chinese LLM - CMMLU | 70.37 | Accuracy (%) | 86.4 |
| Open Portuguese LLM - OAB Exams | 54.44 | Accuracy (%) | 86.3 |
Interactive version: theaggregate.ai/model?slug=blossom-v5-1-34b · How It Works · Data refreshed daily, snapshot 2026-10-20.