Faro-Yi-34B: benchmark results
wenbopan's Faro-Yi-34B, an instruction-tuned version of Yi-34B-200K trained on the Fusang-V1 dataset with a 200K-token context window. Provider: wenbopan. Released 2024-04-01. Access: Open.
Aggregate Rating 1487 ± 23, rank #818 of 1635 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 76.6 | Accuracy (%) (5-shot) | 97.7 |
| Open Chinese LLM - WinoGrande | 70.01 | Accuracy (%) | 91.6 |
| Open Chinese LLM - C-Eval Semantic | 89.42 | Accuracy (%) | 90.8 |
| Open LLM Leaderboard v1 - WinoGrande | 83.66 | Accuracy (%) (5-shot) | 90.1 |
| Open Chinese LLM - ARC Challenge | 60.49 | Accuracy (%) | 88.9 |
| Open Portuguese LLM - OAB Exams | 54.53 | Accuracy (%) | 86.6 |
| Open Chinese LLM - HellaSwag | 67.99 | Accuracy (%) | 85.9 |
| Open Portuguese LLM - BLUEX | 64.81 | Accuracy (%) | 85.7 |
| Open Portuguese LLM - ENEM | 73.2 | Accuracy (%) | 84.5 |
| Open Chinese LLM - CMMLU | 69.97 | Accuracy (%) | 84 |
| Open Chinese LLM Leaderboard | 65.19 | Official Nonzero-Task Mean (%) | 81.1 |
| Open LLM Leaderboard v1 - GSM8K | 65.28 | Accuracy (%) (5-shot) | 79.7 |
Interactive version: theaggregate.ai/model?slug=faro-yi-34b · How It Works · Data refreshed daily, snapshot 2026-11-05.