PlatYi-34B-LoRA: benchmark results
Rank-16 LoRA fine-tune of original 01-ai Yi-34B on Open-Platypus; distinct from rank-64 Llama-format variants. Provider: kyujinpy. Access: Open.
Aggregate Rating 1493 ± 40, rank #2476 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 78.46 | Accuracy (%) (5-shot) | 99.3 |
| Open LLM Leaderboard v1 - WinoGrande | 83.66 | Accuracy (%) (5-shot) | 90.1 |
| Open LLM Leaderboard v1 - ARC Challenge | 67.15 | Normalized accuracy (%) (25-shot) | 75.9 |
| Open LLM Leaderboard v1 - HellaSwag | 85.37 | Normalized accuracy (%) (10-shot) | 74 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 53.32 | MC2 (%) (0-shot) | 59.2 |
| Open LLM Leaderboard v1 - GSM8K | 40.64 | Accuracy (%) (5-shot) | 55.9 |
Interactive version: theaggregate.ai/model?slug=platyi-34b-lora · How It Works · Data refreshed daily, snapshot 2026-11-08.