Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp: benchmark results
invalid-coder's 10.7B model, a LazyMergekit SLERP merge of kyujinpy's Sakura-SOLAR-Instruct and jeonsworld's CarbonVillain-en-10.7B-v2. Provider: invalid-coder. Released 2024-02-01. Access: Open.
Aggregate Rating 1437 ± 14, rank #1062 of 1656 rated models, from 46 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 71.94 | MC2 (%) (0-shot) | 92.7 |
| Open LLM Leaderboard v1 - HellaSwag | 88.42 | Normalized accuracy (%) (10-shot) | 91.7 |
| Open Korean LLM - KorNAT-Harmless | 67.79 | Normalized accuracy (%) | 90.6 |
| Open LLM Leaderboard v1 - WinoGrande | 83.43 | Accuracy (%) (5-shot) | 88.8 |
| Open LLM Leaderboard v1 - ARC Challenge | 71.25 | Normalized accuracy (%) (25-shot) | 88.6 |
| Open Chinese LLM - TruthfulQA MC | 61.13 | Accuracy (%) | 86.7 |
| Open LLM Leaderboard v1 - MMLU | 66.31 | Accuracy (%) (5-shot) | 86.7 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 52.69 | Exact match, flexible extract (%) | 86.1 |
| Open Portuguese LLM - ASSIN2 RTE | 92.93 | Macro F1 (%) | 85 |
| Open Korean LLM - KorNAT-Helpful | 49.78 | Normalized accuracy (%) | 81.7 |
| Open LLM Leaderboard v1 - GSM8K | 65.35 | Accuracy (%) (5-shot) | 79.9 |
| Open Korean LLM - Ko-EQ-Bench | 29.88 | EQ-Bench score | 78.3 |
Interactive version: theaggregate.ai/model?slug=sakura-solar-instruct-carbonvillain-en-10-7b-v2-slerp · How It Works · Data refreshed daily, snapshot 2026-10-20.