CodeNinja-1.0-OpenChat-7B: benchmark results
beowolx's CodeNinja 7B, an SFT fine-tune of OpenChat 3.5 on over 400,000 coding instructions from Glaive and TokenBender datasets. Provider: beowolx. Released 2023-12-20. Access: Open.
Aggregate Rating 1394 ± 15, rank #1248 of 1656 rated models, from 36 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 66.57 | Accuracy (%) (5-shot) | 82 |
| Open Portuguese LLM - FaQuAD NLI | 77.75 | Macro F1 (%) | 80.7 |
| Open Korean LLM - Ko-IFEval | 39.15 | Strict accuracy, prompt/instruction mean (%) | 80.6 |
| Open LLM Leaderboard v1 - WinoGrande | 79.79 | Accuracy (%) (5-shot) | 71 |
| Open Portuguese LLM - ASSIN2 RTE | 91.54 | Macro F1 (%) | 67.9 |
| Open LLM Leaderboard - IFEval | 54.47 | Score | 65.6 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 31.16 | Exact match, flexible extract (%) | 65.5 |
| Open Korean LLM - Ko-EQ-Bench | 18.99 | EQ-Bench score | 65.3 |
| Open LLM Leaderboard - MuSR | 42.43 | Score | 64.7 |
| Open Ko-LLM Leaderboard | 37.77 | Average (%) | 64 |
| Open Korean LLM Leaderboard | 37.77 | Average Score (%) | 64 |
| Open LLM Leaderboard v1 - MMLU | 63.77 | Accuracy (%) (5-shot) | 63.1 |
Interactive version: theaggregate.ai/model?slug=codeninja-1-0-openchat-7b · How It Works · Data refreshed daily, snapshot 2026-10-20.