MSH-v1-Bielik-v2.3-Instruct-MedIT-merge: benchmark results
MedIT Solutions' Polish-language merge of SpeakLeash's Bielik 11B v2.3 Instruct with Bielik 11B v2 using its proprietary checkpoint-merging method. Provider: MedIT Solutions. Released 2024-10-29. Access: Open.
Aggregate Rating 1555 ± 49, rank #1624 of 10916 rated models, from 33 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open PL LLM - PolQA Open Book (generative, 5-shot) | 92.65 | Levenshtein Similarity (%) | 100 |
| Open PL LLM - PolQA Closed Book (generative, 5-shot) | 85.77 | Levenshtein Similarity (%) | 99.4 |
| Open PL LLM - RAG | 75.62 | Average RAG Score (%) | 96.9 |
| Open PL LLM - Multiple Choice | 63.39 | Average Multiple-Choice Score (%) | 95.8 |
| Open PL LLM Leaderboard | 65.8 | Average Score (%) | 94.8 |
| Open PL LLM - 8TAGS (multiple choice, 5-shot) | 79.53 | Accuracy (%) | 94 |
| Open PL LLM - CBD (multiple choice, 5-shot) | 38.79 | Macro F1 (%) | 93.8 |
| Open PL LLM - PSC (multiple choice, 5-shot) | 91.88 | Binary F1 (%) | 93.4 |
| Open PL LLM - PolEmo2-OUT (multiple choice, 5-shot) | 79.35 | Accuracy (%) | 93.1 |
| Open PL LLM - PolQA Reranking (multiple choice, 5-shot) | 83.34 | Accuracy (%) | 93.1 |
| Open PL LLM - PPC (multiple choice, 5-shot) | 79.9 | Accuracy (%) | 92.6 |
| Open PL LLM - PolEmo2-IN (multiple choice, 5-shot) | 86.84 | Accuracy (%) | 92.4 |
Interactive version: theaggregate.ai/model?slug=msh-v1-bielik-v2-3-instruct-medit-merge · How It Works · Data refreshed daily, snapshot 2026-11-08.