MSH-Lite-7B-v1-Bielik-v2.3-Instruct-Llama-Prune: benchmark results
Provider: MedIT Solutions. Released 2024-11-12. Access: Open.
Aggregate Rating 1350 ± 49, rank #7708 of 10916 rated models, from 33 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MuSR | 42.53 | Score | 65.7 |
| Open PL LLM - PolEmo2-OUT (multiple choice, 5-shot) | 70.65 | Accuracy (%) | 60.8 |
| Open LLM Leaderboard - GPQA | 30.29 | Score | 59.3 |
| Open PL LLM - RAG | 56.87 | Average RAG Score (%) | 51.4 |
| Open PL LLM - PoQuAD Open Book (generative, 5-shot) | 62.13 | Levenshtein Similarity (%) | 50.8 |
| Open PL LLM - PolQA Open Book (generative, 5-shot) | 78.64 | Levenshtein Similarity (%) | 46.1 |
| Open PL LLM - PolQA Reranking (multiple choice, 5-shot) | 67.28 | Accuracy (%) | 45.9 |
| Open PL LLM - PSC (multiple choice, 5-shot) | 58.35 | Binary F1 (%) | 43.6 |
| Open PL LLM - PolEmo2-IN (multiple choice, 5-shot) | 74.24 | Accuracy (%) | 40.6 |
| Open PL LLM - PolQA Closed Book (generative, 5-shot) | 41.12 | Levenshtein Similarity (%) | 38.7 |
| Open PL LLM - CBD (multiple choice, 5-shot) | 19.65 | Macro F1 (%) | 38.4 |
| Open PL LLM - KLEJ NER (generative, 5-shot) | 39.36 | Accuracy (%) | 37.1 |
Interactive version: theaggregate.ai/model?slug=msh-lite-7b-v1-bielik-v2-3-instruct-llama-prune · How It Works · Data refreshed daily, snapshot 2026-11-08.