Mistral7B-PairRM-SPPO-Iter3 Benchmark Results

Mistral7B-PairRM-SPPO-Iter3 — benchmark results

Provider: Mistral. Released 2024-05-04. Access: Open.

Unified ELO 1354 ± 18, rank #1399 of 1776 rated models, from 15 benchmark results.

Benchmark	Score	Metric	Percentile
Open Chinese LLM - TruthfulQA MC	60.84	Accuracy (%)	84.9
Open Korean LLM Leaderboard	340.96	Average Score (%)	69.8
Open LLM Leaderboard - MuSR	9.49	Score	46.2
Open LLM Leaderboard - IFEval	43.51	Score	46
Open LLM Leaderboard - BBH	21.82	Score	32.5
Open Chinese LLM - HellaSwag	56.65	Accuracy (%)	30.3
Open LLM Leaderboard - GPQA	3.36	Score	30.2
Open Chinese LLM - ARC Challenge	47.44	Accuracy (%)	28.6
Open LLM Leaderboard - MMLU-Pro	18.42	Score	28.1
Open Chinese LLM - WinoGrande	60.85	Accuracy (%)	20.3
Open Chinese LLM - GSM8K	24.11	Accuracy (%)	17.8
Open Chinese LLM Leaderboard	49.34	Average Score (%)	17.5