Llama 3 8B ProLong 64k Base Benchmark Results

Llama 3 8B ProLong 64k Base — benchmark results

Provider: Meta. Released 2024-07-22. Access: Open.

Unified ELO 1381 ± 17, rank #1309 of 1776 rated models, from 15 benchmark results.

Benchmark	Score	Metric	Percentile
Open Korean LLM Leaderboard	428.13	Average Score (%)	80.3
Open LLM Leaderboard - MuSR	14.62	Score	79.5
Open LLM Leaderboard - IFEval	52.01	Score	62
Open LLM Leaderboard - MMLU-Pro	26.09	Score	47.5
Open LLM Leaderboard - BBH	28.69	Score	47.4
Open Chinese LLM - CMMLU	53.07	Accuracy (%)	47.2
Open Chinese LLM - WinoGrande	62.27	Accuracy (%)	39.9
Open Chinese LLM - C-Eval Semantic	64.76	Accuracy (%)	38.3
Open LLM Leaderboard - MATH Level 5	6.5	Score	35.3
Open Chinese LLM - HellaSwag	56.41	Accuracy (%)	29.7
Open Chinese LLM - GSM8K	34.5	Accuracy (%)	29.1
Open Chinese LLM Leaderboard	52.39	Average Score (%)	26.3