Llama-Phi-3_DoRA: benchmark results

Provider: Meta. Access: Open.

Unified ELO 1500 ± 20, rank #1228 of 2928 rated models, from 12 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard v1 - MMLU69.44Accuracy (%) (5-shot)92.6
Open LLM Leaderboard v1 - GSM8K68.01Accuracy (%) (5-shot)85
Open LLM Leaderboard - GPQA32.63Score78.8
Open LLM Leaderboard - BBH55.15Score73.9
Open LLM Leaderboard - MMLU-Pro39.15Score71.1
Open LLM Leaderboard v1 - TruthfulQA MC254.08MC2 (%) (0-shot)61.5
Open LLM Leaderboard - IFEval51.31Score61.3
Open LLM Leaderboard v1 - ARC Challenge62.29Normalized accuracy (%) (25-shot)57.1
Open LLM Leaderboard - MATH Level 512.16Score54.6
Open LLM Leaderboard - MuSR40.69Score47.5
Open LLM Leaderboard v1 - HellaSwag79.08Normalized accuracy (%) (10-shot)36.7
Open LLM Leaderboard v1 - WinoGrande73.4Accuracy (%) (5-shot)31.9

Interactive version: theaggregate.ai/model?slug=llama-phi-3-dora · How It Works · Data refreshed daily, snapshot 2026-09-23.