Dolly_Shygmalion-6b / README.md
leaderboard-pr-bot's picture
Adding Evaluation Results
3f8d305
|
raw
history blame
813 Bytes

#TODO card. Mix of (GPT-J-6B-Shinen + GPT-J-Dolly LoRA) + Pygmalion-6b

At a ratio of

GPT-J-6B-Shinen - 20%

GPT-J-Dolly LoRA - 20%

Pygmalion-6b - 60%

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 34.9
ARC (25-shot) 41.89
HellaSwag (10-shot) 68.48
MMLU (5-shot) 27.58
TruthfulQA (0-shot) 33.91
Winogrande (5-shot) 65.35
GSM8K (5-shot) 2.12
DROP (3-shot) 4.93