CoolFace
Modelpublic

kyujinpy/Sakura-SOLAR-Instruct

sourceHugging Facecc-by-nc-sa-4.0updated 9mo agoView on Hugging Face
29likes130downloads
Model Card

Sakura-SOLAR-Instruct

<img src='./sakura.png' width=512>

(주)미디어그룹사람과숲과 (주)마커의 LLM 연구 컨소시엄에서 개발된 모델입니다

Model Details

Model Developers Kyujin Han (kyujinpy)

Method Using Mergekit. I shared the information about my model. (training and code) Please see: [⭐Sakura-SOLAR](https://github.com/KyujinHan/Sakura-SOLAR-DPO).

Blog

Model Benchmark

Open leaderboard

  • Follow up as link.
ModelAverageARCHellaSwagMMLUTruthfulQAWinograndeGSM8K
Sakura-SOLRCA-Instruct-DPO74.0571.1688.4966.1772.1082.9563.46
Sakura-SOLAR-Instruct-DPO-v274.1470.9088.4166.4871.8683.4363.76
kyujinpy/Sakura-SOLAR-Instruct74.4070.9988.4266.3371.7983.6665.20
Rank1 2023.12.27 PM 11:50

Implementation Code

python
### KO-Platypus
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

repo = "kyujinpy/Sakura-SOLAR-Instruct"
sakura_solar = AutoModelForCausalLM.from_pretrained(
        repo,
        return_dict=True,
        torch_dtype=torch.float16,
        device_map='auto'
)
sakura_solar_tokenizer = AutoTokenizer.from_pretrained(repo)

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

MetricValue
Avg.74.40
AI2 Reasoning Challenge (25-Shot)70.99
HellaSwag (10-Shot)88.42
MMLU (5-Shot)66.33
TruthfulQA (0-shot)71.79
Winogrande (5-shot)83.66
GSM8k (5-shot)65.20