RichardErkhov/kyujinpy_-_Sakura-SOLRCA-Instruct-DPO-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Sakura-SOLRCA-Instruct-DPO - GGUF

• —Model creator: https://huggingface.co/kyujinpy/
• —Original model: https://huggingface.co/kyujinpy/Sakura-SOLRCA-Instruct-DPO/

Original model description: --- language:

• —en license: cc-by-nc-sa-4.0 datasets:
• —Intel/orcadpopairs pipeline_tag: text-generation model-index:
• —name: Sakura-SOLRCA-Instruct-DPO results:
• —task: type: text-generation name: Text Generation dataset: name: AI2 Reasoning Challenge (25-Shot) type: ai2arc config: ARC-Challenge split: test args: numfew_shot: 25 metrics:
• —type: accnorm value: 71.16 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/openllm_leaderboard?query=kyujinpy/Sakura-SOLRCA-Instruct-DPO name: Open LLM Leaderboard
• —task: type: text-generation name: Text Generation dataset: name: HellaSwag (10-Shot) type: hellaswag split: validation args: numfewshot: 10 metrics:
• —type: accnorm value: 88.49 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/openllm_leaderboard?query=kyujinpy/Sakura-SOLRCA-Instruct-DPO name: Open LLM Leaderboard
• —task: type: text-generation name: Text Generation dataset: name: MMLU (5-Shot) type: cais/mmlu config: all split: test args: numfewshot: 5 metrics:
• —type: acc value: 66.17 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/openllmleaderboard?query=kyujinpy/Sakura-SOLRCA-Instruct-DPO name: Open LLM Leaderboard
• —task: type: text-generation name: Text Generation dataset: name: TruthfulQA (0-shot) type: truthfulqa config: multiplechoice split: validation args: numfewshot: 0 metrics:
• —type: mc2 value: 72.1 source: url: https://huggingface.co/spaces/HuggingFaceH4/openllmleaderboard?query=kyujinpy/Sakura-SOLRCA-Instruct-DPO name: Open LLM Leaderboard
• —task: type: text-generation name: Text Generation dataset: name: Winogrande (5-shot) type: winogrande config: winograndexl split: validation args: numfew_shot: 5 metrics:
• —type: acc value: 82.95 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/openllmleaderboard?query=kyujinpy/Sakura-SOLRCA-Instruct-DPO name: Open LLM Leaderboard
• —task: type: text-generation name: Text Generation dataset: name: GSM8k (5-shot) type: gsm8k config: main split: test args: numfewshot: 5 metrics:
• —type: acc value: 63.46 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/openllmleaderboard?query=kyujinpy/Sakura-SOLRCA-Instruct-DPO name: Open LLM Leaderboard ---

Sakura-SOLRCA-Instruct-DPO

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

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

Model Details

Model Developers Kyujin Han (kyujinpy)

Method Using DPO method. With Intel/orca_dpo_pairs.

I shared the information about my model. (training and code) Please see: ⭐Sakura-SOLAR.

Model Benchmark

Open leaderboard

• —Follow up as link.

Implementation Code

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

repo = "kyujinpy/Sakura-SOLRCA-Instruct-DPO"
OpenOrca = AutoModelForCausalLM.from_pretrained(
        repo,
        return_dict=True,
        torch_dtype=torch.float16,
        device_map='auto'
)
OpenOrca_tokenizer = AutoTokenizer.from_pretrained(repo)

Open LLM Leaderboard Evaluation Results

Detailed results can be found here