PracticeLLM/Custom-KoLLM-13B-v1

⭐My custom LLM 13B⭐

Model Details

Model Developers

• Kyujin Han (kyujinpy)

Model Architecture

• My custom LLM 13B is an auto-regressive language model based on the LLaMA2 transformer architecture.

Base Model

• beomi/llama-2-koen-13b

Training Dataset

• kyujinpy/KOR-OpenOrca-Platypus-v3.

---

Model comparisons1

Ko-LLM leaderboard(11/23; link)

Model	Average	Ko-ARC	Ko-HellaSwag	Ko-MMLU	Ko-TruthfulQA	Ko-CommonGen V2
⭐My custom LLM 13B⭐	50.19	45.99	56.93	41.78	41.66	64.58

---

Model comparisons2

AI-Harness evaluation; link

Model	Copa	Copa	HellaSwag	HellaSwag	BoolQ	BoolQ	Sentineg	Sentineg
	0-shot	5-shot	0-shot	5-shot	0-shot	5-shot	0-shot	5-shot
⭐My custom LLM 13B⭐	0.7987	0.8269	0.4994	0.5660	0.3343	0.5060	0.6984	0.9723
beomi/llama-2-koen-13b	0.7768	0.8128	0.4999	0.5127	0.3988	0.7038	0.5870	0.9748

---

Implementation Code

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

repo = "PracticeLLM/Custom-KoLLM-13B-v1"
OpenOrca = AutoModelForCausalLM.from_pretrained(
        repo,
        return_dict=True,
        torch_dtype=torch.float16,
        device_map='auto'
)
OpenOrca_tokenizer = AutoTokenizer.from_pretrained(repo)

---