File size: 1,278 Bytes
2f04424 164f3d6 2f04424 164f3d6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 |
---
license: cc-by-nc-4.0
language:
- ko
library_name: transformers
pipeline_tag: text-generation
---
**The license is `cc-by-nc-4.0`.**
# **GAI-LLM/OPEN-SOLAR-KO-10.7B-mixed-v15**
## Model Details
**Model Developers** Donghoon Oh, Hanmin Myung, Eunyoung Kim (SK C&C G.AI Eng)
**Input** Models input text only.
**Output** Models generate text only.
**Model Architecture**
GAI-LLM/OPEN-SOLAR-KO-10.7B-mixed-v15 is an auto-regressive language model based on the LLaMA2 transformer architecture.
**Base Model** [beomi/OPEN-SOLAR-KO-10.7B](https://huggingface.co/beomi/OPEN-SOLAR-KO-10.7B)
**Training Dataset**
- We combined Open Korean Dateset using mixed-strategy
- We use A100 GPU 80GB * 8, when training.
# **Model Benchmark**
## KO-LLM leaderboard
- Follow up as [Open KO-LLM LeaderBoard](https://huggingface.co/spaces/upstage/open-ko-llm-leaderboard).
# Implementation Code
```python
### GAI-LLM/OPEN-SOLAR-KO-10.7B-mixed-v15
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
repo = "GAI-LLM/OPEN-SOLAR-KO-10.7B-mixed-v15"
model = AutoModelForCausalLM.from_pretrained(
repo,
return_dict=True,
torch_dtype=torch.float16,
device_map='auto'
)
tokenizer = AutoTokenizer.from_pretrained(repo)
``` |