File size: 3,753 Bytes
3144609 6e2b3c7 cce246a 3144609 6e2b3c7 3142729 3144609 6e2b3c7 3144609 6e2b3c7 3144609 3bcf11d 3144609 6e2b3c7 3142729 3144609 6e2b3c7 ccb6096 6e2b3c7 3144609 6e2b3c7 3144609 fb4a32b 6e2b3c7 4b9cecf 7faa21d 3144609 6e2b3c7 3144609 6e2b3c7 3144609 6e2b3c7 3144609 6e2b3c7 3144609 6e2b3c7 3144609 6e2b3c7 3144609 6e2b3c7 3144609 |
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 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 |
---
tags:
- text-generation
license: cc-by-nc-sa-4.0
language:
- ko
base_model: hyeogi/SOLAR-10.7B-dpo-v0.1
pipeline_tag: text-generation
datasets:
- jojo0217/korean_rlhf_dataset
---
# **DataVortexS-10.7B-v0.3**
<img src="./DataVortex.png" alt="DataVortex" style="height: 8em;">
## Our Team
| Research & Engineering | Product Management |
| :--------------------: | :----------------: |
| Kwangseok Yang | Seunghyun Choi |
| Jeongwon Choi | Hyoseok Choi |
## **Model Details**
### **Base Model**
[hyeogi/SOLAR-10.7B-dpo-v0.1](https://huggingface.co/hyeogi/SOLAR-10.7B-dpo-v0.1)
### **Trained On**
- **OS**: Ubuntu 20.04
- **GPU**: H100 80GB 1ea
- **transformers**: v4.36.2
### **Dataset**
- [jojo0217/korean_rlhf_dataset](https://huggingface.co/datasets/jojo0217/korean_rlhf_dataset)
### **Instruction format**
It follows **Alpaca** format.
E.g.
```python
text = """\
당신은 사람들이 정보를 찾을 수 있도록 도와주는 인공지능 비서입니다.
### Instruction:
대한민국의 수도는 어디야?
### Response:
대한민국의 수도는 서울입니다.
### Instruction:
서울 인구는 총 몇 명이야?
"""
```
## **Model Benchmark**
### **[Ko LM Eval Harness](https://github.com/Beomi/ko-lm-evaluation-harness)**
| Task | 0-shot | 5-shot | 10-shot | 50-shot |
| :--------------- | -------------: | -------------: | ------------: | -------------: |
| kobest_boolq | 0.606754 | 0.553485 | 0.583201 | 0.587602 |
| kobest_copa | 0.603643 | 0.625567 | 0.618533 | 0.627404 |
| kobest_hellaswag | 0.360793 | 0.366002 | 0.37105 | 0.357439 |
| kobest_sentineg | 0.652929 | 0.751097 | 0.742426 | 0.760152 |
| **Average** | **0.55602975** | **0.57403775** | **0.5788025** | **0.58314925** |
### **[Ko-LLM-Leaderboard](https://huggingface.co/spaces/upstage/open-ko-llm-leaderboard)**
| Average | Ko-ARC | Ko-HellaSwag | Ko-MMLU | Ko-TruthfulQA | Ko-CommonGen V2 |
| ------: | -----: | -----------: | ------: | ------------: | --------------: |
| 37.57 | 33.87 | 42.47 | 28.21 | 46.09 | 37.19 |
## **Implementation Code**
This model contains the chat_template instruction format.
You can use the code below.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto
model = AutoModelForCausalLM.from_pretrained("Edentns/DataVortexS-10.7B-v0.3")
tokenizer = AutoTokenizer.from_pretrained("Edentns/DataVortexS-10.7B-v0.3")
messages = [
{"role": "system", "content": "당신은 사람들이 정보를 찾을 수 있도록 도와주는 인공지능 비서입니다."},
{"role": "user", "content": "대한민국의 수도는 어디야?"},
{"role": "assistant", "content": "대한민국의 수도는 서울입니다."},
{"role": "user", "content": "서울 인구는 총 몇 명이야?"}
]
encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")
model_inputs = encodeds.to(device)
model.to(device)
generated_ids = model.generate(model_inputs, max_new_tokens=1000, do_sample=True)
decoded = tokenizer.batch_decode(generated_ids)
print(decoded[0])
```
## **License**
The model is licensed under the [cc-by-nc-sa-4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/) license, which allows others to copy, modify, and share the work non-commercially, as long as they give appropriate credit and distribute any derivative works under the same license.
<div align="center">
<a href="https://edentns.com/">
<img src="./Logo.png" alt="Logo" style="height: 3em;">
</a>
</div>
|