41 lines
800 B
Markdown
41 lines
800 B
Markdown
---
|
|
language:
|
|
- en
|
|
pipeline_tag: text-generation
|
|
license: cc-by-nc-4.0
|
|
---
|
|
|
|
# AISquare-Instruct-llama2-koen-13b-v0.9.24
|
|
|
|
## Model Details
|
|
|
|
**Developed by**
|
|
[Inswave Systems](https://www.inswave.com) UI Platform Team
|
|
|
|
**Method**
|
|
Using DPO method and SFT method
|
|
|
|
**Hardware**
|
|
We utilized an A100x4 * 1 for training our model
|
|
|
|
**Base Model**
|
|
[beomi/llama2-koen-13b](https://huggingface.co/beomi/llama-2-koen-13b)
|
|
|
|
|
|
# Implementation Code
|
|
```python
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
import torch
|
|
|
|
repo = "inswave/AISquare-Instruct-llama2-koen-13b-v0.9.24"
|
|
model = AutoModelForCausalLM.from_pretrained(
|
|
repo,
|
|
return_dict=True,
|
|
torch_dtype=torch.float16,
|
|
device_map='auto'
|
|
)
|
|
tokenizer = AutoTokenizer.from_pretrained(repo)
|
|
```
|
|
|
|
---
|