README.md

---
language:
- fr
- en
license: apache-2.0
library_name: transformers
tags:
- lucie
- dpo
- llama
- math
datasets:
- jpacifico/french-orca-dpo-pairs-revised
---
### Distilucie-7B-Math-Instruct-DPO-v0.1

Post-training optimization of the model [OpenLLM-France/Lucie-7B-Instruct-v1.1](https://huggingface.co/OpenLLM-France/Lucie-7B-Instruct-v1.1)  
DPO fine-tuning using the dataset [argilla/distilabel-math-preference-dpo](https://huggingface.co/datasets/argilla/distilabel-math-preference-dpo)    
Training set to 5 full epochs  
*Lucie-7B has a context size of 32K tokens*  

### OpenLLM Leaderboard

TBD.  

### Usage

You can run the model using this [Colab notebook](https://github.com/jpacifico/Chocolatine-LLM/blob/main/Chocolatine_14B_inference_test_colab.ipynb) 

You can also run Distilucie using the following code:  

```python
import transformers
from transformers import AutoTokenizer

# Format prompt
message = [
    {"role": "system", "content": "You are a helpful assistant chatbot."},
    {"role": "user", "content": "What is a Large Language Model?"}
]
tokenizer = AutoTokenizer.from_pretrained(new_model)
prompt = tokenizer.apply_chat_template(message, add_generation_prompt=True, tokenize=False)

# Create pipeline
pipeline = transformers.pipeline(
    "text-generation",
    model=new_model,
    tokenizer=tokenizer
)

# Generate text
sequences = pipeline(
    prompt,
    do_sample=True,
    temperature=0.7,
    top_p=0.9,
    num_return_sequences=1,
    max_length=200,
)
print(sequences[0]['generated_text'])
```

### Limitations

This Distilucie model is a quick demonstration that the Lucie foundation model can be easily fine-tuned to achieve compelling performance.  
It does not have any moderation mechanism.  

- **Developed by:** Jonathan Pacifico, 2025  
- **Model type:** LLM 
- **Language(s) (NLP):** French, English  
- **License:** Apache-2.0