Vigogne-13B-Instruct: A French Instruction-following LLaMA Model

Vigogne-13B-Instruct is a LLaMA-13B model fine-tuned to follow the French instructions.

For more information, please visit the Github repo: https://github.com/bofenghuang/vigogne

Usage and License Notices: Same as Stanford Alpaca, Vigogne is intended and licensed for research use only. The dataset is CC BY NC 4.0 (allowing only non-commercial use) and models trained using the dataset should not be used outside of research purposes.

Changelog

All versions are available in branches.

• V1.0: Initial release, trained on the translated Stanford Alpaca dataset.
• V1.1: Improved translation quality of the Stanford Alpaca dataset.
• V2.0: Expanded training dataset to 224k for better performance.
• V3.0: Further expanded training dataset to 262k for improved results.

Usage

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig
from vigogne.preprocess import generate_instruct_prompt

model_name_or_path = "bofenghuang/vigogne-13b-instruct"
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, padding_side="right", use_fast=False)
model = AutoModelForCausalLM.from_pretrained(model_name_or_path, torch_dtype=torch.float16, device_map="auto")

user_query = "Expliquez la différence entre DoS et phishing."
prompt = generate_instruct_prompt(user_query)
input_ids = tokenizer(prompt, return_tensors="pt")["input_ids"].to(model.device)
input_length = input_ids.shape[1]

generated_outputs = model.generate(
    input_ids=input_ids,
    generation_config=GenerationConfig(
        temperature=0.1,
        do_sample=True,
        repetition_penalty=1.0,
        max_new_tokens=512,
    ),
    return_dict_in_generate=True,
)
generated_tokens = generated_outputs.sequences[0, input_length:]
generated_text = tokenizer.decode(generated_tokens, skip_special_tokens=True)
print(generated_text)

You can also infer this model by using the following Google Colab Notebook.

Limitations

Vigogne is still under development, and there are many limitations that have to be addressed. Please note that it is possible that the model generates harmful or biased content, incorrect information or generally unhelpful answers.