QuantFactory/Chocolatine-3B-Instruct-DPO-Revised-GGUF

This is quantized version of jpacifico/Chocolatine-3B-Instruct-DPO-Revised created using llama.cpp

Original Model Card

Chocolatine-3B-Instruct-DPO-Revised

DPO fine-tuned of microsoft/Phi-3-mini-4k-instruct (3.82B params)
using the jpacifico/french-orca-dpo-pairs-revised rlhf dataset.
Chocolatine is a general model and can itself be finetuned to be specialized for specific use cases.
Window context = 4k tokens

Benchmarks

The best 3B model on OpenLLM Leaderboard (july 2024)
5th best < 30B params (average benchmarks).

MT-Bench-French

Chocolatine-3B-Instruct-DPO-Revised is outperforming GPT-3.5-Turbo on MT-Bench-French by Bofeng Huang,
used with multilingual-mt-bench

########## First turn ##########
                                           score
model                               turn        
gpt-3.5-turbo                       1     8.1375
Chocolatine-3B-Instruct-DPO-Revised 1     7.9875
Daredevil-8B                        1     7.8875
Daredevil-8B-abliterated            1     7.8375
Chocolatine-3B-Instruct-DPO-v1.0    1     7.6875
NeuralDaredevil-8B-abliterated      1     7.6250
Phi-3-mini-4k-instruct              1     7.2125
Meta-Llama-3-8B-Instruct            1     7.1625
vigostral-7b-chat                   1     6.7875
Mistral-7B-Instruct-v0.3            1     6.7500
Mistral-7B-Instruct-v0.2            1     6.2875
French-Alpaca-7B-Instruct_beta      1     5.6875
vigogne-2-7b-chat                   1     5.6625
vigogne-2-7b-instruct               1     5.1375

########## Second turn ##########
                                             score
model                               turn          
Chocolatine-3B-Instruct-DPO-Revised 2     7.937500
gpt-3.5-turbo                       2     7.679167
Chocolatine-3B-Instruct-DPO-v1.0    2     7.612500
NeuralDaredevil-8B-abliterated      2     7.125000
Daredevil-8B                        2     7.087500
Daredevil-8B-abliterated            2     6.873418
Meta-Llama-3-8B-Instruct            2     6.800000
Mistral-7B-Instruct-v0.2            2     6.512500
Mistral-7B-Instruct-v0.3            2     6.500000
Phi-3-mini-4k-instruct              2     6.487500
vigostral-7b-chat                   2     6.162500
French-Alpaca-7B-Instruct_beta      2     5.487395
vigogne-2-7b-chat                   2     2.775000
vigogne-2-7b-instruct               2     2.240506

########## Average ##########
                                        score
model                                        
Chocolatine-3B-Instruct-DPO-Revised  7.962500
gpt-3.5-turbo                        7.908333
Chocolatine-3B-Instruct-DPO-v1.0     7.650000
Daredevil-8B                         7.487500
NeuralDaredevil-8B-abliterated       7.375000
Daredevil-8B-abliterated             7.358491
Meta-Llama-3-8B-Instruct             6.981250
Phi-3-mini-4k-instruct               6.850000
Mistral-7B-Instruct-v0.3             6.625000
vigostral-7b-chat                    6.475000
Mistral-7B-Instruct-v0.2             6.400000
French-Alpaca-7B-Instruct_beta       5.587866
vigogne-2-7b-chat                    4.218750
vigogne-2-7b-instruct                3.698113

Usage

You can run this model using my Colab notebook

You can also run Chocolatine using the following code:

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

The Chocolatine model is a quick demonstration that a base model can be easily fine-tuned to achieve compelling performance.
It does not have any moderation mechanism.

• Developed by: Jonathan Pacifico, 2024
• Model type: LLM
• Language(s) (NLP): French, English
• License: MIT