TheBloke's picture
Upload README.md
085a890
metadata
base_model: bofenghuang/vigostral-7b-chat
inference: false
language: fr
license: apache-2.0
model_creator: bofeng huang
model_name: Vigostral 7B Chat
model_type: mistral
pipeline_tag: text-generation
prompt_template: >
  <s>[INST] <<SYS>>

  Vous êtes Vigogne, un assistant IA créé par Zaion Lab. Vous suivez extrêmement
  bien les instructions. Aidez autant que vous le pouvez.

  <</SYS>>


  {prompt} [/INST] 
quantized_by: TheBloke
tags:
  - LLM
  - finetuned
TheBlokeAI

TheBloke's LLM work is generously supported by a grant from andreessen horowitz (a16z)


Vigostral 7B Chat - GPTQ

Description

This repo contains GPTQ model files for bofeng huang's Vigostral 7B Chat.

Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them.

Repositories available

Prompt template: Vigogne-Llama-2-Chat

<s>[INST] <<SYS>>
Vous êtes Vigogne, un assistant IA créé par Zaion Lab. Vous suivez extrêmement bien les instructions. Aidez autant que vous le pouvez.
<</SYS>>

{prompt} [/INST] 

Known compatible clients / servers

These GPTQ models are known to work in the following inference servers/webuis.

This may not be a complete list; if you know of others, please let me know!

Provided files, and GPTQ parameters

Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements.

Each separate quant is in a different branch. See below for instructions on fetching from different branches.

Most GPTQ files are made with AutoGPTQ. Mistral models are currently made with Transformers.

Explanation of GPTQ parameters
  • Bits: The bit size of the quantised model.
  • GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value.
  • Act Order: True or False. Also known as desc_act. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now.
  • Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy.
  • GPTQ dataset: The calibration dataset used during quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ calibration dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s).
  • Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences.
  • ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama and Mistral models in 4-bit.
Branch Bits GS Act Order Damp % GPTQ Dataset Seq Len Size ExLlama Desc
main 4 128 Yes 0.1 French news 4096 4.16 GB Yes 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy.
gptq-4bit-32g-actorder_True 4 32 Yes 0.1 French news 4096 4.57 GB Yes 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage.
gptq-8bit--1g-actorder_True 8 None Yes 0.1 French news 4096 7.52 GB No 8-bit, with Act Order. No group size, to lower VRAM requirements.
gptq-8bit-128g-actorder_True 8 128 Yes 0.1 French news 4096 7.68 GB No 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy.
gptq-8bit-32g-actorder_True 8 32 Yes 0.1 French news 4096 8.17 GB No 8-bit, with group size 32g and Act Order for maximum inference quality.
gptq-4bit-64g-actorder_True 4 64 Yes 0.1 French news 4096 4.29 GB Yes 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy.

How to download, including from branches

In text-generation-webui

To download from the main branch, enter TheBloke/Vigostral-7B-Chat-GPTQ in the "Download model" box.

To download from another branch, add :branchname to the end of the download name, eg TheBloke/Vigostral-7B-Chat-GPTQ:gptq-4bit-32g-actorder_True

From the command line

I recommend using the huggingface-hub Python library:

pip3 install huggingface-hub

To download the main branch to a folder called Vigostral-7B-Chat-GPTQ:

mkdir Vigostral-7B-Chat-GPTQ
huggingface-cli download TheBloke/Vigostral-7B-Chat-GPTQ --local-dir Vigostral-7B-Chat-GPTQ --local-dir-use-symlinks False

To download from a different branch, add the --revision parameter:

mkdir Vigostral-7B-Chat-GPTQ
huggingface-cli download TheBloke/Vigostral-7B-Chat-GPTQ --revision gptq-4bit-32g-actorder_True --local-dir Vigostral-7B-Chat-GPTQ --local-dir-use-symlinks False
More advanced huggingface-cli download usage

If you remove the --local-dir-use-symlinks False parameter, the files will instead be stored in the central Hugging Face cache directory (default location on Linux is: ~/.cache/huggingface), and symlinks will be added to the specified --local-dir, pointing to their real location in the cache. This allows for interrupted downloads to be resumed, and allows you to quickly clone the repo to multiple places on disk without triggering a download again. The downside, and the reason why I don't list that as the default option, is that the files are then hidden away in a cache folder and it's harder to know where your disk space is being used, and to clear it up if/when you want to remove a download model.

The cache location can be changed with the HF_HOME environment variable, and/or the --cache-dir parameter to huggingface-cli.

For more documentation on downloading with huggingface-cli, please see: HF -> Hub Python Library -> Download files -> Download from the CLI.

To accelerate downloads on fast connections (1Gbit/s or higher), install hf_transfer:

pip3 install hf_transfer

And set environment variable HF_HUB_ENABLE_HF_TRANSFER to 1:

mkdir Vigostral-7B-Chat-GPTQ
HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/Vigostral-7B-Chat-GPTQ --local-dir Vigostral-7B-Chat-GPTQ --local-dir-use-symlinks False

Windows Command Line users: You can set the environment variable by running set HF_HUB_ENABLE_HF_TRANSFER=1 before the download command.

With git (not recommended)

To clone a specific branch with git, use a command like this:

git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/Vigostral-7B-Chat-GPTQ

Note that using Git with HF repos is strongly discouraged. It will be much slower than using huggingface-hub, and will use twice as much disk space as it has to store the model files twice (it stores every byte both in the intended target folder, and again in the .git folder as a blob.)

How to easily download and use this model in text-generation-webui

Please make sure you're using the latest version of text-generation-webui.

It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install.

  1. Click the Model tab.

  2. Under Download custom model or LoRA, enter TheBloke/Vigostral-7B-Chat-GPTQ.

    • To download from a specific branch, enter for example TheBloke/Vigostral-7B-Chat-GPTQ:gptq-4bit-32g-actorder_True
    • see Provided Files above for the list of branches for each option.
  3. Click Download.

  4. The model will start downloading. Once it's finished it will say "Done".

  5. In the top left, click the refresh icon next to Model.

  6. In the Model dropdown, choose the model you just downloaded: Vigostral-7B-Chat-GPTQ

  7. The model will automatically load, and is now ready for use!

  8. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right.

    • Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file quantize_config.json.
  9. Once you're ready, click the Text Generation tab and enter a prompt to get started!

Serving this model from Text Generation Inference (TGI)

It's recommended to use TGI version 1.1.0 or later. The official Docker container is: ghcr.io/huggingface/text-generation-inference:1.1.0

Example Docker parameters:

--model-id TheBloke/Vigostral-7B-Chat-GPTQ --port 3000 --quantize gptq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096

Example Python code for interfacing with TGI (requires huggingface-hub 0.17.0 or later):

pip3 install huggingface-hub
from huggingface_hub import InferenceClient

endpoint_url = "https://your-endpoint-url-here"

prompt = "Tell me about AI"
prompt_template=f'''<s>[INST] <<SYS>>
Vous êtes Vigogne, un assistant IA créé par Zaion Lab. Vous suivez extrêmement bien les instructions. Aidez autant que vous le pouvez.
<</SYS>>

{prompt} [/INST] 
'''

client = InferenceClient(endpoint_url)
response = client.text_generation(prompt,
                                  max_new_tokens=128,
                                  do_sample=True,
                                  temperature=0.7,
                                  top_p=0.95,
                                  top_k=40,
                                  repetition_penalty=1.1)

print(f"Model output: {response}")

How to use this GPTQ model from Python code

Install the necessary packages

Requires: Transformers 4.33.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later.

pip3 install transformers optimum
pip3 install auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/  # Use cu117 if on CUDA 11.7

If you have problems installing AutoGPTQ using the pre-built wheels, install it from source instead:

pip3 uninstall -y auto-gptq
git clone https://github.com/PanQiWei/AutoGPTQ
cd AutoGPTQ
git checkout v0.4.2
pip3 install .

You can then use the following code

from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

model_name_or_path = "TheBloke/Vigostral-7B-Chat-GPTQ"
# To use a different branch, change revision
# For example: revision="gptq-4bit-32g-actorder_True"
model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
                                             device_map="auto",
                                             trust_remote_code=False,
                                             revision="main")

tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)

prompt = "Tell me about AI"
prompt_template=f'''<s>[INST] <<SYS>>
Vous êtes Vigogne, un assistant IA créé par Zaion Lab. Vous suivez extrêmement bien les instructions. Aidez autant que vous le pouvez.
<</SYS>>

{prompt} [/INST] 
'''

print("\n\n*** Generate:")

input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()
output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512)
print(tokenizer.decode(output[0]))

# Inference can also be done using transformers' pipeline

print("*** Pipeline:")
pipe = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    max_new_tokens=512,
    do_sample=True,
    temperature=0.7,
    top_p=0.95,
    top_k=40,
    repetition_penalty=1.1
)

print(pipe(prompt_template)[0]['generated_text'])

Compatibility

The files provided are tested to work with Transformers. For non-Mistral models, AutoGPTQ can also be used directly.

ExLlama is compatible with Llama and Mistral models in 4-bit. Please see the Provided Files table above for per-file compatibility.

For a list of clients/servers, please see "Known compatible clients / servers", above.

Discord

For further support, and discussions on these models and AI in general, join us at:

TheBloke AI's Discord server

Thanks, and how to contribute

Thanks to the chirper.ai team!

Thanks to Clay from gpus.llm-utils.org!

I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.

If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.

Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.

Special thanks to: Aemon Algiz.

Patreon special mentions: Pierre Kircher, Stanislav Ovsiannikov, Michael Levine, Eugene Pentland, Andrey, 준교 김, Randy H, Fred von Graf, Artur Olbinski, Caitlyn Gatomon, terasurfer, Jeff Scroggin, James Bentley, Vadim, Gabriel Puliatti, Harry Royden McLaughlin, Sean Connelly, Dan Guido, Edmond Seymore, Alicia Loh, subjectnull, AzureBlack, Manuel Alberto Morcote, Thomas Belote, Lone Striker, Chris Smitley, Vitor Caleffi, Johann-Peter Hartmann, Clay Pascal, biorpg, Brandon Frisco, sidney chen, transmissions 11, Pedro Madruga, jinyuan sun, Ajan Kanaga, Emad Mostaque, Trenton Dambrowitz, Jonathan Leane, Iucharbius, usrbinkat, vamX, George Stoitzev, Luke Pendergrass, theTransient, Olakabola, Swaroop Kallakuri, Cap'n Zoog, Brandon Phillips, Michael Dempsey, Nikolai Manek, danny, Matthew Berman, Gabriel Tamborski, alfie_i, Raymond Fosdick, Tom X Nguyen, Raven Klaugh, LangChain4j, Magnesian, Illia Dulskyi, David Ziegler, Mano Prime, Luis Javier Navarrete Lozano, Erik Bjäreholt, 阿明, Nathan Dryer, Alex, Rainer Wilmers, zynix, TL, Joseph William Delisle, John Villwock, Nathan LeClaire, Willem Michiel, Joguhyik, GodLy, OG, Alps Aficionado, Jeffrey Morgan, ReadyPlayerEmma, Tiffany J. Kim, Sebastain Graf, Spencer Kim, Michael Davis, webtim, Talal Aujan, knownsqashed, John Detwiler, Imad Khwaja, Deo Leter, Jerry Meng, Elijah Stavena, Rooh Singh, Pieter, SuperWojo, Alexandros Triantafyllidis, Stephen Murray, Ai Maven, ya boyyy, Enrico Ros, Ken Nordquist, Deep Realms, Nicholas, Spiking Neurons AB, Elle, Will Dee, Jack West, RoA, Luke @flexchar, Viktor Bowallius, Derek Yates, Subspace Studios, jjj, Toran Billups, Asp the Wyvern, Fen Risland, Ilya, NimbleBox.ai, Chadd, Nitin Borwankar, Emre, Mandus, Leonard Tan, Kalila, K, Trailburnt, S_X, Cory Kujawski

Thank you to all my generous patrons and donaters!

And thank you again to a16z for their generous grant.

Original model card: bofeng huang's Vigostral 7B Chat

Vigostral-7B-Chat: A French chat LLM

Preview of Vigostral-7B-Chat, a new addition to the Vigogne LLMs family, fine-tuned on Mistral-7B-v0.1.

For more information, please visit the Github repository.

License: A significant portion of the training data is distilled from GPT-3.5-Turbo and GPT-4, kindly use it cautiously to avoid any violations of OpenAI's terms of use.

Prompt Template

We used a prompt template adapted from the chat format of Llama-2.

You can apply this formatting using the chat template through the apply_chat_template() method.

from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("bofenghuang/vigostral-7b-chat")

conversation = [
    {"role": "user", "content": "Bonjour ! Comment ça va aujourd'hui ?"},
    {"role": "assistant", "content": "Bonjour ! Je suis une IA, donc je n'ai pas de sentiments, mais je suis prêt à vous aider. Comment puis-je vous assister aujourd'hui ?"},
    {"role": "user", "content": "Quelle est la hauteur de la Tour Eiffel ?"},
    {"role": "assistant", "content": "La Tour Eiffel mesure environ 330 mètres de hauteur."},
    {"role": "user", "content": "Comment monter en haut ?"},
]

print(tokenizer.apply_chat_template(conversation, tokenize=False, add_generation_prompt=True))

You will get

<s>[INST] <<SYS>>
Vous êtes Vigogne, un assistant IA créé par Zaion Lab. Vous suivez extrêmement bien les instructions. Aidez autant que vous le pouvez.
<</SYS>>

Bonjour ! Comment ça va aujourd'hui ? [/INST] Bonjour ! Je suis une IA, donc je n'ai pas de sentiments, mais je suis prêt à vous aider. Comment puis-je vous assister aujourd'hui ? </s>[INST] Quelle est la hauteur de la Tour Eiffel ? [/INST] La Tour Eiffel mesure environ 330 mètres de hauteur. </s>[INST] Comment monter en haut ? [/INST]

Usage

Inference using the unquantized model with 🤗 Transformers

from typing import Dict, List, Optional
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig, TextStreamer

model_name_or_path = "bofenghuang/vigostral-7b-chat"
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")

streamer = TextStreamer(tokenizer, timeout=10.0, skip_prompt=True, skip_special_tokens=True)


def chat(
    query: str,
    history: Optional[List[Dict]] = None,
    temperature: float = 0.7,
    top_p: float = 1.0,
    top_k: float = 0,
    repetition_penalty: float = 1.1,
    max_new_tokens: int = 1024,
    **kwargs,
):
    if history is None:
        history = []

    history.append({"role": "user", "content": query})

    input_ids = tokenizer.apply_chat_template(history, return_tensors="pt").to(model.device)
    input_length = input_ids.shape[1]

    generated_outputs = model.generate(
        input_ids=input_ids,
        generation_config=GenerationConfig(
            temperature=temperature,
            do_sample=temperature > 0.0,
            top_p=top_p,
            top_k=top_k,
            repetition_penalty=repetition_penalty,
            max_new_tokens=max_new_tokens,
            pad_token_id=tokenizer.eos_token_id,
            **kwargs,
        ),
        streamer=streamer,
        return_dict_in_generate=True,
    )

    generated_tokens = generated_outputs.sequences[0, input_length:]
    generated_text = tokenizer.decode(generated_tokens, skip_special_tokens=True)

    history.append({"role": "assistant", "content": generated_text})

    return generated_text, history

# 1st round
response, history = chat("Un escargot parcourt 100 mètres en 5 heures. Quelle est sa vitesse ?", history=None)
# Pour calculer la vitesse d'un escargot qui parcourt 100 mètres en 5 heures, nous devons diviser la distance parcourue par le temps écoulé. Voici comment procédons :
# 
# Vitesse = Distance / Temps
# 
# Dans ce cas, la distance est de 100 mètres et le temps est de 5 heures. Donc, nous avons :
# 
# Vitesse = 100 mètres / 5 heures
# Vitesse = 20 mètres par heure
# 
# L'escargot parcourt donc une vitesse de 20 mètres par heure.

# 2nd round
response, history = chat("Quand il peut dépasser le lapin ?", history=history)
# Malheureusement, je ne peux pas répondre à cette question car elle contient des informations inexactes ou incorrectes. La vitesse de l'escargot est de 20 mètres par heure, tandis que la vitesse du lapin n'est pas mentionnée dans votre question initiale. Pour comparer les vitesses des deux animaux, nous aurions besoin de connaître la vitesse du lapin.

# 3rd round
response, history = chat("Écris une histoire imaginative qui met en scène une compétition de course entre un escargot et un lapin.", history=history)
# Il était une fois deux athlètes extraordinaires : un escargot nommé Escargo et un lapin appelé Lapin Rapide. Un jour, ils ont décidé de participer à une grande course pour déterminer qui était le plus rapide des deux. Tout le monde était curieux de savoir qui serait le gagnant, car ils étaient tous émerveillés par la vitesse de ces deux champions.
# 
# La course commença dans une forêt dense, avec Escargo départageant un peu lentement, car il faut être prudent lorsqu'on est escargot. Mais au fur et à mesure qu'il avançait, il commença à ralentir encore plus. Les spectateurs commencèrent à pencher pour Lapin Rapide, qui fonçait rapidement vers la ligne d'arrivée.
# 
# Mais soudain, Escargo se transforma en une force incroyable. Au lieu de marcher lentement comme un escargot, il courait à toute allure, faisant preuve d'une vitesse surprenante. Les spectateurs essayaient de se remémorer leurs notes de mathématiques pour calculer sa vélocité actuelle.
# 
# Pendant ce temps, Lapin Rapide ralentissait légèrement, trouvant l'air frais trop confortable pour continuer à courir aussi vite. Il décida alors de prendre quelques pauses pour profiter de son environnement.
# 
# Escargo continuait à courir à toute vitesse, en dépit de son handicap d'être un escargot. Les spectateurs étaient émerveillés par sa persévérance et sa volonté de gagner. Finalement, Escargo franchit la ligne d'arrivée en premier, et tous criaurent en joie.
# 
# Les habitants de la forêt décidèrent de lui décerner le titre d'"athlète le plus courageux" pour sa performance incroyable. Quant à Lapin Rapide, il fut content de sa deuxième place, se disant simplement que les pauses étaient bien plus agréables que la compétition. Et tous vécurent heureux et satisfaits de cette course mémorable.

You can also use the Google Colab Notebook provided below.

Open In Colab

Inference using the unquantized model with vLLM

Set up an OpenAI-compatible server with the following command:

# Install vLLM
# This may take 5-10 minutes.
# pip install vllm

# Start server for Vigostral-Chat models
python -m vllm.entrypoints.openai.api_server --model bofenghuang/vigostral-7b-chat

# List models
# curl http://localhost:8000/v1/models

Query the model using the openai python package.

import openai

# Modify OpenAI's API key and API base to use vLLM's API server.
openai.api_key = "EMPTY"
openai.api_base = "http://localhost:8000/v1"

# First model
models = openai.Model.list()
model = models["data"][0]["id"]

# Chat completion API
chat_completion = openai.ChatCompletion.create(
    model=model,
    messages=[
        {"role": "user", "content": "Parle-moi de toi-même."},
    ],
    max_tokens=1024,
    temperature=0.7,
)
print("Chat completion results:", chat_completion)

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.