Claire-Mistral-7B-0.1

Claire-Mistral-7B-0.1

Claire-Mistral-7B-0.1 is a 7B parameter causal decoder-only model built by LINAGORA and OpenLLM-France adapted from Mistral-7B on French conversational data.

Claire-Mistral-7B-0.1 is a pretrained language model designed to be attuned to the dynamics of linguistic interactions in dialogue. Without further training, its expected use is to generate continuations of dialogues. Its main purpose is to serve as a base model for fine-tuning on dialogue generation (e.g., chat) and dialogue understanding (e.g., meeting summarization) tasks. Please note that due to its training, the model is prone to generate dialogues with disfluencies and other constructions common to spoken language.

A qualitatively better variant of this model is available under Claire-7B-0.1.

• Typical usage
  • Typical prompts
• Training Details
  • Training Data
  • Training Procedure
• Evaluation
• License
• Acknowledgements
• Contact

Typical usage

import transformers
import torch

model_name = "OpenLLM-France/Claire-Mistral-7B-0.1"

tokenizer = transformers.AutoTokenizer.from_pretrained(model_name)
model = transformers.AutoModelForCausalLM.from_pretrained(model_name,
    device_map="auto",
    torch_dtype=torch.bfloat16,
    load_in_4bit=True                          # For efficient inference, if supported by the GPU card
)

pipeline = transformers.pipeline("text-generation", model=model, tokenizer=tokenizer)
generation_kwargs = dict(
    num_return_sequences=1,                    # Number of variants to generate.
    return_full_text= False,                   # Do not include the prompt in the generated text.
    max_new_tokens=200,                        # Maximum length for the output text.
    do_sample=True, top_k=10, temperature=1.0, # Sampling parameters.
    pad_token_id=tokenizer.eos_token_id,       # Just to avoid a harmless warning.
)

prompt = """\
- Bonjour Dominique, qu'allez-vous nous cuisiner aujourd'hui ?
- Bonjour Camille,\
"""
completions = pipeline(prompt, **generation_kwargs)
for completion in completions:
    print(prompt + " […]" + completion['generated_text'])

This will print something like:

- Bonjour Dominique, qu'allez-vous nous cuisiner aujourd'hui ?
- Bonjour Camille, […] je vous prépare un plat de saison, une daube provençale.
- Ah je ne connais pas cette recette.
- C'est très facile à préparer, vous n'avez qu'à mettre de l'eau dans une marmite, y mettre de l'oignon émincé, des carottes coupées en petits morceaux, et vous allez mettre votre viande de bœuf coupé en petits morceaux également.
- Je n'ai jamais cuisiné de viande de bœuf, mais c'est vrai que ça a l'air bien facile.
- Vous n'avez plus qu'à laisser mijoter, et ensuite il sera temps de servir les clients.
- Très bien.

You will need at least 6GB of VRAM to run inference using 4bit quantization (16GB of VRAM without 4bit quantization).

If you have trouble running this code, make sure you have recent versions of torch, transformers and accelerate (see requirements.txt).

Typical prompts

Claire-Mistral-7B-0.1 was trained on diarized French conversations. During training, the dialogues were normalized in several formats. The possible formats for expected prompts are as follows:

A monologue can be specified as a single line prompt (though keep in mind that the model might still return a dialogue because of its training):

prompt = "Mesdames et messieurs les députés, chers collègues, bonsoir. Vous l'aurez peut-être remarqué, je cite rarement"

A dialogue between two speakers can be specified with one line per speech turn starting with a dash:

prompt = """\
- Bonjour Dominique, qu'allez-vous nous cuisiner aujourd'hui ?
- Bonjour Camille,\
"""

A dialogue or multilogue (with two or more speakers) can be specified with lines that start with [Intervenant X:] where X is a number:

prompt = """\
[Intervenant 1:] Bonjour Dominique, qu'allez-vous nous cuisiner aujourd'hui ?
[Intervenant 2:] Bonjour Camille,\
"""

A dialogue or multilogue with named speakers can be specified with lines that start with [SpeakerName:] where SpeakerName can be a first name, a first and a last name, a nickname, a title…

prompt = """\
[Mme Camille Durand:] Bonjour Dominique, qu'allez-vous nous cuisiner aujourd'hui ?
[Mr. Dominique Petit:] Bonjour Camille,\
"""

Training Details

Training Data

The training dataset is available at OpenLLM-France/Claire-Dialogue-French-0.1 and described in "The Claire French Dialogue Dataset" (2023).

Claire-Mistral-7B-0.1 was tuned from Mistral-7B-v0.1 on the following data distribution:

Training data was augmented with the following techniques:

• varying the format used to indicate speech turns (dashes or [XXX:])
• substituting [Intervenant X:] for [SpeakerName:] or vice versa, where [SpeakerName:] might be a real name or a randomly generated name
• removing punctuation marks and/or casing (to prepare the model for transcripts produced by some Automatic Speech Recognition systems)

Long conversations were truncated at a maximum of 4096 tokens. Where possible, they were split between speaker turns.

While the model has been trained and evaluated only on French dialogues, it may be able to generate conversations in other languages from the original Mistral-7B-v0.1 training data.

Training Procedure

The training code is available at https://github.com/OpenLLM-France/Lit-Claire.

Claire-Mistral-7B-0.1 is a causal decoder-only model trained on a causal language modeling task (i.e., predict the next token). See Mistral-7B for more details.

Claire-Mistral-7B-0.1 was trained on 8 A100 80GB GPUs for about 50 GPU hours.

Hyperparameters were the following:

Evaluation

See the Evaluation section of Claire-7B-0.1.

License

Given that some of the corpora used for training are only available under CC-BY-NC-SA licenses, Claire-Mistral-7B-0.1 is made available under the CC-BY-NC-SA 4.0 license.

Acknowledgements

This work was performed using HPC resources from GENCI–IDRIS (Grant 2023-AD011014561).

Claire-Mistral-7B-0.1 was created by members of LINAGORA (in alphabetical order): Ismaïl Harrando, Julie Hunter, Jean-Pierre Lorré, Jérôme Louradour, Michel-Marie Maudet, Virgile Rennard, Guokan Shang.

Special thanks to partners from the OpenLLM-France community, especially Christophe Cerisara (LORIA), Pierre-Carl Langlais and Anastasia Stasenko (OpSci), and Pierre Colombo, for valuable advice.

Contact

[email protected]