README.md

metadata

pipeline_tag: sentence-similarity
language: fr
license: mit
datasets:
  - unicamp-dl/mmarco
metrics:
  - recall
tags:
  - passage-retrieval
library_name: sentence-transformers
base_model: antoinelouis/camembert-L10
model-index:
  - name: biencoder-camembert-L10-mmarcoFR
    results:
      - task:
          type: sentence-similarity
          name: Passage Retrieval
        dataset:
          type: unicamp-dl/mmarco
          name: mMARCO-fr
          config: french
          split: validation
        metrics:
          - type: recall_at_500
            name: Recall@500
            value: 87.8
          - type: recall_at_100
            name: Recall@100
            value: 76.7
          - type: recall_at_10
            name: Recall@10
            value: 49.5
          - type: mrr_at_10
            name: MRR@10
            value: 27.5
          - type: ndcg_at_10
            name: nDCG@10
            value: 32.5
          - type: map_at_10
            name: MAP@10
            value: 27

biencoder-camembert-L10-mmarcoFR

This is a lightweight dense single-vector bi-encoder model for French. It maps questions and paragraphs 768-dimensional dense vectors and should be used for semantic search. The model uses an CamemBERT-L10 backbone, which is a pruned version of the pre-trained CamemBERT checkpoint with 13% less parameters, obtained by dropping the top-layers from the original model.

Usage

Here are some examples for using this model with Sentence-Transformers, FlagEmbedding, or Huggingface Transformers.

Using Sentence-Transformers

Start by installing the library: pip install -U sentence-transformers. Then, you can use the model like this:

from sentence_transformers import SentenceTransformer

queries = ["Ceci est un exemple de requête.", "Voici un second exemple."]
passages = ["Ceci est un exemple de passage.", "Et voilà un deuxième exemple."]

model = SentenceTransformer('antoinelouis/biencoder-camembert-L10-mmarcoFR')

q_embeddings = model.encode(queries, normalize_embeddings=True)
p_embeddings = model.encode(passages, normalize_embeddings=True)

similarity = q_embeddings @ p_embeddings.T
print(similarity)

Using FlagEmbedding

Start by installing the library: pip install -U FlagEmbedding. Then, you can use the model like this:

from FlagEmbedding import FlagModel

queries = ["Ceci est un exemple de requête.", "Voici un second exemple."]
passages = ["Ceci est un exemple de passage.", "Et voilà un deuxième exemple."]

model = FlagModel('antoinelouis/biencoder-camembert-L10-mmarcoFR')

q_embeddings = model.encode(queries, normalize_embeddings=True)
p_embeddings = model.encode(passages, normalize_embeddings=True)

similarity = q_embeddings @ p_embeddings.T
print(similarity)

Using Transformers

Start by installing the library: pip install -U transformers. Then, you can use the model like this:

import torch
from torch.nn.functional import normalize
from transformers import AutoTokenizer, AutoModel

def mean_pooling(model_output, attention_mask):
    """ Perform mean pooling on-top of the contextualized word embeddings, while ignoring mask tokens in the mean computation."""
    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)


queries = ["Ceci est un exemple de requête.", "Voici un second exemple."]
passages = ["Ceci est un exemple de passage.", "Et voilà un deuxième exemple."]

tokenizer = AutoTokenizer.from_pretrained('antoinelouis/biencoder-camembert-L10-mmarcoFR')
model = AutoModel.from_pretrained('antoinelouis/biencoder-camembert-L10-mmarcoFR')

q_input = tokenizer(queries, padding=True, truncation=True, return_tensors='pt')
p_input = tokenizer(passages, padding=True, truncation=True, return_tensors='pt')
with torch.no_grad():
    q_output = model(**encoded_queries)
    p_output = model(**encoded_passages)
q_embeddings = mean_pooling(q_output, q_input['attention_mask'])
q_embedddings = normalize(q_embeddings, p=2, dim=1)
p_embeddings = mean_pooling(p_output, p_input['attention_mask'])
p_embedddings = normalize(p_embeddings, p=2, dim=1)

similarity = q_embeddings @ p_embeddings.T
print(similarity)

---

Evaluation

We evaluate the model on the smaller development set of mMARCO-fr, which consists of 6,980 queries for a corpus of 8.8M candidate passages. Below, we compare the model performance with other CamemBERT-based biencoder models fine-tuned on the same dataset. We report the mean reciprocal rank (MRR), normalized discounted cumulative gainand (NDCG), mean average precision (MAP), and recall at various cut-offs (R@k).

	model	#Param.	Size	R@500	R@100(↑)	R@10	MRR@10	NDCG@10	MAP@10
1	biencoder-camembert-base-mmarcoFR	111M	445MB	89.1	77.8	51.5	28.5	33.7	27.9
2	biencoder-camembert-L10-mmarcoFR	96M	386MB	87.8	76.7	49.5	27.5	32.5	27.0
3	biencoder-camembert-L8-mmarcoFR	82M	329MB	87.4	75.9	48.9	26.7	31.8	26.2
4	biencoder-camembert-L6-mmarcoFR	68M	272MB	86.7	74.9	46.7	25.7	30.4	25.1
5	biencoder-camembert-L4-mmarcoFR	54M	216MB	85.4	72.1	44.2	23.7	28.3	23.2
6	biencoder-camembert-L2-mmarcoFR	40M	159MB	81.0	66.3	38.5	20.1	24.3	19.7

---

Training

Data

We use the French training samples from the mMARCO dataset, a multilingual machine-translated version of MS MARCO that contains 8.8M passages and 539K training queries. We do not employ the BM25 negatives provided by the official dataset but instead sample harder negatives mined from 12 distinct dense retrievers, using the msmarco-hard-negatives distillation dataset.

Implementation

The model is initialized from the camembert-L10 checkpoint and optimized via the cross-entropy loss (as in DPR) with a temperature of 0.05. It is fine-tuned on one 32GB NVIDIA V100 GPU for 52k steps (or 40 epochs) using the AdamW optimizer with a batch size of 384, a peak learning rate of 2e-5 with warm up along the first 5200 steps and linear scheduling. We set the maximum sequence lengths for both the questions and passages to 128 tokens. We use the cosine similarity to compute relevance scores.

---

Citation

@online{louis2023,
   author    = 'Antoine Louis',
   title     = 'biencoder-camembert-L10-mmarcoFR: A Biencoder Model Trained on French mMARCO',
   publisher = 'Hugging Face',
   month     = 'may',
   year      = '2023',
   url       = 'https://huggingface.co/antoinelouis/biencoder-camembert-L10-mmarcoFR',
}