S13v3n-2/scoring-camembert-v2

SentenceTransformer based on almanach/camembert-base

This is a sentence-transformers model finetuned from almanach/camembert-base. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: almanach/camembert-base
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False, 'architecture': 'CamembertModel'})
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("S13v3n-2/scoring-camembert-v2")
# Run inference
sentences = [
    'Gérer les procédures juridiques et le contentieux',
    'Créer du contenu digital engageant (vidéos, posts, stories)',
    'Créer des dashboards interactifs pour le pilotage décisionnel',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.0931, 0.0454],
#         [0.0931, 1.0000, 0.1108],
#         [0.0454, 0.1108, 1.0000]])

Evaluation

Metrics

Semantic Similarity

• Dataset: eval-referentiel
• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.9919
spearman_cosine	0.9201

Training Details

Training Dataset

Unnamed Dataset

• Size: 2,544 training samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 9 tokens mean: 13.57 tokens max: 32 tokens	min: 9 tokens mean: 13.59 tokens max: 32 tokens	min: 0.1 mean: 0.35 max: 0.9

• Samples:

  sentence1	sentence2	score
  Concevoir des réseaux de neurones profonds pour le deep learning	Utiliser la conteneurisation avec Docker et l'orchestration avec Kubernetes	0.4
  Faire preuve de créativité numérique dans la production de contenu	Être passionné par l'innovation technologique et l'IA	0.1
  Faire preuve d'esprit entrepreneurial et d'initiative	Résoudre des énigmes logiques et des problèmes complexes	0.1

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Evaluation Dataset

Unnamed Dataset

• Size: 636 evaluation samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 636 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 9 tokens mean: 13.8 tokens max: 32 tokens	min: 9 tokens mean: 13.57 tokens max: 32 tokens	min: 0.1 mean: 0.38 max: 0.9

• Samples:

  sentence1	sentence2	score
  Définir et déployer une stratégie de communication interne et externe	Aimer construire des systèmes et voir ses créations fonctionner	0.1
  Gérer les situations de crise médiatique et protéger la réputation	Effectuer une veille réglementaire et juridique continue	0.1
  Rédiger des actes juridiques et documents légaux professionnels	Créer des designs graphiques professionnels et esthétiques	0.1

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• num_train_epochs: 5
• warmup_ratio: 0.1
• fp16: True
• load_best_model_at_end: True

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 5
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.1
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• save_safetensors: True
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• no_cuda: False
• use_cpu: False
• use_mps_device: False
• seed: 42
• data_seed: None
• jit_mode_eval: False
• bf16: False
• fp16: True
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: 0
• ddp_backend: None
• tpu_num_cores: None
• tpu_metrics_debug: False
• debug: []
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• past_index: -1
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: True
• ignore_data_skip: False
• fsdp: []
• fsdp_min_num_params: 0
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• fsdp_transformer_layer_cls_to_wrap: None
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• parallelism_config: None
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch_fused
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• project: huggingface
• trackio_space_id: trackio
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• skip_memory_metrics: True
• use_legacy_prediction_loop: False
• push_to_hub: False
• resume_from_checkpoint: None
• hub_model_id: None
• hub_strategy: every_save
• hub_private_repo: None
• hub_always_push: False
• hub_revision: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• eval_do_concat_batches: True
• fp16_backend: auto
• push_to_hub_model_id: None
• push_to_hub_organization: None
• mp_parameters:
• auto_find_batch_size: False
• full_determinism: False
• torchdynamo: None
• ray_scope: last
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: no
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• liger_kernel_config: None
• eval_use_gather_object: False
• average_tokens_across_devices: True
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	Validation Loss	eval-referentiel_spearman_cosine
0.5	20	0.1629	-	-
1.0	40	0.0555	-	-
1.25	50	-	0.0268	0.8669
1.5	60	0.0272	-	-
2.0	80	0.0168	-	-
2.5	100	0.0085	0.0064	0.9085
3.0	120	0.0066	-	-
3.5	140	0.0046	-	-
3.75	150	-	0.0026	0.9195
4.0	160	0.0036	-	-
4.5	180	0.0041	-	-
5.0	200	0.0035	0.0020	0.9201

Framework Versions

• Python: 3.13.5
• Sentence Transformers: 5.2.0
• Transformers: 4.57.1
• PyTorch: 2.9.1+cu128
• Accelerate: 1.12.0
• Datasets: 4.7.0
• Tokenizers: 0.22.1

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}