pyrac/rse_engagement_des_collaborateurs

MPNet base trained on AllNLI triplets

This is a sentence-transformers model finetuned from sentence-transformers/paraphrase-multilingual-mpnet-base-v2. 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: sentence-transformers/paraphrase-multilingual-mpnet-base-v2
• Maximum Sequence Length: 128 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity
• Language: en
• License: apache-2.0

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': 128, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (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("pyrac/rse_engagement_des_collaborateurs")
# Run inference
sentences = [
    'On aurait aimé plus d’implication de la part des employés, c’était moyen.',
    'Le service était assez médiocre, mais pas désastreux non plus.',
    "On a été forcé d’accepter cette chambre, ce n'était pas du tout ce qu’on avait demandé.",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Triplet

• Datasets: all-nli-dev and all-nli-test
• Evaluated with TripletEvaluator

Metric	all-nli-dev	all-nli-test
cosine_accuracy	1.0	1.0

Training Details

Training Dataset

Unnamed Dataset

• Size: 132,020 training samples
• Columns: anchor, positive, and negative
• Approximate statistics based on the first 1000 samples:

  	anchor	positive	negative
  type	string	string	string
  details	min: 11 tokens mean: 18.6 tokens max: 29 tokens	min: 10 tokens mean: 16.93 tokens max: 27 tokens	min: 6 tokens mean: 18.87 tokens max: 31 tokens

• Samples:

  anchor	positive	negative
  L’équipe était tellement sympa et réactive, c’était vraiment agréable.	Nous avons été agréablement surpris par l’attention portée aux détails par le personnel.	Très bon service de navette depuis le parking.
  C’était un service classique, ni particulièrement bon ni mauvais.	Le sourire et la disponibilité des employés ont illuminé notre séjour.	Cette chambre nous a totalement déçus, c’était tout sauf confortable.
  Le service était correct, rien de plus.	Les employés étaient lents et mal organisés, cela a beaucoup gêné notre séjour.	La sécurité est assurée avec un gardien présent.

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim"
  }
  

Evaluation Dataset

Unnamed Dataset

• Size: 16,502 evaluation samples
• Columns: anchor, positive, and negative
• Approximate statistics based on the first 1000 samples:

  	anchor	positive	negative
  type	string	string	string
  details	min: 10 tokens mean: 18.49 tokens max: 29 tokens	min: 10 tokens mean: 16.97 tokens max: 27 tokens	min: 6 tokens mean: 18.86 tokens max: 31 tokens

• Samples:

  anchor	positive	negative
  Les employés semblaient démotivés et peu impliqués dans leur travail.	L'équipe semblait désorganisée et peu concernée par les besoins des clients.	La chambre n'était pas adaptée à nos attentes, c’était frustrant.
  Le service était correct, mais il manquait un peu de chaleur humaine.	Un service qui n’a pas marqué mais qui reste acceptable.	Une chambre que nous n’avions pas demandée, c'était une vraie déception.
  Les employés semblaient désintéressés, ça a un peu gâché l’expérience.	Le service était fonctionnel, mais pas très personnalisé.	Stationnement facile et rapide, un plaisir.

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• num_train_epochs: 1
• warmup_ratio: 0.1
• bf16: True
• batch_sampler: no_duplicates

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: 1
• 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
• use_ipex: False
• bf16: True
• fp16: False
• 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: False
• 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}
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• 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
• 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
• dispatch_batches: None
• split_batches: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	Validation Loss	all-nli-dev_cosine_accuracy	all-nli-test_cosine_accuracy
0.0485	100	5.2586	4.1160	1.0	-
0.0969	200	4.1542	4.1071	1.0	-
0.1454	300	4.1483	4.1009	1.0	-
0.1939	400	4.1327	4.0772	1.0	-
0.2424	500	4.1122	4.0561	1.0	-
0.2908	600	4.1027	4.0457	1.0	-
0.3393	700	4.0877	4.0345	1.0	-
0.3878	800	4.0863	4.0216	1.0	-
0.4363	900	4.0785	4.0196	1.0	-
0.4847	1000	4.0661	4.0182	1.0	-
0.5332	1100	4.0637	4.0163	1.0	-
0.5817	1200	4.0606	4.0130	1.0	-
0.6302	1300	4.0601	4.0086	1.0	-
0.6786	1400	4.0516	4.0037	1.0	-
0.7271	1500	4.0472	4.0015	1.0	-
0.7756	1600	4.0465	4.0008	1.0	-
0.8240	1700	4.0421	4.0007	1.0	-
0.8725	1800	4.0463	3.9944	1.0	-
0.9210	1900	4.035	3.9919	1.0	-
0.9695	2000	4.0408	3.9909	1.0	-
-1	-1	-	-	-	1.0

Framework Versions

• Python: 3.12.3
• Sentence Transformers: 3.4.1
• Transformers: 4.48.3
• PyTorch: 2.6.0+cu124
• Accelerate: 1.3.0
• Datasets: 3.2.0
• Tokenizers: 0.21.0

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",
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}