ghost-beard-9942/multilingual-e5-base-custom-finetune

SentenceTransformer based on intfloat/multilingual-e5-base

This is a sentence-transformers model finetuned from intfloat/multilingual-e5-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: intfloat/multilingual-e5-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': '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})
  (2): Normalize()
)

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("ghost-beard-9942/multilingual-e5-base-custom-finetune")
# Run inference
sentences = [
    'query: Lidl Toffees de mantequilla Original 200g',
    'passage: Packaged snacks, confectionery, and nuts intended for casual eating between meals. Schokolade, Gummibären, Chips, Kekse, Cracker, Popcorn, Brezel, Nachos. Erdnüsse, Mandeln, Cashews, Walnüsse, Pistazien, Studentenfutter, Nussmischungen. Riegel, Proteinriegel, Müsliriegel, Bonbons, Lollipops, Lakritze, Marshmallow. chocolat, biscuits, bonbons, noix, chips, chocolate, dulces, frutos secos.',
    'passage: Non-food and uncategorised products that do not belong to any food or personal care category. Tiernahrung: Katzenfutter, Hundefutter, Vogelfutter, Tiersnacks, Aquariumbedarf. Haushaltsgegenstände: Batterien, Glühbirnen, Kerzen, Streichhölzer, Kleber, Klebeband. Bürobedarf: Stifte, Notizbücher, Druckerpapier. Blumen, Pflanzen, Erde, Dünger. Spielzeug, Zeitschriften, Bücher, Geschenkpapier. This category is a last resort — only use when no other category fits.',
]
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,  0.7387,  0.0476],
#         [ 0.7387,  1.0000, -0.0439],
#         [ 0.0476, -0.0439,  1.0000]])

Evaluation

Metrics

Information Retrieval

• Dataset: classification-eval
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.9991
cosine_accuracy@3	1.0
cosine_accuracy@5	1.0
cosine_precision@1	0.9991
cosine_precision@3	0.3333
cosine_precision@5	0.2
cosine_precision@10	0.1
cosine_recall@1	0.9991
cosine_recall@3	1.0
cosine_recall@5	1.0
cosine_recall@10	1.0
cosine_ndcg@5	0.9997
cosine_ndcg@10	0.9997
cosine_mrr@5	0.9996
cosine_mrr@10	0.9996
cosine_map@100	0.9996

Training Details

Training Dataset

Unnamed Dataset

• Size: 93,755 training samples
• Columns: sentence_0 and sentence_1
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1
  type	string	string
  details	min: 7 tokens mean: 16.65 tokens max: 32 tokens	min: 129 tokens mean: 157.47 tokens max: 184 tokens

• Samples:

  sentence_0	sentence_1
  query: Traditional poulet Everyday Original	passage: Animal protein products — fresh, chilled, smoked, or cured — including deli meats and processed meat substitutes. Hähnchenbrust, Rindfleisch, Schweinefleisch, Hackfleisch, Lammfleisch. Lachs, Thunfisch, Kabeljau, Garnelen, Muscheln, Fischfilet, Räucherlachs. Wurst, Salami, Schinken, Mortadella, Aufschnitt, Leberwurst, veganer Aufschnitt, Tofu-Wurst. viande, poulet, poisson, jambon, saucisson, carne, pollo, pescado, jamón. Excludes canned fish/meat (→ canned preserved foods) and frozen (→ frozen foods).
  query: Quality Erdbeermarmelade Extra 450g Deluxe	passage: Condiments, spreads, and dips used to flavour or accompany food — not cooking oils or loose spices. Senf, Ketchup, Mayonnaise, Remoulade, BBQ-Sauce, Sriracha, Tabasco, Sojasauce, Teriyaki. Pesto, Tapenade, Hummus, Baba Ganoush, Tzatziki, Guacamole. Marmelade, Konfitüre, Honig, Erdnussbutter, Mandelmus, Nutella, Aufstrich. sauce, confiture, moutarde, mayonnaise, salsa, mermelada, crema de cacahuete.
  query: Classic Seitenbacher Müsli Verwöhnmischung 750g Deluxe	passage: Dry staple carbohydrates and legumes requiring cooking: grains, pasta, rice, pulses, and breakfast cereals. Spaghetti, Penne, Fusilli, Nudeln, Reis, Basmatireis, Wildreis, Risotto-Reis. Haferflocken, Müsli, Granola, Cornflakes, Quinoa, Couscous, Bulgur, Polenta, Grieß. Linsen, Kichererbsen, Kidneybohnen, Erbsen, Sojabohnen (trocken). pâtes, riz, céréales, légumineuses, pasta, arroz, legumbres, cereales.

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• num_train_epochs: 1
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• 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
• num_train_epochs: 1
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: None
• warmup_ratio: None
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• enable_jit_checkpoint: False
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• use_cpu: False
• seed: 42
• data_seed: None
• bf16: False
• fp16: False
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: -1
• ddp_backend: None
• debug: []
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: False
• ignore_data_skip: False
• fsdp: []
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• 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
• 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
• 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_for_metrics: []
• eval_do_concat_batches: True
• auto_find_batch_size: False
• full_determinism: False
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• 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
• use_cache: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	classification-eval_cosine_ndcg@10
0.1365	200	-	0.9554
0.2730	400	-	0.9913
0.3413	500	2.1584	-
0.4096	600	-	0.9977
0.5461	800	-	0.9987
0.6826	1000	1.5900	0.9993
0.8191	1200	-	0.9997

Framework Versions

• Python: 3.12.12
• Sentence Transformers: 5.2.3
• Transformers: 5.0.0
• PyTorch: 2.10.0+cu128
• Accelerate: 1.12.0
• Datasets: 4.8.3
• Tokenizers: 0.22.2

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}
}