Speedsy/turkish-multilingual-e5-small-32768-colbert-1000

PyLate model based on Speedsy/turkish-multilingual-e5-small-32768

This is a PyLate model finetuned from Speedsy/turkish-multilingual-e5-small-32768 on the train dataset. It maps sentences & paragraphs to sequences of 128-dimensional dense vectors and can be used for semantic textual similarity using the MaxSim operator.

Model Details

Model Description

• Model Type: PyLate model
• Base model: Speedsy/turkish-multilingual-e5-small-32768
• Document Length: 180 tokens
• Query Length: 32 tokens
• Output Dimensionality: 128 tokens
• Similarity Function: MaxSim
• Training Dataset:
  • train
• Language: en

Model Sources

• Documentation: PyLate Documentation
• Repository: PyLate on GitHub
• Hugging Face: PyLate models on Hugging Face

Full Model Architecture

ColBERT(
  (0): Transformer({'max_seq_length': 179, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Dense({'in_features': 384, 'out_features': 128, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
)

Usage

First install the PyLate library:

pip install -U pylate

Retrieval

PyLate provides a streamlined interface to index and retrieve documents using ColBERT models. The index leverages the Voyager HNSW index to efficiently handle document embeddings and enable fast retrieval.

Indexing documents

First, load the ColBERT model and initialize the Voyager index, then encode and index your documents:

from pylate import indexes, models, retrieve

# Step 1: Load the ColBERT model
model = models.ColBERT(
    model_name_or_path=pylate_model_id,
)

# Step 2: Initialize the Voyager index
index = indexes.Voyager(
    index_folder="pylate-index",
    index_name="index",
    override=True,  # This overwrites the existing index if any
)

# Step 3: Encode the documents
documents_ids = ["1", "2", "3"]
documents = ["document 1 text", "document 2 text", "document 3 text"]

documents_embeddings = model.encode(
    documents,
    batch_size=32,
    is_query=False,  # Ensure that it is set to False to indicate that these are documents, not queries
    show_progress_bar=True,
)

# Step 4: Add document embeddings to the index by providing embeddings and corresponding ids
index.add_documents(
    documents_ids=documents_ids,
    documents_embeddings=documents_embeddings,
)

Note that you do not have to recreate the index and encode the documents every time. Once you have created an index and added the documents, you can re-use the index later by loading it:

# To load an index, simply instantiate it with the correct folder/name and without overriding it
index = indexes.Voyager(
    index_folder="pylate-index",
    index_name="index",
)

Retrieving top-k documents for queries

Once the documents are indexed, you can retrieve the top-k most relevant documents for a given set of queries. To do so, initialize the ColBERT retriever with the index you want to search in, encode the queries and then retrieve the top-k documents to get the top matches ids and relevance scores:

# Step 1: Initialize the ColBERT retriever
retriever = retrieve.ColBERT(index=index)

# Step 2: Encode the queries
queries_embeddings = model.encode(
    ["query for document 3", "query for document 1"],
    batch_size=32,
    is_query=True,  #  # Ensure that it is set to False to indicate that these are queries
    show_progress_bar=True,
)

# Step 3: Retrieve top-k documents
scores = retriever.retrieve(
    queries_embeddings=queries_embeddings,
    k=10,  # Retrieve the top 10 matches for each query
)

Reranking

If you only want to use the ColBERT model to perform reranking on top of your first-stage retrieval pipeline without building an index, you can simply use rank function and pass the queries and documents to rerank:

from pylate import rank, models

queries = [
    "query A",
    "query B",
]

documents = [
    ["document A", "document B"],
    ["document 1", "document C", "document B"],
]

documents_ids = [
    [1, 2],
    [1, 3, 2],
]

model = models.ColBERT(
    model_name_or_path=pylate_model_id,
)

queries_embeddings = model.encode(
    queries,
    is_query=True,
)

documents_embeddings = model.encode(
    documents,
    is_query=False,
)

reranked_documents = rank.rerank(
    documents_ids=documents_ids,
    queries_embeddings=queries_embeddings,
    documents_embeddings=documents_embeddings,
)

Evaluation

Metrics

Py Late Information Retrieval

• Dataset: ['NanoDBPedia', 'NanoFiQA2018', 'NanoHotpotQA', 'NanoMSMARCO', 'NanoNQ', 'NanoSCIDOCS']
• Evaluated with pylate.evaluation.pylate_information_retrieval_evaluator.PyLateInformationRetrievalEvaluator

Metric	NanoDBPedia	NanoFiQA2018	NanoHotpotQA	NanoMSMARCO	NanoNQ	NanoSCIDOCS
MaxSim_accuracy@1	0.76	0.26	0.76	0.36	0.48	0.34
MaxSim_accuracy@3	0.92	0.46	0.94	0.48	0.66	0.54
MaxSim_accuracy@5	0.96	0.54	0.98	0.64	0.7	0.58
MaxSim_accuracy@10	0.96	0.64	0.98	0.72	0.8	0.66
MaxSim_precision@1	0.76	0.26	0.76	0.36	0.48	0.34
MaxSim_precision@3	0.6333	0.2067	0.4867	0.16	0.2267	0.24
MaxSim_precision@5	0.604	0.16	0.308	0.128	0.144	0.184
MaxSim_precision@10	0.508	0.102	0.17	0.072	0.084	0.128
MaxSim_recall@1	0.0986	0.1372	0.38	0.36	0.46	0.0717
MaxSim_recall@3	0.188	0.2915	0.73	0.48	0.64	0.1487
MaxSim_recall@5	0.2485	0.3707	0.77	0.64	0.67	0.1887
MaxSim_recall@10	0.3532	0.4531	0.85	0.72	0.76	0.2607
MaxSim_ndcg@10	0.6362	0.3393	0.7805	0.517	0.6149	0.2639
MaxSim_mrr@10	0.8367	0.3701	0.849	0.454	0.5769	0.4448
MaxSim_map@100	0.4908	0.2724	0.7118	0.4644	0.5709	0.2017

Pylate Custom Nano BEIR

• Dataset: NanoBEIR_mean
• Evaluated with pylate_nano_beir_evaluator.PylateCustomNanoBEIREvaluator

Metric	Value
MaxSim_accuracy@1	0.4933
MaxSim_accuracy@3	0.6667
MaxSim_accuracy@5	0.7333
MaxSim_accuracy@10	0.7933
MaxSim_precision@1	0.4933
MaxSim_precision@3	0.3256
MaxSim_precision@5	0.2547
MaxSim_precision@10	0.1773
MaxSim_recall@1	0.2512
MaxSim_recall@3	0.413
MaxSim_recall@5	0.4813
MaxSim_recall@10	0.5662
MaxSim_ndcg@10	0.5253
MaxSim_mrr@10	0.5886
MaxSim_map@100	0.452

Training Details

Training Dataset

train

• Dataset: train at b9b0f7f
• Size: 798,036 training samples
• Columns: query_id, document_ids, and scores
• Approximate statistics based on the first 1000 samples:

  	query_id	document_ids	scores
  type	string	list	list
  details	min: 4 tokens mean: 5.82 tokens max: 6 tokens	size: 32 elements	size: 32 elements

• Samples:

  query_id	document_ids	scores
  817836	['2716076', '6741935', '2681109', '5562684', '3507339', ...]	[1.0, 0.7059561610221863, 0.21702419221401215, 0.38270196318626404, 0.20812414586544037, ...]
  1045170	['5088671', '2953295', '8783471', '4268439', '6339935', ...]	[1.0, 0.6493034362792969, 0.0692221149802208, 0.17963139712810516, 0.6697239875793457, ...]
  1154488	['6498614', '3770829', '1060712', '2590533', '7672044', ...]	[0.9497447609901428, 0.6662212610244751, 0.7423420548439026, 1.0, 0.6580896973609924, ...]

• Loss: pylate.losses.distillation.Distillation

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 16
• learning_rate: 3e-05
• num_train_epochs: 1
• bf16: 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: 16
• per_device_eval_batch_size: 8
• 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: 3e-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: 0.0
• 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: batch_sampler
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	NanoDBPedia_MaxSim_ndcg@10	NanoFiQA2018_MaxSim_ndcg@10	NanoHotpotQA_MaxSim_ndcg@10	NanoMSMARCO_MaxSim_ndcg@10	NanoNQ_MaxSim_ndcg@10	NanoSCIDOCS_MaxSim_ndcg@10	NanoBEIR_mean_MaxSim_ndcg@10
0.0004	20	0.0321	-	-	-	-	-	-	-
0.0008	40	0.0316	-	-	-	-	-	-	-
0.0012	60	0.0305	-	-	-	-	-	-	-
0.0016	80	0.0298	-	-	-	-	-	-	-
0.0020	100	0.0291	-	-	-	-	-	-	-
0.0024	120	0.0286	-	-	-	-	-	-	-
0.0028	140	0.0268	-	-	-	-	-	-	-
0.0032	160	0.0262	-	-	-	-	-	-	-
0.0036	180	0.0282	-	-	-	-	-	-	-
0.0040	200	0.0264	-	-	-	-	-	-	-
0.0044	220	0.0257	-	-	-	-	-	-	-
0.0048	240	0.0267	-	-	-	-	-	-	-
0.0052	260	0.026	-	-	-	-	-	-	-
0.0056	280	0.0258	-	-	-	-	-	-	-
0.0060	300	0.025	-	-	-	-	-	-	-
0.0064	320	0.0238	-	-	-	-	-	-	-
0.0068	340	0.0242	-	-	-	-	-	-	-
0.0072	360	0.0256	-	-	-	-	-	-	-
0.0076	380	0.0259	-	-	-	-	-	-	-
0.0080	400	0.0254	-	-	-	-	-	-	-
0.0084	420	0.0245	-	-	-	-	-	-	-
0.0088	440	0.0253	-	-	-	-	-	-	-
0.0092	460	0.0252	-	-	-	-	-	-	-
0.0096	480	0.0238	-	-	-	-	-	-	-
0.0100	500	0.0259	0.6349	0.3320	0.7611	0.5174	0.6534	0.2696	0.5281
0.0104	520	0.0241	-	-	-	-	-	-	-
0.0108	540	0.024	-	-	-	-	-	-	-
0.0112	560	0.0247	-	-	-	-	-	-	-
0.0116	580	0.0234	-	-	-	-	-	-	-
0.0120	600	0.0244	-	-	-	-	-	-	-
0.0124	620	0.0261	-	-	-	-	-	-	-
0.0128	640	0.0267	-	-	-	-	-	-	-
0.0132	660	0.0246	-	-	-	-	-	-	-
0.0136	680	0.0243	-	-	-	-	-	-	-
0.0140	700	0.0231	-	-	-	-	-	-	-
0.0144	720	0.0247	-	-	-	-	-	-	-
0.0148	740	0.0228	-	-	-	-	-	-	-
0.0152	760	0.0239	-	-	-	-	-	-	-
0.0156	780	0.0255	-	-	-	-	-	-	-
0.0160	800	0.0238	-	-	-	-	-	-	-
0.0164	820	0.0249	-	-	-	-	-	-	-
0.0168	840	0.0249	-	-	-	-	-	-	-
0.0172	860	0.0239	-	-	-	-	-	-	-
0.0176	880	0.0247	-	-	-	-	-	-	-
0.0180	900	0.0257	-	-	-	-	-	-	-
0.0184	920	0.0248	-	-	-	-	-	-	-
0.0188	940	0.0235	-	-	-	-	-	-	-
0.0192	960	0.0232	-	-	-	-	-	-	-
0.0196	980	0.0229	-	-	-	-	-	-	-
0.0200	1000	0.024	0.6362	0.3393	0.7805	0.5170	0.6149	0.2639	0.5253

Framework Versions

• Python: 3.11.12
• Sentence Transformers: 3.4.1
• PyLate: 1.1.7
• Transformers: 4.48.2
• PyTorch: 2.6.0+cu124
• Accelerate: 1.6.0
• Datasets: 3.6.0
• Tokenizers: 0.21.1

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"
}

PyLate

@misc{PyLate,
title={PyLate: Flexible Training and Retrieval for Late Interaction Models},
author={Chaffin, Antoine and Sourty, Raphaël},
url={https://github.com/lightonai/pylate},
year={2024}
}