modernbert_xtr_contrastive_multik128-256-512

PyLate model based on answerdotai/ModernBERT-base

This is a PyLate model finetuned from answerdotai/ModernBERT-base on the msmarco-10m-triplets 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: answerdotai/ModernBERT-base <!-- at revision 8949b909ec900327062f0ebf497f51aef5e6f0c8 -->
• Document Length: 512 tokens
• Query Length: 32 tokens
• Output Dimensionality: 128 tokens
• Similarity Function: MaxSim
• Training Dataset:
  • msmarco-10m-triplets

<!-- - **Language:** Unknown --> <!-- - **License:** Unknown -->

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': 512, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
  (1): Dense({'in_features': 768, 'out_features': 128, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity', 'use_residual': False})
)

Usage

First install the PyLate library:

pip install -U pylate

Retrieval

Use this model with PyLate to index and retrieve documents. The index uses FastPLAID for efficient similarity search.

Indexing documents

Load the ColBERT model and initialize the PLAID 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 PLAID index
index = indexes.PLAID(
    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.PLAID(
    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,
)

<!-- ### Direct Usage (Transformers) <details><summary>Click to see the direct usage in Transformers</summary> </details> --> <!-- ### Downstream Usage (Sentence Transformers) You can finetune this model on your own dataset. <details><summary>Click to expand</summary> </details> --> <!-- ### Out-of-Scope Use *List how the model may foreseeably be misused and address what users ought not to do with the model.* -->

Evaluation

Metrics

Py Late Information Retrieval

• Dataset: ['NanoClimateFEVER', 'NanoDBPedia', 'NanoFEVER', 'NanoFiQA2018', 'NanoHotpotQA', 'NanoMSMARCO', 'NanoNFCorpus', 'NanoNQ', 'NanoQuoraRetrieval', 'NanoSCIDOCS', 'NanoArguAna', 'NanoSciFact', 'NanoTouche2020']
• Evaluated with <code>pylate.evaluation.pylate_information_retrieval_evaluator.PyLateInformationRetrievalEvaluator</code>

Nano BEIR

• Dataset: NanoBEIR_mean
• Evaluated with <code>pylate.evaluation.nano_beir_evaluator.NanoBEIREvaluator</code>

<!-- ## Bias, Risks and Limitations *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* --> <!-- ### Recommendations *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* -->

Training Details

Training Dataset

msmarco-10m-triplets

• Dataset: msmarco-10m-triplets at 8c5139a
• Size: 10,000,000 training samples
• Columns: <code>query</code>, <code>positive</code>, and <code>negative</code>
• Approximate statistics based on the first 1000 samples:

  	query	positive	negative
  type	string	string	string
  details	<ul><li>min: 4 tokens</li><li>mean: 9.31 tokens</li><li>max: 31 tokens</li></ul>	<ul><li>min: 20 tokens</li><li>mean: 31.95 tokens</li><li>max: 32 tokens</li></ul>	<ul><li>min: 18 tokens</li><li>mean: 31.91 tokens</li><li>max: 32 tokens</li></ul>

• Samples:

  query	positive	negative
  <code>the most important factor that influences k+ secretion is __________.</code>	<code>The regulation of K+ distribution between the intracellular and extracellular space is referred to as internal K+ balance. The most important factors regulating this movement under normal conditions are insulin and catecholamines (1).</code>	<code>They are both also important for secretion and flow of bile: 1 Cholecystokinin: The name of this hormone describes its effect on the biliary system-cholecysto = gallbladder and kinin = movement. 2 Secretin: This hormone is secreted in response to acid in the duodenum.</code>
  <code>how much did the mackinac bridge cost to build</code>	<code>The cost to design the project was $3,500,000 (Steinman Company). The cost to construct the bridge was $70, 268,500. Two primary contractors were hired to build the bridge: American Bridge for superstructure - $44,532,900; and Merritt-Chapman and Scott of New York for the foundations - $25,735,600.</code>	<code>When your child needs a dental tooth bridge, you need to know the average cost so you can factor the price into your budget. Several factors affect the price of a bridge, which can run between $700 to $1,500 per tooth. If you have insurance or your child is covered by Medicaid, part of the cost may be covered.</code>
  <code>when do concussion symptoms appear</code>	<code>Then you can get advice on what to do next. For milder symptoms, the doctor may recommend rest and ask you to watch your child closely for changes, such as a headache that gets worse. Symptoms of a concussion don't always show up right away, and can develop within 24 to 72 hours after an injury.</code>	<code>Concussion: A traumatic injury to soft tissue, usually the brain, as a result of a violent blow, shaking, or spinning. A brain concussion can cause immediate but temporary impairment of brain functions, such as thinking, vision, equilibrium, and consciousness. After a person has had a concussion, he or she is at increased risk for recurrence. Moreover, after a person has several concussions, less of a blow can cause injury, and the person can require more time to recover.</code>

• Loss: <code>pylate.losses.contrastive.Contrastive</code>

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• learning_rate: 3e-05
• max_steps: 50000
• fp16: True
• dataloader_drop_last: True
• dataloader_num_workers: 8
• ddp_find_unused_parameters: False
• torch_compile: True
• torch_compile_backend: inductor
• eval_on_start: True

All Hyperparameters

<details><summary>Click to expand</summary>

• 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: 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: 3.0
• max_steps: 50000
• 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: 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: True
• dataloader_num_workers: 8
• 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}
• parallelism_config: 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: False
• 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: True
• torch_compile_backend: inductor
• torch_compile_mode: 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: True
• use_liger_kernel: False
• liger_kernel_config: None
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

</details>