Add new SparseEncoder model

74ef59a verified about 2 months ago

21.8 kB

	---
	language:
	- en
	license: apache-2.0
	tags:
	- sentence-transformers
	- sparse-encoder
	- sparse
	- splade
	- generated_from_trainer
	- dataset_size:10000
	- loss:SpladeLoss
	- loss:SparseMultipleNegativesRankingLoss
	- loss:FlopsLoss
	base_model: naver/splade-cocondenser-ensembledistil
	widget:
	- text: Two kids at a ballgame wash their hands.
	- text: Two dogs near a lake, while a person rides by on a horse.
	- text: This mother and her daughter and granddaughter are having car trouble, and
	the poor little girl looks hot out in the heat.
	- text: A young man competes in the Olympics in the pole vaulting competition.
	- text: A man is playing with the brass pots
	datasets:
	- sentence-transformers/all-nli
	pipeline_tag: feature-extraction
	library_name: sentence-transformers
	metrics:
	- pearson_cosine
	- spearman_cosine
	- active_dims
	- sparsity_ratio
	co2_eq_emissions:
	emissions: 0.16583474956305416
	energy_consumed: 0.0029592738907377744
	source: codecarbon
	training_type: fine-tuning
	on_cloud: false
	cpu_model: AMD Ryzen 9 6900HX with Radeon Graphics
	ram_total_size: 30.6114501953125
	hours_used: 0.025
	hardware_used: 1 x NVIDIA GeForce RTX 3070 Ti Laptop GPU
	model-index:
	- name: splade-cocondenser-ensembledistil trained on Natural Language Inference (NLI)
	results:
	- task:
	type: semantic-similarity
	name: Semantic Similarity
	dataset:
	name: sts dev
	type: sts-dev
	metrics:
	- type: pearson_cosine
	value: 0.8553775938865431
	name: Pearson Cosine
	- type: spearman_cosine
	value: 0.8486465022828363
	name: Spearman Cosine
	- type: active_dims
	value: 99.12466812133789
	name: Active Dims
	- type: sparsity_ratio
	value: 0.9967523534459951
	name: Sparsity Ratio
	- task:
	type: semantic-similarity
	name: Semantic Similarity
	dataset:
	name: sts test
	type: sts-test
	metrics:
	- type: pearson_cosine
	value: 0.8223180736705796
	name: Pearson Cosine
	- type: spearman_cosine
	value: 0.8068358333807579
	name: Spearman Cosine
	- type: active_dims
	value: 95.42276763916016
	name: Active Dims
	- type: sparsity_ratio
	value: 0.9968736397470952
	name: Sparsity Ratio
	---

	# splade-cocondenser-ensembledistil trained on Natural Language Inference (NLI)

	This is a [SPLADE Sparse Encoder](https://www.sbert.net/docs/sparse_encoder/usage/usage.html) model finetuned from [naver/splade-cocondenser-ensembledistil](https://huggingface.co/naver/splade-cocondenser-ensembledistil) on the [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) dataset using the [sentence-transformers](https://www.SBERT.net) library. It maps sentences & paragraphs to a 30522-dimensional sparse vector space and can be used for semantic search and sparse retrieval.
	## Model Details

	### Model Description
	- Model Type: SPLADE Sparse Encoder
	- Base model: [naver/splade-cocondenser-ensembledistil](https://huggingface.co/naver/splade-cocondenser-ensembledistil) <!-- at revision 25178a62708a3ab1b5c4b5eb30764d65bfddcfbb -->
	- Maximum Sequence Length: 256 tokens
	- Output Dimensionality: 30522 dimensions
	- Similarity Function: Dot Product
	- Training Dataset:
	- [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli)
	- Language: en
	- License: apache-2.0

	### Model Sources

	- Documentation: [Sentence Transformers Documentation](https://sbert.net)
	- Documentation: [Sparse Encoder Documentation](https://www.sbert.net/docs/sparse_encoder/usage/usage.html)
	- Repository: [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
	- Hugging Face: [Sparse Encoders on Hugging Face](https://huggingface.co/models?library=sentence-transformers&other=sparse-encoder)

	### Full Model Architecture

	```
	SparseEncoder(
	(0): MLMTransformer({'max_seq_length': 256, 'do_lower_case': False}) with MLMTransformer model: BertForMaskedLM
	(1): SpladePooling({'pooling_strategy': 'max', 'activation_function': 'relu', 'word_embedding_dimension': 30522})
	)
	```

	## Usage

	### Direct Usage (Sentence Transformers)

	First install the Sentence Transformers library:

	```bash
	pip install -U sentence-transformers
	```

	Then you can load this model and run inference.
	```python
	from sentence_transformers import SparseEncoder

	# Download from the 🤗 Hub
	model = SparseEncoder("arthurbresnu/example-splade-cocondenser-ensembledistil-nli")
	# Run inference
	sentences = [
	'A man is sitting in on the side of the street with brass pots.',
	'A man is playing with the brass pots',
	'A group of adults are swimming at the beach.',
	]
	embeddings = model.encode(sentences)
	print(embeddings.shape)
	# (3, 30522)

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

	<!--
	### 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

	#### Semantic Similarity

	* Datasets: `sts-dev` and `sts-test`
	* Evaluated with [<code>SparseEmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseEmbeddingSimilarityEvaluator)

	\| Metric \| sts-dev \| sts-test \|
	\|:--------------------\|:-----------\|:-----------\|
	\| pearson_cosine \| 0.8554 \| 0.8223 \|
	\| spearman_cosine \| 0.8486 \| 0.8068 \|
	\| active_dims \| 99.1247 \| 95.4228 \|
	\| sparsity_ratio \| 0.9968 \| 0.9969 \|

	<!--
	## 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

	#### all-nli

	* Dataset: [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) at [d482672](https://huggingface.co/datasets/sentence-transformers/all-nli/tree/d482672c8e74ce18da116f430137434ba2e52fab)
	* Size: 10,000 training samples
	* Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
	* Approximate statistics based on the first 1000 samples:
	\| \| sentence1 \| sentence2 \| score \|
	\|:--------\|:----------------------------------------------------------------------------------\|:---------------------------------------------------------------------------------\|:--------------------------------------------------------------\|
	\| type \| string \| string \| float \|
	\| details \| <ul><li>min: 6 tokens</li><li>mean: 17.38 tokens</li><li>max: 52 tokens</li></ul> \| <ul><li>min: 4 tokens</li><li>mean: 10.7 tokens</li><li>max: 31 tokens</li></ul> \| <ul><li>min: 0.0</li><li>mean: 0.5</li><li>max: 1.0</li></ul> \|
	* Samples:
	\| sentence1 \| sentence2 \| score \|
	\|:--------------------------------------------------------------------\|:---------------------------------------------------------------\|:-----------------\|
	\| <code>A person on a horse jumps over a broken down airplane.</code> \| <code>A person is training his horse for a competition.</code> \| <code>0.5</code> \|
	\| <code>A person on a horse jumps over a broken down airplane.</code> \| <code>A person is at a diner, ordering an omelette.</code> \| <code>0.0</code> \|
	\| <code>A person on a horse jumps over a broken down airplane.</code> \| <code>A person is outdoors, on a horse.</code> \| <code>1.0</code> \|
	* Loss: [<code>SpladeLoss</code>](https://sbert.net/docs/package_reference/sparse_encoder/losses.html#spladeloss) with these parameters:
	```json
	{
	"loss": "SparseMultipleNegativesRankingLoss(scale=1, similarity_fct='dot_score')",
	"lambda_corpus": 0.003
	}
	```

	### Evaluation Dataset

	#### all-nli

	* Dataset: [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) at [d482672](https://huggingface.co/datasets/sentence-transformers/all-nli/tree/d482672c8e74ce18da116f430137434ba2e52fab)
	* Size: 1,000 evaluation samples
	* Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
	* Approximate statistics based on the first 1000 samples:
	\| \| sentence1 \| sentence2 \| score \|
	\|:--------\|:----------------------------------------------------------------------------------\|:----------------------------------------------------------------------------------\|:--------------------------------------------------------------\|
	\| type \| string \| string \| float \|
	\| details \| <ul><li>min: 6 tokens</li><li>mean: 18.44 tokens</li><li>max: 57 tokens</li></ul> \| <ul><li>min: 5 tokens</li><li>mean: 10.57 tokens</li><li>max: 25 tokens</li></ul> \| <ul><li>min: 0.0</li><li>mean: 0.5</li><li>max: 1.0</li></ul> \|
	* Samples:
	\| sentence1 \| sentence2 \| score \|
	\|:-------------------------------------------------------------------\|:---------------------------------------------------------------------------------------------------\|:-----------------\|
	\| <code>Two women are embracing while holding to go packages.</code> \| <code>The sisters are hugging goodbye while holding to go packages after just eating lunch.</code> \| <code>0.5</code> \|
	\| <code>Two women are embracing while holding to go packages.</code> \| <code>Two woman are holding packages.</code> \| <code>1.0</code> \|
	\| <code>Two women are embracing while holding to go packages.</code> \| <code>The men are fighting outside a deli.</code> \| <code>0.0</code> \|
	* Loss: [<code>SpladeLoss</code>](https://sbert.net/docs/package_reference/sparse_encoder/losses.html#spladeloss) with these parameters:
	```json
	{
	"loss": "SparseMultipleNegativesRankingLoss(scale=1, similarity_fct='dot_score')",
	"lambda_corpus": 0.003
	}
	```

	### Training Hyperparameters
	#### Non-Default Hyperparameters

	- `eval_strategy`: steps
	- `per_device_train_batch_size`: 16
	- `per_device_eval_batch_size`: 16
	- `learning_rate`: 4e-06
	- `num_train_epochs`: 1
	- `bf16`: True
	- `load_best_model_at_end`: True
	- `batch_sampler`: no_duplicates

	#### 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`: 16
	- `per_device_eval_batch_size`: 16
	- `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`: 4e-06
	- `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`: 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}
	- `tp_size`: 0
	- `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

	</details>

	### Training Logs
	\| Epoch \| Step \| Training Loss \| Validation Loss \| sts-dev_spearman_cosine \| sts-test_spearman_cosine \|
	\|:--------:\|:-------:\|:-------------:\|:---------------:\|:-----------------------:\|:------------------------:\|
	\| -1 \| -1 \| - \| - \| 0.8366 \| - \|
	\| 0.032 \| 20 \| 1.0832 \| - \| - \| - \|
	\| 0.064 \| 40 \| 0.8212 \| - \| - \| - \|
	\| 0.096 \| 60 \| 0.796 \| - \| - \| - \|
	\| 0.128 \| 80 \| 0.7953 \| - \| - \| - \|
	\| 0.16 \| 100 \| 0.7574 \| - \| - \| - \|
	\| 0.192 \| 120 \| 0.6197 \| 0.6750 \| 0.8443 \| - \|
	\| 0.224 \| 140 \| 0.7125 \| - \| - \| - \|
	\| 0.256 \| 160 \| 0.817 \| - \| - \| - \|
	\| 0.288 \| 180 \| 0.7309 \| - \| - \| - \|
	\| 0.32 \| 200 \| 0.639 \| - \| - \| - \|
	\| 0.352 \| 220 \| 0.6873 \| - \| - \| - \|
	\| 0.384 \| 240 \| 0.6973 \| 0.6253 \| 0.8471 \| - \|
	\| 0.416 \| 260 \| 0.7197 \| - \| - \| - \|
	\| 0.448 \| 280 \| 0.5894 \| - \| - \| - \|
	\| 0.48 \| 300 \| 0.6682 \| - \| - \| - \|
	\| 0.512 \| 320 \| 0.6064 \| - \| - \| - \|
	\| 0.544 \| 340 \| 0.648 \| - \| - \| - \|
	\| 0.576 \| 360 \| 0.6344 \| 0.6071 \| 0.8483 \| - \|
	\| 0.608 \| 380 \| 0.5742 \| - \| - \| - \|
	\| 0.64 \| 400 \| 0.4962 \| - \| - \| - \|
	\| 0.672 \| 420 \| 0.4863 \| - \| - \| - \|
	\| 0.704 \| 440 \| 0.5547 \| - \| - \| - \|
	\| 0.736 \| 460 \| 0.6097 \| - \| - \| - \|
	\| 0.768 \| 480 \| 0.6307 \| 0.6027 \| 0.8471 \| - \|
	\| 0.8 \| 500 \| 0.6226 \| - \| - \| - \|
	\| 0.832 \| 520 \| 0.6607 \| - \| - \| - \|
	\| 0.864 \| 540 \| 0.526 \| - \| - \| - \|
	\| 0.896 \| 560 \| 0.6036 \| - \| - \| - \|
	\| 0.928 \| 580 \| 0.5897 \| - \| - \| - \|
	\| 0.96 \| 600 \| 0.6395 \| 0.5892 \| 0.8486 \| - \|
	\| 0.992 \| 620 \| 0.6069 \| - \| - \| - \|
	\| -1 \| -1 \| - \| - \| - \| 0.8068 \|

	* The bold row denotes the saved checkpoint.

	### Environmental Impact
	Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
	- Energy Consumed: 0.003 kWh
	- Carbon Emitted: 0.000 kg of CO2
	- Hours Used: 0.025 hours

	### Training Hardware
	- On Cloud: No
	- GPU Model: 1 x NVIDIA GeForce RTX 3070 Ti Laptop GPU
	- CPU Model: AMD Ryzen 9 6900HX with Radeon Graphics
	- RAM Size: 30.61 GB

	### Framework Versions
	- Python: 3.12.9
	- Sentence Transformers: 4.2.0.dev0
	- Transformers: 4.50.3
	- PyTorch: 2.6.0+cu124
	- Accelerate: 1.6.0
	- Datasets: 3.5.0
	- Tokenizers: 0.21.1

	## Citation

	### BibTeX

	#### Sentence Transformers
	```bibtex
	@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",
	}
	```

	#### SpladeLoss
	```bibtex
	@misc{formal2022distillationhardnegativesampling,
	title={From Distillation to Hard Negative Sampling: Making Sparse Neural IR Models More Effective},
	author={Thibault Formal and Carlos Lassance and Benjamin Piwowarski and Stéphane Clinchant},
	year={2022},
	eprint={2205.04733},
	archivePrefix={arXiv},
	primaryClass={cs.IR},
	url={https://arxiv.org/abs/2205.04733},
	}
	```

	#### SparseMultipleNegativesRankingLoss
	```bibtex
	@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}
	}
	```

	#### FlopsLoss
	```bibtex
	@article{paria2020minimizing,
	title={Minimizing flops to learn efficient sparse representations},
	author={Paria, Biswajit and Yeh, Chih-Kuan and Yen, Ian EH and Xu, Ning and Ravikumar, Pradeep and P{'o}czos, Barnab{'a}s},
	journal={arXiv preprint arXiv:2004.05665},
	year={2020}
	}
	```

	<!--
	## Glossary

	Clearly define terms in order to be accessible across audiences.
	-->

	<!--
	## Model Card Authors

	Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.
	-->

	<!--
	## Model Card Contact

	Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.
	-->