--- language: - en license: apache-2.0 tags: - sentence-transformers - sparse-encoder - sparse - csr - generated_from_trainer - dataset_size:99000 - loss:CSRLoss - loss:SparseMultipleNegativesRankingLoss base_model: mixedbread-ai/mxbai-embed-large-v1 widget: - text: Saudi Arabia–United Arab Emirates relations However, the UAE and Saudi Arabia continue to take somewhat differing stances on regional conflicts such the Yemeni Civil War, where the UAE opposes Al-Islah, and supports the Southern Movement, which has fought against Saudi-backed forces, and the Syrian Civil War, where the UAE has disagreed with Saudi support for Islamist movements.[4] - text: Economy of New Zealand New Zealand's diverse market economy has a sizable service sector, accounting for 63% of all GDP activity in 2013.[17] Large scale manufacturing industries include aluminium production, food processing, metal fabrication, wood and paper products. Mining, manufacturing, electricity, gas, water, and waste services accounted for 16.5% of GDP in 2013.[17] The primary sector continues to dominate New Zealand's exports, despite accounting for 6.5% of GDP in 2013.[17] - text: who was the first president of indian science congress meeting held in kolkata in 1914 - text: Get Over It (Eagles song) "Get Over It" is a song by the Eagles released as a single after a fourteen-year breakup. It was also the first song written by bandmates Don Henley and Glenn Frey when the band reunited. "Get Over It" was played live for the first time during their Hell Freezes Over tour in 1994. It returned the band to the U.S. Top 40 after a fourteen-year absence, peaking at No. 31 on the Billboard Hot 100 chart. It also hit No. 4 on the Billboard Mainstream Rock Tracks chart. The song was not played live by the Eagles after the "Hell Freezes Over" tour in 1994. It remains the group's last Top 40 hit in the U.S. - text: 'Cornelius the Centurion Cornelius (Greek: Κορνήλιος) was a Roman centurion who is considered by Christians to be one of the first Gentiles to convert to the faith, as related in Acts of the Apostles.' datasets: - sentence-transformers/natural-questions pipeline_tag: feature-extraction library_name: sentence-transformers metrics: - dot_accuracy@1 - dot_accuracy@3 - dot_accuracy@5 - dot_accuracy@10 - dot_precision@1 - dot_precision@3 - dot_precision@5 - dot_precision@10 - dot_recall@1 - dot_recall@3 - dot_recall@5 - dot_recall@10 - dot_ndcg@10 - dot_mrr@10 - dot_map@100 - query_active_dims - query_sparsity_ratio - corpus_active_dims - corpus_sparsity_ratio co2_eq_emissions: emissions: 50.33638897970368 energy_consumed: 0.1294986621620256 source: codecarbon training_type: fine-tuning on_cloud: false cpu_model: 13th Gen Intel(R) Core(TM) i7-13700K ram_total_size: 31.777088165283203 hours_used: 0.335 hardware_used: 1 x NVIDIA GeForce RTX 3090 model-index: - name: Sparse CSR model trained on Natural Questions results: - task: type: sparse-information-retrieval name: Sparse Information Retrieval dataset: name: NanoMSMARCO 4 type: NanoMSMARCO_4 metrics: - type: dot_accuracy@1 value: 0.06 name: Dot Accuracy@1 - type: dot_accuracy@3 value: 0.22 name: Dot Accuracy@3 - type: dot_accuracy@5 value: 0.22 name: Dot Accuracy@5 - type: dot_accuracy@10 value: 0.28 name: Dot Accuracy@10 - type: dot_precision@1 value: 0.06 name: Dot Precision@1 - type: dot_precision@3 value: 0.07333333333333333 name: Dot Precision@3 - type: dot_precision@5 value: 0.04400000000000001 name: Dot Precision@5 - type: dot_precision@10 value: 0.028000000000000008 name: Dot Precision@10 - type: dot_recall@1 value: 0.06 name: Dot Recall@1 - type: dot_recall@3 value: 0.22 name: Dot Recall@3 - type: dot_recall@5 value: 0.22 name: Dot Recall@5 - type: dot_recall@10 value: 0.28 name: Dot Recall@10 - type: dot_ndcg@10 value: 0.1697596420238124 name: Dot Ndcg@10 - type: dot_mrr@10 value: 0.13452380952380952 name: Dot Mrr@10 - type: dot_map@100 value: 0.14616428018874245 name: Dot Map@100 - type: query_active_dims value: 4.0 name: Query Active Dims - type: query_sparsity_ratio value: 0.9990234375 name: Query Sparsity Ratio - type: corpus_active_dims value: 4.0 name: Corpus Active Dims - type: corpus_sparsity_ratio value: 0.9990234375 name: Corpus Sparsity Ratio - task: type: sparse-information-retrieval name: Sparse Information Retrieval dataset: name: NanoMSMARCO 8 type: NanoMSMARCO_8 metrics: - type: dot_accuracy@1 value: 0.08 name: Dot Accuracy@1 - type: dot_accuracy@3 value: 0.2 name: Dot Accuracy@3 - type: dot_accuracy@5 value: 0.34 name: Dot Accuracy@5 - type: dot_accuracy@10 value: 0.4 name: Dot Accuracy@10 - type: dot_precision@1 value: 0.08 name: Dot Precision@1 - type: dot_precision@3 value: 0.06666666666666667 name: Dot Precision@3 - type: dot_precision@5 value: 0.068 name: Dot Precision@5 - type: dot_precision@10 value: 0.04000000000000001 name: Dot Precision@10 - type: dot_recall@1 value: 0.08 name: Dot Recall@1 - type: dot_recall@3 value: 0.2 name: Dot Recall@3 - type: dot_recall@5 value: 0.34 name: Dot Recall@5 - type: dot_recall@10 value: 0.4 name: Dot Recall@10 - type: dot_ndcg@10 value: 0.23173485139840844 name: Dot Ndcg@10 - type: dot_mrr@10 value: 0.17835714285714285 name: Dot Mrr@10 - type: dot_map@100 value: 0.19541026376110485 name: Dot Map@100 - type: query_active_dims value: 8.0 name: Query Active Dims - type: query_sparsity_ratio value: 0.998046875 name: Query Sparsity Ratio - type: corpus_active_dims value: 8.0 name: Corpus Active Dims - type: corpus_sparsity_ratio value: 0.998046875 name: Corpus Sparsity Ratio - task: type: sparse-information-retrieval name: Sparse Information Retrieval dataset: name: NanoMSMARCO 16 type: NanoMSMARCO_16 metrics: - type: dot_accuracy@1 value: 0.28 name: Dot Accuracy@1 - type: dot_accuracy@3 value: 0.48 name: Dot Accuracy@3 - type: dot_accuracy@5 value: 0.54 name: Dot Accuracy@5 - type: dot_accuracy@10 value: 0.62 name: Dot Accuracy@10 - type: dot_precision@1 value: 0.28 name: Dot Precision@1 - type: dot_precision@3 value: 0.15999999999999998 name: Dot Precision@3 - type: dot_precision@5 value: 0.10800000000000001 name: Dot Precision@5 - type: dot_precision@10 value: 0.062 name: Dot Precision@10 - type: dot_recall@1 value: 0.28 name: Dot Recall@1 - type: dot_recall@3 value: 0.48 name: Dot Recall@3 - type: dot_recall@5 value: 0.54 name: Dot Recall@5 - type: dot_recall@10 value: 0.62 name: Dot Recall@10 - type: dot_ndcg@10 value: 0.44021635939536174 name: Dot Ndcg@10 - type: dot_mrr@10 value: 0.3835793650793651 name: Dot Mrr@10 - type: dot_map@100 value: 0.394923411264133 name: Dot Map@100 - type: query_active_dims value: 16.0 name: Query Active Dims - type: query_sparsity_ratio value: 0.99609375 name: Query Sparsity Ratio - type: corpus_active_dims value: 16.0 name: Corpus Active Dims - type: corpus_sparsity_ratio value: 0.99609375 name: Corpus Sparsity Ratio - task: type: sparse-information-retrieval name: Sparse Information Retrieval dataset: name: NanoMSMARCO 32 type: NanoMSMARCO_32 metrics: - type: dot_accuracy@1 value: 0.3 name: Dot Accuracy@1 - type: dot_accuracy@3 value: 0.46 name: Dot Accuracy@3 - type: dot_accuracy@5 value: 0.6 name: Dot Accuracy@5 - type: dot_accuracy@10 value: 0.7 name: Dot Accuracy@10 - type: dot_precision@1 value: 0.3 name: Dot Precision@1 - type: dot_precision@3 value: 0.15333333333333332 name: Dot Precision@3 - type: dot_precision@5 value: 0.12 name: Dot Precision@5 - type: dot_precision@10 value: 0.07 name: Dot Precision@10 - type: dot_recall@1 value: 0.3 name: Dot Recall@1 - type: dot_recall@3 value: 0.46 name: Dot Recall@3 - type: dot_recall@5 value: 0.6 name: Dot Recall@5 - type: dot_recall@10 value: 0.7 name: Dot Recall@10 - type: dot_ndcg@10 value: 0.48681030478639387 name: Dot Ndcg@10 - type: dot_mrr@10 value: 0.4197380952380952 name: Dot Mrr@10 - type: dot_map@100 value: 0.4254614535725137 name: Dot Map@100 - type: query_active_dims value: 32.0 name: Query Active Dims - type: query_sparsity_ratio value: 0.9921875 name: Query Sparsity Ratio - type: corpus_active_dims value: 32.0 name: Corpus Active Dims - type: corpus_sparsity_ratio value: 0.9921875 name: Corpus Sparsity Ratio - task: type: sparse-information-retrieval name: Sparse Information Retrieval dataset: name: NanoMSMARCO 64 type: NanoMSMARCO_64 metrics: - type: dot_accuracy@1 value: 0.36 name: Dot Accuracy@1 - type: dot_accuracy@3 value: 0.52 name: Dot Accuracy@3 - type: dot_accuracy@5 value: 0.6 name: Dot Accuracy@5 - type: dot_accuracy@10 value: 0.76 name: Dot Accuracy@10 - type: dot_precision@1 value: 0.36 name: Dot Precision@1 - type: dot_precision@3 value: 0.1733333333333333 name: Dot Precision@3 - type: dot_precision@5 value: 0.12 name: Dot Precision@5 - type: dot_precision@10 value: 0.07600000000000001 name: Dot Precision@10 - type: dot_recall@1 value: 0.36 name: Dot Recall@1 - type: dot_recall@3 value: 0.52 name: Dot Recall@3 - type: dot_recall@5 value: 0.6 name: Dot Recall@5 - type: dot_recall@10 value: 0.76 name: Dot Recall@10 - type: dot_ndcg@10 value: 0.5372607707612113 name: Dot Ndcg@10 - type: dot_mrr@10 value: 0.4696904761904762 name: Dot Mrr@10 - type: dot_map@100 value: 0.4788022575587187 name: Dot Map@100 - type: query_active_dims value: 64.0 name: Query Active Dims - type: query_sparsity_ratio value: 0.984375 name: Query Sparsity Ratio - type: corpus_active_dims value: 64.0 name: Corpus Active Dims - type: corpus_sparsity_ratio value: 0.984375 name: Corpus Sparsity Ratio - task: type: sparse-information-retrieval name: Sparse Information Retrieval dataset: name: NanoMSMARCO 128 type: NanoMSMARCO_128 metrics: - type: dot_accuracy@1 value: 0.32 name: Dot Accuracy@1 - type: dot_accuracy@3 value: 0.64 name: Dot Accuracy@3 - type: dot_accuracy@5 value: 0.72 name: Dot Accuracy@5 - type: dot_accuracy@10 value: 0.84 name: Dot Accuracy@10 - type: dot_precision@1 value: 0.32 name: Dot Precision@1 - type: dot_precision@3 value: 0.21333333333333335 name: Dot Precision@3 - type: dot_precision@5 value: 0.14400000000000002 name: Dot Precision@5 - type: dot_precision@10 value: 0.08399999999999999 name: Dot Precision@10 - type: dot_recall@1 value: 0.32 name: Dot Recall@1 - type: dot_recall@3 value: 0.64 name: Dot Recall@3 - type: dot_recall@5 value: 0.72 name: Dot Recall@5 - type: dot_recall@10 value: 0.84 name: Dot Recall@10 - type: dot_ndcg@10 value: 0.5822669458808437 name: Dot Ndcg@10 - type: dot_mrr@10 value: 0.5 name: Dot Mrr@10 - type: dot_map@100 value: 0.507267150580799 name: Dot Map@100 - type: query_active_dims value: 128.0 name: Query Active Dims - type: query_sparsity_ratio value: 0.96875 name: Query Sparsity Ratio - type: corpus_active_dims value: 128.0 name: Corpus Active Dims - type: corpus_sparsity_ratio value: 0.96875 name: Corpus Sparsity Ratio - task: type: sparse-information-retrieval name: Sparse Information Retrieval dataset: name: NanoMSMARCO 256 type: NanoMSMARCO_256 metrics: - type: dot_accuracy@1 value: 0.36 name: Dot Accuracy@1 - type: dot_accuracy@3 value: 0.62 name: Dot Accuracy@3 - type: dot_accuracy@5 value: 0.72 name: Dot Accuracy@5 - type: dot_accuracy@10 value: 0.84 name: Dot Accuracy@10 - type: dot_precision@1 value: 0.36 name: Dot Precision@1 - type: dot_precision@3 value: 0.20666666666666667 name: Dot Precision@3 - type: dot_precision@5 value: 0.14400000000000002 name: Dot Precision@5 - type: dot_precision@10 value: 0.08399999999999999 name: Dot Precision@10 - type: dot_recall@1 value: 0.36 name: Dot Recall@1 - type: dot_recall@3 value: 0.62 name: Dot Recall@3 - type: dot_recall@5 value: 0.72 name: Dot Recall@5 - type: dot_recall@10 value: 0.84 name: Dot Recall@10 - type: dot_ndcg@10 value: 0.5958389192755326 name: Dot Ndcg@10 - type: dot_mrr@10 value: 0.5184047619047619 name: Dot Mrr@10 - type: dot_map@100 value: 0.5253233523294375 name: Dot Map@100 - type: query_active_dims value: 256.0 name: Query Active Dims - type: query_sparsity_ratio value: 0.9375 name: Query Sparsity Ratio - type: corpus_active_dims value: 256.0 name: Corpus Active Dims - type: corpus_sparsity_ratio value: 0.9375 name: Corpus Sparsity Ratio --- # Sparse CSR model trained on Natural Questions This is a [CSR Sparse Encoder](https://www.sbert.net/docs/sparse_encoder/usage/usage.html) model finetuned from [mixedbread-ai/mxbai-embed-large-v1](https://huggingface.co/mixedbread-ai/mxbai-embed-large-v1) on the [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions) dataset using the [sentence-transformers](https://www.SBERT.net) library. It maps sentences & paragraphs to a 4096-dimensional sparse vector space with 256 maximum active dimensions and can be used for semantic search and sparse retrieval. ## Model Details ### Model Description - **Model Type:** CSR Sparse Encoder - **Base model:** [mixedbread-ai/mxbai-embed-large-v1](https://huggingface.co/mixedbread-ai/mxbai-embed-large-v1) - **Maximum Sequence Length:** 512 tokens - **Output Dimensionality:** 4096 dimensions (trained with 256 maximum active dimensions) - **Similarity Function:** Dot Product - **Training Dataset:** - [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions) - **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): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, '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): CSRSparsity({'input_dim': 1024, 'hidden_dim': 4096, 'k': 256, 'k_aux': 512, 'normalize': False, 'dead_threshold': 30}) ) ``` ## 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("tomaarsen/csr-mxbai-embed-large-v1-nq-cos-sim-scale-5-gamma-0.1") # Run inference queries = [ "who is cornelius in the book of acts", ] documents = [ 'Cornelius the Centurion Cornelius (Greek: Κορνήλιος) was a Roman centurion who is considered by Christians to be one of the first Gentiles to convert to the faith, as related in Acts of the Apostles.', "Joe Ranft Ranft reunited with Lasseter when he was hired by Pixar in 1991 as their head of story.[1] There he worked on all of their films produced up to 2006; this included Toy Story (for which he received an Academy Award nomination) and A Bug's Life, as the co-story writer and others as story supervisor. His final film was Cars. He also voiced characters in many of the films, including Heimlich the caterpillar in A Bug's Life, Wheezy the penguin in Toy Story 2, and Jacques the shrimp in Finding Nemo.[1]", 'Wonderful Tonight "Wonderful Tonight" is a ballad written by Eric Clapton. It was included on Clapton\'s 1977 album Slowhand. Clapton wrote the song about Pattie Boyd.[1] The female vocal harmonies on the song are provided by Marcella Detroit (then Marcy Levy) and Yvonne Elliman.', ] query_embeddings = model.encode_query(queries) document_embeddings = model.encode_document(documents) print(query_embeddings.shape, document_embeddings.shape) # [1, 4096] [3, 4096] # Get the similarity scores for the embeddings similarities = model.similarity(query_embeddings, document_embeddings) print(similarities) # tensor([[142.1139, 40.4378, 38.2739]]) ``` ## Evaluation ### Metrics #### Sparse Information Retrieval * Dataset: `NanoMSMARCO_4` * Evaluated with [SparseInformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator) with these parameters: ```json { "max_active_dims": 4 } ``` | Metric | Value | |:----------------------|:-----------| | dot_accuracy@1 | 0.06 | | dot_accuracy@3 | 0.22 | | dot_accuracy@5 | 0.22 | | dot_accuracy@10 | 0.28 | | dot_precision@1 | 0.06 | | dot_precision@3 | 0.0733 | | dot_precision@5 | 0.044 | | dot_precision@10 | 0.028 | | dot_recall@1 | 0.06 | | dot_recall@3 | 0.22 | | dot_recall@5 | 0.22 | | dot_recall@10 | 0.28 | | **dot_ndcg@10** | **0.1698** | | dot_mrr@10 | 0.1345 | | dot_map@100 | 0.1462 | | query_active_dims | 4.0 | | query_sparsity_ratio | 0.999 | | corpus_active_dims | 4.0 | | corpus_sparsity_ratio | 0.999 | #### Sparse Information Retrieval * Dataset: `NanoMSMARCO_8` * Evaluated with [SparseInformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator) with these parameters: ```json { "max_active_dims": 8 } ``` | Metric | Value | |:----------------------|:-----------| | dot_accuracy@1 | 0.08 | | dot_accuracy@3 | 0.2 | | dot_accuracy@5 | 0.34 | | dot_accuracy@10 | 0.4 | | dot_precision@1 | 0.08 | | dot_precision@3 | 0.0667 | | dot_precision@5 | 0.068 | | dot_precision@10 | 0.04 | | dot_recall@1 | 0.08 | | dot_recall@3 | 0.2 | | dot_recall@5 | 0.34 | | dot_recall@10 | 0.4 | | **dot_ndcg@10** | **0.2317** | | dot_mrr@10 | 0.1784 | | dot_map@100 | 0.1954 | | query_active_dims | 8.0 | | query_sparsity_ratio | 0.998 | | corpus_active_dims | 8.0 | | corpus_sparsity_ratio | 0.998 | #### Sparse Information Retrieval * Dataset: `NanoMSMARCO_16` * Evaluated with [SparseInformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator) with these parameters: ```json { "max_active_dims": 16 } ``` | Metric | Value | |:----------------------|:-----------| | dot_accuracy@1 | 0.28 | | dot_accuracy@3 | 0.48 | | dot_accuracy@5 | 0.54 | | dot_accuracy@10 | 0.62 | | dot_precision@1 | 0.28 | | dot_precision@3 | 0.16 | | dot_precision@5 | 0.108 | | dot_precision@10 | 0.062 | | dot_recall@1 | 0.28 | | dot_recall@3 | 0.48 | | dot_recall@5 | 0.54 | | dot_recall@10 | 0.62 | | **dot_ndcg@10** | **0.4402** | | dot_mrr@10 | 0.3836 | | dot_map@100 | 0.3949 | | query_active_dims | 16.0 | | query_sparsity_ratio | 0.9961 | | corpus_active_dims | 16.0 | | corpus_sparsity_ratio | 0.9961 | #### Sparse Information Retrieval * Dataset: `NanoMSMARCO_32` * Evaluated with [SparseInformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator) with these parameters: ```json { "max_active_dims": 32 } ``` | Metric | Value | |:----------------------|:-----------| | dot_accuracy@1 | 0.3 | | dot_accuracy@3 | 0.46 | | dot_accuracy@5 | 0.6 | | dot_accuracy@10 | 0.7 | | dot_precision@1 | 0.3 | | dot_precision@3 | 0.1533 | | dot_precision@5 | 0.12 | | dot_precision@10 | 0.07 | | dot_recall@1 | 0.3 | | dot_recall@3 | 0.46 | | dot_recall@5 | 0.6 | | dot_recall@10 | 0.7 | | **dot_ndcg@10** | **0.4868** | | dot_mrr@10 | 0.4197 | | dot_map@100 | 0.4255 | | query_active_dims | 32.0 | | query_sparsity_ratio | 0.9922 | | corpus_active_dims | 32.0 | | corpus_sparsity_ratio | 0.9922 | #### Sparse Information Retrieval * Dataset: `NanoMSMARCO_64` * Evaluated with [SparseInformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator) with these parameters: ```json { "max_active_dims": 64 } ``` | Metric | Value | |:----------------------|:-----------| | dot_accuracy@1 | 0.36 | | dot_accuracy@3 | 0.52 | | dot_accuracy@5 | 0.6 | | dot_accuracy@10 | 0.76 | | dot_precision@1 | 0.36 | | dot_precision@3 | 0.1733 | | dot_precision@5 | 0.12 | | dot_precision@10 | 0.076 | | dot_recall@1 | 0.36 | | dot_recall@3 | 0.52 | | dot_recall@5 | 0.6 | | dot_recall@10 | 0.76 | | **dot_ndcg@10** | **0.5373** | | dot_mrr@10 | 0.4697 | | dot_map@100 | 0.4788 | | query_active_dims | 64.0 | | query_sparsity_ratio | 0.9844 | | corpus_active_dims | 64.0 | | corpus_sparsity_ratio | 0.9844 | #### Sparse Information Retrieval * Dataset: `NanoMSMARCO_128` * Evaluated with [SparseInformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator) with these parameters: ```json { "max_active_dims": 128 } ``` | Metric | Value | |:----------------------|:-----------| | dot_accuracy@1 | 0.32 | | dot_accuracy@3 | 0.64 | | dot_accuracy@5 | 0.72 | | dot_accuracy@10 | 0.84 | | dot_precision@1 | 0.32 | | dot_precision@3 | 0.2133 | | dot_precision@5 | 0.144 | | dot_precision@10 | 0.084 | | dot_recall@1 | 0.32 | | dot_recall@3 | 0.64 | | dot_recall@5 | 0.72 | | dot_recall@10 | 0.84 | | **dot_ndcg@10** | **0.5823** | | dot_mrr@10 | 0.5 | | dot_map@100 | 0.5073 | | query_active_dims | 128.0 | | query_sparsity_ratio | 0.9688 | | corpus_active_dims | 128.0 | | corpus_sparsity_ratio | 0.9688 | #### Sparse Information Retrieval * Dataset: `NanoMSMARCO_256` * Evaluated with [SparseInformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator) with these parameters: ```json { "max_active_dims": 256 } ``` | Metric | Value | |:----------------------|:-----------| | dot_accuracy@1 | 0.36 | | dot_accuracy@3 | 0.62 | | dot_accuracy@5 | 0.72 | | dot_accuracy@10 | 0.84 | | dot_precision@1 | 0.36 | | dot_precision@3 | 0.2067 | | dot_precision@5 | 0.144 | | dot_precision@10 | 0.084 | | dot_recall@1 | 0.36 | | dot_recall@3 | 0.62 | | dot_recall@5 | 0.72 | | dot_recall@10 | 0.84 | | **dot_ndcg@10** | **0.5958** | | dot_mrr@10 | 0.5184 | | dot_map@100 | 0.5253 | | query_active_dims | 256.0 | | query_sparsity_ratio | 0.9375 | | corpus_active_dims | 256.0 | | corpus_sparsity_ratio | 0.9375 | ## Training Details ### Training Dataset #### natural-questions * Dataset: [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions) at [f9e894e](https://huggingface.co/datasets/sentence-transformers/natural-questions/tree/f9e894e1081e206e577b4eaa9ee6de2b06ae6f17) * Size: 99,000 training samples * Columns: query and answer * Approximate statistics based on the first 1000 samples: | | query | answer | |:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------| | type | string | string | | details | | | * Samples: | query | answer | |:--------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | who played the father in papa don't preach | Alex McArthur Alex McArthur (born March 6, 1957) is an American actor. | | where was the location of the battle of hastings | Battle of Hastings The Battle of Hastings[a] was fought on 14 October 1066 between the Norman-French army of William, the Duke of Normandy, and an English army under the Anglo-Saxon King Harold Godwinson, beginning the Norman conquest of England. It took place approximately 7 miles (11 kilometres) northwest of Hastings, close to the present-day town of Battle, East Sussex, and was a decisive Norman victory. | | how many puppies can a dog give birth to | Canine reproduction The largest litter size to date was set by a Neapolitan Mastiff in Manea, Cambridgeshire, UK on November 29, 2004; the litter was 24 puppies.[22] | * Loss: [CSRLoss](https://sbert.net/docs/package_reference/sparse_encoder/losses.html#csrloss) with these parameters: ```json { "beta": 0.1, "gamma": 0.1, "loss": "SparseMultipleNegativesRankingLoss(scale=5.0, similarity_fct='cos_sim')" } ``` ### Evaluation Dataset #### natural-questions * Dataset: [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions) at [f9e894e](https://huggingface.co/datasets/sentence-transformers/natural-questions/tree/f9e894e1081e206e577b4eaa9ee6de2b06ae6f17) * Size: 1,000 evaluation samples * Columns: query and answer * Approximate statistics based on the first 1000 samples: | | query | answer | |:--------|:-----------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------| | type | string | string | | details | | | * Samples: | query | answer | |:-------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | where is the tiber river located in italy | Tiber The Tiber (/ˈtaɪbər/, Latin: Tiberis,[1] Italian: Tevere [ˈteːvere])[2] is the third-longest river in Italy, rising in the Apennine Mountains in Emilia-Romagna and flowing 406 kilometres (252 mi) through Tuscany, Umbria and Lazio, where it is joined by the river Aniene, to the Tyrrhenian Sea, between Ostia and Fiumicino.[3] It drains a basin estimated at 17,375 square kilometres (6,709 sq mi). The river has achieved lasting fame as the main watercourse of the city of Rome, founded on its eastern banks. | | what kind of car does jay gatsby drive | Jay Gatsby At the Buchanan home, Jordan Baker, Nick, Jay, and the Buchanans decide to visit New York City. Tom borrows Gatsby's yellow Rolls Royce to drive up to the city. On the way to New York City, Tom makes a detour at a gas station in "the Valley of Ashes", a run-down part of Long Island. The owner, George Wilson, shares his concern that his wife, Myrtle, may be having an affair. This unnerves Tom, who has been having an affair with Myrtle, and he leaves in a hurry. | | who sings if i can dream about you | I Can Dream About You "I Can Dream About You" is a song performed by American singer Dan Hartman on the soundtrack album of the film Streets of Fire. Released in 1984 as a single from the soundtrack, and included on Hartman's album I Can Dream About You, it reached number 6 on the Billboard Hot 100.[1] | * Loss: [CSRLoss](https://sbert.net/docs/package_reference/sparse_encoder/losses.html#csrloss) with these parameters: ```json { "beta": 0.1, "gamma": 0.1, "loss": "SparseMultipleNegativesRankingLoss(scale=5.0, similarity_fct='cos_sim')" } ``` ### Training Hyperparameters #### Non-Default Hyperparameters - `eval_strategy`: steps - `per_device_train_batch_size`: 64 - `per_device_eval_batch_size`: 64 - `learning_rate`: 4e-05 - `num_train_epochs`: 1 - `bf16`: True - `load_best_model_at_end`: 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`: 4e-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`: 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} - `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 - `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 - `router_mapping`: {} - `learning_rate_mapping`: {}
### Training Logs | Epoch | Step | Training Loss | Validation Loss | NanoMSMARCO_4_dot_ndcg@10 | NanoMSMARCO_8_dot_ndcg@10 | NanoMSMARCO_16_dot_ndcg@10 | NanoMSMARCO_32_dot_ndcg@10 | NanoMSMARCO_64_dot_ndcg@10 | NanoMSMARCO_128_dot_ndcg@10 | NanoMSMARCO_256_dot_ndcg@10 | |:----------:|:-------:|:-------------:|:---------------:|:-------------------------:|:-------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:---------------------------:|:---------------------------:| | -1 | -1 | - | - | 0.1017 | 0.2908 | 0.4027 | 0.5247 | 0.5770 | 0.5697 | 0.6298 | | 0.0646 | 100 | 0.5458 | - | - | - | - | - | - | - | - | | 0.1293 | 200 | 0.5016 | - | - | - | - | - | - | - | - | | **0.1939** | **300** | **0.4845** | **0.4551** | **0.1698** | **0.2317** | **0.4402** | **0.4868** | **0.5373** | **0.5823** | **0.5958** | | 0.2586 | 400 | 0.475 | - | - | - | - | - | - | - | - | | 0.3232 | 500 | 0.4681 | - | - | - | - | - | - | - | - | | 0.3878 | 600 | 0.4624 | 0.4370 | 0.0873 | 0.2026 | 0.3564 | 0.4182 | 0.5077 | 0.5727 | 0.5906 | | 0.4525 | 700 | 0.4589 | - | - | - | - | - | - | - | - | | 0.5171 | 800 | 0.4547 | - | - | - | - | - | - | - | - | | 0.5818 | 900 | 0.4523 | 0.4284 | 0.1323 | 0.1842 | 0.3484 | 0.3763 | 0.4621 | 0.5478 | 0.5690 | | 0.6464 | 1000 | 0.4505 | - | - | - | - | - | - | - | - | | 0.7111 | 1100 | 0.4484 | - | - | - | - | - | - | - | - | | 0.7757 | 1200 | 0.4474 | 0.4238 | 0.1341 | 0.1551 | 0.3362 | 0.3824 | 0.4837 | 0.5492 | 0.5906 | | 0.8403 | 1300 | 0.4458 | - | - | - | - | - | - | - | - | | 0.9050 | 1400 | 0.4451 | - | - | - | - | - | - | - | - | | 0.9696 | 1500 | 0.4454 | 0.4222 | 0.1466 | 0.1904 | 0.3406 | 0.3836 | 0.4802 | 0.5513 | 0.5895 | | -1 | -1 | - | - | 0.1698 | 0.2317 | 0.4402 | 0.4868 | 0.5373 | 0.5823 | 0.5958 | * The bold row denotes the saved checkpoint. ### Environmental Impact Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon). - **Energy Consumed**: 0.129 kWh - **Carbon Emitted**: 0.050 kg of CO2 - **Hours Used**: 0.335 hours ### Training Hardware - **On Cloud**: No - **GPU Model**: 1 x NVIDIA GeForce RTX 3090 - **CPU Model**: 13th Gen Intel(R) Core(TM) i7-13700K - **RAM Size**: 31.78 GB ### Framework Versions - Python: 3.11.6 - Sentence Transformers: 4.2.0.dev0 - Transformers: 4.52.4 - PyTorch: 2.6.0+cu124 - Accelerate: 1.5.1 - Datasets: 2.21.0 - 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", } ``` #### CSRLoss ```bibtex @misc{wen2025matryoshkarevisitingsparsecoding, title={Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation}, author={Tiansheng Wen and Yifei Wang and Zequn Zeng and Zhong Peng and Yudi Su and Xinyang Liu and Bo Chen and Hongwei Liu and Stefanie Jegelka and Chenyu You}, year={2025}, eprint={2503.01776}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2503.01776}, } ``` #### 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} } ```