--- language: - en license: apache-2.0 library_name: sentence-transformers tags: - sentence-transformers - sentence-similarity - feature-extraction - generated_from_trainer - dataset_size:6300 - loss:MatryoshkaLoss - loss:MultipleNegativesRankingLoss base_model: BAAI/bge-base-en-v1.5 datasets: [] metrics: - cosine_accuracy@1 - cosine_accuracy@3 - cosine_accuracy@5 - cosine_accuracy@10 - cosine_precision@1 - cosine_precision@3 - cosine_precision@5 - cosine_precision@10 - cosine_recall@1 - cosine_recall@3 - cosine_recall@5 - cosine_recall@10 - cosine_ndcg@10 - cosine_mrr@10 - cosine_map@10 widget: - source_sentence: The Gross Merchandise Sales (GMS) decreased by 1.2% in 2023 compared to 2022. sentences: - What specific matters did the CFPB investigate concerning Equifax? - What was the percentage decline in GMS for the year ended December 31, 2023 compared to 2022? - What percentage of eBay's 2023 net revenues were attributed to international markets? - source_sentence: Asset management and administration fees vary with changes in the balances of client assets due to market fluctuations and client activity. sentences: - Why was there a net outflow of cash in financing activities in fiscal 2022? - How do asset management and administration fees vary at The Charles Schwab Corporation? - What are some key goals of the corporation related to climate change? - source_sentence: Operating profit margin was 19.3 percent in 2023, compared with 13.3 percent in 2022. sentences: - What was the operating profit margin for 2023? - How do the studios compete in the entertainment industry? - What types of audio products does Garmin's Fusion and JL Audio brands offer? - source_sentence: Subsequent to 2023, on February 12, 2024, AbbVie borrowed $5.0 billion under the term loan credit agreement. sentences: - What percentage of U.S. dialysis patient service revenues in 2023 came from Medicare and Medicare Advantage plans? - What is Peloton Interactive, Inc. known for in the interactive fitness industry? - What was the purpose stated by AbbVie for borrowing $5.0 billion under the term loan credit agreement on February 12, 2024? - source_sentence: Chipotle retains an independent third-party compensation consultant each year to conduct a pay equity analysis of its U.S. and Canadian workforce, including factors of pay such as grade level, tenure in role, and external market conditions like geographic location, to ensure consistency and equitable treatment among employees. sentences: - How does Chipotle ensure pay equity among its employees? - How can one locate information on legal proceedings within the Consolidated Financial Statements? - What criteria did the independent audit use to assess the effectiveness of internal control over financial reporting at the company? pipeline_tag: sentence-similarity model-index: - name: BGE base Financial Matryoshka results: - task: type: information-retrieval name: Information Retrieval dataset: name: dim 768 type: dim_768 metrics: - type: cosine_accuracy@1 value: 0.48714285714285716 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.6428571428571429 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.7028571428571428 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.75 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.48714285714285716 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.21428571428571427 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.14057142857142857 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.075 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.48714285714285716 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.6428571428571429 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.7028571428571428 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.75 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.6189459704659449 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.5768225623582763 name: Cosine Mrr@10 - type: cosine_map@10 value: 0.5768225623582766 name: Cosine Map@10 - task: type: information-retrieval name: Information Retrieval dataset: name: dim 512 type: dim_512 metrics: - type: cosine_accuracy@1 value: 0.4857142857142857 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.6328571428571429 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.6885714285714286 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.7457142857142857 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.4857142857142857 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.2109523809523809 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.13771428571428573 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.07457142857142858 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.4857142857142857 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.6328571428571429 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.6885714285714286 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.7457142857142857 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.6149627471785961 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.5730890022675735 name: Cosine Mrr@10 - type: cosine_map@10 value: 0.5730890022675738 name: Cosine Map@10 - task: type: information-retrieval name: Information Retrieval dataset: name: dim 256 type: dim_256 metrics: - type: cosine_accuracy@1 value: 0.46 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.62 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.69 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.74 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.46 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.20666666666666667 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.13799999999999998 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.074 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.46 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.62 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.69 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.74 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.5987029783221659 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.5533594104308387 name: Cosine Mrr@10 - type: cosine_map@10 value: 0.553359410430839 name: Cosine Map@10 - task: type: information-retrieval name: Information Retrieval dataset: name: dim 128 type: dim_128 metrics: - type: cosine_accuracy@1 value: 0.44857142857142857 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.59 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.6542857142857142 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.7385714285714285 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.44857142857142857 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.19666666666666666 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.13085714285714284 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.07385714285714286 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.44857142857142857 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.59 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.6542857142857142 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.7385714285714285 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.5851556676898599 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.5369790249433104 name: Cosine Mrr@10 - type: cosine_map@10 value: 0.5369790249433106 name: Cosine Map@10 - task: type: information-retrieval name: Information Retrieval dataset: name: dim 64 type: dim_64 metrics: - type: cosine_accuracy@1 value: 0.42 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.58 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.6357142857142857 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.7014285714285714 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.42 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.1933333333333333 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.12714285714285714 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.07014285714285713 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.42 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.58 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.6357142857142857 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.7014285714285714 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.5588909341096171 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.5134659863945576 name: Cosine Mrr@10 - type: cosine_map@10 value: 0.5134659863945579 name: Cosine Map@10 --- # BGE base Financial Matryoshka This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5). 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:** [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5) - **Maximum Sequence Length:** 512 tokens - **Output Dimensionality:** 768 tokens - **Similarity Function:** Cosine Similarity - **Language:** en - **License:** apache-2.0 ### Model Sources - **Documentation:** [Sentence Transformers Documentation](https://sbert.net) - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers) - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers) ### Full Model Architecture ``` SentenceTransformer( (0): Transformer({'max_seq_length': 512, 'do_lower_case': True}) with Transformer model: BertModel (1): Pooling({'word_embedding_dimension': 768, '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): Normalize() ) ``` ## 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 SentenceTransformer # Download from the 🤗 Hub model = SentenceTransformer("Sailesh9999/bge-base-financial-matryoshka_2") # Run inference sentences = [ 'Chipotle retains an independent third-party compensation consultant each year to conduct a pay equity analysis of its U.S. and Canadian workforce, including factors of pay such as grade level, tenure in role, and external market conditions like geographic location, to ensure consistency and equitable treatment among employees.', 'How does Chipotle ensure pay equity among its employees?', 'How can one locate information on legal proceedings within the Consolidated Financial Statements?', ] embeddings = model.encode(sentences) print(embeddings.shape) # [3, 768] # Get the similarity scores for the embeddings similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [3, 3] ``` ## Evaluation ### Metrics #### Information Retrieval * Dataset: `dim_768` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.4871 | | cosine_accuracy@3 | 0.6429 | | cosine_accuracy@5 | 0.7029 | | cosine_accuracy@10 | 0.75 | | cosine_precision@1 | 0.4871 | | cosine_precision@3 | 0.2143 | | cosine_precision@5 | 0.1406 | | cosine_precision@10 | 0.075 | | cosine_recall@1 | 0.4871 | | cosine_recall@3 | 0.6429 | | cosine_recall@5 | 0.7029 | | cosine_recall@10 | 0.75 | | cosine_ndcg@10 | 0.6189 | | cosine_mrr@10 | 0.5768 | | **cosine_map@10** | **0.5768** | #### Information Retrieval * Dataset: `dim_512` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.4857 | | cosine_accuracy@3 | 0.6329 | | cosine_accuracy@5 | 0.6886 | | cosine_accuracy@10 | 0.7457 | | cosine_precision@1 | 0.4857 | | cosine_precision@3 | 0.211 | | cosine_precision@5 | 0.1377 | | cosine_precision@10 | 0.0746 | | cosine_recall@1 | 0.4857 | | cosine_recall@3 | 0.6329 | | cosine_recall@5 | 0.6886 | | cosine_recall@10 | 0.7457 | | cosine_ndcg@10 | 0.615 | | cosine_mrr@10 | 0.5731 | | **cosine_map@10** | **0.5731** | #### Information Retrieval * Dataset: `dim_256` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.46 | | cosine_accuracy@3 | 0.62 | | cosine_accuracy@5 | 0.69 | | cosine_accuracy@10 | 0.74 | | cosine_precision@1 | 0.46 | | cosine_precision@3 | 0.2067 | | cosine_precision@5 | 0.138 | | cosine_precision@10 | 0.074 | | cosine_recall@1 | 0.46 | | cosine_recall@3 | 0.62 | | cosine_recall@5 | 0.69 | | cosine_recall@10 | 0.74 | | cosine_ndcg@10 | 0.5987 | | cosine_mrr@10 | 0.5534 | | **cosine_map@10** | **0.5534** | #### Information Retrieval * Dataset: `dim_128` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:----------| | cosine_accuracy@1 | 0.4486 | | cosine_accuracy@3 | 0.59 | | cosine_accuracy@5 | 0.6543 | | cosine_accuracy@10 | 0.7386 | | cosine_precision@1 | 0.4486 | | cosine_precision@3 | 0.1967 | | cosine_precision@5 | 0.1309 | | cosine_precision@10 | 0.0739 | | cosine_recall@1 | 0.4486 | | cosine_recall@3 | 0.59 | | cosine_recall@5 | 0.6543 | | cosine_recall@10 | 0.7386 | | cosine_ndcg@10 | 0.5852 | | cosine_mrr@10 | 0.537 | | **cosine_map@10** | **0.537** | #### Information Retrieval * Dataset: `dim_64` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.42 | | cosine_accuracy@3 | 0.58 | | cosine_accuracy@5 | 0.6357 | | cosine_accuracy@10 | 0.7014 | | cosine_precision@1 | 0.42 | | cosine_precision@3 | 0.1933 | | cosine_precision@5 | 0.1271 | | cosine_precision@10 | 0.0701 | | cosine_recall@1 | 0.42 | | cosine_recall@3 | 0.58 | | cosine_recall@5 | 0.6357 | | cosine_recall@10 | 0.7014 | | cosine_ndcg@10 | 0.5589 | | cosine_mrr@10 | 0.5135 | | **cosine_map@10** | **0.5135** | ## Training Details ### Training Dataset #### Unnamed Dataset * Size: 6,300 training samples * Columns: positive and anchor * Approximate statistics based on the first 1000 samples: | | positive | anchor | |:--------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------| | type | string | string | | details | | | * Samples: | positive | anchor | |:-------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------| | Americas | $ | 7,631,647 | | | $ | 6,817,454 | | 79.3 | % | 84.1 | % | What was the proportion of Americas' net revenue to the company's total net revenue in 2023, and how did it change from 2022? | | Item 1 Business typically includes detailed information about the organization's operations, the nature of the business, and its strategic direction. | What is the title of the section that potentially discusses the operations or nature of a business in a document? | | Operating expenses as a percentage of total revenues decreased to 15.3% in 2023 compared to 15.9% in 2022. | What was the operating expenses as a percentage of total revenues in 2023? | * Loss: [MatryoshkaLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters: ```json { "loss": "MultipleNegativesRankingLoss", "matryoshka_dims": [ 768, 512, 256, 128, 64 ], "matryoshka_weights": [ 1, 1, 1, 1, 1 ], "n_dims_per_step": -1 } ``` ### Training Hyperparameters #### Non-Default Hyperparameters - `eval_strategy`: epoch - `per_device_train_batch_size`: 32 - `per_device_eval_batch_size`: 16 - `gradient_accumulation_steps`: 16 - `learning_rate`: 0.002 - `num_train_epochs`: 4 - `lr_scheduler_type`: cosine - `warmup_ratio`: 0.1 - `bf16`: True - `tf32`: True - `load_best_model_at_end`: True - `optim`: adamw_torch_fused - `batch_sampler`: no_duplicates #### All Hyperparameters
Click to expand - `overwrite_output_dir`: False - `do_predict`: False - `eval_strategy`: epoch - `prediction_loss_only`: True - `per_device_train_batch_size`: 32 - `per_device_eval_batch_size`: 16 - `per_gpu_train_batch_size`: None - `per_gpu_eval_batch_size`: None - `gradient_accumulation_steps`: 16 - `eval_accumulation_steps`: None - `learning_rate`: 0.002 - `weight_decay`: 0.0 - `adam_beta1`: 0.9 - `adam_beta2`: 0.999 - `adam_epsilon`: 1e-08 - `max_grad_norm`: 1.0 - `num_train_epochs`: 4 - `max_steps`: -1 - `lr_scheduler_type`: cosine - `lr_scheduler_kwargs`: {} - `warmup_ratio`: 0.1 - `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`: True - `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_fused - `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`: False - `hub_always_push`: False - `gradient_checkpointing`: False - `gradient_checkpointing_kwargs`: None - `include_inputs_for_metrics`: False - `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 - `batch_sampler`: no_duplicates - `multi_dataset_batch_sampler`: proportional
### Training Logs | Epoch | Step | Training Loss | dim_128_cosine_map@10 | dim_256_cosine_map@10 | dim_512_cosine_map@10 | dim_64_cosine_map@10 | dim_768_cosine_map@10 | |:----------:|:------:|:-------------:|:---------------------:|:---------------------:|:---------------------:|:--------------------:|:---------------------:| | 0.8122 | 10 | 1.7296 | - | - | - | - | - | | 0.9746 | 12 | - | 0.4001 | 0.4162 | 0.4276 | 0.3764 | 0.4325 | | 1.6244 | 20 | 5.4001 | - | - | - | - | - | | 1.9492 | 24 | - | 0.2783 | 0.2849 | 0.2904 | 0.2511 | 0.2977 | | 2.4365 | 30 | 6.4296 | - | - | - | - | - | | 2.9239 | 36 | - | 0.5106 | 0.5267 | 0.5399 | 0.4879 | 0.5439 | | 3.2487 | 40 | 1.2919 | - | - | - | - | - | | **3.8985** | **48** | **-** | **0.537** | **0.5534** | **0.5731** | **0.5135** | **0.5768** | * The bold row denotes the saved checkpoint. ### Framework Versions - Python: 3.9.18 - Sentence Transformers: 3.0.1 - Transformers: 4.41.2 - PyTorch: 2.1.2+cu121 - Accelerate: 0.29.3 - Datasets: 2.19.1 - Tokenizers: 0.19.1 ## 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", } ``` #### MatryoshkaLoss ```bibtex @misc{kusupati2024matryoshka, title={Matryoshka Representation Learning}, author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi}, year={2024}, eprint={2205.13147}, archivePrefix={arXiv}, primaryClass={cs.LG} } ``` #### MultipleNegativesRankingLoss ```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} } ```