Add new SentenceTransformer model

1_Pooling/config.json

@@ -0,0 +1,10 @@
+{
+  "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
+}

README.md

@@ -0,0 +1,410 @@
+---
+tags:
+- sentence-transformers
+- sentence-similarity
+- feature-extraction
+- generated_from_trainer
+- dataset_size:28050
+- loss:MultipleNegativesRankingLoss
+base_model: BAAI/bge-base-en-v1.5
+widget:
+- source_sentence: What helps an insectivorous plant attract and digest insects?
+  sentences:
+  - This investigation examined the accuracy of several generalizable anthropometric
+    (ANTHRO) and bioelectrical impedance (BIA) regression equations to estimate %
+    body fat (%BF) in women with either upper body (UB) or lower body (LB) fat distribution
+    patterns.
+  - Bacteria can also be chemotrophs. Chemosynthetic bacteria, or chemotrophs , obtain
+    energy by breaking down chemical compounds in their environment. An example of
+    one of these chemicals broken down by bacteria is nitrogen-containing ammonia.
+    These bacteria are important because they help cycle nitrogen through the environment
+    for other living things to use. Nitrogen cannot be made by living organisms, so
+    it must be continually recycled. Organisms need nitrogen to make organic compounds,
+    such as DNA.
+  - Insectivorous Plants An insectivorous plant has specialized leaves to attract
+    and digest insects. The Venus flytrap is popularly known for its insectivorous
+    mode of nutrition, and has leaves that work as traps (Figure 31.16). The minerals
+    it obtains from prey compensate for those lacking in the boggy (low pH) soil of
+    its native North Carolina coastal plains. There are three sensitive hairs in the
+    center of each half of each leaf. The edges of each leaf are covered with long
+    spines. Nectar secreted by the plant attracts flies to the leaf. When a fly touches
+    the sensory hairs, the leaf immediately closes. Next, fluids and enzymes break
+    down the prey and minerals are absorbed by the leaf. Since this plant is popular
+    in the horticultural trade, it is threatened in its original habitat.
+- source_sentence: When carbon atoms are not bonded to as many hydrogen atoms as possible,
+    what kind of hydrocarbon results?
+  sentences:
+  - Unsaturated hydrocarbons have at least one double or triple bond between carbon
+    atoms, so the carbon atoms are not bonded to as many hydrogen atoms as possible.
+    In other words, they are unsaturated with hydrogen atoms.
+  - Endoscopic radiofrequency ablation (RFA) is a promising new treatment of Barrett's
+    esophagus (BE). Adjunctive intra-esophageal pH control with proton pump inhibitors
+    and/or anti-reflux surgery is generally recommended to optimize squamous re-epithelialization
+    after ablation.
+  - The cell wall is located outside the cell membrane. It consists mainly of cellulose
+    and may also contain lignin, which makes it more rigid. The cell wall shapes,
+    supports, and protects the cell. It prevents the cell from absorbing too much
+    water and bursting. It also keeps large, damaging molecules out of the cell.
+- source_sentence: Do comparison of ambulance dispatch protocols for nontraumatic
+    abdominal pain?
+  sentences:
+  - KIOM-79, a combination of four plant extracts, has a preventive effect on diabetic
+    nephropathy and retinopathy in diabetic animal models. In this study, we have
+    investigated the inhibitory effects of KIOM-79 on diabetic cataractogenesis.
+  - To compare rates of undertriage and overtriage of six ambulance dispatch protocols
+    for the presenting complaint of nontraumatic abdominal pain, and to identify the
+    optimal protocol.
+  - a flower is a source of nectar
+- source_sentence: Does altered fractalkine cleavage potentially promote local inflammation
+    in NOD salivary gland?
+  sentences:
+  - In France, when physicians in ambulances take care of patients, they report medical
+    status to the dispatch centre. Then the dispatching physician search for the available
+    and appropriate hospital service to agree in directly receiving the patient. We
+    attempted to evaluate this direct admission dispatch, in a urban area, with many
+    health care facilities.
+  - Despite the high prevalence of cannabis use in schizophrenia, few studies have
+    examined the potential relationship between cannabis exposure and brain structural
+    abnormalities in schizophrenia.
+  - In the nonobese diabetic (NOD) mouse model of Sjögren's syndrome, lymphocytic
+    infiltration is preceded by an accumulation of dendritic cells in the submandibular
+    glands (SMGs). NOD mice also exhibit an increased frequency of mature, fractalkine
+    receptor (CX3C chemokine receptor [CX3CR]1) expressing monocytes, which are considered
+    to be precursors for tissue dendritic cells. To unravel further the role played
+    by fractalkine-CX3CR1 interactions in the salivary gland inflammation, we studied
+    the expression of fractalkine in NOD SMGs.
+- source_sentence: The smallest cyclic ether is called what?
+  sentences:
+  - Most human traits have more complex modes of inheritance than simple Mendelian
+    inheritance. For example, the traits may be controlled by multiple alleles or
+    multiple genes.
+  - Neonatal stress impairs postnatal bone mineralization. Evidence suggests that
+    mechanical tactile stimulation (MTS) in early life decreases stress hormones and
+    improves bone mineralization. Insulin-like growth factor (IGF1) is impacted by
+    stress and essential to bone development. We hypothesized that MTS administered
+    during neonatal stress would improve bone phenotype in later life. We also predicted
+    an increase in bone specific mRNA expression of IGF1 related pathways.
+  - The smallest cyclic ether is called an epoxide. Draw its structure.
+pipeline_tag: sentence-similarity
+library_name: sentence-transformers
+---
+
+# SentenceTransformer based on BAAI/bge-base-en-v1.5
+
+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) <!-- at revision a5beb1e3e68b9ab74eb54cfd186867f64f240e1a -->
+- **Maximum Sequence Length:** 512 tokens
+- **Output Dimensionality:** 768 dimensions
+- **Similarity Function:** Cosine Similarity
+<!-- - **Training Dataset:** Unknown -->
+<!-- - **Language:** Unknown -->
+<!-- - **License:** Unknown -->
+
+### 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("danthepol/mcqa_embedder_v2")
+# Run inference
+sentences = [
+    'The smallest cyclic ether is called what?',
+    'The smallest cyclic ether is called an epoxide. Draw its structure.',
+    'Neonatal stress impairs postnatal bone mineralization. Evidence suggests that mechanical tactile stimulation (MTS) in early life decreases stress hormones and improves bone mineralization. Insulin-like growth factor (IGF1) is impacted by stress and essential to bone development. We hypothesized that MTS administered during neonatal stress would improve bone phenotype in later life. We also predicted an increase in bone specific mRNA expression of IGF1 related pathways.',
+]
+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]
+```
+
+<!--
+### 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.*
+-->
+
+<!--
+## 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
+
+#### Unnamed Dataset
+
+* Size: 28,050 training samples
+* Columns: <code>sentence_0</code> and <code>sentence_1</code>
+* Approximate statistics based on the first 1000 samples:
+  |         | sentence_0                                                                        | sentence_1                                                                         |
+  |:--------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
+  | type    | string                                                                            | string                                                                             |
+  | details | <ul><li>min: 6 tokens</li><li>mean: 23.02 tokens</li><li>max: 63 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 81.53 tokens</li><li>max: 512 tokens</li></ul> |
+* Samples:
+  | sentence_0                                                                                                   | sentence_1                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          |
+  |:-------------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+  | <code>Ectotherms undergo a variety of changes at the cellular level to acclimatize to shifts in what?</code> | <code>There are 44 autosomes and 2 sex chromosomes in the human genome, for a total of 46 chromosomes (23 pairs). Sex chromosomes specify an organism's genetic sex. Humans can have two different sex chromosomes, one called X and the other Y. Normal females possess two X chromosomes and normal males one X and one Y. An autosome is any chromosome other than a sex chromosome. The Figure below shows a representation of the 24 different human chromosomes. Figure below shows a karyotype of the human genome. A karyotype depicts, usually in a photograph, the chromosomal complement of an individual, including the number of chromosomes and any large chromosomal abnormalities. Karyotypes use chromosomes from the metaphase stage of mitosis.</code>                                           |
+  | <code>All polar compounds contain what type of bonds?</code>                                                 | <code>Polar compounds, such as water, are compounds that have a partial negative charge on one side of each molecule and a partial positive charge on the other side. All polar compounds contain polar bonds (although not all compounds that contain polar bonds are polar. ) In a polar bond, two atoms share electrons unequally. One atom attracts the shared electrons more strongly, so it has a partial negative charge. The other atom attracts the shared electrons less strongly, so it is has a partial positive charge. In a water molecule, the oxygen atom attracts the shared electrons more strongly than the hydrogen atoms do. This explains why the oxygen side of the water molecule has a partial negative charge and the hydrogen side of the molecule has a partial positive charge.</code> |
+  | <code>Do lateral cephalometric radiograph for the planning of maxillary implant reconstruction?</code>       | <code>To present a simple and objective method for the planning of maxillary implant reconstruction with autogenous bone graft in maxilla atrophy.</code>                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           |
+* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
+  ```json
+  {
+      "scale": 20.0,
+      "similarity_fct": "cos_sim"
+  }
+  ```
+
+### Training Hyperparameters
+#### Non-Default Hyperparameters
+
+- `per_device_train_batch_size`: 32
+- `per_device_eval_batch_size`: 32
+- `multi_dataset_batch_sampler`: round_robin
+
+#### All Hyperparameters
+<details><summary>Click to expand</summary>
+
+- `overwrite_output_dir`: False
+- `do_predict`: False
+- `eval_strategy`: no
+- `prediction_loss_only`: True
+- `per_device_train_batch_size`: 32
+- `per_device_eval_batch_size`: 32
+- `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`: 5e-05
+- `weight_decay`: 0.0
+- `adam_beta1`: 0.9
+- `adam_beta2`: 0.999
+- `adam_epsilon`: 1e-08
+- `max_grad_norm`: 1
+- `num_train_epochs`: 3
+- `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`: False
+- `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}
+- `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
+- `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`: round_robin
+
+</details>
+
+### Training Logs
+| Epoch  | Step | Training Loss |
+|:------:|:----:|:-------------:|
+| 0.5701 | 500  | 0.064         |
+| 1.1403 | 1000 | 0.0455        |
+| 1.7104 | 1500 | 0.0254        |
+| 2.2805 | 2000 | 0.0189        |
+| 2.8506 | 2500 | 0.0155        |
+
+
+### Framework Versions
+- Python: 3.12.8
+- Sentence Transformers: 3.4.1
+- Transformers: 4.51.3
+- PyTorch: 2.3.0+cu121
+- Accelerate: 1.3.0
+- Datasets: 3.6.0
+- 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",
+}
+```
+
+#### 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}
+}
+```
+
+<!--
+## 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.*
+-->

config.json

@@ -0,0 +1,31 @@
+{
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "LABEL_0"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "LABEL_0": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.51.3",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

config_sentence_transformers.json

@@ -0,0 +1,10 @@
+{
+  "__version__": {
+    "sentence_transformers": "3.4.1",
+    "transformers": "4.51.3",
+    "pytorch": "2.3.0+cu121"
+  },
+  "prompts": {},
+  "default_prompt_name": null,
+  "similarity_fn_name": "cosine"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b90c40c2e43a3c54129cc9371442f91f8f82f5ad9d8a0464f66ac8c9c94d695
+size 437951328

modules.json

@@ -0,0 +1,20 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  },
+  {
+    "idx": 2,
+    "name": "2",
+    "path": "2_Normalize",
+    "type": "sentence_transformers.models.Normalize"
+  }
+]

sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+  "max_seq_length": 512,
+  "do_lower_case": true
+}

special_tokens_map.json

@@ -0,0 +1,37 @@
+{
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "[MASK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,65 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "max_length": 512,
+  "model_max_length": 512,
+  "never_split": null,
+  "pad_to_multiple_of": null,
+  "pad_token": "[PAD]",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "stride": 0,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "[UNK]"
+}