Hugging Face's logo

WeihaoLi
/
icd9

Sentence Similarity
sentence-transformers
Safetensors
bert
feature-extraction
dense
Generated from Trainer
dataset_size:148295
loss:SymmetricLoss
text-embeddings-inference
Model card Files
xet
Community
icd9 / README.md
WeihaoLi's picture
WeihaoLi
Upload model from ../experiments/HiT-biobert-v1.1-icd9-temp/final
a4ed05d verified
preview code
|
raw
history blame contribute delete
19.3 kB
 ---
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- dense
- generated_from_trainer
- dataset_size:148295
- loss:SymmetricLoss
base_model: dmis-lab/biobert-v1.1
widget:
- source_sentence: Complications of pregnancy; childbirth; and the puerperium → Complications
during labor → Forceps delivery
sentences:
- Complications of pregnancy; childbirth; and the puerperium → Complications during
labor
- Complications of pregnancy; childbirth; and the puerperium → Other complications
of birth; puerperium affecting management of mother
- Complications of pregnancy; childbirth; and the puerperium → Normal pregnancy
and/or delivery → Other pregnancy and delivery including normal
- source_sentence: Complications of pregnancy; childbirth; and the puerperium → Complications
mainly related to pregnancy → Early or threatened labor
sentences:
- Complications of pregnancy; childbirth; and the puerperium → Complications mainly
related to pregnancy
- Complications of pregnancy; childbirth; and the puerperium → Abortion-related
disorders → Postabortion complications
- Complications of pregnancy; childbirth; and the puerperium → Indications for care
in pregnancy; labor; and delivery
- source_sentence: Diseases of the respiratory system → Respiratory infections → Acute
bronchitis
sentences:
- Diseases of the respiratory system → Asthma → Asthma
- Diseases of the respiratory system → Lung disease due to external agents
- Diseases of the respiratory system → Respiratory infections
- source_sentence: Diseases of the circulatory system → Diseases of the heart → Cardiac
arrest and ventricular fibrillation
sentences:
- Diseases of the circulatory system → Hypertension → Essential hypertension
- Diseases of the circulatory system → Cerebrovascular disease
- Diseases of the circulatory system → Diseases of the heart
- source_sentence: Infectious and parasitic diseases → Mycoses
sentences:
- Diseases of the skin and subcutaneous tissue → Skin and subcutaneous tissue infections
- Mental illness
- Infectious and parasitic diseases
pipeline_tag: sentence-similarity
library_name: sentence-transformers
---
# HierarchyTransformer based on dmis-lab/biobert-v1.1
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [dmis-lab/biobert-v1.1](https://huggingface.co/dmis-lab/biobert-v1.1). 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:** [dmis-lab/biobert-v1.1](https://huggingface.co/dmis-lab/biobert-v1.1) <!-- at revision 551ca18efd7f052c8dfa0b01c94c2a8e68bc5488 -->
- **Maximum Sequence Length:** 256 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/huggingface/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
HierarchyTransformer(
(0): Transformer({'max_seq_length': 256, 'do_lower_case': False, 'architecture': 'BertModel'})
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)
```
## 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("sentence_transformers_model_id")
# Run inference
sentences = [
'Infectious and parasitic diseases → Mycoses',
'Infectious and parasitic diseases',
'Mental illness',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.6610, 0.3361],
# [0.6610, 1.0000, 0.2730],
# [0.3361, 0.2730, 1.0000]])
```
<!--
### 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: 148,295 training samples
* Columns: <code>child</code>, <code>parent</code>, <code>parent_negative</code>, and <code>child_negative</code>
* Approximate statistics based on the first 1000 samples:
| | child | parent | parent_negative | child_negative |
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
| type | string | string | string | string |
| details | <ul><li>min: 8 tokens</li><li>mean: 25.19 tokens</li><li>max: 65 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 16.22 tokens</li><li>max: 41 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 16.94 tokens</li><li>max: 34 tokens</li></ul> | <ul><li>min: 11 tokens</li><li>mean: 23.48 tokens</li><li>max: 65 tokens</li></ul> |
* Samples:
| child | parent | parent_negative | child_negative |
|:---------------------------------------------------------------------|:-----------------------------------------------|:----------------------------|:----------------------------------------------------------------------------------------------------|
| <code>Infectious and parasitic diseases → Bacterial infection</code> | <code>Infectious and parasitic diseases</code> | <code>Mental illness</code> | <code>Diseases of the nervous system and sense organs → Central nervous system infection</code> |
| <code>Infectious and parasitic diseases → Bacterial infection</code> | <code>Infectious and parasitic diseases</code> | <code>Mental illness</code> | <code>Diseases of the digestive system → Intestinal infection</code> |
| <code>Infectious and parasitic diseases → Bacterial infection</code> | <code>Infectious and parasitic diseases</code> | <code>Mental illness</code> | <code>Diseases of the skin and subcutaneous tissue → Skin and subcutaneous tissue infections</code> |
* Loss: <code>hierarchy_transformers.losses.symmetric_loss.SymmetricLoss</code> with these parameters:
```json
{
"distance_metric": "PoincareBall(c=0.0013021096820011735).dist and dist0",
"HyperbolicChildTriplet": {
"weight": 1.0,
"distance_metric": "PoincareBall(c=0.0013021096820011735).dist",
"margin": 3.0
},
"HyperbolicParentTriplet": {
"weight": 1.0,
"distance_metric": "PoincareBall(c=0.0013021096820011735).dist",
"margin": 3.0
}
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: epoch
- `per_device_train_batch_size`: 128
- `per_device_eval_batch_size`: 512
- `learning_rate`: 1e-05
- `num_train_epochs`: 10
- `warmup_steps`: 500
- `load_best_model_at_end`: True
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: epoch
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 128
- `per_device_eval_batch_size`: 512
- `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`: 1e-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`: 10
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.0
- `warmup_steps`: 500
- `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
- `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`: 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}
- `parallelism_config`: 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
- `project`: huggingface
- `trackio_space_id`: trackio
- `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
- `hub_revision`: None
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: no
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `liger_kernel_config`: None
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: True
- `prompts`: None
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: proportional
- `router_mapping`: {}
- `learning_rate_mapping`: {}
</details>
### Training Logs
<details><summary>Click to expand</summary>
| Epoch | Step | Training Loss |
|:-------:|:--------:|:-------------:|
| 0.0863 | 100 | 2.1613 |
| 0.1726 | 200 | 0.5936 |
| 0.2588 | 300 | 0.1998 |
| 0.3451 | 400 | 0.1107 |
| 0.4314 | 500 | 0.0567 |
| 0.5177 | 600 | 0.0452 |
| 0.6040 | 700 | 0.032 |
| 0.6903 | 800 | 0.0279 |
| 0.7765 | 900 | 0.0218 |
| 0.8628 | 1000 | 0.0235 |
| 0.9491 | 1100 | 0.018 |
| 1.0 | 1159 | - |
| 1.0354 | 1200 | 0.0192 |
| 1.1217 | 1300 | 0.0176 |
| 1.2079 | 1400 | 0.0137 |
| 1.2942 | 1500 | 0.0119 |
| 1.3805 | 1600 | 0.0139 |
| 1.4668 | 1700 | 0.0138 |
| 1.5531 | 1800 | 0.0123 |
| 1.6393 | 1900 | 0.0104 |
| 1.7256 | 2000 | 0.0117 |
| 1.8119 | 2100 | 0.0097 |
| 1.8982 | 2200 | 0.0133 |
| 1.9845 | 2300 | 0.01 |
| **2.0** | **2318** | **-** |
| 2.0708 | 2400 | 0.0109 |
| 2.1570 | 2500 | 0.0074 |
| 2.2433 | 2600 | 0.0072 |
| 2.3296 | 2700 | 0.015 |
| 2.4159 | 2800 | 0.0069 |
| 2.5022 | 2900 | 0.0107 |
| 2.5884 | 3000 | 0.0094 |
| 2.6747 | 3100 | 0.0105 |
| 2.7610 | 3200 | 0.0095 |
| 2.8473 | 3300 | 0.0072 |
| 2.9336 | 3400 | 0.0084 |
| 3.0 | 3477 | - |
| 3.0198 | 3500 | 0.0104 |
| 3.1061 | 3600 | 0.0078 |
| 3.1924 | 3700 | 0.008 |
| 3.2787 | 3800 | 0.0086 |
| 3.3650 | 3900 | 0.0085 |
| 3.4513 | 4000 | 0.0081 |
| 3.5375 | 4100 | 0.0093 |
| 3.6238 | 4200 | 0.0107 |
| 3.7101 | 4300 | 0.008 |
| 3.7964 | 4400 | 0.0099 |
| 3.8827 | 4500 | 0.0058 |
| 3.9689 | 4600 | 0.0084 |
| 4.0 | 4636 | - |
| 4.0552 | 4700 | 0.01 |
| 4.1415 | 4800 | 0.0053 |
| 4.2278 | 4900 | 0.0075 |
| 4.3141 | 5000 | 0.0077 |
| 4.4003 | 5100 | 0.0065 |
| 4.4866 | 5200 | 0.0089 |
| 4.5729 | 5300 | 0.0082 |
| 4.6592 | 5400 | 0.0093 |
| 4.7455 | 5500 | 0.0076 |
| 4.8318 | 5600 | 0.0095 |
| 4.9180 | 5700 | 0.0078 |
| 5.0 | 5795 | - |
| 5.0043 | 5800 | 0.0055 |
| 5.0906 | 5900 | 0.0061 |
| 5.1769 | 6000 | 0.005 |
| 5.2632 | 6100 | 0.0075 |
| 5.3494 | 6200 | 0.0079 |
| 5.4357 | 6300 | 0.006 |
| 5.5220 | 6400 | 0.0095 |
| 5.6083 | 6500 | 0.0099 |
| 5.6946 | 6600 | 0.0084 |
| 5.7808 | 6700 | 0.008 |
| 5.8671 | 6800 | 0.0064 |
| 5.9534 | 6900 | 0.0097 |
| 6.0 | 6954 | - |
| 6.0397 | 7000 | 0.0063 |
| 6.1260 | 7100 | 0.0069 |
| 6.2123 | 7200 | 0.0095 |
| 6.2985 | 7300 | 0.0067 |
| 6.3848 | 7400 | 0.0056 |
| 6.4711 | 7500 | 0.0074 |
| 6.5574 | 7600 | 0.0086 |
| 6.6437 | 7700 | 0.0072 |
| 6.7299 | 7800 | 0.0065 |
| 6.8162 | 7900 | 0.0052 |
| 6.9025 | 8000 | 0.0101 |
| 6.9888 | 8100 | 0.0086 |
| 7.0 | 8113 | - |
| 7.0751 | 8200 | 0.0065 |
| 7.1613 | 8300 | 0.0106 |
| 7.2476 | 8400 | 0.0049 |
| 7.3339 | 8500 | 0.0074 |
| 7.4202 | 8600 | 0.0065 |
| 7.5065 | 8700 | 0.004 |
| 7.5928 | 8800 | 0.0075 |
| 7.6790 | 8900 | 0.009 |
| 7.7653 | 9000 | 0.0059 |
| 7.8516 | 9100 | 0.0063 |
| 7.9379 | 9200 | 0.0095 |
| 8.0 | 9272 | - |
| 8.0242 | 9300 | 0.0082 |
| 8.1104 | 9400 | 0.0067 |
| 8.1967 | 9500 | 0.0063 |
| 8.2830 | 9600 | 0.0071 |
| 8.3693 | 9700 | 0.0064 |
| 8.4556 | 9800 | 0.0072 |
| 8.5418 | 9900 | 0.0059 |
| 8.6281 | 10000 | 0.0085 |
| 8.7144 | 10100 | 0.0083 |
| 8.8007 | 10200 | 0.0046 |
| 8.8870 | 10300 | 0.0055 |
| 8.9733 | 10400 | 0.008 |
| 9.0 | 10431 | - |
| 9.0595 | 10500 | 0.0066 |
| 9.1458 | 10600 | 0.0068 |
| 9.2321 | 10700 | 0.0093 |
| 9.3184 | 10800 | 0.0067 |
| 9.4047 | 10900 | 0.0054 |
| 9.4909 | 11000 | 0.0079 |
| 9.5772 | 11100 | 0.0052 |
| 9.6635 | 11200 | 0.0073 |
| 9.7498 | 11300 | 0.0088 |
| 9.8361 | 11400 | 0.005 |
| 9.9223 | 11500 | 0.0069 |
| 10.0 | 11590 | - |
* The bold row denotes the saved checkpoint.
</details>
### Framework Versions
- Python: 3.10.13
- Sentence Transformers: 5.1.2
- Transformers: 4.57.1
- PyTorch: 2.9.0+cu128
- Accelerate: 1.11.0
- Datasets: 4.3.0
- Tokenizers: 0.22.1
## 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",
}
```
#### SymmetricLoss
```bibtex
@article{he2024language,
title={Language models as hierarchy encoders},
author={He, Yuan and Yuan, Zhangdie and Chen, Jiaoyan and Horrocks, Ian},
journal={arXiv preprint arXiv:2401.11374},
year={2024}
}
```
<!--
## 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.*
-->