dbswldnjs's picture
Upload folder using huggingface_hub
e9bf8b5 verified
---
base_model: google-bert/bert-base-multilingual-cased
datasets: []
language: []
library_name: sentence-transformers
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:1890
- loss:MultipleNegativesRankingLoss
widget:
- source_sentence: 32์„ธ ์—ฌ์ž๊ฐ€ ๋ชฉ์„ ๋งค๋‹ค๊ฐ€ ๊ฐ€์กฑ์—๊ฒŒ ๋ฐœ๊ฒฌ๋˜์–ด ๋ณ‘์›์— ์™”๋‹ค. ์ž„์‹  16์ฃผ์˜€์œผ๋ฉฐ 1๊ฐœ์›” ์ „๋ถ€ํ„ฐ ์‹์‚ฌ๋ฅผ ํ•˜์ง€ ์•Š๊ณ  ๋ˆ„์›Œ๋งŒ
์ง€๋ƒˆ๋‹ค๊ณ  ํ•œ๋‹ค. ๊ธฐ๋ถ„์ด ์šฐ์šธํ•˜๊ณ  ์•„๋ฌด๊ฒƒ๋„ ํ•˜๊ธฐ๊ฐ€ ์‹ซ๋‹ค๊ณ  ํ•œ๋‹ค. ์•„์ด๋ฅผ ์ž˜ ํ‚ค์šธ ์ž์‹ ๋„ ์—†๊ณ  ์‚ด๊ณ  ์‹ถ์ง€ ์•Š์œผ๋‹ˆ ์ฃฝ๊ฒŒ ๋‚ด๋ฒ„๋ ค ๋‘๋ผ๊ณ  ํ•œ๋‹ค. ์น˜๋ฃŒ๋Š”?
sentences:
- ์ „๊ธฐ๊ฒฝ๋ จ์š”๋ฒ•
- ํ•ญ์‘๊ณ ์ œ
- ๊ดœ์ฐฎ๋‹ค๊ณ  ์•ˆ์‹ฌ์‹œํ‚ด
- source_sentence: 59์„ธ ์—ฌ์ž๊ฐ€ ์งˆ๋ถ„๋น„๋ฌผ์ด ์žˆ๊ณ  ์™ธ์Œ๋ถ€๊ฐ€ ๊ฑด์กฐํ•˜๊ณ  ๋”ฐ๊ฐ€์›Œ ๋ณ‘์›์— ์™”๋‹ค. ๋ณด์Šต์ œ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ๋„ ์ฆ์ƒ์ด ์ง€์†๋˜์—ˆ๋‹ค. 40์„ธ์—
์ž๊ถ๊ทผ์ข…์œผ๋กœ ์ž๊ถ์ ˆ์ œ์ˆ ์„ ๋ฐ›์•˜๊ณ  ์™ผ์ชฝ ๋‹ค๋ฆฌ์˜ ๊นŠ์€์ •๋งฅํ˜ˆ์ „์ฆ์œผ๋กœ ์•ฝ๋ฌผ์„ ๋ณต์šฉ ์ค‘์ด๋‹ค. ์•ˆ๋ฉดํ™์กฐ์™€ ๋ถˆ๋ฉด์ฆ์ด 50๋Œ€ ์ดˆ๋ฐ˜์— ์žˆ์—ˆ๋‹ค๊ฐ€ ํ˜„์žฌ๋Š” ์—†๊ณ 
์„ฑ๊ตํ†ต์ด ์žˆ๋‹ค. ๊ณจ๋ฐ˜๊ฒ€์‚ฌ์—์„œ ์™ธ์Œ๋ถ€ ์œ„์ถ•์ด ๊ด€์ฐฐ๋˜์—ˆ๊ณ  ์งˆ๋ถ„๋น„๋ฌผ์˜ ์ –์€ํŽด๋ฐ”๋ฅธํ‘œ๋ณธ๊ฒ€์‚ฌ์—์„œ๋Š” ์ด์ƒ์ด ์—†๋‹ค. ์ฒ˜์น˜๋Š”?
sentences:
- ์‹œ์ƒํ•˜๋ถ€๊ธฐ๋Šฅ์ด์ƒ
- ๊ฒฝ์งˆ ์—์ŠคํŠธ๋กœ๊ฒ
- ๋ฉดํ—ˆ ์ทจ์†Œ์ผ๋ถ€ํ„ฐ 3๋…„ ๊ฒฝ๊ณผ
- source_sentence: '15์„ธ ์—ฌ์ž๊ฐ€ 5์ผ ์ „๋ถ€ํ„ฐ ์—ด์ด ๋‚˜๊ณ  ์˜คํ•œ์ด ๋“ ๋‹ค๋ฉฐ ๋ณ‘์›์— ์™”๋‹ค. ์Œ์‹์„ ์‚ผํ‚ฌ ๋•Œ ๋ชฉ์ด ์•„ํ”„๋‹ค๊ณ  ํ•œ๋‹ค. ํ˜ˆ์•• 100/60
mmHg, ๋งฅ๋ฐ• 75ํšŒ/๋ถ„, ํ˜ธํก 18ํšŒ/๋ถ„, ์ฒด์˜จ 38.0โ„ƒ์ด๋‹ค. ๋ชฉ์˜ ์–‘์ชฝ ์—ฌ๋Ÿฌ ๊ตฐ๋ฐ์—์„œ 1 cm ์ดํ•˜ ํฌ๊ธฐ์˜ ๋ฆผํ”„์ ˆ์ด ๋งŒ์ ธ์ง„๋‹ค. ๋ฆผํ”„์ ˆ์€
์••ํ†ต์ด ์žˆ์œผ๋‚˜ ์ฃผ์œ„ ์กฐ์ง์— ๊ณ ์ •๋˜์–ด ์žˆ์ง€ ์•Š๋‹ค. ๋ชธ์—์„œ ๋ฐœ์ง„์€ ๋ณด์ด์ง€ ์•Š๋Š”๋‹ค. ํ˜ˆ์•ก๊ฒ€์‚ฌ ๊ฒฐ๊ณผ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค. ๋‹ค์Œ ๊ฒ€์‚ฌ๋Š”?๋ฐฑํ˜ˆ๊ตฌ 13,780/mm^3
(์ค‘์„ฑ๊ตฌ 25%, ๋ฆผํ”„๊ตฌ 64%) ํ˜ˆ์ƒ‰์†Œ 13.3 g/dL, ํ˜ˆ์†ŒํŒ 209,000/mm^3 ํ˜ˆ์•ก์š”์†Œ์งˆ์†Œ 7 mg/dL, ํฌ๋ ˆ์•„ํ‹ฐ๋‹Œ 0.5
mg/dL, ์•„์ŠคํŒŒ๋ฅดํ…Œ์ดํŠธ์•„๋ฏธ๋…ธ์ „๋‹ฌํšจ์†Œ 266 U/L ์•Œ๋ผ๋‹Œ์•„๋ฏธ๋…ธ์ „๋‹ฌํšจ์†Œ 298 U/L ์ด๋นŒ๋ฆฌ๋ฃจ๋นˆ 0.7 mg/dL, ์•Œ์นผ๋ฆฌ์ธ์‚ฐ๋ถ„ํ•ดํšจ์†Œ 146
U/L (์ฐธ๊ณ ์น˜, 33๏ฝž96) C-๋ฐ˜์‘๋‹จ๋ฐฑ์งˆ 13 mg/L (์ฐธ๊ณ ์น˜, <10) '
sentences:
- ํ˜ˆ์ฒญ ๋ฐ”์ด๋Ÿฌ์Šค์บก์‹œ๋“œํ•ญ์›(VCA) IgM ํ•ญ์ฒด
- ์ธก์ • ๋ฐ”์ด์–ด์Šค
- ๋‚ ํŠธ๋ ‰์†
- source_sentence: ์ž„์‹ ๋‚˜์ด 27์ฃผ, ์ถœ์ƒ์ฒด์ค‘ 750 g์œผ๋กœ ํƒœ์–ด๋‚œ ์‹ ์ƒ์•„๊ฐ€ ์ƒํ›„ 5์ผ์งธ ๊ฐ‘์ž๊ธฐ ์ฒญ์ƒ‰์ฆ์ด ๋ฐœ์ƒํ•˜์˜€๋‹ค. ์ถœ์ƒ ์งํ›„ ํํ‘œ๋ฉดํ™œ์„ฑ์ œ๋ฅผ
ํˆฌ์—ฌ๋ฐ›์•˜๊ณ , ์ดํ›„ ๊ธฐ๊ณ„ํ™˜๊ธฐ์น˜๋ฃŒ ์ค‘์ด๋‹ค. ์‹ฌ๋ฐ• 170ํšŒ/๋ถ„, ํ˜ธํก 80ํšŒ/๋ถ„, ๊ฒฝํ”ผ์‚ฐ์†Œํฌํ™”๋„๋Š” ์˜ค๋ฅธ์†๊ณผ ์™ผ๋ฐœ์—์„œ ๋ชจ๋‘ 60% ์ด๋‹ค. ์•ž๊ฐ€์Šด์ด
ํŒฝ์ฐฝ๋˜๊ณ , ์˜ค๋ฅธ์ชฝ ๊ฐ€์Šด ์ฒญ์ง„์—์„œ ํ˜ธํก์Œ์ด ์ž˜ ๋“ค๋ฆฌ์ง€ ์•Š๋Š”๋‹ค. ๊ฒ€์‚ฌ๋Š”?
sentences:
- ์š”์ฒญ์— ์‘ํ•จ
- ๋น„์ „ํ˜•์  ์–‘์ƒ ๋™๋ฐ˜ ์ฃผ์š”์šฐ์šธ์žฅ์• 
- ๊ฐ€์Šด X์„ ์‚ฌ์ง„
- source_sentence: '58์„ธ ๋‚จ์ž๊ฐ€ 7์‹œ๊ฐ„ ์ „๋ถ€ํ„ฐ ์œ—๋ฐฐ๊ฐ€ ์•„ํŒŒ์„œ ๋ณ‘์›์— ์™”๋‹ค. ํ‰์†Œ์— ์•Œ์ฝ”์˜ฌ๊ฐ„๊ฒฝํ™”๋กœ ์น˜๋ฃŒ๋ฅผ ๋ฐ›๊ณ  ์žˆ์œผ๋ฉฐ ์†Œํ™”๊ถค์–‘์— ์˜ํ•œ
์ฒœ๊ณต์œผ๋กœ ์ˆ˜์ˆ ์„ ๋ฐ›์„ ์˜ˆ์ •์ด๋‹ค. ํ˜ˆ์•• 130/90 mmHg, ๋งฅ๋ฐ• 95ํšŒ/๋ถ„, ํ˜ธํก 22ํšŒ/๋ถ„, ์ฒด์˜จ 37.5โ„ƒ์ด๋‹ค. ๋ฐฐ ์ „์ฒด๊ฐ€ ๋”ฑ๋”ฑํ•˜๊ณ  ๋ฐฐ์—
์••ํ†ต๊ณผ ๋ฐ˜๋™์••ํ†ต์ด ์žˆ๋‹ค. ํ˜ˆ์•ก๊ฒ€์‚ฌ ๊ฒฐ๊ณผ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค. ์ˆ˜์ˆ  ์ „ ํˆฌ์—ฌํ•ด์•ผ ํ•  ์ œ์ œ๋Š”?ํ˜ˆ์ƒ‰์†Œ 10.3 g/dL, ๋ฐฑํ˜ˆ๊ตฌ 22,000/mm^3,
ํ˜ˆ์†ŒํŒ 120,000/mm^3 ํ”„๋กœํŠธ๋กฌ๋นˆ์‹œ๊ฐ„ 20์ดˆ(์ฐธ๊ณ ์น˜, 12.7๏ฝž15.4) ํ™œ์„ฑํ™”๋ถ€๋ถ„ํŠธ๋กฌ๋ณดํ”Œ๋ผ์Šคํ‹ด์‹œ๊ฐ„ 30์ดˆ(์ฐธ๊ณ ์น˜, 26.3๏ฝž39.4)
์ด๋‹จ๋ฐฑ์งˆ 6.5 g/dL, ์•Œ๋ถ€๋ฏผ 3.0 g/dL,์ด๋นŒ๋ฆฌ๋ฃจ๋นˆ 3.5 mg/dL, '
sentences:
- โ€œ์ „ํŒŒ ๊ฐ€๋Šฅ์„ฑ์ด ์ด๋ ‡๊ฒŒ ๋†’์€๋ฐ๋„ ๋‹ค๋ฅธ ์‚ฌ๋žŒ์—๊ฒŒ ์ „ํŒŒ๋ฅผ ๋งค๊ฐœํ•˜๋Š” ํ–‰์œ„๋ฅผ ํ•˜๋ฉด ํ˜•์‚ฌ์ฒ˜๋ฒŒ์„ ๋ฐ›์„ ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค.โ€
- ์‹ ์„ ๋™๊ฒฐํ˜ˆ์žฅ
- ๋ฉดํ—ˆ์ž๊ฒฉ ์ •์ง€
---
# SentenceTransformer based on google-bert/bert-base-multilingual-cased
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [google-bert/bert-base-multilingual-cased](https://huggingface.co/google-bert/bert-base-multilingual-cased). 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:** [google-bert/bert-base-multilingual-cased](https://huggingface.co/google-bert/bert-base-multilingual-cased) <!-- at revision 3f076fdb1ab68d5b2880cb87a0886f315b8146f8 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 tokens
- **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': False}) with Transformer model: 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 ๐Ÿค— Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
'58์„ธ ๋‚จ์ž๊ฐ€ 7์‹œ๊ฐ„ ์ „๋ถ€ํ„ฐ ์œ—๋ฐฐ๊ฐ€ ์•„ํŒŒ์„œ ๋ณ‘์›์— ์™”๋‹ค. ํ‰์†Œ์— ์•Œ์ฝ”์˜ฌ๊ฐ„๊ฒฝํ™”๋กœ ์น˜๋ฃŒ๋ฅผ ๋ฐ›๊ณ  ์žˆ์œผ๋ฉฐ ์†Œํ™”๊ถค์–‘์— ์˜ํ•œ ์ฒœ๊ณต์œผ๋กœ ์ˆ˜์ˆ ์„ ๋ฐ›์„ ์˜ˆ์ •์ด๋‹ค. ํ˜ˆ์•• 130/90 mmHg, ๋งฅ๋ฐ• 95ํšŒ/๋ถ„, ํ˜ธํก 22ํšŒ/๋ถ„, ์ฒด์˜จ 37.5โ„ƒ์ด๋‹ค. ๋ฐฐ ์ „์ฒด๊ฐ€ ๋”ฑ๋”ฑํ•˜๊ณ  ๋ฐฐ์— ์••ํ†ต๊ณผ ๋ฐ˜๋™์••ํ†ต์ด ์žˆ๋‹ค. ํ˜ˆ์•ก๊ฒ€์‚ฌ ๊ฒฐ๊ณผ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค. ์ˆ˜์ˆ  ์ „ ํˆฌ์—ฌํ•ด์•ผ ํ•  ์ œ์ œ๋Š”?ํ˜ˆ์ƒ‰์†Œ 10.3 g/dL, ๋ฐฑํ˜ˆ๊ตฌ 22,000/mm^3, ํ˜ˆ์†ŒํŒ 120,000/mm^3 ํ”„๋กœํŠธ๋กฌ๋นˆ์‹œ๊ฐ„ 20์ดˆ(์ฐธ๊ณ ์น˜, 12.7๏ฝž15.4) ํ™œ์„ฑํ™”๋ถ€๋ถ„ํŠธ๋กฌ๋ณดํ”Œ๋ผ์Šคํ‹ด์‹œ๊ฐ„ 30์ดˆ(์ฐธ๊ณ ์น˜, 26.3๏ฝž39.4) ์ด๋‹จ๋ฐฑ์งˆ 6.5 g/dL, ์•Œ๋ถ€๋ฏผ 3.0 g/dL,์ด๋นŒ๋ฆฌ๋ฃจ๋นˆ 3.5 mg/dL, ',
'์‹ ์„ ๋™๊ฒฐํ˜ˆ์žฅ',
'๋ฉดํ—ˆ์ž๊ฒฉ ์ •์ง€',
]
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: 1,890 training samples
* Columns: <code>query</code> and <code>answer</code>
* Approximate statistics based on the first 1000 samples:
| | query | answer |
|:--------|:-------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 11 tokens</li><li>mean: 112.75 tokens</li><li>max: 316 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 8.62 tokens</li><li>max: 33 tokens</li></ul> |
* Samples:
| query | answer |
|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------|
| <code>ํ•ญ๋ฌธ์•• ์ธก์ • ๊ฒ€์‚ฌ์—์„œ ํ•ญ๋ฌธ ์••๋ ฅ์ด ์ฆ๊ฐ€ํ•˜๋Š” ๊ฒฝ์šฐ๋Š”?</code> | <code>ํ•ญ๋ฌธ์—ด์ฐฝ(anal fissure)</code> |
| <code>๋ณต๋ถ€๋Œ€๋™๋งฅ(abdominal aorta) ์—์„œ ์ฒ˜์Œ ๋ถ„์ง€(first branch) ๋˜๋Š” ๋™๋งฅ์€?</code> | <code>๋Œ์ž˜๋ก์ฐฝ์ž๋™๋งฅ(ileocolic artery)</code> |
| <code>58์„ธ ๋‚จ์ž๊ฐ€ ๋Œ€๋Ÿ‰ ์žฅ์ ˆ์ œ ํ›„ ์งง์€์ฐฝ์ž์ฆํ›„๊ตฐ(short bowel syndrome) ์œผ๋กœ 4๊ฐœ์›” ๊ฐ„ ์™„์ „๋น„๊ฒฝ๊ตฌ<br>์˜์–‘์š”๋ฒ•์„ ๋ฐ›๊ณ  ์žˆ๋Š” ์ค‘์ด๋‹ค. ์ฑ„ํ˜ˆ ํ›„ ํ”ผ๊ฐ€ ์ž˜ ๋ฉŽ์ง€ ์•Š์•˜๋‹ค. ํ˜ˆ์•ก๊ฒ€์‚ฌ ๊ฒฐ๊ณผ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค.<br>๊ฒฐํ•์ด ์˜์‹ฌ๋˜๋Š” ๊ฒƒ์€?<br>ํ˜ˆ์ƒ‰์†Œ 13.5 g/dL, ๋ฐฑํ˜ˆ๊ตฌ 4,500/mm^3, ํ˜ˆ์†ŒํŒ 220,000/mm^3 <br>์•Œ๋ถ€๋ฏผ 3.7 g/dL, ์ด ๋นŒ๋ฆฌ๋ฃจ๋นˆ 1.0 mg/dL, ์•Œ์นผ๋ฆฌ ์ธ์‚ฐ๋ถ„ํ•ดํšจ์†Œ(ALP) 90 U/L,<br>์•„์ŠคํŒŒ๋ฅดํ…Œ์ดํŠธ ์•„๋ฏธ๋…ธ์ „๋‹ฌํšจ์†Œ(AST) 22 U/L, ์•Œ๋ผ๋‹Œ ์•„๋ฏธ๋…ธ์ „๋‹ฌํšจ์†Œ(ALT) 16 U/L,<br>ํ”„๋กœํŠธ๋กฌ๋นˆ์‹œ๊ฐ„ 30.5์ดˆ (์ฐธ๊ณ ์น˜, 12.7~15.4),<br>ํ™œ์„ฑํ™”๋ถ€๋ถ„ํŠธ๋กฌ๋ณดํ”Œ๋ผ์Šคํ‹ด์‹œ๊ฐ„ 34.5์ดˆ (์ฐธ๊ณ ์น˜, 26.3~39.4) </code> | <code>ํŠธ๋กฌ๋นˆ</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"
}
```
### Evaluation Dataset
#### Unnamed Dataset
* Size: 164 evaluation samples
* Columns: <code>query</code> and <code>answer</code>
* Approximate statistics based on the first 1000 samples:
| | query | answer |
|:--------|:-------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 18 tokens</li><li>mean: 153.24 tokens</li><li>max: 369 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 9.71 tokens</li><li>max: 40 tokens</li></ul> |
* Samples:
| query | answer |
|:-----------------------------------------------------------------------------------------------------------|:----------------------|
| <code>๊ด‘์—ญ์‹œ ์†Œ์žฌ ๋Œ€ํ•™๋ณ‘์›์— ์†Œ์†๋œ ๋‚ด๊ณผ ์ „๋ฌธ์˜ A๊ฐ€ ์ฝœ๋ ˆ๋ผ ํ™˜์ž๋ฅผ ์ง„๋‹จํ–ˆ๋‹ค. A๊ฐ€ ํ•  ์กฐ์น˜๋Š”?</code> | <code>๋ณ‘์›์žฅ์—๊ฒŒ ๋ณด๊ณ </code> |
| <code>A๋Š” ์ œ1๊ธ‰ ๊ฐ์—ผ๋ณ‘์œผ๋กœ ์ง„๋‹จ์„ ๋ฐ›์•˜๋‹ค. B๋Š” ๋งˆ์Šคํฌ๋ฅผ ์ฐฉ์šฉํ•˜์ง€ ์•Š์€ ์ฑ„ A์™€ ๋ฐ€์ ‘ํ•˜๊ฒŒ ์ ‘์ด‰ํ–ˆ๋‹ค. B๋Š” ์ฆ์ƒ์ด ์—†๋‹ค. ์—ญํ•™์กฐ์‚ฌ๊ด€์€ ์ด ๋‹จ๊ณ„์—์„œ B๋ฅผ ๋ฌด์—‡์œผ๋กœ ๋ถ„๋ฅ˜ํ•˜๋Š”๊ฐ€?</code> | <code>๊ฐ์—ผ๋ณ‘ ์˜์‹ฌ์ž</code> |
| <code>๊ฒ€์—ญ์†Œ ๋‚ด ๊ฒฉ๋ฆฌ๋ณ‘๋™์— ๊ฒฉ๋ฆฌ๋˜์–ด ์žˆ๋˜ ์ฝœ๋ ˆ๋ผ ํ™˜์ž A์˜ ๊ฐ์—ผ๋ ฅ์ด ์—†์–ด์ง„ ๊ฒƒ์ด ํ™•์ธ๋˜์—ˆ๋‹ค. A์— ๋Œ€ํ•œ ์กฐ์น˜๋Š”?</code> | <code>๊ฒฉ๋ฆฌ ํ•ด์ œ</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
- `eval_strategy`: epoch
- `per_device_eval_batch_size`: 16
- `learning_rate`: 3e-05
- `warmup_ratio`: 0.1
- `fp16`: True
- `load_best_model_at_end`: True
- `ddp_find_unused_parameters`: False
#### 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`: 8
- `per_device_eval_batch_size`: 16
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `learning_rate`: 3e-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`: 3
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `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`: False
- `fp16`: True
- `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`: False
- `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
- `eval_on_start`: False
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: proportional
</details>
### Training Logs
| Epoch | Step | Training Loss | loss |
|:------:|:----:|:-------------:|:------:|
| 0.1055 | 25 | 2.4397 | - |
| 0.2110 | 50 | 1.986 | - |
| 0.3165 | 75 | 1.881 | - |
| 0.4219 | 100 | 1.8105 | - |
| 0.5274 | 125 | 1.7378 | - |
| 0.6329 | 150 | 1.5942 | - |
| 0.7384 | 175 | 1.4586 | - |
| 0.8439 | 200 | 1.3904 | - |
| 0.9494 | 225 | 1.4707 | - |
| 1.0 | 237 | - | 1.3109 |
| 1.0549 | 250 | 1.234 | - |
| 1.1603 | 275 | 1.1867 | - |
| 1.2658 | 300 | 1.0103 | - |
| 1.3713 | 325 | 1.088 | - |
| 1.4768 | 350 | 1.1066 | - |
| 1.5823 | 375 | 1.049 | - |
| 1.6878 | 400 | 1.0639 | - |
| 1.7932 | 425 | 1.1133 | - |
| 1.8987 | 450 | 0.9188 | - |
| 2.0 | 474 | - | 1.0434 |
### Framework Versions
- Python: 3.10.14
- Sentence Transformers: 3.0.1
- Transformers: 4.42.2
- PyTorch: 2.3.0
- Accelerate: 0.31.0
- 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",
}
```
#### 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.*
-->