---
language:
- tr
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:482091
- loss:MatryoshkaLoss
- loss:MultipleNegativesRankingLoss
- loss:CoSENTLoss
base_model: Alibaba-NLP/gte-multilingual-base
widget:
- source_sentence: Ya da dışarı çıkıp yürü ya da biraz koşun. Bunu düzenli olarak
yapmıyorum ama Washington bunu yapmak için harika bir yer.
sentences:
- “Washington's yürüyüş ya da koşu için harika bir yer.”
- H-2A uzaylılar Amerika Birleşik Devletleri'nde zaman kısa süreleri var.
- “Washington'da düzenli olarak yürüyüşe ya da koşuya çıkıyorum.”
- source_sentence: Orta yaylalar ve güney kıyıları arasındaki kontrast daha belirgin
olamazdı.
sentences:
- İşitme Yardımı Uyumluluğu Müzakere Kuralları Komitesi, Federal İletişim Komisyonu'nun
bir ürünüdür.
- Dağlık ve sahil arasındaki kontrast kolayca işaretlendi.
- Kontrast işaretlenemedi.
- source_sentence: Bir 1997 Henry J. Kaiser Aile Vakfı anket yönetilen bakım planlarında
Amerikalılar temelde kendi bakımı ile memnun olduğunu bulundu.
sentences:
- Kaplanları takip ederken çok sessiz olmalısın.
- Henry Kaiser vakfı insanların sağlık hizmetlerinden hoşlandığını gösteriyor.
- Henry Kaiser Vakfı insanların sağlık hizmetlerinden nefret ettiğini gösteriyor.
- source_sentence: Eminim yapmışlardır.
sentences:
- Eminim öyle yapmışlardır.
- Batı Teksas'ta 100 10 dereceydi.
- Eminim yapmamışlardır.
- source_sentence: Ve gerçekten, baba haklıydı, oğlu zaten her şeyi tecrübe etmişti,
her şeyi denedi ve daha az ilgileniyordu.
sentences:
- Oğlu her şeye olan ilgisini kaybediyordu.
- Pek bir şey yapmadım.
- Baba oğlunun tecrübe için hala çok şey olduğunu biliyordu.
datasets:
- emrecan/all-nli-tr
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
- cosine_accuracy
- pearson_cosine
- spearman_cosine
model-index:
- name: SentenceTransformer based on Alibaba-NLP/gte-multilingual-base
results:
- task:
type: triplet
name: Triplet
dataset:
name: all nli tr test
type: all-nli-tr-test
metrics:
- type: cosine_accuracy
value: 0.8966145437983908
name: Cosine Accuracy
- type: cosine_accuracy
value: 0.9351753453772582
name: Cosine Accuracy
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: sts test
type: sts-test
metrics:
- type: pearson_cosine
value: 0.8043925123766598
name: Pearson Cosine
- type: spearman_cosine
value: 0.804133282756889
name: Spearman Cosine
- type: pearson_cosine
value: 0.8133873820848544
name: Pearson Cosine
- type: spearman_cosine
value: 0.8199552151367876
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: sts22 test
type: sts22-test
metrics:
- type: pearson_cosine
value: 0.647912337747937
name: Pearson Cosine
- type: spearman_cosine
value: 0.6694072470896322
name: Spearman Cosine
- type: pearson_cosine
value: 0.6514085062457564
name: Pearson Cosine
- type: spearman_cosine
value: 0.6827342891126081
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: sts dev gte multilingual base
type: sts-dev-gte-multilingual-base
metrics:
- type: pearson_cosine
value: 0.838717139426684
name: Pearson Cosine
- type: spearman_cosine
value: 0.8428367492381358
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: sts test gte multilingual base
type: sts-test-gte-multilingual-base
metrics:
- type: pearson_cosine
value: 0.8133873820848544
name: Pearson Cosine
- type: spearman_cosine
value: 0.8199552151367876
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb dev 768
type: stsb-dev-768
metrics:
- type: pearson_cosine
value: 0.870311456444647
name: Pearson Cosine
- type: spearman_cosine
value: 0.8747522169942328
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb dev 512
type: stsb-dev-512
metrics:
- type: pearson_cosine
value: 0.8696934286998554
name: Pearson Cosine
- type: spearman_cosine
value: 0.8753487201891684
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb dev 256
type: stsb-dev-256
metrics:
- type: pearson_cosine
value: 0.8644706498119142
name: Pearson Cosine
- type: spearman_cosine
value: 0.873468734899321
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb dev 128
type: stsb-dev-128
metrics:
- type: pearson_cosine
value: 0.8591309130178328
name: Pearson Cosine
- type: spearman_cosine
value: 0.8700377378574327
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb dev 64
type: stsb-dev-64
metrics:
- type: pearson_cosine
value: 0.8479124810212979
name: Pearson Cosine
- type: spearman_cosine
value: 0.8655596653561272
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb test 768
type: stsb-test-768
metrics:
- type: pearson_cosine
value: 0.8455412308380735
name: Pearson Cosine
- type: spearman_cosine
value: 0.8535290217691063
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb test 512
type: stsb-test-512
metrics:
- type: pearson_cosine
value: 0.8464773608783734
name: Pearson Cosine
- type: spearman_cosine
value: 0.8553900248212041
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb test 256
type: stsb-test-256
metrics:
- type: pearson_cosine
value: 0.8443046458551826
name: Pearson Cosine
- type: spearman_cosine
value: 0.8550098621393595
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb test 128
type: stsb-test-128
metrics:
- type: pearson_cosine
value: 0.8363964421208214
name: Pearson Cosine
- type: spearman_cosine
value: 0.8511193715667303
name: Spearman Cosine
- task:
type: semantic-similarity
name: Semantic Similarity
dataset:
name: stsb test 64
type: stsb-test-64
metrics:
- type: pearson_cosine
value: 0.8235450515966374
name: Pearson Cosine
- type: spearman_cosine
value: 0.8460761238725121
name: Spearman Cosine
---
# SentenceTransformer based on Alibaba-NLP/gte-multilingual-base
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Alibaba-NLP/gte-multilingual-base](https://huggingface.co/Alibaba-NLP/gte-multilingual-base) on the [all-nli-tr](https://huggingface.co/datasets/emrecan/all-nli-tr) dataset. 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:** [Alibaba-NLP/gte-multilingual-base](https://huggingface.co/Alibaba-NLP/gte-multilingual-base)
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 dimensions
- **Similarity Function:** Cosine Similarity
- **Training Dataset:**
- [all-nli-tr](https://huggingface.co/datasets/emrecan/all-nli-tr)
- **Language:** tr
### 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: NewModel
(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("sentence_transformers_model_id")
# Run inference
sentences = [
'Ve gerçekten, baba haklıydı, oğlu zaten her şeyi tecrübe etmişti, her şeyi denedi ve daha az ilgileniyordu.',
'Oğlu her şeye olan ilgisini kaybediyordu.',
'Baba oğlunun tecrübe için hala çok şey olduğunu biliyordu.',
]
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
#### Triplet
* Dataset: `all-nli-tr-test`
* Evaluated with [TripletEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
| Metric | Value |
|:--------------------|:-----------|
| **cosine_accuracy** | **0.8966** |
#### Semantic Similarity
* Datasets: `sts-test`, `sts22-test`, `sts-dev-gte-multilingual-base`, `sts-test-gte-multilingual-base`, `sts-test`, `sts22-test`, `stsb-dev-768`, `stsb-dev-512`, `stsb-dev-256`, `stsb-dev-128`, `stsb-dev-64`, `stsb-test-768`, `stsb-test-512`, `stsb-test-256`, `stsb-test-128` and `stsb-test-64`
* Evaluated with [EmbeddingSimilarityEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
| Metric | sts-test | sts22-test | sts-dev-gte-multilingual-base | sts-test-gte-multilingual-base | stsb-dev-768 | stsb-dev-512 | stsb-dev-256 | stsb-dev-128 | stsb-dev-64 | stsb-test-768 | stsb-test-512 | stsb-test-256 | stsb-test-128 | stsb-test-64 |
|:--------------------|:---------|:-----------|:------------------------------|:-------------------------------|:-------------|:-------------|:-------------|:-------------|:------------|:--------------|:--------------|:--------------|:--------------|:-------------|
| pearson_cosine | 0.8134 | 0.6514 | 0.8387 | 0.8134 | 0.8703 | 0.8697 | 0.8645 | 0.8591 | 0.8479 | 0.8455 | 0.8465 | 0.8443 | 0.8364 | 0.8235 |
| **spearman_cosine** | **0.82** | **0.6827** | **0.8428** | **0.82** | **0.8748** | **0.8753** | **0.8735** | **0.87** | **0.8656** | **0.8535** | **0.8554** | **0.855** | **0.8511** | **0.8461** |
#### Triplet
* Dataset: `all-nli-tr-test`
* Evaluated with [TripletEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
| Metric | Value |
|:--------------------|:-----------|
| **cosine_accuracy** | **0.9352** |
## Training Details
### Training Dataset
#### all-nli-tr
* Dataset: [all-nli-tr](https://huggingface.co/datasets/emrecan/all-nli-tr) at [daeabfb](https://huggingface.co/datasets/emrecan/all-nli-tr/tree/daeabfbc01f82757ab998bd23ce0ddfceaa5e24d)
* Size: 482,091 training samples
* Columns: sentence1, sentence2, and score
* Approximate statistics based on the first 1000 samples:
| | sentence1 | sentence2 | score |
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:---------------------------------------------------------------|
| type | string | string | float |
| details |
Bir uçak kalkıyor. | Bir hava uçağı kalkıyor. | 5.0 |
| Bir adam büyük bir flüt çalıyor. | Bir adam flüt çalıyor. | 3.8 |
| Bir adam pizzaya rendelenmiş peynir yayıyor. | Bir adam pişmemiş pizzaya rendelenmiş peynir yayıyor. | 3.8 |
* Loss: [MatryoshkaLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "CoSENTLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64
],
"matryoshka_weights": [
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Evaluation Dataset
#### all-nli-tr
* Dataset: [all-nli-tr](https://huggingface.co/datasets/emrecan/all-nli-tr) at [daeabfb](https://huggingface.co/datasets/emrecan/all-nli-tr/tree/daeabfbc01f82757ab998bd23ce0ddfceaa5e24d)
* Size: 6,567 evaluation samples
* Columns: sentence1, sentence2, and score
* Approximate statistics based on the first 1000 samples:
| | sentence1 | sentence2 | score |
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:--------------------------------------------------------------|
| type | string | string | float |
| details | Şapkalı bir adam dans ediyor. | Sert şapka takan bir adam dans ediyor. | 5.0 |
| Küçük bir çocuk ata biniyor. | Bir çocuk ata biniyor. | 4.75 |
| Bir adam yılana fare yediriyor. | Adam yılana fare yediriyor. | 5.0 |
* Loss: [MatryoshkaLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "CoSENTLoss",
"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`: steps
- `per_device_train_batch_size`: 32
- `per_device_eval_batch_size`: 32
- `learning_rate`: 1e-05
- `weight_decay`: 0.01
- `num_train_epochs`: 10
- `warmup_ratio`: 0.1
- `warmup_steps`: 144
- `bf16`: True
#### All Hyperparameters