---
language:
- en
library_name: sentence-transformers
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:67190
- loss:AdaptiveLayerLoss
- loss:MultipleNegativesRankingLoss
base_model: microsoft/deberta-v3-small
datasets:
- stanfordnlp/snli
metrics:
- cosine_accuracy
- cosine_accuracy_threshold
- cosine_f1
- cosine_f1_threshold
- cosine_precision
- cosine_recall
- cosine_ap
- dot_accuracy
- dot_accuracy_threshold
- dot_f1
- dot_f1_threshold
- dot_precision
- dot_recall
- dot_ap
- manhattan_accuracy
- manhattan_accuracy_threshold
- manhattan_f1
- manhattan_f1_threshold
- manhattan_precision
- manhattan_recall
- manhattan_ap
- euclidean_accuracy
- euclidean_accuracy_threshold
- euclidean_f1
- euclidean_f1_threshold
- euclidean_precision
- euclidean_recall
- euclidean_ap
- max_accuracy
- max_accuracy_threshold
- max_f1
- max_f1_threshold
- max_precision
- max_recall
- max_ap
widget:
- source_sentence: A person in a red shirt is mowing the grass with a green riding
mower.
sentences:
- A person in red is moving grass on a John Deer motor.
- An angry military veteran watches as people protest the war.
- A man is sitting on a truck.
- source_sentence: Some dogs are running on a deserted beach.
sentences:
- daddy taught her
- There are multiple dogs present.
- a woman at a beach
- source_sentence: Two street people and a dog sitting on the ground and one is holding
an "out of luck" sign.
sentences:
- A person biking.
- The man and woman are married.
- the dog is a chihuahua
- source_sentence: One tan girl with a wool hat is running and leaning over an object,
while another person in a wool hat is sitting on the ground.
sentences:
- A tan girl runs leans over an object
- A man and his daughter are petting a pony.
- A man with a baby is petting a pony.
- source_sentence: These girls are having a great time looking for seashells.
sentences:
- The girls are happy.
- Two woman are trying to finish orders from a doctor
- A girl is standing outside.
pipeline_tag: sentence-similarity
model-index:
- name: SentenceTransformer based on microsoft/deberta-v3-small
results:
- task:
type: binary-classification
name: Binary Classification
dataset:
name: Unknown
type: unknown
metrics:
- type: cosine_accuracy
value: 0.6652580742529429
name: Cosine Accuracy
- type: cosine_accuracy_threshold
value: 0.6691544055938721
name: Cosine Accuracy Threshold
- type: cosine_f1
value: 0.7050935184095989
name: Cosine F1
- type: cosine_f1_threshold
value: 0.5757889747619629
name: Cosine F1 Threshold
- type: cosine_precision
value: 0.5903092377388222
name: Cosine Precision
- type: cosine_recall
value: 0.8752920560747663
name: Cosine Recall
- type: cosine_ap
value: 0.7023886827641951
name: Cosine Ap
- type: dot_accuracy
value: 0.6308481738605494
name: Dot Accuracy
- type: dot_accuracy_threshold
value: 127.05267333984375
name: Dot Accuracy Threshold
- type: dot_f1
value: 0.6983614124163396
name: Dot F1
- type: dot_f1_threshold
value: 101.77250671386719
name: Dot F1 Threshold
- type: dot_precision
value: 0.5772605875619993
name: Dot Precision
- type: dot_recall
value: 0.8837616822429907
name: Dot Recall
- type: dot_ap
value: 0.6558335483108544
name: Dot Ap
- type: manhattan_accuracy
value: 0.6675218834892847
name: Manhattan Accuracy
- type: manhattan_accuracy_threshold
value: 210.99388122558594
name: Manhattan Accuracy Threshold
- type: manhattan_f1
value: 0.7107997100748973
name: Manhattan F1
- type: manhattan_f1_threshold
value: 252.65306091308594
name: Manhattan F1 Threshold
- type: manhattan_precision
value: 0.6060980634528225
name: Manhattan Precision
- type: manhattan_recall
value: 0.8592289719626168
name: Manhattan Recall
- type: manhattan_ap
value: 0.709424985473672
name: Manhattan Ap
- type: euclidean_accuracy
value: 0.6619378207063085
name: Euclidean Accuracy
- type: euclidean_accuracy_threshold
value: 11.227606773376465
name: Euclidean Accuracy Threshold
- type: euclidean_f1
value: 0.7073199115559177
name: Euclidean F1
- type: euclidean_f1_threshold
value: 12.850802421569824
name: Euclidean F1 Threshold
- type: euclidean_precision
value: 0.587928032501451
name: Euclidean Precision
- type: euclidean_recall
value: 0.8875584112149533
name: Euclidean Recall
- type: euclidean_ap
value: 0.7037559902823934
name: Euclidean Ap
- type: max_accuracy
value: 0.6675218834892847
name: Max Accuracy
- type: max_accuracy_threshold
value: 210.99388122558594
name: Max Accuracy Threshold
- type: max_f1
value: 0.7107997100748973
name: Max F1
- type: max_f1_threshold
value: 252.65306091308594
name: Max F1 Threshold
- type: max_precision
value: 0.6060980634528225
name: Max Precision
- type: max_recall
value: 0.8875584112149533
name: Max Recall
- type: max_ap
value: 0.709424985473672
name: Max Ap
---
# SentenceTransformer based on microsoft/deberta-v3-small
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [microsoft/deberta-v3-small](https://huggingface.co/microsoft/deberta-v3-small) on the [stanfordnlp/snli](https://huggingface.co/datasets/stanfordnlp/snli) 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:** [microsoft/deberta-v3-small](https://huggingface.co/microsoft/deberta-v3-small)
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 tokens
- **Similarity Function:** Cosine Similarity
- **Training Dataset:**
- [stanfordnlp/snli](https://huggingface.co/datasets/stanfordnlp/snli)
- **Language:** en
### 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: DebertaV2Model
(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("bobox/DeBERTaV3-small-ST-AdaptiveLayer-3L-ep2")
# Run inference
sentences = [
'These girls are having a great time looking for seashells.',
'The girls are happy.',
'A girl is standing outside.',
]
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
#### Binary Classification
* Evaluated with [BinaryClassificationEvaluator
](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.BinaryClassificationEvaluator)
| Metric | Value |
|:-----------------------------|:-----------|
| cosine_accuracy | 0.6653 |
| cosine_accuracy_threshold | 0.6692 |
| cosine_f1 | 0.7051 |
| cosine_f1_threshold | 0.5758 |
| cosine_precision | 0.5903 |
| cosine_recall | 0.8753 |
| cosine_ap | 0.7024 |
| dot_accuracy | 0.6308 |
| dot_accuracy_threshold | 127.0527 |
| dot_f1 | 0.6984 |
| dot_f1_threshold | 101.7725 |
| dot_precision | 0.5773 |
| dot_recall | 0.8838 |
| dot_ap | 0.6558 |
| manhattan_accuracy | 0.6675 |
| manhattan_accuracy_threshold | 210.9939 |
| manhattan_f1 | 0.7108 |
| manhattan_f1_threshold | 252.6531 |
| manhattan_precision | 0.6061 |
| manhattan_recall | 0.8592 |
| manhattan_ap | 0.7094 |
| euclidean_accuracy | 0.6619 |
| euclidean_accuracy_threshold | 11.2276 |
| euclidean_f1 | 0.7073 |
| euclidean_f1_threshold | 12.8508 |
| euclidean_precision | 0.5879 |
| euclidean_recall | 0.8876 |
| euclidean_ap | 0.7038 |
| max_accuracy | 0.6675 |
| max_accuracy_threshold | 210.9939 |
| max_f1 | 0.7108 |
| max_f1_threshold | 252.6531 |
| max_precision | 0.6061 |
| max_recall | 0.8876 |
| **max_ap** | **0.7094** |
## Training Details
### Training Dataset
#### stanfordnlp/snli
* Dataset: [stanfordnlp/snli](https://huggingface.co/datasets/stanfordnlp/snli) at [cdb5c3d](https://huggingface.co/datasets/stanfordnlp/snli/tree/cdb5c3d5eed6ead6e5a341c8e56e669bb666725b)
* Size: 67,190 training samples
* Columns: sentence1
, sentence2
, and label
* Approximate statistics based on the first 1000 samples:
| | sentence1 | sentence2 | label |
|:--------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:-----------------------------|
| type | string | string | int |
| details |
Without a placebo group, we still won't know if any of the treatments are better than nothing and therefore worth giving.
| It is necessary to use a controlled method to ensure the treatments are worthwhile.
| 0
|
| It was conducted in silence.
| It was done silently.
| 0
|
| oh Lewisville any decent food in your cafeteria up there
| Is there any decent food in your cafeteria up there in Lewisville?
| 0
|
* Loss: [AdaptiveLayerLoss
](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#adaptivelayerloss) with these parameters:
```json
{
"loss": "MultipleNegativesRankingLoss",
"n_layers_per_step": 3,
"last_layer_weight": 1,
"prior_layers_weight": 0.3,
"kl_div_weight": 1,
"kl_temperature": 1
}
```
### Evaluation Dataset
#### stanfordnlp/snli
* Dataset: [stanfordnlp/snli](https://huggingface.co/datasets/stanfordnlp/snli) at [cdb5c3d](https://huggingface.co/datasets/stanfordnlp/snli/tree/cdb5c3d5eed6ead6e5a341c8e56e669bb666725b)
* Size: 6,626 evaluation samples
* Columns: premise
, hypothesis
, and label
* Approximate statistics based on the first 1000 samples:
| | premise | hypothesis | label |
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:------------------------------------------------|
| type | string | string | int |
| details | This church choir sings to the masses as they sing joyous songs from the book at a church.
| The church has cracks in the ceiling.
| 0
|
| This church choir sings to the masses as they sing joyous songs from the book at a church.
| The church is filled with song.
| 1
|
| A woman with a green headscarf, blue shirt and a very big grin.
| The woman is young.
| 0
|
* Loss: [AdaptiveLayerLoss
](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#adaptivelayerloss) with these parameters:
```json
{
"loss": "MultipleNegativesRankingLoss",
"n_layers_per_step": 3,
"last_layer_weight": 1,
"prior_layers_weight": 0.3,
"kl_div_weight": 1,
"kl_temperature": 1
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 45
- `per_device_eval_batch_size`: 22
- `learning_rate`: 3e-06
- `weight_decay`: 1e-09
- `num_train_epochs`: 2
- `lr_scheduler_type`: cosine
- `warmup_ratio`: 0.5
- `save_safetensors`: False
- `fp16`: True
- `push_to_hub`: True
- `hub_model_id`: bobox/DeBERTaV3-small-ST-AdaptiveLayer-3L-ep2-n
- `hub_strategy`: checkpoint
- `batch_sampler`: no_duplicates
#### All Hyperparameters