metadata

base_model: microsoft/deberta-v3-small
datasets:
  - tals/vitaminc
language:
  - en
library_name: sentence-transformers
metrics:
  - pearson_cosine
  - spearman_cosine
  - pearson_manhattan
  - spearman_manhattan
  - pearson_euclidean
  - spearman_euclidean
  - pearson_dot
  - spearman_dot
  - pearson_max
  - spearman_max
  - 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
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:225247
  - loss:CachedGISTEmbedLoss
widget:
  - source_sentence: what is exfo toolbox
    sentences:
      - >-
        Eye dilation from eye drops used for examination of the eye usually
        lasts from 4 to 24 hours, depending upon the strength of the drop and
        upon the individual patient.
      - >-
        Garden Grove is a city in northern Orange County in the U.S. state of
        California, 34 miles (55 km) south of Los Angeles. The population was
        170,883 at the 2010 United States Census. State Route 22, also known as
        the Garden Grove Freeway, passes through the city in an east-west
        direction.
      - >-
        EXFO ToolBox Office is a product that offers you a collection of viewers
        and analyzers. It enables you to manage and analyze results acquired
        from fiber optic test modules and instruments.
  - source_sentence: >-
      More than 273 people have died from the 2019-20 coronavirus outside
      mainland China .
    sentences:
      - >-
        More than 3,700 people have died : around 3,100 in mainland China and
        around 550 in all other countries combined .
      - >-
        More than 3,200 people have died : almost 3,000 in mainland China and
        around 275 in other countries .
      - more than 4,900 deaths have been attributed to COVID-19 .
  - source_sentence: >-
      Ultrasound, a diagnostic technology, uses high-frequency vibrations
      transmitted into any tissue in contact with the transducer.
    sentences:
      - >-
        What diagnostic technology uses high-frequency vibrations transmitted
        into any tissue in contact with the transducer?
      - The abnormal cells cannot carry oxygen properly and can get stuck where?
      - What type of organism is a bacteria?
  - source_sentence: >-
      When you add moles of gas to a baloon by blowing it up, the volume
      increases.
    sentences:
      - What shape is the lens of the eye?
      - >-
        What happens to the volume of a balloon when you add moles of gas to it
        by blowing up?
      - >-
        Most turtle bodies are covered by a special bony or cartilaginous shell
        developed from their what?
  - source_sentence: >-
      What was the name of eleven rulers of the 19th and 20th Egyptian
      dynasties?
    sentences:
      - >-
        Airlines Yugoslavia 1968 - 1968 Renamed ^ Comments : Aviogenex was
        formed on 21May1968 as Genex Airlines. Restarted under current name on
        30Apr1969 & liquidated in Feb2015 ^ Genealogy : Genex Airlines
        >Aviogenex 1968 - 1986 Renamed ^ Comments : Adria Airways was formed on
        14Mar1961 & operations started on 30Jun1961 as Adria Airways, renamed to
        Inex in 1968 and back to Adria again in 1986. National airline of
        Slovenia ^ Genealogy : Adria Airways >Inex Adria Airways >Adria Airways
        JAT (Jugoslovenski Aerotransport) 1947 - 2003 Renamed ^ Comments : Air
        Serbia was founded as Aeroput on 17Jun1927, renamed to JAT on 01Apr1947.
        Started ops on 15Apr1947, Renamed again on 08Aug2003 to JAT Airways &
        reformed as Air Serbia on 26Oct2013 ^ Genealogy : Aeroput >JAT
        (Jugoslovenski Aerotransport) >JAT Airways >Air Serbia Jugoslovenski
        Aerotransport
      - >-
        List of Rulers of Ancient Egypt and Nubia | Lists of Rulers | Heilbrunn
        Timeline of Art History | The Metropolitan Museum of Art The
        Metropolitan Museum of Art List of Rulers of Ancient Egypt and Nubia See
        works of art 30.8.234 52.127.4 Our knowledge of the succession of
        Egyptian kings is based on kinglists kept by the ancient Egyptians
        themselves. The most famous are the Palermo Stone, which covers the
        period from the earliest dynasties to the middle of Dynasty 5; the
        Abydos Kinglist, which Seti I had carved on his temple at Abydos; and
        the Turin Canon, a papyrus that covers the period from the earliest
        dynasties to the reign of Ramesses II. All are incomplete or
        fragmentary. We also rely on the History of Egypt written by Manetho in
        the third century B.C. A priest in the temple at Heliopolis, Manetho had
        access to many original sources and it was he who divided the kings into
        the thirty dynasties we use today. It is to this structure of dynasties
        and listed kings that we now attempt to link an absolute chronology of
        dates in terms of our own calendrical system. The process is made
        difficult by the fragmentary condition of the kinglists and by
        differences in the calendrical years used at various times. Some
        astronomical observations from the ancient Egyptians have survived,
        allowing us to calculate absolute dates within a margin of error.
        Synchronisms with the other civilizations of the ancient world are also
        of limited use.
      - >-
        What is the "Jack Sprat" nursery rhyme? | Reference.com What is the
        "Jack Sprat" nursery rhyme? A: Quick Answer "Jack Sprat" is a
        traditional English nursery rhyme whose main verse says, "Jack Sprat
        could eat no fat. His wife could eat no lean. And so between them both,
        you see, they licked the platter clean." Though it was likely sung by
        children long before, "Jack Sprat" was first published around 1765 in
        the compilation "Mother Goose's Melody." Full Answer According to
        Rhymes.org, a U.K. website devoted to nursery rhyme lyrics and origins,
        the "Jack Sprat" nursery rhyme has its origins in British history. In
        one interpretation, Jack Sprat was King Charles I, who ruled England in
        the early part of the 17th century, and his wife was Queen Henrietta
        Maria. Parliament refused to finance the king's war with Spain, which
        made him lean. However, the queen fattened the coffers by levying an
        illegal war tax. In an alternative version, the "Jack Sprat" nursery
        rhyme is linked to King Richard and his brother John of the Robin Hood
        legend. Jack Sprat was King John, the usurper who tried to take over the
        crown when King Richard went off to fight in the Crusades in the 12th
        century. When King Richard was captured, John had to raise a ransom to
        rescue him, leaving the country lean. The wife was Joan, daughter of the
        Earl of Gloucester, the greedy wife of King John. However, after King
        Richard died and John became king, he had his marriage with Joan
        annulled.
model-index:
  - name: SentenceTransformer based on microsoft/deberta-v3-small
    results:
      - task:
          type: semantic-similarity
          name: Semantic Similarity
        dataset:
          name: sts test
          type: sts-test
        metrics:
          - type: pearson_cosine
            value: 0.7673854808079448
            name: Pearson Cosine
          - type: spearman_cosine
            value: 0.7776198286738142
            name: Spearman Cosine
          - type: pearson_manhattan
            value: 0.782368447545155
            name: Pearson Manhattan
          - type: spearman_manhattan
            value: 0.7720687033298573
            name: Spearman Manhattan
          - type: pearson_euclidean
            value: 0.7882638792170585
            name: Pearson Euclidean
          - type: spearman_euclidean
            value: 0.7775073687564514
            name: Spearman Euclidean
          - type: pearson_dot
            value: 0.7669147371310585
            name: Pearson Dot
          - type: spearman_dot
            value: 0.7762894632049069
            name: Spearman Dot
          - type: pearson_max
            value: 0.7882638792170585
            name: Pearson Max
          - type: spearman_max
            value: 0.7776198286738142
            name: Spearman Max
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: allNLI dev
          type: allNLI-dev
        metrics:
          - type: cosine_accuracy
            value: 0.708984375
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.8714957237243652
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.5913043478260869
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.7768557071685791
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.4738675958188153
            name: Cosine Precision
          - type: cosine_recall
            value: 0.7861271676300579
            name: Cosine Recall
          - type: cosine_ap
            value: 0.5644305887001508
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.7109375
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 674.426025390625
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.5913043478260869
            name: Dot F1
          - type: dot_f1_threshold
            value: 603.435302734375
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.4738675958188153
            name: Dot Precision
          - type: dot_recall
            value: 0.7861271676300579
            name: Dot Recall
          - type: dot_ap
            value: 0.5664868031504724
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.7109375
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 294.4728088378906
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.5935483870967742
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 401.1482849121094
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.4726027397260274
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.7976878612716763
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.5642688421649988
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.7109375
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 14.565500259399414
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.5913043478260869
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 18.60409164428711
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.4738675958188153
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.7861271676300579
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.5645557227019772
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.7109375
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 674.426025390625
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.5935483870967742
            name: Max F1
          - type: max_f1_threshold
            value: 603.435302734375
            name: Max F1 Threshold
          - type: max_precision
            value: 0.4738675958188153
            name: Max Precision
          - type: max_recall
            value: 0.7976878612716763
            name: Max Recall
          - type: max_ap
            value: 0.5664868031504724
            name: Max Ap
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: Qnli dev
          type: Qnli-dev
        metrics:
          - type: cosine_accuracy
            value: 0.6796875
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.7726649045944214
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.6925675675675677
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.7317887544631958
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.5758426966292135
            name: Cosine Precision
          - type: cosine_recall
            value: 0.8686440677966102
            name: Cosine Recall
          - type: cosine_ap
            value: 0.7302564198016936
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.67578125
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 598.0419921875
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.6912751677852348
            name: Dot F1
          - type: dot_f1_threshold
            value: 565.4718017578125
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.5722222222222222
            name: Dot Precision
          - type: dot_recall
            value: 0.8728813559322034
            name: Dot Recall
          - type: dot_ap
            value: 0.7300462025003271
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.6796875
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 404.8309020996094
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.6933333333333332
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 444.99224853515625
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.5714285714285714
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.8813559322033898
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.7369214156436785
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.6796875
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 18.790739059448242
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.6934306569343065
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 19.35132598876953
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.6089743589743589
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.8050847457627118
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.7307381840067684
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.6796875
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 598.0419921875
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.6934306569343065
            name: Max F1
          - type: max_f1_threshold
            value: 565.4718017578125
            name: Max F1 Threshold
          - type: max_precision
            value: 0.6089743589743589
            name: Max Precision
          - type: max_recall
            value: 0.8813559322033898
            name: Max Recall
          - type: max_ap
            value: 0.7369214156436785
            name: Max Ap

SentenceTransformer based on microsoft/deberta-v3-small

This is a sentence-transformers model finetuned from microsoft/deberta-v3-small. 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
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 tokens
Similarity Function: Cosine Similarity
Language: en

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: DebertaV2Model 
  (1): AdvancedWeightedPooling(
    (linear_cls): Linear(in_features=768, out_features=768, bias=True)
    (linear_mean): Linear(in_features=768, out_features=768, bias=True)
    (mha): MultiheadAttention(
      (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
    )
    (layernorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (layernorm2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (layernorm_cls): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (layernorm_mean): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
  )
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("bobox/DeBERTa3-s-CustomPoolin-v3-step1")
# Run inference
sentences = [
    'What was the name of eleven rulers of the 19th and 20th Egyptian dynasties?',
    'List of Rulers of Ancient Egypt and Nubia | Lists of Rulers | Heilbrunn Timeline of Art History | The Metropolitan Museum of Art The Metropolitan Museum of Art List of Rulers of Ancient Egypt and Nubia See works of art 30.8.234 52.127.4 Our knowledge of the succession of Egyptian kings is based on kinglists kept by the ancient Egyptians themselves. The most famous are the Palermo Stone, which covers the period from the earliest dynasties to the middle of Dynasty 5; the Abydos Kinglist, which Seti I had carved on his temple at Abydos; and the Turin Canon, a papyrus that covers the period from the earliest dynasties to the reign of Ramesses II. All are incomplete or fragmentary. We also rely on the History of Egypt written by Manetho in the third century B.C. A priest in the temple at Heliopolis, Manetho had access to many original sources and it was he who divided the kings into the thirty dynasties we use today. It is to this structure of dynasties and listed kings that we now attempt to link an absolute chronology of dates in terms of our own calendrical system. The process is made difficult by the fragmentary condition of the kinglists and by differences in the calendrical years used at various times. Some astronomical observations from the ancient Egyptians have survived, allowing us to calculate absolute dates within a margin of error. Synchronisms with the other civilizations of the ancient world are also of limited use.',
    'What is the "Jack Sprat" nursery rhyme? | Reference.com What is the "Jack Sprat" nursery rhyme? A: Quick Answer "Jack Sprat" is a traditional English nursery rhyme whose main verse says, "Jack Sprat could eat no fat. His wife could eat no lean. And so between them both, you see, they licked the platter clean." Though it was likely sung by children long before, "Jack Sprat" was first published around 1765 in the compilation "Mother Goose\'s Melody." Full Answer According to Rhymes.org, a U.K. website devoted to nursery rhyme lyrics and origins, the "Jack Sprat" nursery rhyme has its origins in British history. In one interpretation, Jack Sprat was King Charles I, who ruled England in the early part of the 17th century, and his wife was Queen Henrietta Maria. Parliament refused to finance the king\'s war with Spain, which made him lean. However, the queen fattened the coffers by levying an illegal war tax. In an alternative version, the "Jack Sprat" nursery rhyme is linked to King Richard and his brother John of the Robin Hood legend. Jack Sprat was King John, the usurper who tried to take over the crown when King Richard went off to fight in the Crusades in the 12th century. When King Richard was captured, John had to raise a ransom to rescue him, leaving the country lean. The wife was Joan, daughter of the Earl of Gloucester, the greedy wife of King John. However, after King Richard died and John became king, he had his marriage with Joan annulled.',
]
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

Semantic Similarity

Dataset: sts-test
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.7674
spearman_cosine	0.7776
pearson_manhattan	0.7824
spearman_manhattan	0.7721
pearson_euclidean	0.7883
spearman_euclidean	0.7775
pearson_dot	0.7669
spearman_dot	0.7763
pearson_max	0.7883
spearman_max	0.7776

Binary Classification

Dataset: allNLI-dev
Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.709
cosine_accuracy_threshold	0.8715
cosine_f1	0.5913
cosine_f1_threshold	0.7769
cosine_precision	0.4739
cosine_recall	0.7861
cosine_ap	0.5644
dot_accuracy	0.7109
dot_accuracy_threshold	674.426
dot_f1	0.5913
dot_f1_threshold	603.4353
dot_precision	0.4739
dot_recall	0.7861
dot_ap	0.5665
manhattan_accuracy	0.7109
manhattan_accuracy_threshold	294.4728
manhattan_f1	0.5935
manhattan_f1_threshold	401.1483
manhattan_precision	0.4726
manhattan_recall	0.7977
manhattan_ap	0.5643
euclidean_accuracy	0.7109
euclidean_accuracy_threshold	14.5655
euclidean_f1	0.5913
euclidean_f1_threshold	18.6041
euclidean_precision	0.4739
euclidean_recall	0.7861
euclidean_ap	0.5646
max_accuracy	0.7109
max_accuracy_threshold	674.426
max_f1	0.5935
max_f1_threshold	603.4353
max_precision	0.4739
max_recall	0.7977
max_ap	0.5665

Binary Classification

Dataset: Qnli-dev
Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.6797
cosine_accuracy_threshold	0.7727
cosine_f1	0.6926
cosine_f1_threshold	0.7318
cosine_precision	0.5758
cosine_recall	0.8686
cosine_ap	0.7303
dot_accuracy	0.6758
dot_accuracy_threshold	598.042
dot_f1	0.6913
dot_f1_threshold	565.4718
dot_precision	0.5722
dot_recall	0.8729
dot_ap	0.73
manhattan_accuracy	0.6797
manhattan_accuracy_threshold	404.8309
manhattan_f1	0.6933
manhattan_f1_threshold	444.9922
manhattan_precision	0.5714
manhattan_recall	0.8814
manhattan_ap	0.7369
euclidean_accuracy	0.6797
euclidean_accuracy_threshold	18.7907
euclidean_f1	0.6934
euclidean_f1_threshold	19.3513
euclidean_precision	0.609
euclidean_recall	0.8051
euclidean_ap	0.7307
max_accuracy	0.6797
max_accuracy_threshold	598.042
max_f1	0.6934
max_f1_threshold	565.4718
max_precision	0.609
max_recall	0.8814
max_ap	0.7369

Training Details

Evaluation Dataset

vitaminc-pairs

Dataset: vitaminc-pairs at be6febb
Size: 128 evaluation samples
Columns: claim and evidence
Approximate statistics based on the first 128 samples:
claim evidence
type string string
details
min: 9 tokens
mean: 21.42 tokens
max: 41 tokens

min: 11 tokens
mean: 35.55 tokens
max: 79 tokens

	claim	evidence
type	string	string
details	min: 9 tokens mean: 21.42 tokens max: 41 tokens	min: 11 tokens mean: 35.55 tokens max: 79 tokens

Samples:

claim	evidence
`Dragon Con had over 5000 guests .`	`Among the more than 6000 guests and musical performers at the 2009 convention were such notables as Patrick Stewart , William Shatner , Leonard Nimoy , Terry Gilliam , Bruce Boxleitner , James Marsters , and Mary McDonnell .`
`COVID-19 has reached more than 185 countries .`	`As of , more than cases of COVID-19 have been reported in more than 190 countries and 200 territories , resulting in more than deaths .`
`In March , Italy had 3.6x times more cases of coronavirus than China .`	`As of 12 March , among nations with at least one million citizens , Italy has the world 's highest per capita rate of positive coronavirus cases at 206.1 cases per million people ( 3.6x times the rate of China ) and is the country with the second-highest number of positive cases as well as of deaths in the world , after China .`

Loss: CachedGISTEmbedLoss with these parameters:

{'guide': 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': 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()
), 'temperature': 0.025}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 100
per_device_eval_batch_size: 256
gradient_accumulation_steps: 2
lr_scheduler_type: cosine_with_min_lr
lr_scheduler_kwargs: {'num_cycles': 0.5, 'min_lr': 1.6666666666666667e-05}
warmup_ratio: 0.33
save_safetensors: False
fp16: True
push_to_hub: True
hub_model_id: bobox/DeBERTa3-s-CustomPoolin-v3-step1-checkpoints-tmp
hub_strategy: all_checkpoints
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
prediction_loss_only: True
per_device_train_batch_size: 100
per_device_eval_batch_size: 256
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 2
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: 1.0
num_train_epochs: 3
max_steps: -1
lr_scheduler_type: cosine_with_min_lr
lr_scheduler_kwargs: {'num_cycles': 0.5, 'min_lr': 1.6666666666666667e-05}
warmup_ratio: 0.33
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
save_safetensors: False
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: 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}
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: True
resume_from_checkpoint: None
hub_model_id: bobox/DeBERTa3-s-CustomPoolin-v3-step1-checkpoints-tmp
hub_strategy: all_checkpoints
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
eval_use_gather_object: False
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Click to expand

Epoch	Step	Training Loss	vitaminc-pairs loss	negation-triplets loss	scitail-pairs-pos loss	scitail-pairs-qa loss	xsum-pairs loss	sciq pairs loss	qasc pairs loss	openbookqa pairs loss	msmarco pairs loss	nq pairs loss	trivia pairs loss	gooaq pairs loss	paws-pos loss	global dataset loss	sts-test_spearman_cosine	allNLI-dev_max_ap	Qnli-dev_max_ap
0.0168	8	10.2928	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.0336	16	9.2166	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.0504	24	9.4858	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.0672	32	10.6143	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.0840	40	8.7553	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1008	48	10.9939	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1176	56	7.6039	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1345	64	5.9498	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1513	72	7.3051	3.2988	3.9604	1.9818	2.1997	6.0515	0.6095	6.3199	4.8391	6.4886	6.6406	6.4894	6.1527	2.0082	4.9577	0.3066	0.3444	0.5627
0.1681	80	8.3034	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.1849	88	7.6669	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2017	96	6.6415	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2185	104	5.7797	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2353	112	5.8361	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2521	120	5.3339	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2689	128	5.5908	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.2857	136	5.3209	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3025	144	5.5359	3.3310	3.8580	1.4769	1.6994	5.4819	0.5385	5.2021	4.4410	5.3419	5.5506	5.6972	5.3376	1.4170	3.9169	0.2954	0.3795	0.6317
0.3193	152	5.4713	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3361	160	4.9368	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3529	168	4.6594	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3697	176	4.8392	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.3866	184	4.414	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4034	192	4.891	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4202	200	4.4553	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4370	208	3.9729	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4538	216	3.7705	3.2468	3.6435	0.7890	0.7356	3.9327	0.4082	3.7175	3.5404	3.5351	4.0506	3.9953	3.6074	0.4195	2.4726	0.3791	0.4133	0.6779
0.4706	224	3.8409	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.4874	232	3.7894	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5042	240	3.3523	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5210	248	3.2407	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5378	256	3.3203	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5546	264	2.8457	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5714	272	2.4181	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.5882	280	3.4589	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6050	288	2.8203	3.1119	3.1485	0.4531	0.2652	2.6895	0.2656	2.5542	2.7523	2.6600	3.1773	3.2099	2.7316	0.2006	1.6342	0.5257	0.4717	0.7078
0.6218	296	2.4697	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6387	304	2.4654	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6555	312	2.4236	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6723	320	2.2879	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.6891	328	2.2145	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7059	336	1.8464	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7227	344	2.0086	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7395	352	2.0635	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7563	360	1.8584	3.3202	2.5793	0.3434	0.1618	1.6759	0.1834	1.6454	2.1257	2.1938	2.5316	2.4558	2.0596	0.0984	1.2206	0.6610	0.5199	0.7119
0.7731	368	2.0286	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.7899	376	1.9389	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8067	384	1.7453	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8235	392	1.6629	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8403	400	1.2724	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8571	408	1.7824	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8739	416	1.5826	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.8908	424	1.1971	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9076	432	1.5228	3.3624	2.1952	0.3006	0.1223	1.1091	0.1582	1.2383	1.8664	1.7434	2.3959	2.0697	1.7563	0.0766	1.0193	0.7292	0.5194	0.7126
0.9244	440	1.3323	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9412	448	1.5124	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9580	456	1.5565	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9748	464	1.3672	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
0.9916	472	1.0382	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.0084	480	1.0626	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.0252	488	1.3539	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.0420	496	1.1723	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.0588	504	1.4235	3.4031	1.9759	0.2554	0.0814	0.9034	0.1378	1.1603	1.7589	1.5608	2.1230	1.7719	1.6633	0.0720	0.9380	0.7523	0.5297	0.7129
1.0756	512	1.2283	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.0924	520	1.2455	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.1092	528	1.4265	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.1261	536	1.296	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.1429	544	0.8763	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.1597	552	1.5678	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.1765	560	1.2548	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.1933	568	1.3731	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.2101	576	1.3023	3.3815	1.8740	0.2373	0.0769	0.7711	0.1237	0.9432	1.6871	1.5070	1.9947	1.6041	1.5579	0.0721	0.8661	0.7642	0.5412	0.7159
1.2269	584	0.8135	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.2437	592	1.0259	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.2605	600	1.1896	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.2773	608	1.0532	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.2941	616	1.3221	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.3109	624	1.3136	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.3277	632	1.2238	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.3445	640	1.2407	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.3613	648	1.2245	3.4717	1.7962	0.2242	0.0488	0.7472	0.1108	0.9272	1.6692	1.3845	1.9117	1.3410	1.4387	0.0701	0.8505	0.7680	0.5471	0.7227
1.3782	656	1.0428	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.3950	664	1.1391	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.4118	672	1.2632	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.4286	680	0.9403	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.4454	688	0.7571	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.4622	696	0.9436	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.4790	704	1.1239	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.4958	712	0.9499	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.5126	720	1.0945	3.6495	1.6693	0.2157	0.0492	0.6830	0.1049	0.9140	1.5967	1.4397	1.7394	1.3303	1.4334	0.0603	0.8185	0.7815	0.5606	0.7098
1.5294	728	1.1161	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.5462	736	1.0056	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.5630	744	1.1743	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.5798	752	0.9153	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.5966	760	1.1589	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.6134	768	0.9187	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.6303	776	0.6937	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.6471	784	0.9704	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.6639	792	0.7343	3.5442	1.6493	0.2208	0.0249	0.6152	0.0969	0.7111	1.5369	1.4058	1.7066	1.2784	1.3419	0.0585	0.7827	0.7749	0.5627	0.7284
1.6807	800	1.2878	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.6975	808	0.9898	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.7143	816	0.7613	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.7311	824	0.9612	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.7479	832	1.1524	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.7647	840	0.827	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.7815	848	1.1898	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.7983	856	1.0117	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.8151	864	0.7019	3.4544	1.6149	0.2035	0.0181	0.5525	0.0999	0.6641	1.5456	1.3911	1.7188	1.2547	1.3517	0.0562	0.7473	0.7684	0.5697	0.7329
1.8319	872	0.8352	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.8487	880	0.7836	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.8655	888	1.0187	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.8824	896	0.74	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.8992	904	0.7263	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.9160	912	0.8073	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.9328	920	0.8185	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.9496	928	1.0992	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
1.9664	936	0.9973	3.5110	1.5776	0.2035	0.0250	0.5881	0.0934	0.6719	1.5059	1.2970	1.6186	1.1815	1.2714	0.0564	0.7213	0.7799	0.5544	0.7341
1.9832	944	0.6662	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.0	952	0.533	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.0168	960	0.7712	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.0336	968	0.6879	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.0504	976	0.7975	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.0672	984	0.873	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.0840	992	0.7995	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.1008	1000	1.0119	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.1176	1008	0.6317	3.6778	1.5845	0.2102	0.0228	0.5851	0.0977	0.6411	1.4752	1.2992	1.6314	1.1260	1.2683	0.0556	0.7329	0.7693	0.5614	0.7274
2.1345	1016	0.72	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.1513	1024	0.9418	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.1681	1032	0.7848	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.1849	1040	0.6965	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.2017	1048	1.0447	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.2185	1056	0.6361	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.2353	1064	0.6837	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.2521	1072	0.5713	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.2689	1080	0.8193	3.6399	1.5565	0.2069	0.0213	0.5440	0.0904	0.6057	1.4815	1.2856	1.6441	1.1469	1.2540	0.0543	0.7216	0.7765	0.5599	0.7322
2.2857	1088	0.9754	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.3025	1096	0.8932	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.3193	1104	0.8716	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.3361	1112	0.8787	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.3529	1120	0.9529	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.3697	1128	0.775	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.3866	1136	0.6178	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.4034	1144	0.8384	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.4202	1152	0.9425	3.5672	1.5244	0.2111	0.0162	0.5593	0.0893	0.5759	1.4933	1.2703	1.5815	1.1202	1.2132	0.0531	0.7058	0.7730	0.5635	0.7350
2.4370	1160	0.4551	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.4538	1168	0.6392	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.4706	1176	0.8341	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.4874	1184	0.7392	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.5042	1192	0.7646	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.5210	1200	0.8613	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.5378	1208	0.7585	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.5546	1216	1.0611	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.5714	1224	0.6506	3.6439	1.5040	0.2125	0.0162	0.5282	0.0863	0.5858	1.5073	1.2444	1.5493	1.1014	1.2073	0.0532	0.7022	0.7774	0.5647	0.7328
2.5882	1232	0.8525	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.6050	1240	0.6304	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.6218	1248	0.6354	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.6387	1256	0.6583	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.6555	1264	0.5964	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.6723	1272	0.818	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.6891	1280	0.8635	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.7059	1288	0.6389	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.7227	1296	0.6819	3.6131	1.5104	0.2084	0.0148	0.5229	0.0854	0.5588	1.4963	1.2766	1.5679	1.0982	1.2203	0.0529	0.7059	0.7762	0.5659	0.7355
2.7395	1304	0.7878	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.7563	1312	0.7638	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.7731	1320	0.8885	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.7899	1328	0.8184	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.8067	1336	0.7472	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.8235	1344	0.7012	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.8403	1352	0.4622	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.8571	1360	0.846	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.8739	1368	0.8308	3.6224	1.5088	0.2084	0.0148	0.5118	0.0858	0.5523	1.4941	1.2756	1.5808	1.0925	1.2114	0.0521	0.7022	0.7765	0.5662	0.7366
2.8908	1376	0.5334	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.9076	1384	0.7893	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.9244	1392	0.6897	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.9412	1400	0.7803	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.9580	1408	0.841	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.9748	1416	0.787	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
2.9916	1424	0.5861	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
3.0	1428	-	3.6139	1.5071	0.2084	0.0150	0.5124	0.0862	0.5532	1.4924	1.2700	1.5806	1.0905	1.2081	0.0519	0.6997	0.7776	0.5665	0.7369

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.2.0
Transformers: 4.44.2
PyTorch: 2.4.1+cu121
Accelerate: 0.34.2
Datasets: 3.0.1
Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers

@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",
}