checkpoint-305/README.md

metadata

base_model: microsoft/deberta-v3-small
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:32500
  - loss:GISTEmbedLoss
widget:
  - source_sentence: A picture of a white gas range with figurines above.
    sentences:
      - A nerdy woman brushing her teeth with a friend nearby.
      - a white stove turned off with a digital clock
      - >-
        The plasma membrane also contains other molecules, primarily other
        lipids and proteins. The green molecules in Figure above , for example,
        are the lipid cholesterol. Molecules of cholesterol help the plasma
        membrane keep its shape. Many of the proteins in the plasma membrane
        assist other substances in crossing the membrane.
  - source_sentence: who makes the kentucky derby garland of roses
    sentences:
      - >-
        Accrington strengthened their position in the play-off places with a
        hard-fought win over struggling Dagenham.
      - >-
        tidal energy can be used to produce electricity. Ocean thermal is energy
        derived from waves and also from tidal waves. 
         Ocean thermal energy can be used to produce electricity.
      - >-
        Kentucky Derby Trophy The Kroger Company has been the official florist
        of the Kentucky Derby since 1987. After taking over the duties from the
        Kingsley Walker florist, Kroger began constructing the prestigious
        garland in one of its local stores for the public to view on Derby Eve.
        The preservation of the garland and crowds of spectators watching its
        construction are a testament to the prestige and mystique of the Garland
        of Roses.
  - source_sentence: what is the difference between a general sense and a special sense?
    sentences:
      - >-
        Ian Curtis ( of Touching from a distance) Ian Kevin Curtis was an
        English musician and singer-songwriter. He is best known as the lead
        singer and lyricist of the post-punk band Joy Division. Joy Division
        released its debut album, Unknown Pleasures, in 1979 and recorded its
        follow-up, Closer, in 1980. Curtis, who suffered from epilepsy and
        depression, committed suicide on 18 May 1980, on the eve of Joy
        Division's first North American tour, resulting in the band's
        dissolution and the subsequent formation of New Order. Curtis was known
        for his baritone voice, dance style, and songwriting filled with imagery
        of desolation, emptiness and alienation. In 1995, Curtis's widow Deborah
        published Touching from a Distance: Ian Curtis and Joy Division, a
        biography of the singer. His life and death Ian Kevin Curtis was an
        English musician and singer-songwriter. He is best known as the lead
        singer and lyricist of the post-punk band Joy Division. Joy Division
        released its debut album, Unknown Pleasures, in 1979 and recorded its
        follow-up, Closer, in 1980. Curtis, who suffered from epilepsy and
        depression, committed suicide on 18 May 1980, on the eve of Joy
        Division's first North American tour, resulting in the band's
        dissolution and the subsequent formation of New Order. Curtis was known
        for his baritone voice, dance style, and songwriting filled with imagery
        of desolation, emptiness and alienation. In 1995, Curtis's widow Deborah
        published Touching from a Distance: Ian Curtis and Joy Division, a
        biography of the singer. His life and death have been dramatised in the
        films 24 Hour Party People (2002) and Control (2007). ...more
      - >-
        The human body has two basic types of senses, called special senses and
        general senses. Special senses have specialized sense organs that gather
        sensory information and change it into nerve impulses. ... General
        senses, in contrast, are all associated with the sense of touch. They
        lack special sense organs.
      - >-
        Captain Hook Barrie states in the novel that "Hook was not his true
        name. To reveal who he really was would even at this date set the
        country in a blaze", and relates that Peter Pan began their rivalry by
        feeding the pirate's hand to the crocodile. He is said to be
        "Blackbeard's bo'sun" and "the only man of whom Barbecue was afraid".[5]
        (In Robert Louis Stevenson's Treasure Island, one of the names Long John
        Silver goes by is Barbecue.)[6]
  - source_sentence: >-
      Retzius was born in Stockholm , son of the anatomist Anders Jahan Retzius
      ( and grandson of the naturalist and chemist Anders Retzius ) .
    sentences:
      - >-
        Retzius was born in Stockholm , the son of anatomist Anders Jahan
        Retzius ( and grandson of the naturalist and chemist Anders Retzius ) .
      - >-
        As of 14 March , over 156,000 cases of COVID-19 have been reported in
        around 140 countries and territories ; more than 5,800 people have died
        from the disease and around 75,000 have recovered .
      - A person sitting on a stool on the street.
  - source_sentence: who was the first person who made the violin
    sentences:
      - >-
        Alice in Chains Alice in Chains is an American rock band from Seattle,
        Washington, formed in 1987 by guitarist and vocalist Jerry Cantrell and
        drummer Sean Kinney,[1] who recruited bassist Mike Starr[1] and lead
        vocalist Layne Staley.[1][2][3] Starr was replaced by Mike Inez in
        1993.[4] After Staley's death in 2002, William DuVall joined in 2006 as
        co-lead vocalist and rhythm guitarist. The band took its name from
        Staley's previous group, the glam metal band Alice N' Chains.[5][2]
      - as distance from an object decreases , that object will appear larger
      - >-
        Violin The first makers of violins probably borrowed from various
        developments of the Byzantine lira. These included the rebec;[13] the
        Arabic rebab; the vielle (also known as the fidel or viuola); and the
        lira da braccio[11][14] The violin in its present form emerged in early
        16th-century northern Italy. The earliest pictures of violins, albeit
        with three strings, are seen in northern Italy around 1530, at around
        the same time as the words "violino" and "vyollon" are seen in Italian
        and French documents. One of the earliest explicit descriptions of the
        instrument, including its tuning, is from the Epitome musical by Jambe
        de Fer, published in Lyon in 1556.[15] By this time, the violin had
        already begun to spread throughout Europe.
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: 0.1561600438268545
            name: Pearson Cosine
          - type: spearman_cosine
            value: 0.22356441354815124
            name: Spearman Cosine
          - type: pearson_manhattan
            value: 0.2216924674035587
            name: Pearson Manhattan
          - type: spearman_manhattan
            value: 0.24997065610359018
            name: Spearman Manhattan
          - type: pearson_euclidean
            value: 0.1908690981304929
            name: Pearson Euclidean
          - type: spearman_euclidean
            value: 0.22363767136304896
            name: Spearman Euclidean
          - type: pearson_dot
            value: 0.15588248423807516
            name: Pearson Dot
          - type: spearman_dot
            value: 0.22337189362164545
            name: Spearman Dot
          - type: pearson_max
            value: 0.2216924674035587
            name: Pearson Max
          - type: spearman_max
            value: 0.24997065610359018
            name: Spearman Max
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: allNLI dev
          type: allNLI-dev
        metrics:
          - type: cosine_accuracy
            value: 0.666015625
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.9797871112823486
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.504258943781942
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.8929213285446167
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.357487922705314
            name: Cosine Precision
          - type: cosine_recall
            value: 0.8554913294797688
            name: Cosine Recall
          - type: cosine_ap
            value: 0.4008449937025217
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.666015625
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 752.6634521484375
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.504258943781942
            name: Dot F1
          - type: dot_f1_threshold
            value: 685.9220581054688
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.357487922705314
            name: Dot Precision
          - type: dot_recall
            value: 0.8554913294797688
            name: Dot Recall
          - type: dot_ap
            value: 0.40071344979441287
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.66796875
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 144.52613830566406
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.5075987841945289
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 267.046875
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.3443298969072165
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.9653179190751445
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.4008700157620745
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.666015625
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 5.572628974914551
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.504258943781942
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 12.826179504394531
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.357487922705314
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.8554913294797688
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.40083962142052487
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.66796875
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 752.6634521484375
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.5075987841945289
            name: Max F1
          - type: max_f1_threshold
            value: 685.9220581054688
            name: Max F1 Threshold
          - type: max_precision
            value: 0.357487922705314
            name: Max Precision
          - type: max_recall
            value: 0.9653179190751445
            name: Max Recall
          - type: max_ap
            value: 0.4008700157620745
            name: Max Ap
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: Qnli dev
          type: Qnli-dev
        metrics:
          - type: cosine_accuracy
            value: 0.591796875
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.9479926824569702
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.6291834002677376
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.7761930823326111
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.4598825831702544
            name: Cosine Precision
          - type: cosine_recall
            value: 0.9957627118644068
            name: Cosine Recall
          - type: cosine_ap
            value: 0.5658036772817674
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.59375
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 724.091064453125
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.6291834002677376
            name: Dot F1
          - type: dot_f1_threshold
            value: 596.2498779296875
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.4598825831702544
            name: Dot Precision
          - type: dot_recall
            value: 0.9957627118644068
            name: Dot Recall
          - type: dot_ap
            value: 0.5657459555147606
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.6171875
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 202.07958984375
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.6291834002677376
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 307.9236145019531
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.4598825831702544
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.9957627118644068
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.5891966424964378
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.591796875
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 8.938886642456055
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.6291834002677376
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 18.542938232421875
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.4598825831702544
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.9957627118644068
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.5658036772817674
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.6171875
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 724.091064453125
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.6291834002677376
            name: Max F1
          - type: max_f1_threshold
            value: 596.2498779296875
            name: Max F1 Threshold
          - type: max_precision
            value: 0.4598825831702544
            name: Max Precision
          - type: max_recall
            value: 0.9957627118644068
            name: Max Recall
          - type: max_ap
            value: 0.5891966424964378
            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

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_pj): Linear(in_features=768, out_features=768, bias=True)
    (linear_cls_Qpj): Linear(in_features=768, out_features=768, bias=True)
    (linear_mean_pj): Linear(in_features=768, out_features=768, bias=True)
    (linear_attnOut): Linear(in_features=768, out_features=768, bias=True)
    (mha): MultiheadAttention(
      (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
    )
    (layernorm_output): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (layernorm_weightedPooing): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (layernorm_pjCls): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (layernorm_pjMean): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
    (layernorm_attnOut): 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-toytest-step1-checkpoints-tmp")
# Run inference
sentences = [
    'who was the first person who made the violin',
    'Violin The first makers of violins probably borrowed from various developments of the Byzantine lira. These included the rebec;[13] the Arabic rebab; the vielle (also known as the fidel or viuola); and the lira da braccio[11][14] The violin in its present form emerged in early 16th-century northern Italy. The earliest pictures of violins, albeit with three strings, are seen in northern Italy around 1530, at around the same time as the words "violino" and "vyollon" are seen in Italian and French documents. One of the earliest explicit descriptions of the instrument, including its tuning, is from the Epitome musical by Jambe de Fer, published in Lyon in 1556.[15] By this time, the violin had already begun to spread throughout Europe.',
    "Alice in Chains Alice in Chains is an American rock band from Seattle, Washington, formed in 1987 by guitarist and vocalist Jerry Cantrell and drummer Sean Kinney,[1] who recruited bassist Mike Starr[1] and lead vocalist Layne Staley.[1][2][3] Starr was replaced by Mike Inez in 1993.[4] After Staley's death in 2002, William DuVall joined in 2006 as co-lead vocalist and rhythm guitarist. The band took its name from Staley's previous group, the glam metal band Alice N' Chains.[5][2]",
]
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	0.1562
spearman_cosine	0.2236
pearson_manhattan	0.2217
spearman_manhattan	0.25
pearson_euclidean	0.1909
spearman_euclidean	0.2236
pearson_dot	0.1559
spearman_dot	0.2234
pearson_max	0.2217
spearman_max	0.25

Binary Classification

• Dataset: allNLI-dev
• Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.666
cosine_accuracy_threshold	0.9798
cosine_f1	0.5043
cosine_f1_threshold	0.8929
cosine_precision	0.3575
cosine_recall	0.8555
cosine_ap	0.4008
dot_accuracy	0.666
dot_accuracy_threshold	752.6635
dot_f1	0.5043
dot_f1_threshold	685.9221
dot_precision	0.3575
dot_recall	0.8555
dot_ap	0.4007
manhattan_accuracy	0.668
manhattan_accuracy_threshold	144.5261
manhattan_f1	0.5076
manhattan_f1_threshold	267.0469
manhattan_precision	0.3443
manhattan_recall	0.9653
manhattan_ap	0.4009
euclidean_accuracy	0.666
euclidean_accuracy_threshold	5.5726
euclidean_f1	0.5043
euclidean_f1_threshold	12.8262
euclidean_precision	0.3575
euclidean_recall	0.8555
euclidean_ap	0.4008
max_accuracy	0.668
max_accuracy_threshold	752.6635
max_f1	0.5076
max_f1_threshold	685.9221
max_precision	0.3575
max_recall	0.9653
max_ap	0.4009

Binary Classification

• Dataset: Qnli-dev
• Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.5918
cosine_accuracy_threshold	0.948
cosine_f1	0.6292
cosine_f1_threshold	0.7762
cosine_precision	0.4599
cosine_recall	0.9958
cosine_ap	0.5658
dot_accuracy	0.5938
dot_accuracy_threshold	724.0911
dot_f1	0.6292
dot_f1_threshold	596.2499
dot_precision	0.4599
dot_recall	0.9958
dot_ap	0.5657
manhattan_accuracy	0.6172
manhattan_accuracy_threshold	202.0796
manhattan_f1	0.6292
manhattan_f1_threshold	307.9236
manhattan_precision	0.4599
manhattan_recall	0.9958
manhattan_ap	0.5892
euclidean_accuracy	0.5918
euclidean_accuracy_threshold	8.9389
euclidean_f1	0.6292
euclidean_f1_threshold	18.5429
euclidean_precision	0.4599
euclidean_recall	0.9958
euclidean_ap	0.5658
max_accuracy	0.6172
max_accuracy_threshold	724.0911
max_f1	0.6292
max_f1_threshold	596.2499
max_precision	0.4599
max_recall	0.9958
max_ap	0.5892

Training Details

Training Dataset

Unnamed Dataset

• Size: 32,500 training samples
• Columns: sentence1 and sentence2
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	min: 4 tokens mean: 29.3 tokens max: 343 tokens	min: 2 tokens mean: 57.53 tokens max: 512 tokens

• Samples:

  sentence1	sentence2
  A Slippery Dick is what type of creature?	The Slippery Dick (Juvenile) - Whats That Fish! Description Also known as Sand-reef Wrasses and Slippery Dick Wrasse. Found singly or in pairs or in groups constantly circling around reefs, sea grass beds and sandy areas. Colours highly variable especially between juvenile to adult. They feed on hard shell invertebrates. Length - 18cm Depth - 2-12m Widespread Western Atlantic & Caribbean Most reef fish seen by divers during the day are grazers, that cruise around just above the surface of the coral or snoop into crevices looking for algae, worms and small crustaceans. Wrasses have small protruding teeth and graze the bottom taking in a variety of snails, worms, crabs, shrimps and eggs. Any hard coats or thick shells are then ground down by their pharyngeal jaws and the delicacies inside digested. From juvenile to adult wrasses dramatically alter their colour and body shapes. Wrasses are always on the go during the day, but are the first to go to bed and the last to rise. Small wrasses dive below the sand to sleep and larger wrasses wedge themselves in crevasses. Related creatures Heads up! Many creatures change during their life. Juvenile fish become adults and some change shape or their colour. Some species change sex and others just get older. The following creature(s) are known relatives of the Slippery Dick (Juvenile). Click the image(s) to explore further or hover over to get a better view! Slippery Dick
  e. in solids the atoms are closely locked in position and can only vibrate, in liquids the atoms and molecules are more loosely connected and can collide with and move past one another, while in gases the atoms or molecules are free to move independently, colliding frequently.	Within a substance, atoms that collide frequently and move independently of one another are most likely in a gas
  In December 2015 , the film was ranked # 192 on IMDb .	As of December 2015 , it is the # 192 highest rated film on IMDb.

• Loss: GISTEmbedLoss 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}
  

Evaluation Dataset

Unnamed Dataset

• Size: 1,664 evaluation samples
• Columns: sentence1 and sentence2
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	min: 4 tokens mean: 28.74 tokens max: 330 tokens	min: 2 tokens mean: 56.55 tokens max: 512 tokens

• Samples:

  sentence1	sentence2
  What component of an organism, made up of many cells, in turn makes up an organ?
  Diffusion Diffusion is a process where atoms or molecules move from areas of high concentration to areas of low concentration.	Diffusion is the process in which a substance naturally moves from an area of higher to lower concentration.
  In the 1966 movie The Good, The Bad And The Ugly, Clint Eastwood played the Good" and Lee van Cleef played "the Bad", but who played "the Ugly"?	View All Photos (10) Movie Info In the last and the best installment of his so-called "Dollars" trilogy of Sergio Leone-directed "spaghetti westerns," Clint Eastwood reprised the role of a taciturn, enigmatic loner. Here he searches for a cache of stolen gold against rivals the Bad (Lee Van Cleef), a ruthless bounty hunter, and the Ugly (Eli Wallach), a Mexican bandit. Though dubbed "the Good," Eastwood's character is not much better than his opponents -- he is just smarter and shoots faster. The film's title reveals its ironic attitude toward the canonized heroes of the classical western. "The real West was the world of violence, fear, and brutal instincts," claimed Leone. "In pursuit of profit there is no such thing as good and evil, generosity or deviousness; everything depends on chance, and not the best wins but the luckiest." Immensely entertaining and beautifully shot in Techniscope by Tonino Delli Colli, the movie is a virtually definitive "spaghetti western," rivaled only by Leone's own Once Upon a Time in the West (1968). The main musical theme by Ennio Morricone hit #1 on the British pop charts. Originally released in Italy at 177 minutes, the movie was later cut for its international release. ~ Yuri German, Rovi Rating:

• Loss: GISTEmbedLoss 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: 32
• per_device_eval_batch_size: 256
• lr_scheduler_type: cosine_with_min_lr
• lr_scheduler_kwargs: {'num_cycles': 0.5, 'min_lr': 3.3333333333333337e-06}
• warmup_ratio: 0.33
• save_safetensors: False
• fp16: True
• push_to_hub: True
• hub_model_id: bobox/DeBERTa3-s-CustomPoolin-toytest-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: 32
• per_device_eval_batch_size: 256
• 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: 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': 3.3333333333333337e-06}
• 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-toytest-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	Validation Loss	sts-test_spearman_cosine	allNLI-dev_max_ap	Qnli-dev_max_ap
0.0010	1	4.9603	-	-	-	-
0.0020	2	28.2529	-	-	-	-
0.0030	3	27.6365	-	-	-	-
0.0039	4	6.1387	-	-	-	-
0.0049	5	5.5753	-	-	-	-
0.0059	6	5.6951	-	-	-	-
0.0069	7	6.3533	-	-	-	-
0.0079	8	27.3848	-	-	-	-
0.0089	9	3.8501	-	-	-	-
0.0098	10	27.911	-	-	-	-
0.0108	11	4.9042	-	-	-	-
0.0118	12	6.8003	-	-	-	-
0.0128	13	5.7317	-	-	-	-
0.0138	14	20.261	-	-	-	-
0.0148	15	27.9051	-	-	-	-
0.0157	16	5.5959	-	-	-	-
0.0167	17	5.8052	-	-	-	-
0.0177	18	4.5088	-	-	-	-
0.0187	19	7.3472	-	-	-	-
0.0197	20	5.8668	-	-	-	-
0.0207	21	6.4083	-	-	-	-
0.0217	22	6.011	-	-	-	-
0.0226	23	5.2394	-	-	-	-
0.0236	24	4.2966	-	-	-	-
0.0246	25	26.605	-	-	-	-
0.0256	26	6.2067	-	-	-	-
0.0266	27	6.0346	-	-	-	-
0.0276	28	5.4676	-	-	-	-
0.0285	29	6.4292	-	-	-	-
0.0295	30	26.6452	-	-	-	-
0.0305	31	18.8401	-	-	-	-
0.0315	32	7.4531	-	-	-	-
0.0325	33	4.8286	-	-	-	-
0.0335	34	5.0078	-	-	-	-
0.0344	35	5.4115	-	-	-	-
0.0354	36	5.4196	-	-	-	-
0.0364	37	4.5023	-	-	-	-
0.0374	38	5.376	-	-	-	-
0.0384	39	5.2303	-	-	-	-
0.0394	40	5.6694	-	-	-	-
0.0404	41	4.7825	-	-	-	-
0.0413	42	4.6507	-	-	-	-
0.0423	43	24.2072	-	-	-	-
0.0433	44	4.9285	-	-	-	-
0.0443	45	6.326	-	-	-	-
0.0453	46	4.5724	-	-	-	-
0.0463	47	4.754	-	-	-	-
0.0472	48	5.5443	-	-	-	-
0.0482	49	4.5764	-	-	-	-
0.0492	50	5.1434	-	-	-	-
0.0502	51	22.6991	-	-	-	-
0.0512	52	5.4277	-	-	-	-
0.0522	53	5.0178	-	-	-	-
0.0531	54	4.8779	-	-	-	-
0.0541	55	4.2884	-	-	-	-
0.0551	56	16.0994	-	-	-	-
0.0561	57	21.31	-	-	-	-
0.0571	58	4.9721	-	-	-	-
0.0581	59	5.143	-	-	-	-
0.0591	60	3.5933	-	-	-	-
0.0600	61	5.2559	-	-	-	-
0.0610	62	4.0757	-	-	-	-
0.0620	63	3.6612	-	-	-	-
0.0630	64	4.7505	-	-	-	-
0.0640	65	4.1979	-	-	-	-
0.0650	66	3.9982	-	-	-	-
0.0659	67	4.7065	-	-	-	-
0.0669	68	5.3413	-	-	-	-
0.0679	69	3.6964	-	-	-	-
0.0689	70	17.8774	-	-	-	-
0.0699	71	4.8154	-	-	-	-
0.0709	72	4.8356	-	-	-	-
0.0719	73	4.568	-	-	-	-
0.0728	74	4.0898	-	-	-	-
0.0738	75	3.4502	-	-	-	-
0.0748	76	3.7733	-	-	-	-
0.0758	77	4.5204	-	-	-	-
0.0768	78	4.2526	-	-	-	-
0.0778	79	4.4398	-	-	-	-
0.0787	80	4.0988	-	-	-	-
0.0797	81	3.9704	-	-	-	-
0.0807	82	4.3343	-	-	-	-
0.0817	83	4.2587	-	-	-	-
0.0827	84	15.0149	-	-	-	-
0.0837	85	14.6599	-	-	-	-
0.0846	86	4.0623	-	-	-	-
0.0856	87	3.7597	-	-	-	-
0.0866	88	4.3433	-	-	-	-
0.0876	89	4.0287	-	-	-	-
0.0886	90	4.6257	-	-	-	-
0.0896	91	13.4689	-	-	-	-
0.0906	92	4.6583	-	-	-	-
0.0915	93	4.2682	-	-	-	-
0.0925	94	4.468	-	-	-	-
0.0935	95	3.4333	-	-	-	-
0.0945	96	12.7654	-	-	-	-
0.0955	97	3.5577	-	-	-	-
0.0965	98	12.5875	-	-	-	-
0.0974	99	4.2206	-	-	-	-
0.0984	100	3.5981	-	-	-	-
0.0994	101	3.5575	-	-	-	-
0.1004	102	4.0271	-	-	-	-
0.1014	103	4.0803	-	-	-	-
0.1024	104	4.0886	-	-	-	-
0.1033	105	4.176	-	-	-	-
0.1043	106	4.6653	-	-	-	-
0.1053	107	4.3076	-	-	-	-
0.1063	108	8.7282	-	-	-	-
0.1073	109	3.4192	-	-	-	-
0.1083	110	10.6027	-	-	-	-
0.1093	111	4.0959	-	-	-	-
0.1102	112	4.2785	-	-	-	-
0.1112	113	3.9945	-	-	-	-
0.1122	114	10.0652	-	-	-	-
0.1132	115	3.8621	-	-	-	-
0.1142	116	4.3975	-	-	-	-
0.1152	117	9.7899	-	-	-	-
0.1161	118	4.3812	-	-	-	-
0.1171	119	3.8715	-	-	-	-
0.1181	120	3.8327	-	-	-	-
0.1191	121	3.5103	-	-	-	-
0.1201	122	9.3158	-	-	-	-
0.1211	123	3.7201	-	-	-	-
0.1220	124	3.4311	-	-	-	-
0.1230	125	3.7946	-	-	-	-
0.1240	126	4.0456	-	-	-	-
0.125	127	3.482	-	-	-	-
0.1260	128	3.1901	-	-	-	-
0.1270	129	3.414	-	-	-	-
0.1280	130	3.4967	-	-	-	-
0.1289	131	3.6594	-	-	-	-
0.1299	132	8.066	-	-	-	-
0.1309	133	3.7872	-	-	-	-
0.1319	134	4.0023	-	-	-	-
0.1329	135	3.7728	-	-	-	-
0.1339	136	3.1893	-	-	-	-
0.1348	137	3.3635	-	-	-	-
0.1358	138	4.0195	-	-	-	-
0.1368	139	4.1097	-	-	-	-
0.1378	140	3.7903	-	-	-	-
0.1388	141	3.5748	-	-	-	-
0.1398	142	3.8104	-	-	-	-
0.1407	143	8.0411	-	-	-	-
0.1417	144	3.4819	-	-	-	-
0.1427	145	3.452	-	-	-	-
0.1437	146	3.5861	-	-	-	-
0.1447	147	3.4324	-	-	-	-
0.1457	148	3.521	-	-	-	-
0.1467	149	3.8868	-	-	-	-
0.1476	150	8.1191	-	-	-	-
0.1486	151	3.6447	-	-	-	-
0.1496	152	2.9436	-	-	-	-
0.1506	153	8.1535	2.2032	0.2236	0.4009	0.5892
0.1516	154	3.9619	-	-	-	-
0.1526	155	3.1301	-	-	-	-
0.1535	156	3.0478	-	-	-	-
0.1545	157	3.2986	-	-	-	-
0.1555	158	3.2847	-	-	-	-
0.1565	159	3.6599	-	-	-	-
0.1575	160	3.2238	-	-	-	-
0.1585	161	2.8897	-	-	-	-
0.1594	162	3.9443	-	-	-	-
0.1604	163	3.3733	-	-	-	-
0.1614	164	3.7444	-	-	-	-
0.1624	165	3.4813	-	-	-	-
0.1634	166	2.6865	-	-	-	-
0.1644	167	2.7587	-	-	-	-
0.1654	168	3.3628	-	-	-	-
0.1663	169	3.0035	-	-	-	-
0.1673	170	10.1591	-	-	-	-
0.1683	171	3.5366	-	-	-	-
0.1693	172	8.4047	-	-	-	-
0.1703	173	3.8643	-	-	-	-
0.1713	174	3.3529	-	-	-	-
0.1722	175	3.7143	-	-	-	-
0.1732	176	3.3323	-	-	-	-
0.1742	177	3.1206	-	-	-	-
0.1752	178	3.1348	-	-	-	-
0.1762	179	7.6011	-	-	-	-
0.1772	180	3.7025	-	-	-	-
0.1781	181	10.5662	-	-	-	-
0.1791	182	8.966	-	-	-	-
0.1801	183	9.426	-	-	-	-
0.1811	184	3.0025	-	-	-	-
0.1821	185	7.0984	-	-	-	-
0.1831	186	7.3808	-	-	-	-
0.1841	187	2.8657	-	-	-	-
0.1850	188	6.5636	-	-	-	-
0.1860	189	3.4702	-	-	-	-
0.1870	190	5.9302	-	-	-	-
0.1880	191	3.2406	-	-	-	-
0.1890	192	3.4459	-	-	-	-
0.1900	193	5.269	-	-	-	-
0.1909	194	4.8605	-	-	-	-
0.1919	195	2.9891	-	-	-	-
0.1929	196	3.6681	-	-	-	-
0.1939	197	3.1589	-	-	-	-
0.1949	198	3.1835	-	-	-	-
0.1959	199	3.7561	-	-	-	-
0.1969	200	4.0891	-	-	-	-
0.1978	201	3.563	-	-	-	-
0.1988	202	3.7433	-	-	-	-
0.1998	203	3.3813	-	-	-	-
0.2008	204	5.2311	-	-	-	-
0.2018	205	3.3494	-	-	-	-
0.2028	206	3.3533	-	-	-	-
0.2037	207	3.688	-	-	-	-
0.2047	208	3.5342	-	-	-	-
0.2057	209	4.9381	-	-	-	-
0.2067	210	3.1839	-	-	-	-
0.2077	211	3.0465	-	-	-	-
0.2087	212	3.1232	-	-	-	-
0.2096	213	4.6297	-	-	-	-
0.2106	214	2.9834	-	-	-	-
0.2116	215	4.2231	-	-	-	-
0.2126	216	3.1458	-	-	-	-
0.2136	217	3.2525	-	-	-	-
0.2146	218	3.5971	-	-	-	-
0.2156	219	3.5616	-	-	-	-
0.2165	220	3.2378	-	-	-	-
0.2175	221	2.9075	-	-	-	-
0.2185	222	3.0391	-	-	-	-
0.2195	223	3.5573	-	-	-	-
0.2205	224	3.2092	-	-	-	-
0.2215	225	3.2646	-	-	-	-
0.2224	226	3.0886	-	-	-	-
0.2234	227	3.5241	-	-	-	-
0.2244	228	3.0111	-	-	-	-
0.2254	229	3.707	-	-	-	-
0.2264	230	5.3822	-	-	-	-
0.2274	231	3.2646	-	-	-	-
0.2283	232	2.7021	-	-	-	-
0.2293	233	3.5131	-	-	-	-
0.2303	234	3.103	-	-	-	-
0.2313	235	2.9535	-	-	-	-
0.2323	236	2.9631	-	-	-	-
0.2333	237	2.8068	-	-	-	-
0.2343	238	3.4251	-	-	-	-
0.2352	239	2.8495	-	-	-	-
0.2362	240	2.9972	-	-	-	-
0.2372	241	3.3509	-	-	-	-
0.2382	242	2.9234	-	-	-	-
0.2392	243	2.4086	-	-	-	-
0.2402	244	3.1282	-	-	-	-
0.2411	245	2.3352	-	-	-	-
0.2421	246	2.4706	-	-	-	-
0.2431	247	3.5449	-	-	-	-
0.2441	248	2.8963	-	-	-	-
0.2451	249	2.773	-	-	-	-
0.2461	250	2.355	-	-	-	-
0.2470	251	2.656	-	-	-	-
0.2480	252	2.6221	-	-	-	-
0.2490	253	8.6739	-	-	-	-
0.25	254	10.8242	-	-	-	-
0.2510	255	2.3408	-	-	-	-
0.2520	256	2.1221	-	-	-	-
0.2530	257	3.295	-	-	-	-
0.2539	258	2.5896	-	-	-	-
0.2549	259	2.1215	-	-	-	-
0.2559	260	9.4851	-	-	-	-
0.2569	261	2.1982	-	-	-	-
0.2579	262	3.0568	-	-	-	-
0.2589	263	2.6269	-	-	-	-
0.2598	264	2.4792	-	-	-	-
0.2608	265	1.9445	-	-	-	-
0.2618	266	2.4061	-	-	-	-
0.2628	267	8.3116	-	-	-	-
0.2638	268	8.0804	-	-	-	-
0.2648	269	2.1674	-	-	-	-
0.2657	270	7.1975	-	-	-	-
0.2667	271	5.9104	-	-	-	-
0.2677	272	2.498	-	-	-	-
0.2687	273	2.5249	-	-	-	-
0.2697	274	2.7152	-	-	-	-
0.2707	275	2.7904	-	-	-	-
0.2717	276	2.7745	-	-	-	-
0.2726	277	2.9741	-	-	-	-
0.2736	278	1.8215	-	-	-	-
0.2746	279	4.6844	-	-	-	-
0.2756	280	2.8613	-	-	-	-
0.2766	281	2.7147	-	-	-	-
0.2776	282	2.814	-	-	-	-
0.2785	283	2.3569	-	-	-	-
0.2795	284	2.672	-	-	-	-
0.2805	285	3.2052	-	-	-	-
0.2815	286	2.8056	-	-	-	-
0.2825	287	2.6268	-	-	-	-
0.2835	288	2.5641	-	-	-	-
0.2844	289	2.4475	-	-	-	-
0.2854	290	2.7377	-	-	-	-
0.2864	291	2.3831	-	-	-	-
0.2874	292	8.8069	-	-	-	-
0.2884	293	2.186	-	-	-	-
0.2894	294	2.3389	-	-	-	-
0.2904	295	1.9744	-	-	-	-
0.2913	296	2.4491	-	-	-	-
0.2923	297	2.5668	-	-	-	-
0.2933	298	2.1939	-	-	-	-
0.2943	299	2.2832	-	-	-	-
0.2953	300	2.7508	-	-	-	-
0.2963	301	2.5206	-	-	-	-
0.2972	302	2.3522	-	-	-	-
0.2982	303	2.7186	-	-	-	-
0.2992	304	2.1369	-	-	-	-
0.3002	305	9.7972	-	-	-	-

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.2.1
• Transformers: 4.44.2
• PyTorch: 2.5.0+cu121
• Accelerate: 0.34.2
• Datasets: 3.0.2
• 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",
}

GISTEmbedLoss

@misc{solatorio2024gistembed,
    title={GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning},
    author={Aivin V. Solatorio},
    year={2024},
    eprint={2402.16829},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}