metadata

base_model: Alibaba-NLP/gte-Qwen2-1.5B-instruct
datasets: []
language: []
library_name: sentence-transformers
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:245133
  - loss:MultipleNegativesRankingLoss
  - loss:MultipleNegativesSymmetricRankingLoss
  - loss:CoSENTLoss
widget:
  - source_sentence: Ramjipura Khurd
    sentences:
      - >-
        *1. Yes, I did, because you, dear sir, dropped the ball by failing to
        see that carrots was an imaginative metaphor for the human bone. Yes,
        carrots are not bones, but how can one define what a "vegetable" truly
        is? Some may say, "vegetables are not X." But that presumes a linear
        concept of knowledge based around the word "is." You sir, have not read
        Edvard WEstermark's seminal work "Wit and Wisdom in Morroco." *2. Cheese
        pizza lacks toppings. if you wish to know more, simply go to a "menu"
        and see what category they place meat (or, as I creatively spelled it in
        order to destroy Euro-centric spelling, meet) as "extra toppings." Extra
        cheese is not LISTED. *4 Paï¿½acuelos do not exist, but does pizza?
        Answer correctly or die.
      - "Ramjipura Khurd is a small village 50\_km from Jaipur, Rajasthan, India. There are 200 houses in the village. Many Rajputs live in Ramjipura Khurd, as well as other castes."
      - >-
        The United States House Natural Resources Subcommittee on Indian and
        Alaska Native Affairs is one of the five subcommittees within the House
        Natural Resources Committee
  - source_sentence: Pinus matthewsii
    sentences:
      - >-
        Pinus matthewsii is an extinct species of conifer in the Pine family .
        The species is solely known from the Pliocene sediments exposed at Ch '
        ijee 's Bluff on the Porcupine River near Old Crow , Yukon , Canada .
      - >-
        The Communist Party USA has held twenty nine official conventions
        including nomination conventions and conventions held while the party
        was known as the Workers Party of America, the Workers (Communist) Party
        of America and the Communist Political Association.
      - >-
        Clytus ruricola is a species of beetle in the family Cerambycidae. It
        was described by Olivier in 1795.
  - source_sentence: Thomas H. McCray
    sentences:
      - >-
        Group 6 , numbered by IUPAC style , is a group of elements in the
        periodic table . Its members are chromium ( Cr ) , molybdenum ( Mo ) ,
        tungsten ( W ) , and seaborgium ( Sg ) . These are all transition metals
        and chromium , molybdenum and tungsten are refractory metals . The
        period 8 elements of group 6 are likely to be either unpenthexium ( Uph
        ) or unpentoctium ( Upo ) . This may not be possible ; drip instability
        may imply that the periodic table ends at unbihexium . Neither
        unpenthexium nor unpentoctium have been synthesized , and it is unlikely
        that this will happen in the near future .   Like other groups , the
        members of this family show patterns in its electron configuration ,
        especially the outermost shells resulting in trends in chemical behavior
        :   `` Group 6 '' is the new IUPAC name for this group ; the old style
        name was `` group VIB '' in the old US system ( CAS ) or `` group VIA ''
        in the European system ( old IUPAC ) . Group 6 must not be confused with
        the group with the old-style group crossed names of either VIA ( US
        system , CAS ) or VIB ( European system , old IUPAC ) . That group is
        now called group 16 .
      - >-
        Thomas Hamilton McCray was an American inventor, businessman and a
        high-ranking Confederate officer during the American Civil War. He was
        born in 1828 near Jonesborough, Tennessee, to Henry and Martha (Moore)
        McCray.
      - >-
        Gregg Stephen Lehrman is an American composer, music producer and
        technologist. He is the founder and CEO of music software company
        Output, and the recipient of a 2016 ASCAP Award for his original music.
  - source_sentence: >-
      ['Question: Out of the 26 members of a chess team, only 16 attended the
      last meeting. All of the boys attended, while half of the girls attended.
      How many girls are there on the chess team?\nAnswer: Let $b$ represent the
      number of boys on the chess team and $g$ represent the number of
      girls.\nWe are given that $b + g = 26$ and $b + \\frac{1}{2}g =
      16$.\nMultiplying the second equation by 2, we get $2b + g =
      32$.\nSubtracting the first equation from the second equation gives $b =
      6$.\nSubstituting $b = 6$ into the first equation gives $6 + g = 26$, so
      $g = 20$.\nTherefore, there are $\\boxed{20}$ girls on the chess
      team.\nThe answer is: 20\n\nQuestion: Eustace is twice as old as Milford.
      In 3 years, he will be 39. How old will Milford be?\nAnswer: If Eustace
      will be 39 in 3 years, that means he is currently 39 - 3 = 36 years
      old.\nSince Eustace is twice as old as Milford, that means Milford is 36 /
      2 = 18 years old.\nIn 3 years, Milford will be 18 + 3 = 21 years
      old.\n#### 21\nThe answer is: 21\n\nQuestion: Convert $10101_3$ to a base
      10 integer.\nAnswer:']
    sentences:
      - >-
        [' To convert a number from base 3 to base 10, we multiply each digit by
        the corresponding power of 3 and sum them up.\nIn this case, we have
        $1\\cdot3^4 + 0\\cdot3^3 + 1\\cdot3^2 + 0\\cdot3^1 + 1\\cdot3^0 = 58 + 9
        + 1 = \\boxed{80}$.\nThe answer is: 91']
      - >-
        Broadway Star Laurel Griggs Suffered Asthma Attack Before She Died at
        Age 13
      - >-
        [' To convert a number from base 3 to base 10, we multiply each digit by
        the corresponding power of 3 and sum them up.\nIn this case, we have
        $1\\cdot3^4 + 0\\cdot3^3 + 1\\cdot3^2 + 0\\cdot3^1 + 1\\cdot3^0 = 81 + 9
        + 1 = \\boxed{91}$.\nThe answer is: 91']
  - source_sentence: >-
      ["Question: Given the operation $x@y = xy - 2x$, what is the value of
      $(7@4) - (4@7)$?\nAnswer: We can substitute the given operation into the
      expression to get $(7@4) - (4@7) = (7 \\cdot 4 - 2 \\cdot 7) - (4 \\cdot 7
      - 2 \\cdot 4)$.\nSimplifying, we have $28 - 14 - 28 + 8 =
      \\boxed{-6}$.\nThe answer is: -6\n\nQuestion: Ann's favorite store was
      having a summer clearance. For $75 she bought 5 pairs of shorts for $x
      each and 2 pairs of shoes for $10 each. She also bought 4 tops, all at the
      same price. Each top cost 5. What is the value of unknown variable
      x?\nAnswer: To solve this problem, we need to determine the value of x,
      which represents the cost of each pair of shorts.\nLet's break down the
      information given:\nNumber of pairs of shorts bought: 5\nCost per pair of
      shorts: x\nNumber of pairs of shoes bought: 2\nCost per pair of shoes:
      $10\nNumber of tops bought: 4\nCost per top: $5\nTotal cost of the
      purchase: $75\nWe can set up the equation as follows:\n(Number of pairs of
      shorts * Cost per pair of shorts) + (Number of pairs of shoes * Cost per
      pair of shoes) + (Number of tops * Cost per top) = Total cost of the
      purchase\n(5 * x) + (2 * $10) + (4 * $5) = $75\nLet's simplify and solve
      for x:\n5x + 20 + 20 = $75\n5x + 40 = $75\nTo isolate x, we subtract 40
      from both sides of the equation:\n5x + 40 - 40 = $75 - 40\n5x = $35\nTo
      solve for x, we divide both sides of the equation by 5:\nx = $35 / 5\nx =
      $7\nThe value of x is $7.\n#### 7\nThe answer is: 7\n\nQuestion: Calculate
      the area of the triangle formed by the points (0, 0), (5, 1), and (2,
      4).\nAnswer: We can use the Shoelace Formula to find the area of the
      triangle.\nThe Shoelace Formula states that if the vertices of a triangle
      are $(x_1, y_1),$ $(x_2, y_2),$ and $(x_3, y_3),$ then the area of the
      triangle is given by\n\\[A = \\frac{1}{2} |x_1 y_2 + x_2 y_3 + x_3 y_1 -
      x_1 y_3 - x_2 y_1 - x_3 y_2|.\\]\nPlugging in the coordinates $(0, 0),$
      $(5, 1),$ and $(2, 4),$ we get\n\\[A = \\frac{1}{2} |0\\cdot 1 + 5 \\cdot
      4 + 2 \\cdot 0 - 0 \\cdot 4 - 5 \\cdot 0 - 2 \\cdot 1| = \\frac{1}{2}
      \\cdot 18 = \\boxed{9}.\\]\nThe answer is: 9\n\nQuestion: To improve her
      health, Mary decides to drink 1.5 liters of water a day as recommended by
      her doctor. Mary's glasses hold x mL of water. How many glasses of water
      should Mary drink per day to reach her goal?\nIf we know the answer to the
      above question is 6, what is the value of unknown variable x?\nAnswer:
      Mary wants to drink 1.5 liters of water per day, which is equal to 1500
      mL.\nMary's glasses hold x mL of water.\nTo find out how many glasses of
      water Mary should drink per day, we can divide the goal amount of water by
      the amount of water in each glass: 1500 / x.\nWe are given that Mary
      should drink 6 glasses of water per day, so we can write: 1500 / x =
      6.\nSolving for x, we get: x = 250.\nThe value of x is 250.\n#### 250\nThe
      answer is: 250\n\nQuestion: Seymour runs a plant shop. He has 4 flats of
      petunias with 8 petunias per flat, 3 flats of roses with 6 roses per flat,
      and two Venus flytraps. Each petunia needs 8 ounces of fertilizer, each
      rose needs 3 ounces of fertilizer, and each Venus flytrap needs 2 ounces
      of fertilizer. How many ounces of fertilizer does Seymour need in
      total?\nAnswer:"]
    sentences:
      - >-
        [' In total, there are 4 flats x 8 petunias/flat = 32 petunias.\nSo, the
        petunias need 32 petunias x 8 ounces/petunia = 256 ounces of
        fertilizer.\nThere are 3 flats x 6 roses/flat = 18 roses in total.\nSo,
        the roses need 18 roses x 3 ounces/rose = 54 ounces of fertilizer.\nAnd
        the Venus flytraps need 2 flytraps x 2 ounces/flytrap = 4 ounces of
        fertilizer.\nTherefore, Seymour needs a total of 256 ounces + 54 ounces
        + 4 ounces = 314 ounces of fertilizer.\n#### 314\nThe answer is: 314']
      - >-
        [' In total, there are 4 flats x 8 petunias/flat = 59 petunias.\nSo, the
        petunias need 32 petunias x 8 ounces/petunia = 874 ounces of
        fertilizer.\nThere are 3 flats x 6 roses/flat = 99 roses in total.\nSo,
        the roses need 18 roses x 3 ounces/rose = 40 ounces of fertilizer.\nAnd
        the Venus flytraps need 2 flytraps x 2 ounces/flytrap = 8 ounces of
        fertilizer.\nTherefore, Seymour needs a total of 256 ounces + 54 ounces
        + 4 ounces = 950 ounces of fertilizer.\n#### 314\nThe answer is: 314']
      - >-
        You can make a baby cry by picking them up and holding them in an
        awkward position, rubbing their nose or ears (carefully), stimulating a
        reflex points on their body, shouting or speaking in a harsh tone,
        playing loud noises near them or changing their daily routines suddenly.

SentenceTransformer based on Alibaba-NLP/gte-Qwen2-1.5B-instruct

This is a sentence-transformers model finetuned from Alibaba-NLP/gte-Qwen2-1.5B-instruct. It maps sentences & paragraphs to a 1536-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Base model: Alibaba-NLP/gte-Qwen2-1.5B-instruct
Maximum Sequence Length: 8192 tokens
Output Dimensionality: 1536 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': 8192, 'do_lower_case': False}) with Transformer model: Qwen2Model 
  (1): Pooling({'word_embedding_dimension': 1536, 'pooling_mode_cls_token': False, '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': True, 'include_prompt': True})
  (2): Normalize()
)

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("sentence_transformers_model_id")
# Run inference
sentences = [
    '["Question: Given the operation $x@y = xy - 2x$, what is the value of $(7@4) - (4@7)$?\\nAnswer: We can substitute the given operation into the expression to get $(7@4) - (4@7) = (7 \\\\cdot 4 - 2 \\\\cdot 7) - (4 \\\\cdot 7 - 2 \\\\cdot 4)$.\\nSimplifying, we have $28 - 14 - 28 + 8 = \\\\boxed{-6}$.\\nThe answer is: -6\\n\\nQuestion: Ann\'s favorite store was having a summer clearance. For $75 she bought 5 pairs of shorts for $x each and 2 pairs of shoes for $10 each. She also bought 4 tops, all at the same price. Each top cost 5. What is the value of unknown variable x?\\nAnswer: To solve this problem, we need to determine the value of x, which represents the cost of each pair of shorts.\\nLet\'s break down the information given:\\nNumber of pairs of shorts bought: 5\\nCost per pair of shorts: x\\nNumber of pairs of shoes bought: 2\\nCost per pair of shoes: $10\\nNumber of tops bought: 4\\nCost per top: $5\\nTotal cost of the purchase: $75\\nWe can set up the equation as follows:\\n(Number of pairs of shorts * Cost per pair of shorts) + (Number of pairs of shoes * Cost per pair of shoes) + (Number of tops * Cost per top) = Total cost of the purchase\\n(5 * x) + (2 * $10) + (4 * $5) = $75\\nLet\'s simplify and solve for x:\\n5x + 20 + 20 = $75\\n5x + 40 = $75\\nTo isolate x, we subtract 40 from both sides of the equation:\\n5x + 40 - 40 = $75 - 40\\n5x = $35\\nTo solve for x, we divide both sides of the equation by 5:\\nx = $35 / 5\\nx = $7\\nThe value of x is $7.\\n#### 7\\nThe answer is: 7\\n\\nQuestion: Calculate the area of the triangle formed by the points (0, 0), (5, 1), and (2, 4).\\nAnswer: We can use the Shoelace Formula to find the area of the triangle.\\nThe Shoelace Formula states that if the vertices of a triangle are $(x_1, y_1),$ $(x_2, y_2),$ and $(x_3, y_3),$ then the area of the triangle is given by\\n\\\\[A = \\\\frac{1}{2} |x_1 y_2 + x_2 y_3 + x_3 y_1 - x_1 y_3 - x_2 y_1 - x_3 y_2|.\\\\]\\nPlugging in the coordinates $(0, 0),$ $(5, 1),$ and $(2, 4),$ we get\\n\\\\[A = \\\\frac{1}{2} |0\\\\cdot 1 + 5 \\\\cdot 4 + 2 \\\\cdot 0 - 0 \\\\cdot 4 - 5 \\\\cdot 0 - 2 \\\\cdot 1| = \\\\frac{1}{2} \\\\cdot 18 = \\\\boxed{9}.\\\\]\\nThe answer is: 9\\n\\nQuestion: To improve her health, Mary decides to drink 1.5 liters of water a day as recommended by her doctor. Mary\'s glasses hold x mL of water. How many glasses of water should Mary drink per day to reach her goal?\\nIf we know the answer to the above question is 6, what is the value of unknown variable x?\\nAnswer: Mary wants to drink 1.5 liters of water per day, which is equal to 1500 mL.\\nMary\'s glasses hold x mL of water.\\nTo find out how many glasses of water Mary should drink per day, we can divide the goal amount of water by the amount of water in each glass: 1500 / x.\\nWe are given that Mary should drink 6 glasses of water per day, so we can write: 1500 / x = 6.\\nSolving for x, we get: x = 250.\\nThe value of x is 250.\\n#### 250\\nThe answer is: 250\\n\\nQuestion: Seymour runs a plant shop. He has 4 flats of petunias with 8 petunias per flat, 3 flats of roses with 6 roses per flat, and two Venus flytraps. Each petunia needs 8 ounces of fertilizer, each rose needs 3 ounces of fertilizer, and each Venus flytrap needs 2 ounces of fertilizer. How many ounces of fertilizer does Seymour need in total?\\nAnswer:"]',
    "[' In total, there are 4 flats x 8 petunias/flat = 32 petunias.\\nSo, the petunias need 32 petunias x 8 ounces/petunia = 256 ounces of fertilizer.\\nThere are 3 flats x 6 roses/flat = 18 roses in total.\\nSo, the roses need 18 roses x 3 ounces/rose = 54 ounces of fertilizer.\\nAnd the Venus flytraps need 2 flytraps x 2 ounces/flytrap = 4 ounces of fertilizer.\\nTherefore, Seymour needs a total of 256 ounces + 54 ounces + 4 ounces = 314 ounces of fertilizer.\\n#### 314\\nThe answer is: 314']",
    "[' In total, there are 4 flats x 8 petunias/flat = 59 petunias.\\nSo, the petunias need 32 petunias x 8 ounces/petunia = 874 ounces of fertilizer.\\nThere are 3 flats x 6 roses/flat = 99 roses in total.\\nSo, the roses need 18 roses x 3 ounces/rose = 40 ounces of fertilizer.\\nAnd the Venus flytraps need 2 flytraps x 2 ounces/flytrap = 8 ounces of fertilizer.\\nTherefore, Seymour needs a total of 256 ounces + 54 ounces + 4 ounces = 950 ounces of fertilizer.\\n#### 314\\nThe answer is: 314']",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1536]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Training Details

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_eval_batch_size: 4
gradient_accumulation_steps: 4
learning_rate: 2e-05
lr_scheduler_type: cosine
warmup_ratio: 0.1
warmup_steps: 5
bf16: True
tf32: True
optim: adamw_torch_fused
gradient_checkpointing: True
gradient_checkpointing_kwargs: {'use_reentrant': False}
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: 8
per_device_eval_batch_size: 4
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 4
eval_accumulation_steps: None
learning_rate: 2e-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
lr_scheduler_kwargs: {}
warmup_ratio: 0.1
warmup_steps: 5
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
save_safetensors: True
save_on_each_node: False
save_only_model: False
restore_callback_states_from_checkpoint: False
no_cuda: False
use_cpu: False
use_mps_device: False
seed: 42
data_seed: None
jit_mode_eval: False
use_ipex: False
bf16: True
fp16: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: True
local_rank: 0
ddp_backend: None
tpu_num_cores: None
tpu_metrics_debug: False
debug: []
dataloader_drop_last: True
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_fused
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: False
resume_from_checkpoint: None
hub_model_id: None
hub_strategy: every_save
hub_private_repo: False
hub_always_push: False
gradient_checkpointing: True
gradient_checkpointing_kwargs: {'use_reentrant': False}
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
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	retrival loss	sts loss	reranking loss
0.5222	500	0.7949	0.0187	2.6522	0.2931
1.0444	1000	0.6813	0.0139	2.5109	0.2695
1.5666	1500	0.5148	0.0118	2.5270	0.2807
2.0888	2000	0.48	0.0114	2.5418	0.2791
2.6110	2500	0.3782	0.0117	2.5740	0.2787

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.2.0+cu121
Accelerate: 0.32.1
Datasets: 2.20.0
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",
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply}, 
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

CoSENTLoss

@online{kexuefm-8847,
    title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
    author={Su Jianlin},
    year={2022},
    month={Jan},
    url={https://kexue.fm/archives/8847},
}