README.md

metadata

tags:
  - Data-Science
  - Machine-Learning
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:67416
  - loss:MultipleNegativesRankingLoss
widget:
  - source_sentence: ', k on their diagonal and zero elsewhere.'
    sentences:
      - |-
        The sample ACF and PACF of the data for that time period in Figure 6.17

        466
        TRANSFER FUNCTIONS AND INTERVENTION MODELS
        100
        80
        60
        Week
        Week 88
        40
        20
        0
        100,000
        160,000
        140,000
        Sales
        200,000
        180,000
        220,000
        120,000
        FIGURE 6.16
        Time series plot of the weekly sales data.
      - |-
        Just as in equation 6.10, we can write
        X = u1a1vT
        1 + u2a2vT
        2 + · · · + ukakvT
        k
        (6.29)
        We can ignore the corresponding ui, vi of very small, though nonzero,
        ai and can still reconstruct X without too much error.
      - |-
        13.8
        Multiple Kernel Learning
        It is possible to construct new kernels by combining simpler kernels.
  - source_sentence: >-
      The main difference is that a node appears at most once as a neighbor of
      an-

      other node, whereas a word might appear more than once in the context of
      another word.
    sentences:
      - >-
        3.2.6

        A Decoupled View of Vector-Centric Backpropagation

        In the previous discussion, two equivalent ways of computing the updates
        based on Equa-

        tions 3.12 and 3.18 were provided.
      - |-
        (4.59),
        we obtain
        eT −(1 −𝜆)eT−1 = (yT −̂yT−1) −(1 −𝜆)(yT−1 −̂yT−2)
        = yT −yT−1 −̂yT−1 + 𝜆yT−1 + (1 −𝜆)̂yT−2
        ⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟
        =̂yT−1
        = yT −yT−1 −̂yT−1 + ̂yT−1
        = yT −yT−1.
      - 8This fact is not evident in the toy example of Figure 2.17.
  - source_sentence: |-
      This influence is specified by the conditional probability
      P(Y|X).
    sentences:
      - >-
        Seasonality

        is the component of time series behavior that repeats on a regular
        basis,

        such as each year.
      - >-
        Note that one of the

        classes is defined by strongly non-zero values in the first and third
        dimensions, whereas the

        second class is defined by strongly non-zero values in the second and
        fourth dimensions.
      - >-
        The nodes and the arcs between the nodes define the struc-

        ture of the network, and the conditional probabilities are the
        parameters

        given the structure.
  - source_sentence: |-
      238
      9
      Decision Trees
      Rokach, L., and O. Maimon.
    sentences:
      - >-
        “Top-Down Induction of Decision Trees

        Classifiers—A Survey.” IEEE Transactions on Systems, Man, and
        Cybernetics–

        Part C 35:476–487.
      - |-
        The only feedback is at the
        end of the game when we win or lose the game.
      - >-
        Subsequently, this computation is propagated

        in the backwards direction with dynamic programming updates (similar to
        Equation 3.8).
  - source_sentence: |-
      Therefore, one can use L1-regularization
      to estimate which features are predictive to the application at hand.
    sentences:
      - Blumer, A., A. Ehrenfeucht, D. Haussler, and M. K. Warmuth.
      - >-
        What about the connections in the hidden layers whose weights are set to
        0?
      - >-
        In cases where computational complexity is important, such

        as in a production setting where thousands of models are being fit, it
        may not be

        worth the extra computational effort.
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - pearson_cosine
  - spearman_cosine
model-index:
  - name: SentenceTransformer
    results:
      - task:
          type: semantic-similarity
          name: Semantic Similarity
        dataset:
          name: val
          type: val
        metrics:
          - type: pearson_cosine
            value: null
            name: Pearson Cosine
          - type: spearman_cosine
            value: null
            name: Spearman Cosine
license: apache-2.0
language:
  - en
base_model:
  - sentence-transformers/all-mpnet-base-v2
datasets:
  - DigitalAsocial/ds-tb-5-raw

SentenceTransformer

This is a sentence-transformers model trained. 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
• Maximum Sequence Length: 384 tokens
• Output Dimensionality: 768 dimensions
• 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': 384, 'do_lower_case': False, 'architecture': 'MPNetModel'})
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (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 = [
    'Therefore, one can use L1-regularization\nto estimate which features are predictive to the application at hand.',
    'What about the connections in the hidden layers whose weights are set to 0?',
    'In cases where computational complexity is important, such\nas in a production setting where thousands of models are being fit, it may not be\nworth the extra computational effort.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000,  0.4198,  0.2089],
#         [ 0.4198,  1.0000, -0.0369],
#         [ 0.2089, -0.0369,  1.0000]])

Evaluation

Metrics

Semantic Similarity

• Dataset: val
• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	nan
spearman_cosine	nan

Training Details

Training Dataset

Unnamed Dataset

• Size: 67,416 training samples
• Columns: sentence_0 and sentence_1
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1
  type	string	string
  details	min: 7 tokens mean: 39.68 tokens max: 384 tokens	min: 8 tokens mean: 39.93 tokens max: 384 tokens

• Samples:

  sentence_0	sentence_1
  Leveraging Redundancies in Weights It was shown in [94] that the vast majority of the weights in a neural network are redundant.	Furthermore, it is assumed that k ≪min{m1, m2}.
  Aran, O., O. T. Yıldız, and E. Alpaydın.	“An Incremental Framework Based on Cross-Validation for Estimating the Architecture of a Multilayer Percep- tron.” International Journal of Pattern Recognition and Artificial Intelligence 23:159–190.
  (a) (d) (e) (f) 29 Code is life input_decoder = Input(shape=(latent_dim,), name="decoder_input") decoder_h = Dense(intermediate_dim, activation='relu', name="decoder_h")(input_decoder) x_decoded = Dense(original_dim, activation='sigmoid', name="flat_decoded")(decoder_h) decoder = Model(input_decoder, x_decoded, name="decoder") We can now combine the encoder and the decoder into a single VAE model.	output_combined = decoder(encoder(x)[2]) vae = Model(x, output_combined) vae.summary() Next, we get to the more familiar parts of machine learning: defining a loss function so our autoencoder can train.

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "gather_across_devices": false
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• num_train_epochs: 6
• fp16: True
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: no
• prediction_loss_only: True
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• 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
• num_train_epochs: 6
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• warmup_steps: 0
• 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
• 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}
• parallelism_config: None
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• project: huggingface
• trackio_space_id: trackio
• 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: None
• hub_always_push: False
• hub_revision: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• 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
• include_tokens_per_second: False
• include_num_input_tokens_seen: no
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• liger_kernel_config: None
• eval_use_gather_object: False
• average_tokens_across_devices: True
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	val_spearman_cosine
0.1187	500	1.5671	-
0.2373	1000	1.2804	-
0.3560	1500	1.1256	-
0.4746	2000	0.9789	-
0.5933	2500	0.8839	-
0.7119	3000	0.7748	-
0.8306	3500	0.73	-
0.9492	4000	0.698	-
1.0	4214	-	nan

Framework Versions

• Python: 3.11.7
• Sentence Transformers: 5.1.1
• Transformers: 4.57.0
• PyTorch: 2.5.1+cu121
• Accelerate: 1.12.0
• Datasets: 4.4.1
• Tokenizers: 0.22.1

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}
}