Hugging Face's logo

srikarvar
/
fine_tuned_model_6

Sentence Similarity
sentence-transformers
Safetensors
bert
feature-extraction
Generated from Trainer
dataset_size:1972
loss:OnlineContrastiveLoss
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
Community
fine_tuned_model_6 / README.md
srikarvar's picture
srikarvar
Add new SentenceTransformer model.
268c227 verified
preview code
|
raw
history blame contribute delete
26.1 kB
metadata 
base_model: intfloat/multilingual-e5-small
datasets: []
language: []
library_name: sentence-transformers
metrics:
  - cosine_accuracy
  - cosine_accuracy_threshold
  - cosine_f1
  - cosine_f1_threshold
  - cosine_precision
  - cosine_recall
  - cosine_ap
  - dot_accuracy
  - dot_accuracy_threshold
  - dot_f1
  - dot_f1_threshold
  - dot_precision
  - dot_recall
  - dot_ap
  - manhattan_accuracy
  - manhattan_accuracy_threshold
  - manhattan_f1
  - manhattan_f1_threshold
  - manhattan_precision
  - manhattan_recall
  - manhattan_ap
  - euclidean_accuracy
  - euclidean_accuracy_threshold
  - euclidean_f1
  - euclidean_f1_threshold
  - euclidean_precision
  - euclidean_recall
  - euclidean_ap
  - max_accuracy
  - max_accuracy_threshold
  - max_f1
  - max_f1_threshold
  - max_precision
  - max_recall
  - max_ap
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:1972
  - loss:OnlineContrastiveLoss
widget:
  - source_sentence: Who invented the World Wide Web?
    sentences:
      - >-
        What universities does Chart Industries recruit new grads from? What
        majors are they looking for?
      - Who invented the internet?
      - What are the benefits of a balanced diet?
  - source_sentence: Who was the second President of the United States?
    sentences:
      - Second leader of the USA
      - How many people live in Germany?
      - >-
        How do I get over someone I loved now that we broke up last year and I
        still miss her?
  - source_sentence: How did you first come across porn?
    sentences:
      - How were you first introduced to porn?
      - Date of signing the Declaration of Independence
      - Do we need the IPC section 375?
  - source_sentence: How to invest in cryptocurrency?
    sentences:
      - What is the cheapest toothpaste?
      - What are the environmental advantages of recycling?
      - How to trade cryptocurrency?
  - source_sentence: What is the speed of a racing drone?
    sentences:
      - Who was the first person to swim across the Atlantic?
      - >-
        People say "don't try to please others." Does being nice to others mean
        pleasing them?
      - What is the speed of a racing car?
model-index:
  - name: SentenceTransformer based on intfloat/multilingual-e5-small
    results:
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: pair class dev
          type: pair-class-dev
        metrics:
          - type: cosine_accuracy
            value: 0.8772727272727273
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.8647407293319702
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.8682926829268292
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.8647407293319702
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.8725490196078431
            name: Cosine Precision
          - type: cosine_recall
            value: 0.8640776699029126
            name: Cosine Recall
          - type: cosine_ap
            value: 0.9227827652550092
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.8772727272727273
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 0.8647407293319702
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.8682926829268292
            name: Dot F1
          - type: dot_f1_threshold
            value: 0.8647407293319702
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.8725490196078431
            name: Dot Precision
          - type: dot_recall
            value: 0.8640776699029126
            name: Dot Recall
          - type: dot_ap
            value: 0.9227827652550092
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.8772727272727273
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 8.025869369506836
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.8703703703703703
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 9.006706237792969
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.831858407079646
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.912621359223301
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.9221498446893291
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.8772727272727273
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 0.5201112031936646
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.8682926829268292
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 0.5201112031936646
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.8725490196078431
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.8640776699029126
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.9227827652550092
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.8772727272727273
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 8.025869369506836
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.8703703703703703
            name: Max F1
          - type: max_f1_threshold
            value: 9.006706237792969
            name: Max F1 Threshold
          - type: max_precision
            value: 0.8725490196078431
            name: Max Precision
          - type: max_recall
            value: 0.912621359223301
            name: Max Recall
          - type: max_ap
            value: 0.9227827652550092
            name: Max Ap
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: pair class test
          type: pair-class-test
        metrics:
          - type: cosine_accuracy
            value: 0.8772727272727273
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.8647407293319702
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.8682926829268292
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.8647407293319702
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.8725490196078431
            name: Cosine Precision
          - type: cosine_recall
            value: 0.8640776699029126
            name: Cosine Recall
          - type: cosine_ap
            value: 0.9227827652550092
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.8772727272727273
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 0.8647407293319702
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.8682926829268292
            name: Dot F1
          - type: dot_f1_threshold
            value: 0.8647407293319702
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.8725490196078431
            name: Dot Precision
          - type: dot_recall
            value: 0.8640776699029126
            name: Dot Recall
          - type: dot_ap
            value: 0.9227827652550092
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.8772727272727273
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 8.025869369506836
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.8703703703703703
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 9.006706237792969
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.831858407079646
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.912621359223301
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.9221498446893291
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.8772727272727273
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 0.5201112031936646
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.8682926829268292
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 0.5201112031936646
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.8725490196078431
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.8640776699029126
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.9227827652550092
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.8772727272727273
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 8.025869369506836
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.8703703703703703
            name: Max F1
          - type: max_f1_threshold
            value: 9.006706237792969
            name: Max F1 Threshold
          - type: max_precision
            value: 0.8725490196078431
            name: Max Precision
          - type: max_recall
            value: 0.912621359223301
            name: Max Recall
          - type: max_ap
            value: 0.9227827652550092
            name: Max Ap

SentenceTransformer based on intfloat/multilingual-e5-small

This is a sentence-transformers model finetuned from intfloat/multilingual-e5-small. It maps sentences & paragraphs to a 384-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: intfloat/multilingual-e5-small
  • Maximum Sequence Length: 512 tokens
  • Output Dimensionality: 384 tokens
  • Similarity Function: Cosine Similarity

Model Sources

Full Model Architecture 

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 384, '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("srikarvar/fine_tuned_model_6")
# Run inference
sentences = [
    'What is the speed of a racing drone?',
    'What is the speed of a racing car?',
    'Who was the first person to swim across the Atlantic?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

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

Evaluation

Metrics

Binary Classification

Metric Value
cosine_accuracy 0.8773
cosine_accuracy_threshold 0.8647
cosine_f1 0.8683
cosine_f1_threshold 0.8647
cosine_precision 0.8725
cosine_recall 0.8641
cosine_ap 0.9228
dot_accuracy 0.8773
dot_accuracy_threshold 0.8647
dot_f1 0.8683
dot_f1_threshold 0.8647
dot_precision 0.8725
dot_recall 0.8641
dot_ap 0.9228
manhattan_accuracy 0.8773
manhattan_accuracy_threshold 8.0259
manhattan_f1 0.8704
manhattan_f1_threshold 9.0067
manhattan_precision 0.8319
manhattan_recall 0.9126
manhattan_ap 0.9221
euclidean_accuracy 0.8773
euclidean_accuracy_threshold 0.5201
euclidean_f1 0.8683
euclidean_f1_threshold 0.5201
euclidean_precision 0.8725
euclidean_recall 0.8641
euclidean_ap 0.9228
max_accuracy 0.8773
max_accuracy_threshold 8.0259
max_f1 0.8704
max_f1_threshold 9.0067
max_precision 0.8725
max_recall 0.9126
max_ap 0.9228

Binary Classification

Metric Value
cosine_accuracy 0.8773
cosine_accuracy_threshold 0.8647
cosine_f1 0.8683
cosine_f1_threshold 0.8647
cosine_precision 0.8725
cosine_recall 0.8641
cosine_ap 0.9228
dot_accuracy 0.8773
dot_accuracy_threshold 0.8647
dot_f1 0.8683
dot_f1_threshold 0.8647
dot_precision 0.8725
dot_recall 0.8641
dot_ap 0.9228
manhattan_accuracy 0.8773
manhattan_accuracy_threshold 8.0259
manhattan_f1 0.8704
manhattan_f1_threshold 9.0067
manhattan_precision 0.8319
manhattan_recall 0.9126
manhattan_ap 0.9221
euclidean_accuracy 0.8773
euclidean_accuracy_threshold 0.5201
euclidean_f1 0.8683
euclidean_f1_threshold 0.5201
euclidean_precision 0.8725
euclidean_recall 0.8641
euclidean_ap 0.9228
max_accuracy 0.8773
max_accuracy_threshold 8.0259
max_f1 0.8704
max_f1_threshold 9.0067
max_precision 0.8725
max_recall 0.9126
max_ap 0.9228

Training Details

Training Dataset

Unnamed Dataset

  • Size: 1,972 training samples
  • Columns: sentence1, sentence2, and label
  • Approximate statistics based on the first 1000 samples:
    sentence1 sentence2 label
    type string string int
    details
    • min: 6 tokens
    • mean: 12.22 tokens
    • max: 53 tokens
    • min: 4 tokens
    • mean: 11.89 tokens
    • max: 48 tokens
    • 0: ~51.60%
    • 1: ~48.40%
  • Samples:
    sentence1 sentence2 label
    What is the distance between the Earth and Mars? What is the distance between the Earth and Saturn? 0
    Tell me a joke Make me laugh with a joke 1
    How can I make money online with free of cost? How do I to make money online? 1
  • Loss: OnlineContrastiveLoss

Evaluation Dataset

Unnamed Dataset

  • Size: 220 evaluation samples
  • Columns: sentence1, sentence2, and label
  • Approximate statistics based on the first 1000 samples:
    sentence1 sentence2 label
    type string string int
    details
    • min: 6 tokens
    • mean: 12.44 tokens
    • max: 44 tokens
    • min: 5 tokens
    • mean: 12.4 tokens
    • max: 55 tokens
    • 0: ~53.18%
    • 1: ~46.82%
  • Samples:
    sentence1 sentence2 label
    Who discovered the structure of DNA? Scientist who identified the double helix 1
    How to create a website from scratch? How to create a blog from scratch? 0
    What is the population of New York City? What is the population of Chicago? 0
  • Loss: OnlineContrastiveLoss

Training Hyperparameters

Non-Default Hyperparameters

  • eval_strategy: epoch
  • per_device_train_batch_size: 32
  • per_device_eval_batch_size: 32
  • gradient_accumulation_steps: 2
  • warmup_ratio: 0.1
  • load_best_model_at_end: True
  • optim: adamw_torch_fused
  • batch_sampler: no_duplicates

All Hyperparameters

Click to expand
  • overwrite_output_dir: False
  • do_predict: False
  • eval_strategy: epoch
  • prediction_loss_only: True
  • per_device_train_batch_size: 32
  • per_device_eval_batch_size: 32
  • per_gpu_train_batch_size: None
  • per_gpu_eval_batch_size: None
  • gradient_accumulation_steps: 2
  • eval_accumulation_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: linear
  • lr_scheduler_kwargs: {}
  • warmup_ratio: 0.1
  • 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
  • use_ipex: False
  • bf16: False
  • fp16: False
  • 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: True
  • 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: 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
  • batch_sampler: no_duplicates
  • multi_dataset_batch_sampler: proportional

Training Logs

Epoch Step Training Loss loss pair-class-dev_max_ap pair-class-test_max_ap
0 0 - - 0.6615 -
0.3226 10 1.7113 - - -
0.6452 20 0.9588 - - -
0.9677 30 0.9243 - - -
1.0 31 - 0.8485 0.8985 -
1.2903 40 0.689 - - -
1.6129 50 0.4289 - - -
1.9355 60 0.4655 - - -
2.0 62 - 0.8143 0.9203 -
2.2581 70 0.4183 - - -
2.5806 80 0.3038 - - -
2.9032 90 0.2979 - - -
3.0 93 - 0.8121 0.9228 0.9228
  • The bold row denotes the saved checkpoint.

Framework Versions

  • Python: 3.10.12
  • Sentence Transformers: 3.0.1
  • Transformers: 4.41.2
  • PyTorch: 2.1.2+cu121
  • Accelerate: 0.32.1
  • Datasets: 2.19.1
  • 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",
}