checkpoint-933/README.md

metadata

base_model: colorfulscoop/sbert-base-ja
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:7464
  - loss:CosineSimilarityLoss
widget:
  - source_sentence: なにか思い出せることは？
    sentences:
      - なんて言った？
      - なにが欲しい？
      - いつ失くしたの？
  - source_sentence: 木にスカーフがひっかかってる?
    sentences:
      - 物の姿を変えられる魔法
      - 木にスカーフがひっかかってる?
      - あの木の上のやつ、スカーフ？
  - source_sentence: どんなスカーフ？
    sentences:
      - お気に入りの食べ物は？
      - くさい
      - もしもし
  - source_sentence: どうやったら猫が話せる？
    sentences:
      - どっちがいいと思う？
      - ワゴンにスカーフはあった？
      - 井戸へ行ったことある？
  - source_sentence: なにか思い出せることは？
    sentences:
      - 村人はどんな呪文が使えるの？
      - 町って？
      - キミならどっちがいい？
model-index:
  - name: SentenceTransformer based on colorfulscoop/sbert-base-ja
    results:
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: custom arc semantics data jp
          type: custom-arc-semantics-data-jp
        metrics:
          - type: cosine_accuracy
            value: 0.9705409748259239
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.5118279457092285
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.9102773246329527
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.45031607151031494
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.8857142857142857
            name: Cosine Precision
          - type: cosine_recall
            value: 0.9362416107382551
            name: Cosine Recall
          - type: cosine_ap
            value: 0.9236294738598163
            name: Cosine Ap
          - type: dot_accuracy
            value: 0.9694697375468666
            name: Dot Accuracy
          - type: dot_accuracy_threshold
            value: 251.2455596923828
            name: Dot Accuracy Threshold
          - type: dot_f1
            value: 0.9060955518945634
            name: Dot F1
          - type: dot_f1_threshold
            value: 246.36648559570312
            name: Dot F1 Threshold
          - type: dot_precision
            value: 0.889967637540453
            name: Dot Precision
          - type: dot_recall
            value: 0.9228187919463087
            name: Dot Recall
          - type: dot_ap
            value: 0.9196731890884118
            name: Dot Ap
          - type: manhattan_accuracy
            value: 0.9716122121049813
            name: Manhattan Accuracy
          - type: manhattan_accuracy_threshold
            value: 514.571533203125
            name: Manhattan Accuracy Threshold
          - type: manhattan_f1
            value: 0.9132569558101473
            name: Manhattan F1
          - type: manhattan_f1_threshold
            value: 514.571533203125
            name: Manhattan F1 Threshold
          - type: manhattan_precision
            value: 0.8913738019169329
            name: Manhattan Precision
          - type: manhattan_recall
            value: 0.9362416107382551
            name: Manhattan Recall
          - type: manhattan_ap
            value: 0.9255015709844487
            name: Manhattan Ap
          - type: euclidean_accuracy
            value: 0.9721478307445099
            name: Euclidean Accuracy
          - type: euclidean_accuracy_threshold
            value: 23.195274353027344
            name: Euclidean Accuracy Threshold
          - type: euclidean_f1
            value: 0.9147540983606558
            name: Euclidean F1
          - type: euclidean_f1_threshold
            value: 23.195274353027344
            name: Euclidean F1 Threshold
          - type: euclidean_precision
            value: 0.8942307692307693
            name: Euclidean Precision
          - type: euclidean_recall
            value: 0.9362416107382551
            name: Euclidean Recall
          - type: euclidean_ap
            value: 0.9259440018381992
            name: Euclidean Ap
          - type: max_accuracy
            value: 0.9721478307445099
            name: Max Accuracy
          - type: max_accuracy_threshold
            value: 514.571533203125
            name: Max Accuracy Threshold
          - type: max_f1
            value: 0.9147540983606558
            name: Max F1
          - type: max_f1_threshold
            value: 514.571533203125
            name: Max F1 Threshold
          - type: max_precision
            value: 0.8942307692307693
            name: Max Precision
          - type: max_recall
            value: 0.9362416107382551
            name: Max Recall
          - type: max_ap
            value: 0.9259440018381992
            name: Max Ap

SentenceTransformer based on colorfulscoop/sbert-base-ja

This is a sentence-transformers model finetuned from colorfulscoop/sbert-base-ja. 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: colorfulscoop/sbert-base-ja
• 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: BertModel 
  (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})
)

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 = [
    'なにか思い出せることは？',
    '村人はどんな呪文が使えるの？',
    'キミならどっちがいい？',
]
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

Binary Classification

• Dataset: custom-arc-semantics-data-jp
• Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.9705
cosine_accuracy_threshold	0.5118
cosine_f1	0.9103
cosine_f1_threshold	0.4503
cosine_precision	0.8857
cosine_recall	0.9362
cosine_ap	0.9236
dot_accuracy	0.9695
dot_accuracy_threshold	251.2456
dot_f1	0.9061
dot_f1_threshold	246.3665
dot_precision	0.89
dot_recall	0.9228
dot_ap	0.9197
manhattan_accuracy	0.9716
manhattan_accuracy_threshold	514.5715
manhattan_f1	0.9133
manhattan_f1_threshold	514.5715
manhattan_precision	0.8914
manhattan_recall	0.9362
manhattan_ap	0.9255
euclidean_accuracy	0.9721
euclidean_accuracy_threshold	23.1953
euclidean_f1	0.9148
euclidean_f1_threshold	23.1953
euclidean_precision	0.8942
euclidean_recall	0.9362
euclidean_ap	0.9259
max_accuracy	0.9721
max_accuracy_threshold	514.5715
max_f1	0.9148
max_f1_threshold	514.5715
max_precision	0.8942
max_recall	0.9362
max_ap	0.9259

Training Details

Training Dataset

Unnamed Dataset

• Size: 7,464 training samples
• Columns: text1, text2, and label
• Approximate statistics based on the first 1000 samples:

  	text1	text2	label
  type	string	string	int
  details	min: 4 tokens mean: 8.32 tokens max: 15 tokens	min: 4 tokens mean: 7.93 tokens max: 15 tokens	0: ~86.30% 1: ~13.70%

• Samples:

  text1	text2	label
  昨日夕飯にチキンヌードル食べた？	何か企んでる？	0
  どっちも欲しくない	お気に入りの食べ物は？	0
  見た目を変える魔法	物の姿を変えられる魔法	1

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Evaluation Dataset

Unnamed Dataset

• Size: 1,867 evaluation samples
• Columns: text1, text2, and label
• Approximate statistics based on the first 1000 samples:

  	text1	text2	label
  type	string	string	int
  details	min: 4 tokens mean: 8.24 tokens max: 15 tokens	min: 4 tokens mean: 8.1 tokens max: 15 tokens	0: ~83.40% 1: ~16.60%

• Samples:

  text1	text2	label
  例えば？	どうも	0
  何を作ったの？	君は何でここにいるの？	0
  昨日夕飯にビーフシチュー食べた？	屋根裏って？	0

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• learning_rate: 2e-05
• num_train_epochs: 1
• warmup_ratio: 0.4
• fp16: True
• 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: 8
• per_device_eval_batch_size: 8
• 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: 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: 1
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.4
• 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: 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: 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
• eval_on_start: False
• eval_use_gather_object: False
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss	custom-arc-semantics-data-jp_max_ap
1.0	933	0.0601	0.0303	0.9259

Framework Versions

• Python: 3.10.14
• Sentence Transformers: 3.1.1
• Transformers: 4.44.2
• PyTorch: 2.4.1+cu121
• Accelerate: 0.34.2
• 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",
}