semantic_text_sim-v1

SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L6-v2 on the prep-manga-recom dataset. 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: sentence-transformers/all-MiniLM-L6-v2 <!-- at revision c9745ed1d9f207416be6d2e6f8de32d1f16199bf -->
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • prep-manga-recom

<!-- - **Language:** Unknown --> <!-- - **License:** Unknown -->

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': 256, 'do_lower_case': False, 'architecture': '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("tonnnnnnnnnnnnn/semantic_text_sim-v1")
# Run inference
sentences = [
    "Yamato Akitsuki has recently moved to Tokyo, and in order to maintain his less than wealthy lifestyle, he currently works at his aunt's bathhouse. While walking by the school one evening he sees a girl by the name of Suzuka practicing the high jump, and is instantly in love. Even better is the realization that Suzuka lives next door! Determined to prove himself worthy of her affections, Yamato decides to join the school's track team and show her what he's got, but things won't be so easy; for Suzuka has a love interest of her own, and it isn't Yamato...",
    'What is the name of the manga where the protagonist is involved in a car crash and wakes up in life after death?',
    'What is the name of the manga with the description "As Japan and the rest of the world begins the process of rebuilding after the fall of \'Friend\', Kenji and his friends must try to uncover the identity of the second \'Friend\' and other unresolved mysteries. Before the world is once again thrown into turmoil, they must search deep into their childhood memories to find the key to save the world one more time from the threat of \'Friend\'; some mysteries cannot be left unsolved."',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000, -0.0386,  0.1505],
#         [-0.0386,  1.0000,  0.1664],
#         [ 0.1505,  0.1664,  1.0000]])

<!-- ### Direct Usage (Transformers) <details><summary>Click to see the direct usage in Transformers</summary> </details> --> <!-- ### Downstream Usage (Sentence Transformers) You can finetune this model on your own dataset. <details><summary>Click to expand</summary> </details> --> <!-- ### Out-of-Scope Use *List how the model may foreseeably be misused and address what users ought not to do with the model.* -->

Evaluation

Metrics

Semantic Similarity

• Datasets: manga-dev and manga-test
• Evaluated with <code>EmbeddingSimilarityEvaluator</code>

<!-- ## Bias, Risks and Limitations *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* --> <!-- ### Recommendations *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* -->

Training Details

Training Dataset

prep-manga-recom

• Dataset: prep-manga-recom at 99869dd
• Size: 8,100 training samples
• Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	<ul><li>min: 7 tokens</li><li>mean: 98.51 tokens</li><li>max: 256 tokens</li></ul>	<ul><li>min: 7 tokens</li><li>mean: 34.25 tokens</li><li>max: 228 tokens</li></ul>	<ul><li>min: -0.2</li><li>mean: 0.09</li><li>max: 0.64</li></ul>

• Samples:

  sentence1	sentence2	score
  <code>A girl struggles to survive in a zombie apocalypse and ends up finding herself a boyfriend! But there is a problem...he is undead!</code>	<code>What is the name of a manga about a man with no money and no place to belong who tries to end his life, however is unable to take the last step?</code>	<code>0.2546822726726532</code>
  <code>The second season of Terror Man.</code>	<code>What is the name of a manga about a 30-year-old man who meets a stoic graphic designer and falls in love?</code>	<code>0.0785623267292976</code>
  <code>After the loss of his loyal dog, Winter, Ji Seungwoo is wounded by an irreplaceable loss and sadness of his 22 year-long companionship. But one day he finds an adorable little kitten waiting in a cardboard box next to his front door. Being the gentle and kind-hearted man he was, he decides to bring him home. However, upon waking up the next morning he is no longer in bed with a cute baby kitten, but a grown, handsome man?! “I am this person’s owner!” he says! Stay tuned to watch how Seungwoo and Mr. Cat’s relationship develops…!</code>	<code>What is the name of the manga about a fraudulent exorcist and his student assistant?</code>	<code>-0.021251555532217</code>

• Loss: <code>CosineSimilarityLoss</code> with these parameters:

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

Evaluation Dataset

prep-manga-recom

• Dataset: prep-manga-recom at 99869dd
• Size: 900 evaluation samples
• Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
• Approximate statistics based on the first 900 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	<ul><li>min: 7 tokens</li><li>mean: 98.39 tokens</li><li>max: 233 tokens</li></ul>	<ul><li>min: 7 tokens</li><li>mean: 34.49 tokens</li><li>max: 234 tokens</li></ul>	<ul><li>min: -0.22</li><li>mean: 0.09</li><li>max: 0.69</li></ul>

• Samples:

  sentence1	sentence2	score
  <code>Shuichiro used to secretly go to the terrace of his school, during lunch, when skipping classes and when he didn't want to go home. There he knew Konno. Some say he was difficult to talk, that he is intimidating and scary... but there is who say he is a camera fanatic and an unexpectedly nice guy. Respect, envy, jealousy and inferiority complex will mark their friendship. A masked kid who is hiding behind a fake smile all his tormented thought and a simple-hearted friend who start to mobilize all the conflicts without realizing.</code>	<code>What is Midori’s relationship with Hinata?</code>	<code>0.1419442445039749</code>
  <code>Baki Hanma is a generally happy student with a rather odd hobby; he likes fighting. Specifically, he likes fighting in a secret martial arts tournament that gathers the greatest fighters of the world and pits them against each other in really nasty combat. New Grappler Baki takes off where the original series leaves off. As the new Tournament champion, he's generally taken it easy until he recieves the news of five deadly, murderous, martial artist who have escaped their prisons and are now headed to Japan. Baki and his friends must deal with them before they are killed themselves. </code>	<code>What is the name of a manga about a girl who must face her past to get revenge?</code>	<code>0.1093111932277679</code>
  <code>“Hey, don’t get confused. I’ve never thought of you as my little brother. You don’t even know your right place.” That was the dagger that hurt the most. Han Myoung Woo. A child of a wealthy family with a quick mind. But god didn’t give him everything. A near incurable genetic heart disease. He did his best in order to earn the respect of the people around him. But his father, the chairman, and his family members never took him seriously. That was when an accident suddenly struck him. And… He woke up in the body of the severely injured student, Kim Cheol Min.</code>	<code>What is a manga about a few thousand years has passed since an alchemist created Winter. He is now living with Jane learning what it means to be alive as a human.</code>	<code>-0.086394652724266</code>

• Loss: <code>CosineSimilarityLoss</code> with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• num_train_epochs: 4
• warmup_ratio: 0.1
• fp16: True

All Hyperparameters

<details><summary>Click to expand</summary>

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• 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: 4
• 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
• 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: proportional
• router_mapping: {}
• learning_rate_mapping: {}

</details>

Training Logs

Framework Versions

• Python: 3.10.0
• Sentence Transformers: 5.1.2
• Transformers: 4.57.1
• PyTorch: 2.5.1+cu121
• Accelerate: 1.11.0
• 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",
}

<!-- ## Glossary *Clearly define terms in order to be accessible across audiences.* --> <!-- ## Model Card Authors *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.* --> <!-- ## Model Card Contact *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.* -->