philtoms/minilm-alice-base-rsft-v2

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

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L6-v2. 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
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 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': 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("philtoms/minilm-alice-base-rsft-v2")
# Run inference
sentences = [
    'The King of Hearts dismisses the Hatter. As the Hatter leaves, the Queen of Hearts orders his execution, but he is already out of sight.',
    '‘ you may go, ’ said the king, and the hatter hurriedly left the court, without even waiting to put his shoes on.',
    'the hatter shook his head mournfully. ‘ not i! ’ he replied. ‘ we quarrelled last march — just before he went mad, you know — ’ ( pointing with his tea spoon at the march hare, ) ‘ — it was at the great concert given by the queen of hearts, and i had to sing',
]
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, 0.6118, 0.1128],
#         [0.6118, 1.0000, 0.2371],
#         [0.1128, 0.2371, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

• Size: 272 training samples
• Columns: sentence_0, sentence_1, and sentence_2
• Approximate statistics based on the first 272 samples:

  	sentence_0	sentence_1	sentence_2
  type	string	string	string
  details	min: 20 tokens mean: 36.27 tokens max: 72 tokens	min: 10 tokens mean: 73.69 tokens max: 222 tokens	min: 28 tokens mean: 110.76 tokens max: 188 tokens

• Samples:

  sentence_0	sentence_1	sentence_2
  The Mock Turtle, with tears running down his cheeks, told Alice she could have no idea what a delightful thing a Lobster Quadrille is.	‘ you may not have lived much under the sea — ’ ( ’ i haven ’ t, ’ said alice ) — ‘ and perhaps you were never even introduced to a lobster — ’ ( alice began to say ‘ i once tasted — ’ but checked herself hastily, and said ‘ no, never ’ ) ‘ — so you can have no idea what a delightful thing a lobster quadrille is! ’	the long grass rustled at her feet as the white rabbit hurried by — the frightened mouse splashed his way through the neighbouring pool — she could hear the rattle of the teacups as the march hare and his friends shared their never - ending meal, and the shrill voice of the queen ordering off her unfortunate guests to execution — once more the pig - baby was sneezing on the duchess ’ s knee, while plates and dishes crashed around it — once more the shriek of the gryphon, the squeaking of the lizard ’ s slate - pencil, and the choking of the suppressed guinea - pigs, filled the air, mixed up with the distant sobs of the miserable mock turtle.
  When the King of Hearts suggests a line from the verses refers to the Queen of Hearts having fits, she furiously denies it and throws an inkstand at the Lizard.	‘ never! ’ said the queen furiously, throwing an inkstand at the lizard as she spoke. ( the unfortunate little bill had left off writing on his slate with one finger, as he found it made no mark ; but he now hastily began again, using the ink, that was trickling down his face, as long as it lasted. )	the hatter shook his head mournfully. ‘ not i! ’ he replied. ‘ we quarrelled last march — just before he went mad, you know — ’ ( pointing with his tea spoon at the march hare, ) ‘ — it was at the great concert given by the queen of hearts, and i had to sing
  The argument continues as Alice admits to eating eggs, leading the Pigeon to conclude that little girls are a type of serpent before telling her to leave.	‘ i don ’ t believe it, ’ said the pigeon ; ‘ but if they do, why then they ’ re a kind of serpent, that ’ s all i can say. ’	‘ curiouser and curiouser! ’ cried alice ( she was so much surprised, that for the moment she quite forgot how to speak good english ) ; ‘ now i ’ m opening out like the largest telescope that ever was! good - bye, feet! ’ ( for when she looked down at her feet, they seemed to be almost out of sight, they were getting so far off ). ‘ oh, my poor little feet, i wonder who will put on your shoes and stockings for you now, dears? i ’ m sure i shan ’ t be able! i shall be a great deal too far off to trouble myself about you : you must manage the best way you can ; — but i must be kind to them, ’ thought alice, ‘ or perhaps they won ’ t walk the way i want to go! let me see : i ’ ll give them a new pair of boots every christmas. ’

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 4
• per_device_eval_batch_size: 4
• num_train_epochs: 5
• 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: 4
• per_device_eval_batch_size: 4
• 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: 5
• 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
• 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: 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: 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: False
• 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: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

Framework Versions

• Python: 3.13.5
• Sentence Transformers: 5.0.0
• Transformers: 4.53.1
• PyTorch: 2.7.1
• Accelerate: 1.8.1
• Datasets: 4.0.0
• Tokenizers: 0.21.2

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}
}