codersan/FaMiniLM

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

This model, FaMiniLM, was developed alongside FaLaBSE as part of the research paper "MetaRAG and WikiFaQA: A Co-designed Framework and Benchmark for Advancing Persian Long-Context RAG". It serves as a lightweight Persian sentence encoder. FaMiniLM was created by fine-tuning the all-MiniLM-L6-v2 model—which had no prior Persian knowledge—on the custom PersianSimilarSentences dataset. The training was specifically designed to build Persian semantic understanding from the ground up.

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

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 256, '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("codersan/validadted_all-MiniLM_onV9")
# Run inference
sentences = [
    'برای تبدیل شدن به نویسنده برتر Quora ، چند بازدید و پاسخ لازم است؟',
    'چگونه می توانم نویسنده برتر Quora شوم ، از صعود بیشتر و آمار بهتر استفاده کنم؟',
    'من به دنبال خرید دوچرخه جدید هستم.Suzuki Gixxer 155 یا Honda Hornet 160r.کدام یک را بخرید؟',
]
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]

Training Details

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 64
• learning_rate: 2e-05
• weight_decay: 0.01
• num_train_epochs: 15
• warmup_ratio: 0.1
• batch_sampler: no_duplicates

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.3.1
• Transformers: 4.47.0
• PyTorch: 2.5.1+cu121
• Accelerate: 1.2.1
• Datasets: 3.2.0
• Tokenizers: 0.21.0

Citation

BibTeX

WikiFaQA paper