SentenceTransformer based on meta-llama/Llama-3.2-1B

This is a sentence-transformers model finetuned from meta-llama/Llama-3.2-1B. It maps sentences & paragraphs to a 2048-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: meta-llama/Llama-3.2-1B
Maximum Sequence Length: 512 tokens
Output Dimensionality: 2048 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: LlamaModel 
  (1): Pooling({'word_embedding_dimension': 2048, '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("roboepicss/merged_product_stage_1_llama1B")
# Run inference
sentences = [
    'The weather is lovely today.',
    "It's so sunny outside!",
    'He drove to the stadium.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 2048]

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

Evaluation

Metrics

Information Retrieval

Dataset: ir
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.7918
cosine_accuracy@3	0.9296
cosine_accuracy@5	0.9589
cosine_accuracy@10	0.9825
cosine_precision@1	0.7918
cosine_precision@3	0.3099
cosine_precision@5	0.1918
cosine_precision@10	0.0983
cosine_recall@1	0.0855
cosine_recall@3	0.1003
cosine_recall@5	0.1035
cosine_recall@10	0.106
cosine_ndcg@10	0.2064
cosine_mrr@10	0.8654
cosine_map@100	0.0935
dot_accuracy@1	0.7918
dot_accuracy@3	0.9296
dot_accuracy@5	0.9589
dot_accuracy@10	0.9825
dot_precision@1	0.7918
dot_precision@3	0.3099
dot_precision@5	0.1918
dot_precision@10	0.0983
dot_recall@1	0.0855
dot_recall@3	0.1003
dot_recall@5	0.1035
dot_recall@10	0.106
dot_ndcg@10	0.2064
dot_mrr@10	0.8654
dot_map@100	0.0935

Training Details

Training Logs

Epoch	Step	ir_cosine_map@100
0	0	0.0935

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.2.0
Transformers: 4.44.2
PyTorch: 2.4.1+cu121
Accelerate: 0.34.2
Datasets: 3.0.1
Tokenizers: 0.19.1

roboepicss
/

merged_product_stage_1_llama1B