BEE-spoke-data/bert-plus-L8-v1.0-syntheticSTS-4k

BEE-spoke-data/bert-plus-L8-v1.0-syntheticSTS-4k

This is a sentence-transformers model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.

• Continued-tune of BEE-spoke-data/bert-plus-L8-v1.0-allNLI_matryoshka
• ctx 4096 on synthetic text similarity dataset of text1, text2, label
• Matryoshka dims: [768, 512, 256, 128, 64]

Usage (Sentence-Transformers)

Using this model becomes easy when you have sentence-transformers installed:

pip install -U sentence-transformers

Then you can use the model like this:

from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]

model = SentenceTransformer('BEE-spoke-data/bert-plus-L8-v1.0-syntheticSTS-4k')
embeddings = model.encode(sentences)
print(embeddings)

Usage (HuggingFace Transformers)

Without sentence-transformers, you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.

from transformers import AutoTokenizer, AutoModel
import torch

#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)

# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']

# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('BEE-spoke-data/bert-plus-L8-v1.0-syntheticSTS-4k')
model = AutoModel.from_pretrained('BEE-spoke-data/bert-plus-L8-v1.0-syntheticSTS-4k')

# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')

# Compute token embeddings
with torch.no_grad():
    model_output = model(**encoded_input)

# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])

print("Sentence embeddings:")
print(sentence_embeddings)

Training

The model was trained with the parameters:

Loss:

sentence_transformers.losses.MatryoshkaLoss.MatryoshkaLoss with parameters:

{'loss': 'CosineSimilarityLoss', 'matryoshka_dims': [768, 512, 256, 128, 64], 'matryoshka_weights': [1, 1, 1, 1, 1], 'n_dims_per_step': -1}

See more details at the training run on wandb