Sentence Similarity
sentence-transformers
PyTorch
English
bert
feature-extraction
mteb
custom_code
Eval Results (legacy)
text-embeddings-inference
Instructions to use Hum-Works/lodestone-base-4096-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- sentence-transformers
How to use Hum-Works/lodestone-base-4096-v1 with sentence-transformers:
from sentence_transformers import SentenceTransformer model = SentenceTransformer("Hum-Works/lodestone-base-4096-v1", trust_remote_code=True) sentences = [ "That is a happy person", "That is a happy dog", "That is a very happy person", "Today is a sunny day" ] embeddings = model.encode(sentences) similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [4, 4] - Notebooks
- Google Colab
- Kaggle
| # Copyright 2022 MosaicML Examples authors | |
| # SPDX-License-Identifier: Apache-2.0 | |
| from transformers import BertConfig as TransformersBertConfig | |
| class BertConfig(TransformersBertConfig): | |
| def __init__( | |
| self, | |
| alibi_starting_size: int = 512, | |
| attention_probs_dropout_prob: float = 0.0, | |
| **kwargs, | |
| ): | |
| """Configuration class for MosaicBert. | |
| Args: | |
| alibi_starting_size (int): Use `alibi_starting_size` to determine how large of an alibi tensor to | |
| create when initializing the model. You should be able to ignore this parameter in most cases. | |
| Defaults to 512. | |
| attention_probs_dropout_prob (float): By default, turn off attention dropout in Mosaic BERT | |
| (otherwise, Flash Attention will be off by default). Defaults to 0.0. | |
| """ | |
| super().__init__(attention_probs_dropout_prob=attention_probs_dropout_prob, **kwargs) | |
| self.alibi_starting_size = alibi_starting_size | |