AscendEmbeddings

class langchain_community.embeddings.ascend.AscendEmbeddings

Bases: Embeddings, BaseModel

Ascend NPU accelerate Embedding model

Please ensure that you have installed CANN and torch_npu.

Example:

from langchain_community.embeddings import AscendEmbeddings model = AscendEmbeddings(model_path=<path_to_model>,

device_id=0, query_instruction=”Represent this sentence for searching relevant passages: “

)

param device_id: int = 0

Unstruntion to used for embedding query.

param document_instruction: str = ''

param model: Any = None

param model_path: str [Required]

Ascend NPU device id.

param pooling_method: str | None = 'cls'

param query_instruction: str = ''

Unstruntion to used for embedding document.

param tokenizer: Any = None

param use_fp16: bool = True

async aembed_documents(texts: List[str]) → List[List[float]]

Asynchronous Embed search docs.

Parameters:

texts (List[str]) – List of text to embed.

Returns:

List of embeddings.

Return type:

List[List[float]]

async aembed_query(text: str) → List[float]

Asynchronous Embed query text.

Parameters:

text (str) – Text to embed.

Returns:

Embedding.

Return type:

List[float]

embed_documents(texts: List[str]) → List[List[float]]

Embed search docs.

Parameters:

texts (List[str]) – List of text to embed.

Returns:

List of embeddings.

Return type:

List[List[float]]

embed_query(text: str) → List[float]

Embed query text.

Parameters:

text (str) – Text to embed.

Returns:

Embedding.

Return type:

List[float]

encode(sentences: Any) → Any

Parameters:

sentences (Any) –

Return type:

Any

pooling(last_hidden_state: Any, attention_mask: Any = None) → Any

Parameters:

• last_hidden_state (Any) –

• attention_mask (Any) –

Return type:

Any

Examples using AscendEmbeddings

• # Related

• Ascend