ModelScope Embeddings

In this notebook, we show how to use the ModelScope Embeddings in LlamaIndex. Check out the ModelScope site.

If you’re opening this Notebook on colab, you will need to install LlamaIndex 🦙 and the modelscope.

!pip install llama-index-embeddings-modelscope

Basic Usage

import sys
from llama_index.embeddings.modelscope.base import ModelScopeEmbedding

model = ModelScopeEmbedding(
    model_name="iic/nlp_gte_sentence-embedding_chinese-base",
    model_revision="master",
)

rsp = model.get_query_embedding("Hello, who are you?")
print(rsp)

rsp = model.get_text_embedding("Hello, who are you?")
print(rsp)

Generate Batch Embedding

from llama_index.embeddings.modelscope.base import ModelScopeEmbedding

model = ModelScopeEmbedding(
    model_name="iic/nlp_gte_sentence-embedding_chinese-base",
    model_revision="master",
)

rsp = model.get_text_embedding_batch(
    ["Hello, who are you?", "I am a student."]
)
print(rsp)