How to set the embedding of VectorStoreIndex of ObjectIndex.from_objects to a custom embedding, such as DashScopeEmbedding?

[lifu963]

obj_index = ObjectIndex.from_objects(
    table_schema_objs,
    table_node_mapping,
    VectorStoreIndex,
)

i want to set VectorStoreIndex.embed_model as DashScopeEmbedding(
model_name='text-embedding-v2',
api_key=DASHSCOPE_API_KEY
)

[dosubot[bot]]

To set the embedding of VectorStoreIndex in ObjectIndex.from_objects to a custom embedding such as DashScopeEmbedding, you need to pass the embed_model parameter when initializing the VectorStoreIndex. Here is how you can do it:

from llama_index.core.indices.vector_store.base import VectorStoreIndex
from llama_index.embeddings.dashscope import DashScopeEmbedding
from llama_index.core.objects.base import ObjectIndex

# Initialize your custom embedding
custom_embedding = DashScopeEmbedding(model_name='text-embedding-v2', api_key=DASHSCOPE_API_KEY)

# Create the VectorStoreIndex with the custom embedding
vector_store_index = VectorStoreIndex(embed_model=custom_embedding)

# Use ObjectIndex.from_objects with the created VectorStoreIndex
object_index = ObjectIndex.from_objects(
    objects=table_schema_objs,
    object_mapping=table_node_mapping,
    index_cls=VectorStoreIndex,
    embed_model=custom_embedding
)

In this example, replace table_schema_objs and table_node_mapping with your actual objects and mappings. The embed_model parameter is passed to the VectorStoreIndex to use the custom DashScopeEmbedding ^[1].

To continue talking to Dosu, mention @dosu.