multilingual e5 large instruct embedding model from intfloat
embedding
735 Pulls Updated 2 months ago
Updated 2 months ago
2 months ago
5f3d8f379743 · 1.1GB
model
archbert
·
parameters559M
·
quantizationF16
1.1GB
license
MIT LICENSE
Permission is hereby granted, free of charge, to any person obtaining a copy of this so
1.0kB
Readme
Usage
After you pull the image, you can simply use the openai compatible embedding endpoint:
curl -vvvv -H 'Content-Type: application/json' "http://localhost:11434/v1/embeddings" -d '{
"input": ["passage: my text 1", "passage: my text 2"],
"model": "jeffh/intfloat-multilingual-e5-large-instruct"
}'
Where model
can also include the tag as necessary (such as jeffh/intfloat-multilingual-e5-large-instruct:f16
)
Model Notes
You should use the structure for queries:
Instruct: {task_description}
Query: {query}
Documents do not need any special structure to be embedded for indexing.