Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

FIX: Limit Azure OpenAI embeddings chunk size #13425

Merged
merged 2 commits into from
Nov 20, 2023
Merged

FIX: Limit Azure OpenAI embeddings chunk size #13425

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
23e9507
fix(embeddings): limit Azure OpenAI chunk size
mspronesti Nov 15, 2023
6707103
test Azure OpenAI embeddings chunk size setting
mspronesti Nov 17, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions libs/langchain/langchain/embeddings/azure_openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,6 +83,10 @@ def validate_environment(cls, values: Dict) -> Dict:
values["azure_ad_token"] = values["azure_ad_token"] or os.getenv(
"AZURE_OPENAI_AD_TOKEN"
)
# Azure OpenAI embedding models allow a maximum of 16 texts
# at a time in each batch
# See: https://learn.microsoft.com/en-us/azure/ai-services/openai/reference#embeddings
values["chunk_size"] = min(values["chunk_size"], 16)
try:
import openai

Expand Down
2 changes: 1 addition & 1 deletion libs/langchain/langchain/embeddings/openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -291,7 +291,7 @@ def validate_environment(cls, values: Dict) -> Dict:
# Azure OpenAI embedding models allow a maximum of 16 texts
# at a time in each batch
# See: https://learn.microsoft.com/en-us/azure/ai-services/openai/reference#embeddings
values["chunk_size"] = max(values["chunk_size"], 16)
values["chunk_size"] = min(values["chunk_size"], 16)
else:
default_api_version = ""
values["openai_api_version"] = get_from_dict_or_env(
Expand Down
13 changes: 13 additions & 0 deletions libs/langchain/tests/integration_tests/embeddings/test_azure_openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,12 +30,25 @@ def test_azure_openai_embedding_documents_multiple() -> None:
embedding = _get_embeddings(chunk_size=2)
embedding.embedding_ctx_length = 8191
output = embedding.embed_documents(documents)
assert embedding.chunk_size == 2
assert len(output) == 3
assert len(output[0]) == 1536
assert len(output[1]) == 1536
assert len(output[2]) == 1536


def test_azure_openai_embedding_documents_chunk_size() -> None:
"""Test openai embeddings."""
documents = ["foo bar"] * 20
embedding = _get_embeddings()
embedding.embedding_ctx_length = 8191
output = embedding.embed_documents(documents)
# Max 16 chunks per batch on Azure OpenAI embeddings
assert embedding.chunk_size == 16
assert len(output) == 20
assert all([len(out) == 1536 for out in output])


@pytest.mark.asyncio
async def test_azure_openai_embedding_documents_async_multiple() -> None:
"""Test openai embeddings."""
Expand Down