phoenix/litellm
1
0
Fork 0
forked from phoenix/litellm-mirror
Code Pull requests Activity

(feat) 10x faster embeddings

Browse source
This commit is contained in:
ishaan-jaff 2023-11-24 17:02:57 -08:00
parent 6738044170
commit 23466107a7
1 changed files with 5 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

21
litellm/llms/openai.py
View file

@ -309,8 +309,8 @@ class OpenAIChatCompletion(BaseLLM):
timeout: float, timeout: float,
api_key: Optional[str] = None, api_key: Optional[str] = None,
api_base: Optional[str] = None, api_base: Optional[str] = None,
model_response: litellm.utils.EmbeddingResponse = None,
logging_obj=None, logging_obj=None,
model_response=None,
optional_params=None, optional_params=None,
): ):
super().embedding() super().embedding()
@ -342,21 +342,10 @@ class OpenAIChatCompletion(BaseLLM):
additional_args={"complete_input_dict": data}, additional_args={"complete_input_dict": data},
original_response=response, original_response=response,
) )
model_response.data = response.data
embedding_response = json.loads(response.model_dump_json()) model_response.model = model
output_data = [] model_response.usage = response.usage
for idx, embedding in enumerate(embedding_response["data"]): model_response.object = "list"
output_data.append(
{
"object": embedding["object"],
"index": embedding["index"],
"embedding": embedding["embedding"]
}
)
model_response["object"] = "list"
model_response["data"] = output_data
model_response["model"] = model
model_response["usage"] = embedding_response["usage"]
return model_response return model_response
except OpenAIError as e: except OpenAIError as e:
exception_mapping_worked = True exception_mapping_worked = True