phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-27 19:54:13 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

perf(azure+openai-files): use model_dump instead of json.loads + model_dump_json

Browse source
This commit is contained in:
Krrish Dholakia 2024-01-06 15:50:05 +05:30
parent 837bf5da9c
commit 807b64e68e
3 changed files with 33 additions and 33 deletions
Download patch file Download diff file Expand all files Collapse all files

18
litellm/llms/azure.py
View file

@ -248,7 +248,7 @@ class AzureChatCompletion(BaseLLM):
else:
azure_client = client
response = azure_client.chat.completions.create(**data, timeout=timeout) # type: ignore
stringified_response = response.model_dump_json()
stringified_response = response.model_dump()
## LOGGING
logging_obj.post_call(
input=messages,
@ -261,7 +261,7 @@ class AzureChatCompletion(BaseLLM):
},
)
return convert_to_model_response_object(
response_object=json.loads(stringified_response),
response_object=stringified_response,
model_response_object=model_response,
)
except AzureOpenAIError as e:
@ -323,7 +323,7 @@ class AzureChatCompletion(BaseLLM):
**data, timeout=timeout
)
return convert_to_model_response_object(
response_object=json.loads(response.model_dump_json()),
response_object=response.model_dump(),
model_response_object=model_response,
)
except AzureOpenAIError as e:
@ -465,7 +465,7 @@ class AzureChatCompletion(BaseLLM):
else:
openai_aclient = client
response = await openai_aclient.embeddings.create(**data, timeout=timeout)
stringified_response = response.model_dump_json()
stringified_response = response.model_dump()
## LOGGING
logging_obj.post_call(
input=input,
@ -474,7 +474,7 @@ class AzureChatCompletion(BaseLLM):
original_response=stringified_response,
)
return convert_to_model_response_object(
response_object=json.loads(stringified_response),
response_object=stringified_response,
model_response_object=model_response,
response_type="embedding",
)
@ -564,7 +564,7 @@ class AzureChatCompletion(BaseLLM):
original_response=response,
)
return convert_to_model_response_object(response_object=json.loads(response.model_dump_json()), model_response_object=model_response, response_type="embedding") # type: ignore
return convert_to_model_response_object(response_object=response.model_dump(), model_response_object=model_response, response_type="embedding") # type: ignore
except AzureOpenAIError as e:
exception_mapping_worked = True
raise e
@ -599,7 +599,7 @@ class AzureChatCompletion(BaseLLM):
else:
openai_aclient = client
response = await openai_aclient.images.generate(**data, timeout=timeout)
stringified_response = response.model_dump_json()
stringified_response = response.model_dump()
## LOGGING
logging_obj.post_call(
input=input,
@ -608,7 +608,7 @@ class AzureChatCompletion(BaseLLM):
original_response=stringified_response,
)
return convert_to_model_response_object(
response_object=json.loads(stringified_response),
response_object=stringified_response,
model_response_object=model_response,
response_type="image_generation",
)
@ -697,7 +697,7 @@ class AzureChatCompletion(BaseLLM):
original_response=response,
)
# return response
return convert_to_model_response_object(response_object=json.loads(response.model_dump_json()), model_response_object=model_response, response_type="image_generation") # type: ignore
return convert_to_model_response_object(response_object=response.model_dump(), model_response_object=model_response, response_type="image_generation") # type: ignore
except AzureOpenAIError as e:
exception_mapping_worked = True
raise e