test/dify
1
0
Fork 0
mirror of https://gitee.com/dify_ai/dify.git synced 2024-11-30 02:08:37 +08:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

fix: ark token usage is none (#8351)

Browse Source
This commit is contained in:
sino 2024-09-13 14:19:24 +08:00 committed by GitHub
parent 80a322aaa2
commit a45ac6ab98
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 7 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
api/core/model_runtime/model_providers/volcengine_maas/client.py
View File

@ -208,11 +208,9 @@ class ArkClientV3:
presence_penalty=presence_penalty, presence_penalty=presence_penalty,
top_p=top_p, top_p=top_p,
temperature=temperature, temperature=temperature,
stream_options={"include_usage": True},
) )
for chunk in chunks: yield from chunks
if not chunk.choices:
continue
yield chunk
def embeddings(self, texts: list[str]) -> CreateEmbeddingResponse: def embeddings(self, texts: list[str]) -> CreateEmbeddingResponse:
return self.ark.embeddings.create(model=self.endpoint_id, input=texts) return self.ark.embeddings.create(model=self.endpoint_id, input=texts)

12
api/core/model_runtime/model_providers/volcengine_maas/llm/llm.py
View File

@ -239,16 +239,14 @@ class VolcengineMaaSLargeLanguageModel(LargeLanguageModel):
def _handle_stream_chat_response(chunks: Generator[ChatCompletionChunk]) -> Generator: def _handle_stream_chat_response(chunks: Generator[ChatCompletionChunk]) -> Generator:
for chunk in chunks: for chunk in chunks:
if not chunk.choices:
continue
choice = chunk.choices[0]
yield LLMResultChunk( yield LLMResultChunk(
model=model, model=model,
prompt_messages=prompt_messages, prompt_messages=prompt_messages,
delta=LLMResultChunkDelta( delta=LLMResultChunkDelta(
index=choice.index, index=0,
message=AssistantPromptMessage(content=choice.delta.content, tool_calls=[]), message=AssistantPromptMessage(
content=chunk.choices[0].delta.content if chunk.choices else "", tool_calls=[]
),
usage=self._calc_response_usage( usage=self._calc_response_usage(
model=model, model=model,
credentials=credentials, credentials=credentials,
@ -257,7 +255,7 @@ class VolcengineMaaSLargeLanguageModel(LargeLanguageModel):
) )
if chunk.usage if chunk.usage
else None, else None,
finish_reason=choice.finish_reason, finish_reason=chunk.choices[0].finish_reason if chunk.choices else None,
), ),
) )