phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-27 11:43:54 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix(vertex_httpx.py): support tool calling w/ streaming for vertex ai + gemini

Browse source
This commit is contained in:
Krrish Dholakia 2024-07-06 14:02:25 -07:00
parent 2452753e08
commit faa88a1ab1
6 changed files with 111 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

25
litellm/llms/vertex_httpx.py
View file

@ -1330,17 +1330,30 @@ class ModelResponseIterator:
gemini_chunk = processed_chunk["candidates"][0]
if (
"content" in gemini_chunk
and "text" in gemini_chunk["content"]["parts"][0]
):
text = gemini_chunk["content"]["parts"][0]["text"]
if "content" in gemini_chunk:
if "text" in gemini_chunk["content"]["parts"][0]:
text = gemini_chunk["content"]["parts"][0]["text"]
elif "functionCall" in gemini_chunk["content"]["parts"][0]:
function_call = ChatCompletionToolCallFunctionChunk(
name=gemini_chunk["content"]["parts"][0]["functionCall"][
"name"
],
arguments=json.dumps(
gemini_chunk["content"]["parts"][0]["functionCall"]["args"]
),
)
tool_use = ChatCompletionToolCallChunk(
id=str(uuid.uuid4()),
type="function",
function=function_call,
index=0,
)
if "finishReason" in gemini_chunk:
finish_reason = map_finish_reason(
finish_reason=gemini_chunk["finishReason"]
)
## DO NOT SET 'finish_reason' = True
## DO NOT SET 'is_finished' = True
## GEMINI SETS FINISHREASON ON EVERY CHUNK!
if "usageMetadata" in processed_chunk: