biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Misc][OpenAI] deprecate max_tokens in favor of new max_completion_tokens field for chat completion endpoint (#9837)

Browse Source
This commit is contained in:
Guillaume Calmettes
2024-10-31 02:15:56 +01:00
committed by GitHub
parent 64384bbcdf
commit abbfb6134d
14 changed files with 140 additions and 118 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
tests/entrypoints/openai/test_audio.py
View File

@@ -68,11 +68,12 @@ async def test_single_chat_session_audio(client: openai.AsyncOpenAI,
}]
# test single completion
chat_completion = await client.chat.completions.create(model=model_name,
messages=messages,
max_tokens=10,
logprobs=True,
top_logprobs=5)
chat_completion = await client.chat.completions.create(
model=model_name,
messages=messages,
max_completion_tokens=10,
logprobs=True,
top_logprobs=5)
assert len(chat_completion.choices) == 1
choice = chat_completion.choices[0]
@@ -91,7 +92,7 @@ async def test_single_chat_session_audio(client: openai.AsyncOpenAI,
chat_completion = await client.chat.completions.create(
model=model_name,
messages=messages,
max_tokens=10,
max_completion_tokens=10,
)
message = chat_completion.choices[0].message
assert message.content is not None and len(message.content) >= 0
@@ -123,11 +124,12 @@ async def test_single_chat_session_audio_base64encoded(
}]
# test single completion
chat_completion = await client.chat.completions.create(model=model_name,
messages=messages,
max_tokens=10,
logprobs=True,
top_logprobs=5)
chat_completion = await client.chat.completions.create(
model=model_name,
messages=messages,
max_completion_tokens=10,
logprobs=True,
top_logprobs=5)
assert len(chat_completion.choices) == 1
choice = chat_completion.choices[0]
@@ -146,7 +148,7 @@ async def test_single_chat_session_audio_base64encoded(
chat_completion = await client.chat.completions.create(
model=model_name,
messages=messages,
max_tokens=10,
max_completion_tokens=10,
)
message = chat_completion.choices[0].message
assert message.content is not None and len(message.content) >= 0
@@ -178,7 +180,7 @@ async def test_chat_streaming_audio(client: openai.AsyncOpenAI,
chat_completion = await client.chat.completions.create(
model=model_name,
messages=messages,
max_tokens=10,
max_completion_tokens=10,
temperature=0.0,
)
output = chat_completion.choices[0].message.content
@@ -188,7 +190,7 @@ async def test_chat_streaming_audio(client: openai.AsyncOpenAI,
stream = await client.chat.completions.create(
model=model_name,
messages=messages,
max_tokens=10,
max_completion_tokens=10,
temperature=0.0,
stream=True,
)
@@ -242,7 +244,7 @@ async def test_multi_audio_input(client: openai.AsyncOpenAI, model_name: str,
await client.chat.completions.create(
model=model_name,
messages=messages,
max_tokens=10,
max_completion_tokens=10,
temperature=0.0,
)