model: support mllama4 (#5144)

This commit is contained in:
Mick
2025-04-10 00:28:44 +08:00
committed by GitHub
parent 87eddedfa2
commit fbebcb7aa4
7 changed files with 145 additions and 65 deletions

View File

@@ -307,7 +307,6 @@ class TestOpenAIVisionServer(CustomTestCase):
self.assertGreater(len(video_response), 0)
def test_regex(self):
return
client = openai.Client(api_key=self.api_key, base_url=self.base_url)
regex = (
@@ -683,6 +682,31 @@ class TestJanusProServer(TestOpenAIVisionServer):
pass
class TestLlama4Server(TestOpenAIVisionServer):
@classmethod
def setUpClass(cls):
cls.model = "meta-llama/Llama-4-Scout-17B-16E-Instruct"
cls.base_url = DEFAULT_URL_FOR_TEST
cls.api_key = "sk-123456"
cls.process = popen_launch_server(
cls.model,
cls.base_url,
timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
other_args=[
"--chat-template",
"llama-4",
"--mem-fraction-static",
"0.8",
"--tp-size=8",
"--context-length=8192",
],
)
cls.base_url += "/v1"
def test_video_chat_completion(self):
pass
class TestGemma3itServer(TestOpenAIVisionServer):
@classmethod
def setUpClass(cls):