86 lines
2.5 KiB
Python
86 lines
2.5 KiB
Python
import unittest
|
|
from types import SimpleNamespace
|
|
|
|
from sglang.srt.utils import kill_process_tree
|
|
from sglang.test.ci.ci_register import register_cuda_ci
|
|
from sglang.test.few_shot_gsm8k import run_eval as run_eval_few_shot_gsm8k
|
|
from sglang.test.send_one import BenchArgs, send_one_prompt
|
|
from sglang.test.test_utils import (
|
|
DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
|
|
DEFAULT_URL_FOR_TEST,
|
|
CustomTestCase,
|
|
is_in_ci,
|
|
popen_launch_server,
|
|
write_github_step_summary,
|
|
)
|
|
|
|
register_cuda_ci(est_time=160, stage="base-c", runner_config="8-gpu-h200")
|
|
|
|
MINIMAX_M25_MODEL_PATH = "MiniMaxAI/MiniMax-M2.5"
|
|
|
|
|
|
class TestMiniMaxM25Basic(CustomTestCase):
|
|
@classmethod
|
|
def setUpClass(cls):
|
|
cls.model = MINIMAX_M25_MODEL_PATH
|
|
cls.base_url = DEFAULT_URL_FOR_TEST
|
|
other_args = [
|
|
"--trust-remote-code",
|
|
"--tp",
|
|
"8",
|
|
"--ep-size",
|
|
"8",
|
|
"--mem-fraction-static",
|
|
"0.85",
|
|
"--reasoning-parser",
|
|
"minimax-append-think",
|
|
"--model-loader-extra-config",
|
|
'{"enable_multithread_load": true, "num_threads": 64}',
|
|
"--weight-loader-prefetch-checkpoints",
|
|
]
|
|
cls.process = popen_launch_server(
|
|
cls.model,
|
|
cls.base_url,
|
|
timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
|
|
other_args=other_args,
|
|
)
|
|
|
|
@classmethod
|
|
def tearDownClass(cls):
|
|
kill_process_tree(cls.process.pid)
|
|
|
|
def test_a_gsm8k(self):
|
|
args = SimpleNamespace(
|
|
num_shots=20,
|
|
data_path=None,
|
|
num_questions=200,
|
|
parallel=200,
|
|
max_new_tokens=512,
|
|
host="http://127.0.0.1",
|
|
port=int(self.base_url.split(":")[-1]),
|
|
)
|
|
metrics = run_eval_few_shot_gsm8k(args)
|
|
print(f"{metrics=}")
|
|
|
|
if is_in_ci():
|
|
write_github_step_summary(
|
|
f"### test_gsm8k (minimax-m25)\n" f'{metrics["accuracy"]=:.3f}\n'
|
|
)
|
|
self.assertGreater(metrics["accuracy"], 0.900)
|
|
|
|
def test_bs_1_speed(self):
|
|
args = BenchArgs(port=int(self.base_url.split(":")[-1]), max_new_tokens=2048)
|
|
acc_length, speed = send_one_prompt(args)
|
|
|
|
print(f"{speed=:.2f}")
|
|
|
|
if is_in_ci():
|
|
write_github_step_summary(
|
|
f"### test_bs_1_speed (minimax-m25)\n" f"{speed=:.2f} token/s\n"
|
|
)
|
|
self.assertGreater(speed, 90)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
unittest.main()
|