Unify spec v2's naming manner. (#15990)
This commit is contained in:
@@ -1,9 +1,9 @@
|
||||
import os
|
||||
import unittest
|
||||
from types import SimpleNamespace
|
||||
|
||||
import requests
|
||||
|
||||
from sglang.srt.environ import envs
|
||||
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
|
||||
@@ -26,7 +26,6 @@ SERVER_LAUNCH_TIMEOUT = 1200
|
||||
class TestDeepseekV3FP4MTP(CustomTestCase):
|
||||
@classmethod
|
||||
def setUpClass(cls):
|
||||
os.environ["SGLANG_ENABLE_SPEC_V2"] = "1"
|
||||
cls.model = FULL_DEEPSEEK_V3_FP4_MODEL_PATH
|
||||
cls.base_url = DEFAULT_URL_FOR_TEST
|
||||
other_args = [
|
||||
@@ -51,18 +50,17 @@ class TestDeepseekV3FP4MTP(CustomTestCase):
|
||||
"--model-loader-extra-config",
|
||||
'{"enable_multithread_load": true,"num_threads": 64}',
|
||||
]
|
||||
cls.process = popen_launch_server(
|
||||
cls.model,
|
||||
cls.base_url,
|
||||
timeout=SERVER_LAUNCH_TIMEOUT,
|
||||
other_args=other_args,
|
||||
)
|
||||
with envs.SGLANG_ENABLE_SPEC_V2.override(True):
|
||||
cls.process = popen_launch_server(
|
||||
cls.model,
|
||||
cls.base_url,
|
||||
timeout=SERVER_LAUNCH_TIMEOUT,
|
||||
other_args=other_args,
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def tearDownClass(cls):
|
||||
kill_process_tree(cls.process.pid)
|
||||
if "SGLANG_ENABLE_SPEC_V2" in os.environ:
|
||||
del os.environ["SGLANG_ENABLE_SPEC_V2"]
|
||||
|
||||
def test_a_gsm8k(
|
||||
self,
|
||||
|
||||
@@ -30,7 +30,7 @@ class TestEagleConstrainedDecoding(
|
||||
model = DEFAULT_TARGET_MODEL_EAGLE
|
||||
draft_model = DEFAULT_DRAFT_MODEL_EAGLE
|
||||
grammar_backend = "xgrammar"
|
||||
eagle_v2 = False
|
||||
spec_v2 = False
|
||||
|
||||
@classmethod
|
||||
def setUpClass(cls):
|
||||
@@ -59,7 +59,7 @@ class TestEagleConstrainedDecoding(
|
||||
cls.grammar_backend,
|
||||
]
|
||||
launch_args.extend(cls.other_launch_args)
|
||||
with envs.SGLANG_ENABLE_SPEC_V2.override(cls.eagle_v2):
|
||||
with envs.SGLANG_ENABLE_SPEC_V2.override(cls.spec_v2):
|
||||
cls.process = popen_launch_server(
|
||||
cls.model,
|
||||
cls.base_url,
|
||||
@@ -73,7 +73,7 @@ class TestEagleConstrainedDecoding(
|
||||
|
||||
|
||||
class TestEagleConstrainedDecodingV2(TestEagleConstrainedDecoding):
|
||||
eagle_v2 = True
|
||||
spec_v2 = True
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
|
||||
@@ -1,8 +1,8 @@
|
||||
import os
|
||||
import unittest
|
||||
from types import SimpleNamespace
|
||||
from urllib.parse import urlparse
|
||||
|
||||
from sglang.srt.environ import envs
|
||||
from sglang.srt.utils import kill_process_tree
|
||||
from sglang.test.few_shot_gsm8k import run_eval as run_eval_few_shot_gsm8k
|
||||
from sglang.test.test_utils import (
|
||||
@@ -54,12 +54,9 @@ class TestAscendDeepSeekMTP(CustomTestCase):
|
||||
2,
|
||||
]
|
||||
|
||||
cls.extra_envs = {
|
||||
"SGLANG_NPU_USE_MLAPO": "1",
|
||||
"SGLANG_ENABLE_SPEC_V2": "1",
|
||||
"SGLANG_ENABLE_OVERLAP_PLAN_STREAM": "1",
|
||||
}
|
||||
os.environ.update(cls.extra_envs)
|
||||
envs.SGLANG_NPU_USE_MLAPO.set(True)
|
||||
envs.SGLANG_ENABLE_SPEC_V2.set(True)
|
||||
envs.SGLANG_ENABLE_OVERLAP_PLAN_STREAM.set(True)
|
||||
|
||||
def test_a_gsm8k(self):
|
||||
for model in self.models:
|
||||
|
||||
Reference in New Issue
Block a user