Higher priority for user input of max_prefill_tokens & format (#540)
This commit is contained in:
@@ -2,6 +2,8 @@ import unittest
|
||||
|
||||
from sglang import OpenAI, set_default_backend
|
||||
from sglang.test.test_programs import (
|
||||
test_chat_completion_speculative,
|
||||
test_completion_speculative,
|
||||
test_decode_int,
|
||||
test_decode_json,
|
||||
test_expert_answer,
|
||||
@@ -14,8 +16,6 @@ from sglang.test.test_programs import (
|
||||
test_select,
|
||||
test_stream,
|
||||
test_tool_use,
|
||||
test_completion_speculative,
|
||||
test_chat_completion_speculative
|
||||
)
|
||||
|
||||
|
||||
@@ -97,4 +97,4 @@ if __name__ == "__main__":
|
||||
# global_config.verbosity = 2
|
||||
# t = TestOpenAIBackend()
|
||||
# t.setUp()
|
||||
# t.test_chat_completion_speculative()
|
||||
# t.test_chat_completion_speculative()
|
||||
|
||||
Reference in New Issue
Block a user