diff --git a/python/sglang/srt/managers/detokenizer_manager.py b/python/sglang/srt/managers/detokenizer_manager.py index 73054bf8f..18f77424d 100644 --- a/python/sglang/srt/managers/detokenizer_manager.py +++ b/python/sglang/srt/managers/detokenizer_manager.py @@ -25,8 +25,6 @@ from sglang.srt.managers.io_struct import ( BatchEmbeddingOut, BatchStrOut, BatchTokenIDOut, - GetMemPoolSizeReqOutput, - UpdateWeightReqOutput, ) from sglang.srt.managers.schedule_batch import FINISH_MATCHED_STR, FINISH_MATCHED_TOKEN from sglang.srt.server_args import PortArgs, ServerArgs diff --git a/test/srt/test_pytorch_sampling_backend.py b/test/srt/test_pytorch_sampling_backend.py index 5257740c4..9aa6c3300 100644 --- a/test/srt/test_pytorch_sampling_backend.py +++ b/test/srt/test_pytorch_sampling_backend.py @@ -51,7 +51,7 @@ class TestPyTorchSamplingBackend(unittest.TestCase): response_single = requests.post( self.base_url + "/generate", json={ - "text": "The capital of France is", + "text": "The capital of Germany is", "sampling_params": { "temperature": 0, "max_new_tokens": 32, @@ -69,7 +69,7 @@ class TestPyTorchSamplingBackend(unittest.TestCase): response_batch = requests.post( self.base_url + "/generate", json={ - "text": ["The capital of France is"] * 10, + "text": ["The capital of Germany is"] * 10, "sampling_params": { "temperature": 0, "max_new_tokens": 32,