Reduce excessively long logs caused by transformer version updates. (#25737)

Co-authored-by: Adarsh Shirawalmath <114558126+adarshxs@users.noreply.github.com>

test/registered/ascend/basic_function/HiCache/test_npu_hierarchical_cache_mla.py

@@ -1,3 +1,4 @@
+import os
 import unittest
 
 from sglang.test.ascend.test_ascend_utils import (
@@ -65,6 +66,8 @@ class TestNpuHierarchicalCacheMla(CustomTestCase):
                 "write_back",
             ],
         ]
+
+        os.environ["TRANSFORMERS_VERBOSITY"] = "error"
         for common_arg in common_args:
             other_args = common_arg + (
                 [

test/registered/ascend/basic_function/parallel_strategy/expert_parallelism/test_npu_deepep.py

@@ -59,6 +59,7 @@ class TestAscendDeepEP(CustomTestCase):
             "HCCL_BUFFSIZE": "1000",
             "SGLANG_DEEPEP_NUM_MAX_DISPATCH_TOKENS_PER_RANK": "32",
             "SGLANG_NPU_USE_MLAPO": "1",
+            "TRANSFORMERS_VERBOSITY": "error",
         }
         os.environ.update(cls.extra_envs)
 

test/registered/ascend/basic_function/parallel_strategy/expert_parallelism/test_npu_deepep_auto_deepseek_v3_2_w8a8.py

@@ -53,6 +53,7 @@ class TestDeepEpDeepseekV32(GSM8KAscendMixin, TestMMLU, CustomTestCase):
         "SGLANG_NPU_USE_MLAPO": "0",
         "SGLANG_NPU_USE_MULTI_STREAM": "1",
         "TASK_QUEUE_ENABLE": "0",
+        "TRANSFORMERS_VERBOSITY": "error",
     }
 
     accuracy = 0.95  # Test GSM8K accuracy ≥0.95

test/registered/ascend/llm_models/test_npu_qwq_32b_w8a8.py

@@ -1,3 +1,4 @@
+import os
 import unittest
 
 from sglang.test.ascend.gsm8k_ascend_mixin import GSM8KAscendMixin
@@ -29,6 +30,7 @@ class TestQWQ32BW8A8(GSM8KAscendMixin, CustomTestCase):
         "--quantization",
         "modelslim",
     ]
+    os.environ["TRANSFORMERS_VERBOSITY"] = "error"
 
 
 if __name__ == "__main__":

test/registered/ascend/test_npu_memory_consumption.py

@@ -63,6 +63,9 @@ class TestMemoryConsumptionAscend(CustomTestCase):
                 "--disable-radix-cache",
                 "--disable-cuda-graph",
             ],
+            env={
+                "TRANSFORMERS_VERBOSITY": "error",
+            },
         )
 
         ### Calculate initial used memory