[CI] Rename basic CI stage-a/b/c -> base-a/b/c for symmetry with extra CI (#25420)

test/README.md

@@ -48,14 +48,14 @@ python3 test/registered/core/test_srt_endpoint.py TestSRTEndpoint.test_simple_de
 python3 python/sglang/jit_kernel/tests/test_add_constant.py
 
 # Run a suite
-python3 test/run_suite.py --hw cpu --suite stage-a-test-cpu
-python3 test/run_suite.py --hw cuda --suite stage-a-test-1-gpu-small
+python3 test/run_suite.py --hw cpu --suite base-a-test-cpu
+python3 test/run_suite.py --hw cuda --suite base-a-test-1-gpu-small
 
 # Nightly tests
 python3 test/run_suite.py --hw cuda --suite nightly-1-gpu --nightly
 
 # With auto-partitioning (for parallel CI jobs)
-python3 test/run_suite.py --hw cuda --suite stage-b-test-1-gpu-small \
+python3 test/run_suite.py --hw cuda --suite base-b-test-1-gpu-small \
     --auto-partition-id 0 --auto-partition-size 4
 ```
 
@@ -66,7 +66,7 @@ Every CI-discovered test file must call a registration function at module level:
 ```python
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=80, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=80, stage="base-b", runner_config="1-gpu-small")
 ```
 
 Parameters: `est_time` (seconds), `stage` + `runner_config` (target stage and runner pool from `scripts/ci/runner_configs.yml`), `nightly=True` (nightly-only), `disabled="reason"` (temporarily disable).
@@ -74,8 +74,8 @@ Parameters: `est_time` (seconds), `stage` + `runner_config` (target stage and ru
 Keep `est_time`, `stage`, `runner_config` as **literal values** — `run_suite.py` collects them by AST parsing.
 
 JIT kernel files live outside `test/registered/` but still use registration:
-- Correctness tests: `python/sglang/jit_kernel/tests/test_*.py` → `stage-b-kernel-unit-1-gpu-large`
-- Benchmarks: `python/sglang/jit_kernel/benchmark/bench_*.py` → `stage-b-kernel-benchmark-1-gpu-large`
+- Correctness tests: `python/sglang/jit_kernel/tests/test_*.py` → `base-b-kernel-unit-1-gpu-large`
+- Benchmarks: `python/sglang/jit_kernel/benchmark/bench_*.py` → `base-b-kernel-benchmark-1-gpu-large`
 
 ## Choosing a Suite
 
@@ -83,12 +83,12 @@ Use the lightest suite that meets your test's needs. Full suite tables are in th
 
 | Need | Suite |
 |------|-------|
-| No GPU required | `stage-a-test-cpu` |
-| Small GPU (fits 5090, 32GB) | `stage-b-test-1-gpu-small` (most tests go here) |
-| Large GPU memory or Hopper features | `stage-b-test-1-gpu-large` |
-| JIT kernel correctness | `stage-b-kernel-unit-1-gpu-large` |
-| JIT kernel benchmarks | `stage-b-kernel-benchmark-1-gpu-large` |
-| Multi-GPU (2/4/8) | `stage-b-test-2-gpu-large`, `stage-c-test-*` |
+| No GPU required | `base-a-test-cpu` |
+| Small GPU (fits 5090, 32GB) | `base-b-test-1-gpu-small` (most tests go here) |
+| Large GPU memory or Hopper features | `base-b-test-1-gpu-large` |
+| JIT kernel correctness | `base-b-kernel-unit-1-gpu-large` |
+| JIT kernel benchmarks | `base-b-kernel-benchmark-1-gpu-large` |
+| Multi-GPU (2/4/8) | `base-b-test-2-gpu-large`, `base-c-test-*` |
 | Long-running or experimental | `nightly-*` suites |
 
 ## Steps for Adding a Test

test/registered/4-gpu-models/test_deepseek_v3_cutedsl_4gpu.py

@@ -14,7 +14,7 @@ from sglang.test.test_utils import (
     try_cached_model,
 )
 
-register_cuda_ci(est_time=1800, stage="stage-c", runner_config="4-gpu-gb200")
+register_cuda_ci(est_time=1800, stage="base-c", runner_config="4-gpu-gb200")
 
 
 class TestDeepseekR1Nvfp4CuteDSLDeepEP(CustomTestCase):

test/registered/4-gpu-models/test_gpt_oss_4gpu.py

@@ -3,8 +3,8 @@ import unittest
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.gpt_oss_common import BaseTestGptOss
 
-register_cuda_ci(est_time=392, stage="stage-c", runner_config="4-gpu-h100")
-register_cuda_ci(est_time=350, stage="stage-c", runner_config="4-gpu-b200")
+register_cuda_ci(est_time=392, stage="base-c", runner_config="4-gpu-h100")
+register_cuda_ci(est_time=350, stage="base-c", runner_config="4-gpu-b200")
 
 
 class TestGptOss4Gpu(BaseTestGptOss):

test/registered/4-gpu-models/test_qwen35_fp4_mtp_v2.py

@@ -15,7 +15,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=340, stage="stage-c", runner_config="4-gpu-b200")
+register_cuda_ci(est_time=340, stage="base-c", runner_config="4-gpu-b200")
 
 QWEN35_FP4_MODEL = "nvidia/Qwen3.5-397B-A17B-NVFP4"
 ACC_THRESHOLDS = {QWEN35_FP4_MODEL: {"gsm8k": 0.95}}

test/registered/4-gpu-models/test_qwen35_models.py

@@ -17,7 +17,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=370, stage="stage-c", runner_config="4-gpu-b200")
+register_cuda_ci(est_time=370, stage="base-c", runner_config="4-gpu-b200")
 
 QWEN35_FP4_MODEL = "nvidia/Qwen3.5-397B-A17B-NVFP4"
 ACC_THRESHOLDS = {QWEN35_FP4_MODEL: {"gsm8k": 0.95}}

test/registered/4-gpu-models/test_qwen3_30b.py

@@ -11,7 +11,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=261, stage="stage-c", runner_config="4-gpu-h100")
+register_cuda_ci(est_time=261, stage="base-c", runner_config="4-gpu-h100")
 
 QWEN3_30B_MODEL_PATH = "Qwen/Qwen3-30B-A3B-FP8"
 

test/registered/4-gpu-models/test_qwen3_next_models_mtp.py

@@ -6,7 +6,7 @@ from sglang.test.kits.kl_divergence_kit import KLDivergenceMixin
 from sglang.test.kits.prefix_cache_branching_kit import PrefixCacheBranchingMixin
 from sglang.test.server_fixtures.default_fixture import DefaultServerBase
 
-register_cuda_ci(est_time=290, stage="stage-c", runner_config="4-gpu-h100")
+register_cuda_ci(est_time=290, stage="base-c", runner_config="4-gpu-h100")
 
 QWEN3_NEXT_MODEL = "Qwen/Qwen3-Next-80B-A3B-Instruct"
 

test/registered/8-gpu-models/test_deepseek_v3_mtp.py

@@ -17,7 +17,7 @@ from sglang.test.test_utils import (
     write_github_step_summary,
 )
 
-register_cuda_ci(est_time=300, stage="stage-c", runner_config="8-gpu-h200")
+register_cuda_ci(est_time=300, stage="base-c", runner_config="8-gpu-h200")
 
 FULL_DEEPSEEK_V3_MODEL_PATH = "deepseek-ai/DeepSeek-V3-0324"
 

test/registered/8-gpu-models/test_dsa_models_hisparse.py

@@ -12,9 +12,7 @@ from sglang.test.test_utils import (
     write_github_step_summary,
 )
 
-register_cuda_ci(
-    est_time=720, stage="stage-c", runner_config="8-gpu-h200", nightly=True
-)
+register_cuda_ci(est_time=720, stage="base-c", runner_config="8-gpu-h200", nightly=True)
 
 GLM5_MODEL_PATH = "zai-org/GLM-5-FP8"
 

test/registered/8-gpu-models/test_mimo_models.py

@@ -6,7 +6,7 @@ from sglang.test.kits.spec_decoding_kit import SpecDecodingMixin
 from sglang.test.server_fixtures.default_fixture import DefaultServerBase
 from sglang.test.server_fixtures.mmmu_fixture import MMMUServerBase
 
-register_cuda_ci(est_time=500, stage="stage-c", runner_config="8-gpu-h200")
+register_cuda_ci(est_time=500, stage="base-c", runner_config="8-gpu-h200")
 
 
 class TestMiMoV2Flash(GSM8KMixin, SpecDecodingMixin, DefaultServerBase):

test/registered/8-gpu-models/test_minimax_m25_basic.py

@@ -14,7 +14,7 @@ from sglang.test.test_utils import (
     write_github_step_summary,
 )
 
-register_cuda_ci(est_time=290, stage="stage-c", runner_config="8-gpu-h200")
+register_cuda_ci(est_time=290, stage="base-c", runner_config="8-gpu-h200")
 
 MINIMAX_M25_MODEL_PATH = "MiniMaxAI/MiniMax-M2.5"
 

test/registered/attention/test_chunk_gated_delta_rule.py

@@ -8,7 +8,7 @@ from sglang.srt.layers.attention.fla.fused_recurrent import (
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=11, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=11, stage="base-b", runner_config="1-gpu-large")
 
 
 @unittest.skipIf(not torch.cuda.is_available(), "Test requires CUDA")

test/registered/attention/test_create_kvindices.py

@@ -9,7 +9,7 @@ from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
 # Triton kernel unit test for KV indices creation
-register_cuda_ci(est_time=7, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=7, stage="base-b", runner_config="1-gpu-small")
 register_amd_ci(est_time=10, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/attention/test_flash_attention_4.py

@@ -13,7 +13,7 @@ from sglang.test.test_utils import (
 )
 
 # FlashAttention4 integration test (requires SM 100+ / Blackwell B200)
-register_cuda_ci(est_time=260, stage="stage-b", runner_config="4-gpu-b200")
+register_cuda_ci(est_time=260, stage="base-b", runner_config="4-gpu-b200")
 
 
 @unittest.skipIf(get_device_sm() < 100, "Test requires CUDA SM 100 or higher")

test/registered/attention/test_gdn_noncontiguous_stride.py

@@ -14,7 +14,7 @@ from sglang.srt.layers.attention.fla.fused_sigmoid_gating_recurrent import (
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=7, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=7, stage="base-b", runner_config="1-gpu-large")
 
 
 def _make_noncontiguous_ab(batch, num_heads, dtype=torch.bfloat16, device="cuda"):

test/registered/attention/test_kda_kernels.py

@@ -14,7 +14,7 @@ from sglang.srt.layers.attention.fla.kda import (
 from sglang.srt.utils.common import get_device
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=12, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=12, stage="base-b", runner_config="1-gpu-large")
 
 
 @unittest.skipIf(

test/registered/attention/test_normal_decode_set_metadata.py

@@ -20,8 +20,8 @@ from sglang.srt.mem_cache.swa_memory_pool import SWAKVPool
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
-# Register this test for CUDA CI in stage-b (fast attention/kernel tests)
-register_cuda_ci(est_time=11, stage="stage-b", runner_config="1-gpu-large")
+# Register this test for CUDA CI in base-b (fast attention/kernel tests)
+register_cuda_ci(est_time=11, stage="base-b", runner_config="1-gpu-large")
 
 
 def reference_normal_decode_set_metadata(

test/registered/attention/test_triton_attention_backend.py

@@ -20,7 +20,7 @@ from sglang.test.test_utils import (
 )
 
 # Triton attention backend integration test with latency benchmark and MMLU eval
-register_cuda_ci(est_time=177, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=177, stage="base-b", runner_config="1-gpu-large")
 register_amd_ci(est_time=1400, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/attention/test_triton_attention_kernels.py

@@ -23,7 +23,7 @@ from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 from sglang.test.test_utils import CustomTestCase, is_in_amd_ci
 
 # Triton attention kernel unit tests (decode, extend, prefill)
-register_cuda_ci(est_time=19, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=19, stage="base-b", runner_config="1-gpu-large")
 register_amd_ci(est_time=30, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/bench_fn/test_bench_serving_reasoning_stream.py

@@ -24,7 +24,7 @@ from sglang.bench_serving import (
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu")
+register_cpu_ci(est_time=10, suite="base-a-test-cpu")
 
 
 def _free_port() -> int:

test/registered/bench_fn/test_benchmark_datasets_api.py

@@ -26,7 +26,7 @@ from sglang.benchmark.datasets.random import sample_random_requests
 from sglang.benchmark.datasets.sharegpt import sample_sharegpt_requests
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=6, suite="stage-a-test-cpu")
+register_cpu_ci(est_time=6, suite="base-a-test-cpu")
 
 
 class _DummyTokenTensor:

test/registered/breakable_cuda_graph/test_breakable_cuda_graph.py

@@ -23,7 +23,7 @@ from sglang.test.test_utils import (
 )
 
 # CI Registration — large suite to fit the integration test's server startup.
-register_cuda_ci(est_time=79, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=79, stage="base-b", runner_config="1-gpu-large")
 
 
 def _skip_if_no_cuda(test_func):

test/registered/constrained_decoding/test_constrained_decoding.py

@@ -13,7 +13,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=120, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=120, stage="base-b", runner_config="1-gpu-small")
 register_amd_ci(est_time=179, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/core/test_deterministic.py

@@ -16,7 +16,7 @@ from sglang.test.test_deterministic_utils import (
 )
 from sglang.test.test_utils import is_in_amd_ci
 
-register_cuda_ci(est_time=207, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=207, stage="base-b", runner_config="1-gpu-large")
 register_amd_ci(est_time=278, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/core/test_engine_child_pids.py

@@ -20,7 +20,7 @@ from sglang.test.test_utils import (
     CustomTestCase,
 )
 
-register_cuda_ci(est_time=77, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=77, stage="base-b", runner_config="1-gpu-small")
 
 
 class TestEngineChildPids(CustomTestCase):

test/registered/core/test_gemma4_moe_deterministic.py

@@ -23,7 +23,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=107, stage="stage-b", runner_config="2-gpu-large")
+register_cuda_ci(est_time=107, stage="base-b", runner_config="2-gpu-large")
 
 
 PROMPT = (

test/registered/core/test_hidden_states.py

@@ -8,7 +8,7 @@ from sglang.srt.utils import get_device, is_hip
 from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 from sglang.test.test_utils import DEFAULT_SMALL_MODEL_NAME_FOR_TEST, CustomTestCase
 
-register_cuda_ci(est_time=45, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=45, stage="base-b", runner_config="1-gpu-small")
 register_amd_ci(est_time=55, suite="stage-b-test-1-gpu-small-amd")
 
 _is_hip = is_hip()

test/registered/core/test_mm_process_config.py

@@ -4,7 +4,7 @@ from unittest.mock import MagicMock, patch
 from sglang.srt.server_args import ServerArgs
 from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 
-register_cuda_ci(est_time=9, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=9, stage="base-b", runner_config="1-gpu-small")
 register_amd_ci(est_time=1, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/core/test_request_queue_validation.py

@@ -17,7 +17,7 @@ from sglang.test.test_utils import (
     send_generate_requests,
 )
 
-register_cuda_ci(est_time=53, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=53, stage="base-b", runner_config="1-gpu-small")
 register_amd_ci(est_time=70, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/core/test_srt_endpoint.py

@@ -28,7 +28,7 @@ from sglang.test.test_utils import (
     run_logprob_check,
 )
 
-register_cuda_ci(est_time=134, stage="stage-b", runner_config="1-gpu-small")
+register_cuda_ci(est_time=134, stage="base-b", runner_config="1-gpu-small")
 register_amd_ci(est_time=130, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/core/test_srt_engine.py

@@ -22,7 +22,7 @@ from sglang.test.test_utils import (
     CustomTestCase,
 )
 
-register_cuda_ci(est_time=387, stage="stage-b", runner_config="1-gpu-large")
+register_cuda_ci(est_time=387, stage="base-b", runner_config="1-gpu-large")
 register_amd_ci(est_time=261, suite="stage-b-test-1-gpu-small-amd")
 
 

test/registered/cpu/test_activation.py

@@ -8,7 +8,7 @@ from sglang.srt.server_args import ServerArgs, set_global_server_args_for_schedu
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_binding.py

@@ -8,7 +8,7 @@ kernel = torch.ops.sgl_kernel
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 
 class TestBinding(CustomTestCase):

test/registered/cpu/test_bmm.py

@@ -10,7 +10,7 @@ from sglang.srt.layers.quantization.fp8_utils import input_to_float8
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_causal_conv1d.py

@@ -9,7 +9,7 @@ from utils import parametrize, precision
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 causal_conv1d_weight_pack = torch.ops.sgl_kernel.causal_conv1d_weight_pack
 causal_conv1d_fwd = torch.ops.sgl_kernel.causal_conv1d_fwd_cpu

test/registered/cpu/test_cpu_graph.py

@@ -21,7 +21,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 
 class TestCPUGraph(CustomTestCase):

test/registered/cpu/test_decode.py

@@ -6,7 +6,7 @@ from torch.nn.functional import scaled_dot_product_attention
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_extend.py

@@ -6,7 +6,7 @@ from torch.nn.functional import scaled_dot_product_attention
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_flash_attn.py

@@ -8,7 +8,7 @@ from utils import parametrize, precision
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 flash_attn_varlen_func = torch.ops.sgl_kernel.flash_attn_varlen_func
 

test/registered/cpu/test_gemm.py

@@ -16,7 +16,7 @@ from utils import (
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_intel_amx_attention_backend_a.py

@@ -20,7 +20,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 
 class TestIntelAMXAttnBackend(CustomTestCase):

test/registered/cpu/test_intel_amx_attention_backend_b.py

@@ -14,7 +14,7 @@ from sglang.test.test_utils import (
     intel_amx_benchmark,
 )
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 
 class TestIntelAMXAttnBackendQuant(CustomTestCase):

test/registered/cpu/test_intel_amx_attention_backend_c.py

@@ -14,7 +14,7 @@ from sglang.test.test_utils import (
     intel_amx_benchmark,
 )
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 
 class TestIntelAMXAttnBackendQuant(CustomTestCase):

test/registered/cpu/test_mamba.py

@@ -8,7 +8,7 @@ from utils import precision
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_mla.py

@@ -7,7 +7,7 @@ from utils import precision
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_moe.py

@@ -28,7 +28,7 @@ from utils import (
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 
 def fused_moe(a, w1, w2, score, topk, renormalize, prepack):

test/registered/cpu/test_norm.py

@@ -8,7 +8,7 @@ from utils import make_non_contiguous, parametrize, precision
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_qkv_proj_with_rope.py

@@ -13,7 +13,7 @@ from sglang.srt.layers.rotary_embedding.utils import apply_rotary_emb
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 convert_weight_packed = torch.ops.sgl_kernel.convert_weight_packed
 qkv_proj_with_rope = torch.ops.sgl_kernel.qkv_proj_with_rope

test/registered/cpu/test_qwen3.py

@@ -6,7 +6,7 @@ from utils import precision
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_rope.py

@@ -15,7 +15,7 @@ from sglang.srt.server_args import ServerArgs, set_global_server_args_for_schedu
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_server_args_backend.py

@@ -4,7 +4,7 @@ from unittest.mock import patch
 from sglang.srt.server_args import ServerArgs
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 
 class TestServerArgsCPUBackend(unittest.TestCase):

test/registered/cpu/test_shared_expert.py

@@ -19,7 +19,7 @@ from utils import (
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/cpu/test_topk.py

@@ -11,7 +11,7 @@ from sglang.srt.models.llama4 import Llama4MoE
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=10, suite="stage-b-test-cpu")
+register_cpu_ci(est_time=10, suite="base-b-test-cpu")
 
 torch.manual_seed(1234)
 

test/registered/debug_utils/comparator/aligner/entrypoint/test_executor.py

@@ -28,7 +28,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import ParallelAxis, TokenLayou
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 class TestExecuteSubPlans:

test/registered/debug_utils/comparator/aligner/entrypoint/test_planner.py

@@ -25,7 +25,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import TokenLayout
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 def _make_meta(

test/registered/debug_utils/comparator/aligner/reorderer/test_executor.py

@@ -22,7 +22,7 @@ from sglang.srt.debug_utils.comparator.aligner.unsharder.types import (
 from sglang.srt.debug_utils.comparator.dims_spec import ParallelAxis
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 def _zigzag_order(cp_size: int) -> list[int]:

test/registered/debug_utils/comparator/aligner/reorderer/test_planner.py

@@ -24,7 +24,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 class TestComputeReordererPlans:

test/registered/debug_utils/comparator/aligner/test_axis_aligner.py

@@ -13,7 +13,7 @@ from sglang.srt.debug_utils.comparator.log_sink import log_sink
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 class TestComputeAxisAlignerPlan:

test/registered/debug_utils/comparator/aligner/token_aligner/test_aux_loader.py

@@ -19,7 +19,7 @@ from sglang.srt.debug_utils.comparator.log_sink import LogSink
 from sglang.srt.debug_utils.comparator.output_types import ErrorLog, InfoLog
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 _sglang_plugin = _SGLangPlugin()
 _megatron_plugin = _MegatronPlugin()

test/registered/debug_utils/comparator/aligner/token_aligner/test_aux_plugins.py

@@ -16,7 +16,7 @@ from sglang.srt.debug_utils.comparator.aligner.token_aligner.smart.types import
 from sglang.srt.debug_utils.comparator.dims_spec import TokenLayout
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 _sglang_plugin = _SGLangPlugin()
 _megatron_plugin = _MegatronPlugin()

test/registered/debug_utils/comparator/aligner/token_aligner/test_concat_steps.py

@@ -9,7 +9,7 @@ from sglang.srt.debug_utils.comparator.aligner.token_aligner.concat_steps import
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 class TestExecuteConcat:

test/registered/debug_utils/comparator/aligner/token_aligner/test_executor.py

@@ -25,7 +25,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import TokenLayout
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 def _named(tensor: torch.Tensor, names: list[str]) -> torch.Tensor:

test/registered/debug_utils/comparator/aligner/token_aligner/test_planner.py

@@ -23,7 +23,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import TokenLayout
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=30, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=30, suite="base-a-test-cpu", nightly=True)
 
 
 class TestBuildTokenIndexSGLangThd:

test/registered/debug_utils/comparator/aligner/token_aligner/test_thd_seq_lens_loader.py

@@ -14,7 +14,7 @@ from sglang.srt.debug_utils.comparator.aligner.token_aligner.smart.aux_plugins i
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 def _save_pt(

test/registered/debug_utils/comparator/aligner/unsharder/test_executor.py

@@ -27,7 +27,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import (
 from sglang.srt.debug_utils.comparator.output_types import ReplicatedCheckResult
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 def _name_tensors(

test/registered/debug_utils/comparator/aligner/unsharder/test_parallel_info.py

@@ -9,7 +9,7 @@ from sglang.srt.debug_utils.comparator.aligner.unsharder.types import AxisInfo
 from sglang.srt.debug_utils.comparator.dims_spec import ParallelAxis
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 class TestNormalizeParallelInfo:

test/registered/debug_utils/comparator/aligner/unsharder/test_planner.py

@@ -19,7 +19,7 @@ from sglang.srt.debug_utils.comparator.aligner.unsharder.types import (
 from sglang.srt.debug_utils.comparator.dims_spec import ParallelAxis, parse_dims
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 class TestComputeUnsharderPlan:

test/registered/debug_utils/comparator/dims_spec/test_dim_parser.py

@@ -12,7 +12,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=5, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=5, suite="base-a-test-cpu", nightly=True)
 
 
 class TestParseDim:

test/registered/debug_utils/comparator/dims_spec/test_dims_parser.py

@@ -15,7 +15,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=5, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=5, suite="base-a-test-cpu", nightly=True)
 
 
 class TestSingletonDimUtilFilterOut:

test/registered/debug_utils/comparator/dims_spec/test_tensor_naming.py

@@ -13,7 +13,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=5, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=5, suite="base-a-test-cpu", nightly=True)
 
 
 class TestFindDimIndex:

test/registered/debug_utils/comparator/dims_spec/test_types.py

@@ -9,7 +9,7 @@ from sglang.srt.debug_utils.comparator.dims_spec import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=5, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=5, suite="base-a-test-cpu", nightly=True)
 
 
 class TestDimConstants:

test/registered/debug_utils/comparator/tensor_comparator/test_comparator.py

@@ -14,7 +14,7 @@ from sglang.srt.debug_utils.comparator.tensor_comparator.comparator import (
 from sglang.srt.debug_utils.comparator.tensor_comparator.types import DiffInfo
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=20, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=20, suite="base-a-test-cpu", nightly=True)
 
 
 class TestComputeTensorInfo:

test/registered/debug_utils/comparator/tensor_comparator/test_formatter.py

@@ -58,7 +58,7 @@ from sglang.srt.debug_utils.comparator.tensor_comparator.types import (
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 _DEFAULT_PERCENTILE_LINES: list[str] = [
     "      [blue]p1        [/]    -1.8000      -1.8000   [dim]+0.00e+00[/]",

test/registered/debug_utils/comparator/tensor_comparator/test_types.py

@@ -21,7 +21,7 @@ from sglang.srt.debug_utils.comparator.tensor_comparator.types import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 def _make_stats(**overrides) -> TensorStats:

test/registered/debug_utils/comparator/test_bundle_comparator.py

@@ -15,7 +15,7 @@ from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.srt.debug_utils.dump_loader import ValueWithMeta
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 def _save_tensor(

test/registered/debug_utils/comparator/test_bundle_matcher.py

@@ -13,7 +13,7 @@ from sglang.srt.debug_utils.comparator.bundle_matcher import (
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 def _make_row(

test/registered/debug_utils/comparator/test_display.py

@@ -21,7 +21,7 @@ from sglang.srt.debug_utils.comparator.output_types import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 def _render_rich(renderable: object) -> str:

test/registered/debug_utils/comparator/test_dp_utils.py

@@ -12,7 +12,7 @@ from sglang.srt.debug_utils.comparator.dp_utils import (
 from sglang.srt.debug_utils.dump_loader import ValueWithMeta
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=15, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=15, suite="base-a-test-cpu", nightly=True)
 
 
 def _make_sglang_meta(

test/registered/debug_utils/comparator/test_dump_loader.py

@@ -7,7 +7,7 @@ import torch
 from sglang.srt.debug_utils.dump_loader import read_tokenizer_path
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 def _save_pt(

test/registered/debug_utils/comparator/test_entrypoint.py

@@ -30,7 +30,7 @@ from sglang.srt.debug_utils.comparator.output_types import (
 from sglang.srt.debug_utils.dumper import DumperConfig, _Dumper, _RecomputeStatus
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=30, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=30, suite="base-a-test-cpu", nightly=True)
 
 _FIXED_EXP_NAME = "my_exp_name"
 

test/registered/debug_utils/comparator/test_log_sink.py

@@ -11,7 +11,7 @@ from sglang.srt.debug_utils.comparator.output_types import (
 from sglang.srt.debug_utils.comparator.report_sink import report_sink
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 def _make_error_log(**overrides) -> ErrorLog:

test/registered/debug_utils/comparator/test_manually_verify.py

@@ -23,7 +23,7 @@ import torch
 
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=60, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=60, suite="base-a-test-cpu", nightly=True)
 
 _PUBLISH_DIR: Path = Path("/tmp/comparator_manual_verify")
 _PNG_MAGIC: bytes = b"\x89PNG"

test/registered/debug_utils/comparator/test_meta_overrider.py

@@ -16,7 +16,7 @@ from sglang.srt.debug_utils.comparator.meta_overrider import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 # ───────────────────── Unit: MetaOverrideRule ─────────────────────

test/registered/debug_utils/comparator/test_model_validation.py

@@ -44,7 +44,7 @@ from sglang.srt.debug_utils.comparator.tensor_comparator.types import (
 from sglang.srt.debug_utils.comparator.utils import Pair, _check_equal_lengths
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 class TestCheckEqualLengths:

test/registered/debug_utils/comparator/test_output_types.py

@@ -55,7 +55,7 @@ from sglang.srt.debug_utils.comparator.output_types import (
 from sglang.srt.debug_utils.comparator.utils import Pair
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 def _render_rich(renderable: object) -> str:

test/registered/debug_utils/comparator/test_per_token_visualizer.py

@@ -15,7 +15,7 @@ from sglang.srt.debug_utils.comparator.tensor_comparator.comparator import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=30, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=30, suite="base-a-test-cpu", nightly=True)
 
 _PNG_MAGIC: bytes = b"\x89PNG"
 

test/registered/debug_utils/comparator/test_preset.py

@@ -3,7 +3,7 @@ import pytest
 from sglang.srt.debug_utils.comparator.preset import PRESETS, expand_preset
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=5, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=5, suite="base-a-test-cpu", nightly=True)
 
 
 class TestExpandPreset:

test/registered/debug_utils/comparator/test_utils.py

@@ -17,7 +17,7 @@ from sglang.srt.debug_utils.comparator.utils import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 class TestCalcRelDiff:

test/registered/debug_utils/comparator/test_visualizer.py

@@ -10,7 +10,7 @@ from sglang.srt.debug_utils.comparator.visualizer.preprocessing import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=30, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=30, suite="base-a-test-cpu", nightly=True)
 
 
 class TestPreprocessTensor:

test/registered/debug_utils/comparator/testing_helpers.py

@@ -12,7 +12,7 @@ from sglang.test.ci.ci_register import register_cpu_ci
 
 register_cpu_ci(
     est_time=0,
-    suite="stage-a-test-cpu",
+    suite="base-a-test-cpu",
     nightly=True,
     disabled="helper module, no tests",
 )

test/registered/debug_utils/source_patcher/test_code_patcher.py

@@ -10,7 +10,7 @@ from sglang.srt.debug_utils.source_patcher.code_patcher import (
 from sglang.srt.debug_utils.source_patcher.types import EditSpec, PatchSpec
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 SAMPLE_MODULE_NAME = "_source_patcher_test_fixtures.sample_module"
 

test/registered/debug_utils/source_patcher/test_dumper_integration.py

@@ -8,7 +8,7 @@ import yaml
 from sglang.srt.debug_utils.dumper import DumperConfig, _Dumper
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 SAMPLE_MODULE_NAME = "_source_patcher_test_fixtures.sample_module"
 

test/registered/debug_utils/source_patcher/test_source_editor.py

@@ -5,7 +5,7 @@ from sglang.srt.debug_utils.source_patcher.source_editor import apply_edits
 from sglang.srt.debug_utils.source_patcher.types import EditSpec, PatchApplicationError
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=10, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=10, suite="base-a-test-cpu", nightly=True)
 
 
 class TestApplyEdits:

test/registered/debug_utils/test_cuda_coredump.py

@@ -12,7 +12,7 @@ from sglang.test.ci.ci_register import register_cuda_ci
 
 register_cuda_ci(
     est_time=10,
-    stage="stage-a",
+    stage="base-a",
     runner_config="1-gpu-small",
     disabled="Manual only: triggers intentional CUDA crash for coredump verification",
 )

test/registered/debug_utils/test_dump_comparator.py

@@ -9,7 +9,7 @@ from sglang.srt.debug_utils.dump_comparator import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=30, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=30, suite="base-a-test-cpu", nightly=True)
 
 
 # ----------------------------- Unit tests -----------------------------

test/registered/debug_utils/test_dump_loader.py

@@ -15,7 +15,7 @@ from sglang.srt.debug_utils.dump_loader import (
 )
 from sglang.test.ci.ci_register import register_cpu_ci
 
-register_cpu_ci(est_time=30, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=30, suite="base-a-test-cpu", nightly=True)
 
 
 class TestReadMeta:

test/registered/debug_utils/test_schedule_simulator.py

@@ -25,7 +25,7 @@ from sglang.srt.debug_utils.schedule_simulator import (
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=120, suite="stage-a-test-cpu", nightly=True)
+register_cpu_ci(est_time=120, suite="base-a-test-cpu", nightly=True)
 
 
 # ==================== Non-E2E Tests ====================

test/registered/debug_utils/test_tensor_dump_forward_hook.py

@@ -19,7 +19,7 @@ from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 
 register_cuda_ci(
     est_time=9,
-    stage="stage-b",
+    stage="base-b",
     runner_config="1-gpu-small",
     disabled="Test uses pytest-style function without TestCase class - see #17145",
 )

test/registered/disaggregation/test_disaggregation_basic.py

@@ -21,7 +21,7 @@ from sglang.test.test_utils import (
     DEFAULT_TARGET_MODEL_EAGLE3,
 )
 
-register_cuda_ci(est_time=509, stage="stage-b", runner_config="2-gpu-large")
+register_cuda_ci(est_time=509, stage="base-b", runner_config="2-gpu-large")
 
 
 class TestDisaggregationAccuracy(PauseResumeInPlaceMixin, PDDisaggregationServerBase):

test/registered/disaggregation/test_disaggregation_decode_offload.py

@@ -19,7 +19,7 @@ from sglang.test.test_utils import (
 # Increasing estimated time since we run evaluation twice
 register_cuda_ci(
     est_time=600,
-    stage="stage-b",
+    stage="base-b",
     runner_config="2-gpu-large",
     disabled="Temporarily disable the flaky test.",
 )