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

python/sglang/jit_kernel/tests/deepseek_v4/test_c128_v2.py

@@ -18,7 +18,7 @@ from sglang.jit_kernel.tests.deepseek_v4.common import (
     to_seq_extend,
 )
 
-register_cuda_ci(est_time=30, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=30, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=30, suite="nightly-kernel-1-gpu", nightly=True)
 
 Context = Union[LegacyContext, PagedContext]

python/sglang/jit_kernel/tests/deepseek_v4/test_c4_v2.py

@@ -18,7 +18,7 @@ from sglang.jit_kernel.tests.deepseek_v4.common import (
     to_seq_extend,
 )
 
-register_cuda_ci(est_time=30, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=30, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=30, suite="nightly-kernel-1-gpu", nightly=True)
 
 Context = Union[LegacyContext, PagedContext]

python/sglang/jit_kernel/tests/diffusion/test_diffusion_modelopt_fp8_scaled_mm.py

@@ -14,7 +14,7 @@ from sglang.srt.layers.quantization.fp8_utils import (
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=20, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=20, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=80, suite="nightly-kernel-1-gpu", nightly=True)
 
 DEVICE = "cuda"

python/sglang/jit_kernel/tests/diffusion/test_diffusion_nvfp4_scaled_mm.py

@@ -17,7 +17,7 @@ from sglang.srt.layers.quantization.modelopt_quant import pad_nvfp4_weight
 from sglang.test.ci.ci_register import register_cuda_ci
 
 # B200-only correctness coverage for diffusion NVFP4 scaled mm.
-register_cuda_ci(est_time=15, suite="stage-b-kernel-unit-1-gpu-b200")
+register_cuda_ci(est_time=15, suite="base-b-kernel-unit-1-gpu-b200")
 
 DEVICE = "cuda"
 DTYPE = torch.bfloat16

python/sglang/jit_kernel/tests/diffusion/test_fused_norm_scale_shift.py

@@ -13,7 +13,7 @@ from sglang.jit_kernel.diffusion.cutedsl.scale_residual_norm_scale_shift import
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=28, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=28, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 DEVICE = "cuda"

python/sglang/jit_kernel/tests/diffusion/test_group_norm_silu.py

@@ -9,7 +9,7 @@ from sglang.jit_kernel.diffusion.group_norm_silu import apply_group_norm_silu
 from sglang.jit_kernel.diffusion.triton.group_norm_silu import triton_group_norm_silu
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=8, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=8, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 DEVICE = "cuda"

python/sglang/jit_kernel/tests/diffusion/test_qknorm_rope.py

@@ -8,7 +8,7 @@ import triton
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=44, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=44, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=176, suite="nightly-kernel-1-gpu", nightly=True)
 
 DEVICE = "cuda"

python/sglang/jit_kernel/tests/diffusion/test_qwen_image_modulation.py

@@ -12,7 +12,7 @@ from sglang.jit_kernel.diffusion.triton.scale_shift import (
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=15, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=15, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 DEVICE = "cuda"

python/sglang/jit_kernel/tests/test_activation.py

@@ -8,7 +8,7 @@ from sglang.jit_kernel.activation import SUPPORTED_ACTIVATIONS, run_activation
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=20, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=20, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=30, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_add_constant.py

@@ -6,7 +6,7 @@ import torch
 from sglang.jit_kernel.add_constant import add_constant
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=45, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=45, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=180, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_awq_dequantize.py

@@ -7,7 +7,7 @@ import torch
 from sglang.jit_kernel.awq_dequantize import awq_dequantize as jit_awq_dequantize
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=9, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=9, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 try:

python/sglang/jit_kernel/tests/test_awq_marlin_moe_repack.py

@@ -11,7 +11,7 @@ from sglang.jit_kernel.awq_marlin_repack import (
 from sglang.srt.layers.quantization.utils import pack_cols, quantize_weights
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=10, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=10, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_awq_marlin_repack.py

@@ -12,7 +12,7 @@ from sglang.srt.layers.quantization.utils import pack_cols, quantize_weights
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_marlin_utils import get_weight_perm, marlin_weights
 
-register_cuda_ci(est_time=10, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=10, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_clamp_position.py

@@ -6,7 +6,7 @@ import torch
 from sglang.jit_kernel.clamp_position import clamp_position_cuda
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=12, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=12, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_concat_mla.py

@@ -7,7 +7,7 @@ import triton
 
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=17, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=17, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_custom_all_reduce.py

@@ -38,7 +38,7 @@ from sglang.test.ci.ci_register import register_cuda_ci
 
 register_cuda_ci(
     est_time=300,
-    suite="stage-b-kernel-unit-8-gpu-h200",
+    suite="base-b-kernel-unit-8-gpu-h200",
 )
 register_cuda_ci(
     est_time=300,

python/sglang/jit_kernel/tests/test_cutedsl_gdn.py

@@ -29,7 +29,7 @@ try:
 except ImportError:
     TRITON_AVAILABLE = False
 
-register_cuda_ci(est_time=5, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=5, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_flash_attention_4.py

@@ -14,7 +14,7 @@ from einops import rearrange, repeat
 from sglang.jit_kernel.flash_attention import flash_attn_varlen_func
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=120, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=120, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=900, suite="nightly-kernel-1-gpu", nightly=True)
 
 # Skip this test on Hopper machine

python/sglang/jit_kernel/tests/test_fused_add_rmsnorm.py

@@ -7,7 +7,7 @@ import torch
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=5, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=5, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_fused_metadata_copy.py

@@ -16,7 +16,7 @@ import torch
 
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=100, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=100, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=400, suite="nightly-kernel-1-gpu", nightly=True)
 
 # =============================================================================

python/sglang/jit_kernel/tests/test_fused_store_index_cache.py

@@ -48,7 +48,7 @@ try:
 except ImportError:
     _is_fp8_fnuz = False
 
-register_cuda_ci(est_time=24, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=24, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 PAGE_SIZE = 64

python/sglang/jit_kernel/tests/test_fused_verify_triton_gdn.py

@@ -26,7 +26,7 @@ try:
 except ImportError:
     KERNELS_AVAILABLE = False
 
-register_cuda_ci(est_time=6, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=6, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_gptq_marlin.py

@@ -9,7 +9,7 @@ from sglang.srt.layers.quantization.marlin_utils import marlin_make_workspace
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_marlin_utils import awq_marlin_quantize, marlin_quantize
 
-register_cuda_ci(est_time=13, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=13, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 MNK_FACTORS = [

python/sglang/jit_kernel/tests/test_gptq_marlin_repack.py

@@ -13,7 +13,7 @@ from sglang.srt.layers.quantization.utils import (
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_marlin_utils import get_weight_perm, marlin_weights
 
-register_cuda_ci(est_time=16, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=16, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 MARLIN_K_CHUNKS = [128]

python/sglang/jit_kernel/tests/test_grouped_topk.py

@@ -9,7 +9,7 @@ from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.srt.layers.moe.topk import biased_grouped_topk_impl
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=30, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=30, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_hadamard_jit.py

@@ -16,7 +16,7 @@ from sglang.jit_kernel.hadamard import (
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=128, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=128, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=512, suite="nightly-kernel-1-gpu", nightly=True)
 
 # Exact M×N Hadamard matrices (±1 entries) copied from

python/sglang/jit_kernel/tests/test_hicache.py

@@ -13,7 +13,7 @@ from sglang.srt.mem_cache.memory_pool_host import (
 from sglang.srt.utils import is_cuda, is_hip, is_npu, is_xpu
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=10, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=10, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 pytestmark = pytest.mark.skipif(

python/sglang/jit_kernel/tests/test_hisparse.py

@@ -7,7 +7,7 @@ from sglang.jit_kernel.hisparse import load_cache_to_device_buffer_mla
 from sglang.srt.utils import is_cuda, is_hip, is_npu, is_xpu
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=10, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=10, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 pytestmark = pytest.mark.skipif(

python/sglang/jit_kernel/tests/test_mla_kv_pack_quantize_fp8.py

@@ -7,7 +7,7 @@ from sglang.jit_kernel.mla_kv_pack_quantize_fp8 import mla_kv_pack_quantize_fp8
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=60, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=60, suite="base-b-kernel-unit-1-gpu-large")
 
 DEVICE = "cuda"
 

python/sglang/jit_kernel/tests/test_moe_align_block_size.py

@@ -9,7 +9,7 @@ import triton.language as tl
 from sglang.jit_kernel.moe_align import moe_align_block_size
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=28, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=28, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_moe_lora_align_block_size.py

@@ -11,7 +11,7 @@ import torch
 from sglang.jit_kernel.moe_lora_align import moe_lora_align_block_size
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=28, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=28, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_moe_wna16_marlin.py

@@ -10,7 +10,7 @@ from sglang.srt.layers.moe.fused_moe_triton import moe_align_block_size
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_marlin_utils import awq_marlin_quantize, marlin_quantize
 
-register_cuda_ci(est_time=10, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=10, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_mxfp8_moe.py

@@ -10,7 +10,7 @@ from sglang.jit_kernel.mxfp8 import (
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=5, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=5, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_nvfp4_blockwise_moe.py

@@ -10,7 +10,7 @@ from sglang.jit_kernel.nvfp4 import (
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=5, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=5, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 FLOAT4_E2M1_MAX = 6.0

python/sglang/jit_kernel/tests/test_nvfp4_gemm.py

@@ -6,7 +6,7 @@ import torch
 from sglang.jit_kernel.nvfp4 import cutlass_scaled_fp4_mm, scaled_fp4_quant
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=5, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=5, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_nvfp4_quant.py

@@ -16,7 +16,7 @@ except Exception:
 
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=5, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=5, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_per_tensor_quant_fp8.py

@@ -8,7 +8,7 @@ import torch
 from sglang.jit_kernel.per_tensor_quant_fp8 import per_tensor_quant_fp8
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=16, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=16, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 try:

python/sglang/jit_kernel/tests/test_per_token_group_quant_8bit.py

@@ -25,7 +25,7 @@ from sglang.srt.layers.quantization.fp8_kernel import (
 )
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=16, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=16, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 configs = list(

python/sglang/jit_kernel/tests/test_pos_enc.py

@@ -10,7 +10,7 @@ import triton.language as tl
 from sglang.jit_kernel.rope import rotary_embedding
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=18, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=18, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_qknorm.py

@@ -8,7 +8,7 @@ import triton
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=37, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=37, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=148, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_qknorm_across_heads.py

@@ -8,7 +8,7 @@ import triton
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=15, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=15, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_renorm.py

@@ -9,7 +9,7 @@ import torch
 
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=6, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=6, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_resolve_future_token_ids.py

@@ -6,7 +6,7 @@ import torch
 from sglang.jit_kernel.resolve_future_token_ids import resolve_future_token_ids_cuda
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=9, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=9, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_rmsnorm.py

@@ -7,7 +7,7 @@ import torch
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=45, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=45, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=240, suite="nightly-kernel-1-gpu", nightly=True)
 
 

python/sglang/jit_kernel/tests/test_rmsnorm_hf.py

@@ -13,7 +13,7 @@ from sglang.jit_kernel.rmsnorm_hf import (
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=30, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=30, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 EPS = 1e-5

python/sglang/jit_kernel/tests/test_rope.py

@@ -7,7 +7,7 @@ import triton
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=64, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=64, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=256, suite="nightly-kernel-1-gpu", nightly=True)
 
 DEVICE = "cuda"

python/sglang/jit_kernel/tests/test_set_mla_kv_buffer.py

@@ -10,7 +10,7 @@ from sglang.jit_kernel.set_mla_kv_buffer import (
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=30, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=30, suite="base-b-kernel-unit-1-gpu-large")
 
 DEVICE = "cuda"
 CACHE_SIZE = 4096

python/sglang/jit_kernel/tests/test_store_cache.py

@@ -8,7 +8,7 @@ from sglang.jit_kernel.kvcache import can_use_store_cache, store_cache
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 
-register_cuda_ci(est_time=28, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=28, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 register_amd_ci(est_time=55, suite="jit-kernel-unit-test-amd")
 

python/sglang/jit_kernel/tests/test_timestep_embedding.py

@@ -16,7 +16,7 @@ from sglang.jit_kernel.timestep_embedding import (
 from sglang.jit_kernel.utils import get_ci_test_range
 from sglang.test.ci.ci_register import register_cuda_ci
 
-register_cuda_ci(est_time=16, suite="stage-b-kernel-unit-1-gpu-large")
+register_cuda_ci(est_time=16, suite="base-b-kernel-unit-1-gpu-large")
 register_cuda_ci(est_time=120, suite="nightly-kernel-1-gpu", nightly=True)
 
 CORRECTNESS_BATCH_SIZES = get_ci_test_range(

python/sglang/jit_kernel/tests/test_tp_qknorm.py

@@ -16,7 +16,7 @@ from sglang.test.ci.ci_register import register_cuda_ci
 
 register_cuda_ci(
     est_time=300,
-    suite="stage-b-kernel-unit-8-gpu-h200",
+    suite="base-b-kernel-unit-8-gpu-h200",
 )
 register_cuda_ci(
     est_time=300,