ci: unify PR test suite naming (#21187)

test/registered/quant/test_autoround.py

@@ -17,7 +17,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=77, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=77, suite="stage-b-test-1-gpu-large")
 
 
 class TestAutoRound(CustomTestCase):

test/registered/quant/test_awq.py

@@ -13,8 +13,8 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=700, suite="stage-b-test-large-1-gpu")
-register_amd_ci(est_time=200, suite="stage-b-test-large-1-gpu-amd")
+register_cuda_ci(est_time=700, suite="stage-b-test-1-gpu-large")
+register_amd_ci(est_time=200, suite="stage-b-test-1-gpu-large-amd")
 
 
 class TestAWQ(CustomTestCase):

test/registered/quant/test_awq_dequant.py

@@ -21,7 +21,7 @@ from sglang.srt.utils import get_device
 from sglang.test.ci.ci_register import register_amd_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_amd_ci(est_time=2, suite="stage-a-test-small-1-gpu-amd")
+register_amd_ci(est_time=2, suite="stage-a-test-1-gpu-small-amd")
 
 device = get_device()
 

test/registered/quant/test_block_int8.py

@@ -10,8 +10,8 @@ from sglang.srt.server_args import ServerArgs, set_global_server_args_for_schedu
 from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cuda_ci(est_time=44, suite="stage-b-test-small-1-gpu")
-register_amd_ci(est_time=22, suite="stage-b-test-small-1-gpu-amd")
+register_cuda_ci(est_time=44, suite="stage-b-test-1-gpu-small")
+register_amd_ci(est_time=22, suite="stage-b-test-1-gpu-small-amd")
 
 
 # For test

test/registered/quant/test_bnb.py

@@ -22,7 +22,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=5, suite="stage-b-test-small-1-gpu")
+register_cuda_ci(est_time=5, suite="stage-b-test-1-gpu-small")
 
 VISION_MODELS = [
     "unsloth/Qwen2.5-VL-7B-Instruct-bnb-4bit",

test/registered/quant/test_eval_fp8_accuracy.py

@@ -14,8 +14,8 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=250, suite="stage-b-test-large-1-gpu")
-register_amd_ci(est_time=600, suite="stage-b-test-small-1-gpu-amd")
+register_cuda_ci(est_time=250, suite="stage-b-test-1-gpu-large")
+register_amd_ci(est_time=600, suite="stage-b-test-1-gpu-small-amd")
 
 
 class TestEvalFP8Accuracy(CustomTestCase):

test/registered/quant/test_fp8_kernel.py

@@ -9,7 +9,7 @@ from sglang.srt.layers.quantization.fp8_kernel import (
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cuda_ci(est_time=132, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=132, suite="stage-b-test-1-gpu-large")
 
 
 class TestFP8Base(CustomTestCase):

test/registered/quant/test_fp8_utils.py

@@ -10,7 +10,7 @@ from sglang.srt.layers.quantization.fp8_utils import (
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cuda_ci(est_time=9, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=9, suite="stage-b-test-1-gpu-large")
 
 
 class TestInverseTransformScaleUe8m0(CustomTestCase):

test/registered/quant/test_fp8kv_triton.py

@@ -12,7 +12,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=520, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=520, suite="stage-b-test-1-gpu-large")
 
 
 class TestFP8KVCacheTritonBackend(CustomTestCase):

test/registered/quant/test_fused_rms_fp8_group_quant.py

@@ -7,7 +7,7 @@ import torch.nn.functional as F
 from sglang.test.ci.ci_register import register_amd_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_amd_ci(est_time=10, suite="stage-a-test-small-1-gpu-amd")
+register_amd_ci(est_time=10, suite="stage-a-test-1-gpu-small-amd")
 
 
 def _fp8_available() -> bool:

test/registered/quant/test_gguf.py

@@ -6,7 +6,7 @@ import sglang as sgl
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cuda_ci(est_time=96, suite="stage-b-test-small-1-gpu")
+register_cuda_ci(est_time=96, suite="stage-b-test-1-gpu-small")
 
 
 class TestGGUF(CustomTestCase):

test/registered/quant/test_gptqmodel_dynamic.py

@@ -14,7 +14,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=102, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=102, suite="stage-b-test-1-gpu-large")
 
 
 def check_quant_method(model_path: str, use_marlin_kernel: bool):

test/registered/quant/test_int4fp8_moe.py

@@ -9,7 +9,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_amd_ci(est_time=313, suite="stage-b-test-small-1-gpu-amd")
+register_amd_ci(est_time=313, suite="stage-b-test-1-gpu-small-amd")
 
 
 class TestMixtralAccuracy(CustomTestCase):

test/registered/quant/test_int8_kernel.py

@@ -11,7 +11,7 @@ from sglang.srt.server_args import ServerArgs, set_global_server_args_for_schedu
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cuda_ci(est_time=8, suite="stage-b-test-small-1-gpu")
+register_cuda_ci(est_time=8, suite="stage-b-test-1-gpu-small")
 
 
 def native_w8a8_per_token_matmul(A, B, As, Bs, output_dtype=torch.float16):

test/registered/quant/test_marlin_moe.py

@@ -12,7 +12,7 @@ from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_marlin_utils import awq_marlin_quantize, marlin_quantize
 from sglang.test.test_utils import CustomTestCase
 
-register_cuda_ci(est_time=200, suite="stage-b-test-small-1-gpu")
+register_cuda_ci(est_time=200, suite="stage-b-test-1-gpu-small")
 
 set_global_server_args_for_scheduler(object.__new__(ServerArgs))
 

test/registered/quant/test_modelopt_fp8.py

@@ -12,7 +12,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=120, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=120, suite="stage-b-test-1-gpu-large")
 
 
 class TestModeloptFP8(CustomTestCase):

test/registered/quant/test_quant_config_parsing.py

@@ -5,7 +5,7 @@ from sglang.srt.configs.model_config import ModelConfig
 from sglang.test.ci.ci_register import register_cpu_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cpu_ci(est_time=5, suite="stage-a-cpu-only")
+register_cpu_ci(est_time=5, suite="stage-a-test-cpu")
 
 
 class TestQuantLogString(CustomTestCase):

test/registered/quant/test_quantization.py

@@ -16,7 +16,7 @@ from sglang.test.test_utils import (
     write_results_to_json,
 )
 
-register_cuda_ci(est_time=185, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=185, suite="stage-b-test-1-gpu-large")
 
 MODEL_SCORE_THRESHOLDS = {
     "hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4": 0.825,

test/registered/quant/test_torchao.py

@@ -6,8 +6,8 @@ import requests
 from sglang import Engine
 from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 
-register_cuda_ci(est_time=103, suite="stage-b-test-small-1-gpu")
-register_amd_ci(est_time=230, suite="stage-b-test-small-1-gpu-amd")
+register_cuda_ci(est_time=103, suite="stage-b-test-1-gpu-small")
+register_amd_ci(est_time=230, suite="stage-b-test-1-gpu-small-amd")
 from sglang.lang.chat_template import get_chat_template_by_model_path
 from sglang.srt.utils import kill_process_tree
 from sglang.test.run_eval import run_eval

test/registered/quant/test_triton_scaled_mm.py

@@ -8,8 +8,8 @@ from sglang.srt.layers.quantization.fp8_kernel import triton_scaled_mm
 from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
 from sglang.test.test_utils import CustomTestCase
 
-register_cuda_ci(est_time=8, suite="stage-b-test-small-1-gpu")
-register_amd_ci(est_time=12, suite="stage-b-test-small-1-gpu-amd")
+register_cuda_ci(est_time=8, suite="stage-b-test-1-gpu-small")
+register_amd_ci(est_time=12, suite="stage-b-test-1-gpu-small-amd")
 
 
 def torch_scaled_mm(

test/registered/quant/test_w8a8_quantization.py

@@ -14,7 +14,7 @@ from sglang.test.test_utils import (
     popen_launch_server,
 )
 
-register_cuda_ci(est_time=160, suite="stage-b-test-large-1-gpu")
+register_cuda_ci(est_time=160, suite="stage-b-test-1-gpu-large")
 
 
 class BaseW8A8Test(CustomTestCase):