A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 12, 03:30 UTC
Reverting re-direction to amd_mi355_X. (#29914)
Signed-off-by: Alexei V. Ivanov <alexei.ivanov@amd.com>
Alexei-V-Ivanov-AMD committed
Dec 2, 2025 at 18:40 UTC
5f67361fd12851bfe8faad4cc173ca24565611e4
1 file changed
+11
-11
.buildkite/test-amd.yaml
+11
-11
@@ -51,7 +51,7 @@ steps:
51
- label: Async Engine, Inputs, Utils, Worker Test # 10min
52
timeout_in_minutes: 15
53
mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
54
- agent_pool: mi355_1
54
+ agent_pool: mi325_1
55
grade: Blocking
56
source_file_dependencies:
57
- vllm/
@@ -64,7 +64,7 @@ steps:
64
- label: Async Engine, Inputs, Utils, Worker, Config Test (CPU) # 15min
65
timeout_in_minutes: 20
66
mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
67
- agent_pool: mi355_1
67
+ agent_pool: mi325_1
68
grade: Blocking
69
source_file_dependencies:
70
- vllm/
@@ -99,7 +99,7 @@ steps:
99
- label: Basic Correctness Test # 20min
100
timeout_in_minutes: 30
101
mirror_hardwares: [amdexperimental, amdproduction]
102
- agent_pool: mi355_1
102
+ agent_pool: mi325_1
103
# grade: Blocking
104
fast_check: true
105
torch_nightly: true
@@ -116,7 +116,7 @@ steps:
116
117
- label: Entrypoints Unit Tests # 5min
118
mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
119
- agent_pool: mi355_1
119
+ agent_pool: mi325_1
120
grade: Blocking
121
timeout_in_minutes: 10
122
working_dir: "/vllm-workspace/tests"
@@ -131,7 +131,7 @@ steps:
131
- label: Entrypoints Integration Test (LLM) # 30min
132
timeout_in_minutes: 40
133
mirror_hardwares: [amdexperimental, amdproduction]
134
- agent_pool: mi355_1
134
+ agent_pool: mi325_1
135
# grade: Blocking
136
working_dir: "/vllm-workspace/tests"
137
fast_check: true
@@ -254,7 +254,7 @@ steps:
254
255
- label: EPLB Algorithm Test # 5min
256
mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
257
- agent_pool: mi355_1
257
+ agent_pool: mi325_1
258
grade: Blocking
259
timeout_in_minutes: 15
260
working_dir: "/vllm-workspace/tests"
@@ -266,7 +266,7 @@ steps:
266
267
- label: EPLB Execution Test # 10min
268
mirror_hardwares: [amdexperimental, amdproduction]
269
- agent_pool: mi355_4
269
+ agent_pool: mi325_4
270
# grade: Blocking
271
timeout_in_minutes: 20
272
working_dir: "/vllm-workspace/tests"
@@ -281,7 +281,7 @@ steps:
281
- label: Metrics, Tracing Test # 12min
282
timeout_in_minutes: 20
283
mirror_hardwares: [amdexperimental, amdproduction]
284
- agent_pool: mi355_2
284
+ agent_pool: mi325_2
285
# grade: Blocking
286
num_gpus: 2
287
source_file_dependencies:
@@ -301,7 +301,7 @@ steps:
301
- label: Regression Test # 7min
302
timeout_in_minutes: 20
303
mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
304
- agent_pool: mi355_1
304
+ agent_pool: mi325_1
305
grade: Blocking
306
source_file_dependencies:
307
- vllm/
@@ -343,7 +343,7 @@ steps:
343
- label: V1 Test entrypoints # 35min
344
timeout_in_minutes: 50
345
mirror_hardwares: [amdexperimental, amdproduction, amdtentative]
346
- agent_pool: mi355_1
346
+ agent_pool: mi325_1
347
grade: Blocking
348
source_file_dependencies:
349
- vllm/
@@ -544,7 +544,7 @@ steps:
544
- label: PyTorch Fullgraph Test # 27min
545
timeout_in_minutes: 40
546
mirror_hardwares: [amdexperimental, amdproduction]
547
- agent_pool: mi355_1
547
+ agent_pool: mi325_1
548
# grade: Blocking
549
torch_nightly: true
550
source_file_dependencies: