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 7, 02:45 UTC
[ROCm][CI] Fixing yaml file for external amd-ci signal (#36284)
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Andreas Karatzas committed
Mar 6, 2026 at 18:30 UTC
b5e34e1fcaefaf1d28249b6db17c99084ea25b5e
1 file changed
+3
-3
.buildkite/test-amd.yaml
+3
-3
@@ -2801,7 +2801,7 @@ steps:
2801
- vllm/v1/attention/selector.py
2802
- vllm/platforms/cuda.py
2803
commands:
2804
- rocm-smi
2804
+ - rocm-smi
2805
- python3 examples/offline_inference/basic/chat.py
2806
# Attention
2807
# num_heads2 broken by https://github.com/flashinfer-ai/flashinfer/issues/1353
@@ -3283,7 +3283,7 @@ steps:
3283
commands:
3284
- bash .buildkite/scripts/scheduled_integration_test/deepseek_v2_lite_ep_eplb.sh 0.25 200 8010
3285
3286
-- label: Qwen3-30B-A3B-FP8-block Accuracy (B200/MI355)
3286
+- label: Qwen3-30B-A3B-FP8-block Accuracy (B200-MI355)
3287
mirror_hardwares: [amdexperimental, amdproduction, amdmi355]
3288
agent_pool: mi355_2
3289
timeout_in_minutes: 60
@@ -3305,7 +3305,7 @@ steps:
3305
commands:
3306
- bash .buildkite/scripts/scheduled_integration_test/qwen3_next_mtp_async_eplb.sh 0.8 1319 8040
3307
3308
-- label: Attention Benchmarks Smoke Test (B200/MI355)
3308
+- label: Attention Benchmarks Smoke Test (B200-MI355)
3309
device: b200
3310
mirror_hardwares: [amdexperimental, amdmi355]
3311
agent_pool: mi355_2