We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
1 parent b3b3476 commit 56d9dc5Copy full SHA for 56d9dc5
1 file changed
.github/configs/amd-master.yaml
@@ -434,7 +434,7 @@ qwen3.5-fp4-mi355x-sglang:
434
- { tp: 4, conc-start: 4, conc-end: 16 }
435
436
# target
437
-qwen3.5-fp4-mi355x-sglang-hicache:
+qwen3.5-fp4-mi355x-sglang-agentic-hicache:
438
image: lmsysorg/sglang:v0.5.12-rocm720-mi35x
439
model: amd/Qwen3.5-397B-A17B-MXFP4
440
model-prefix: qwen3.5
@@ -2617,8 +2617,8 @@ minimaxm2.5-fp8-mi355x-vllm-agentic-lmcache:
2617
# AMD uses native OffloadingConnector (NOT SimpleCPUOffloadConnector).
2618
- duration: 1800
2619
search-space:
2620
- - { tp: 1, ep: 1, offloading: none, conc-list: [1, 2, 4, 8, 16, 32, 40, 48] }
2621
- - { tp: 1, ep: 1, offloading: lmcache, conc-list: [1, 2, 4, 8, 16, 32, 40, 48] }
+ - { tp: 2, ep: 1, offloading: none, conc-list: [1, 2, 4, 8, 16, 32, 40, 48] }
+ - { tp: 2, ep: 1, offloading: lmcache, conc-list: [1, 2, 4, 8, 16, 32, 40, 48] }
2622
2623
minimaxm2.5-fp8-mi300x-vllm-agentic:
2624
image: vllm/vllm-openai-rocm:v0.22.0
0 commit comments