Skip to content

Commit 05e127b

Browse files
[Klaud Cold] Update dsv4-fp4-b300-vllm vLLM image to v0.21.0 (#1456)
* Update dsv4-fp4-b300-vllm vLLM image to v0.21.0 * chore: fill pr-link for #1456
1 parent bce10f7 commit 05e127b

2 files changed

Lines changed: 7 additions & 1 deletion

File tree

.github/configs/nvidia-master.yaml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -2964,7 +2964,7 @@ dsv4-fp8-h200-sglang-mtp:
29642964
# field, so dp-attn=true is used as the existing vLLM script switch for DP4
29652965
# layouts on 4 allocated GPUs.
29662966
dsv4-fp4-b300-vllm:
2967-
image: vllm/vllm-openai:v0.20.0-cu130
2967+
image: vllm/vllm-openai:v0.21.0
29682968
model: deepseek-ai/DeepSeek-V4-Pro
29692969
model-prefix: dsv4
29702970
runner: b300

perf-changelog.yaml

Lines changed: 6 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2802,3 +2802,9 @@
28022802
description:
28032803
- "Update vLLM image from v0.20.0-cu130 (14d old) to v0.21.0"
28042804
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1457
2805+
2806+
- config-keys:
2807+
- dsv4-fp4-b300-vllm
2808+
description:
2809+
- "Update vLLM image from v0.20.0-cu130 (18d old) to v0.21.0"
2810+
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1456

0 commit comments

Comments
 (0)