fix: update GB200 fp8 MTP image to nightly-6f91edf

Author: xinli-sw

benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb200-fp8/8k1k/mtp/1p2d-dep4-tep8-eagle3-8k1k.yaml

@@ -2,7 +2,7 @@ name: "minimax-m3-vllm-disagg-gb200-1p2d-dep4-tep8-fp8-8k1k-eagle3"
 
 model:
   path: "minimax-m3-mxfp8"
-  container: "vllm/vllm-openai:v0.26.0"
+  container: "vllm/vllm-openai:nightly-6f91edf96d3f3272945809c04702380053bff4de"
   precision: "fp8"
 
 dynamo:

benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb200-fp8/8k1k/mtp/2p1d-dep4-dep16-eagle3-8k1k.yaml

@@ -2,7 +2,7 @@ name: "minimax-m3-vllm-disagg-gb200-2p1d-dep4-dep16-fp8-8k1k-eagle3"
 
 model:
   path: "minimax-m3-mxfp8"
-  container: "vllm/vllm-openai:v0.26.0"
+  container: "vllm/vllm-openai:nightly-6f91edf96d3f3272945809c04702380053bff4de"
   precision: "fp8"
 
 dynamo:

configs/nvidia-master.yaml

@@ -8169,7 +8169,7 @@ minimaxm3-fp8-gb200-dynamo-vllm:
           dp-attn: true
 
 minimaxm3-fp8-gb200-dynamo-vllm-mtp:
-  image: vllm/vllm-openai:v0.26.0
+  image: vllm/vllm-openai:nightly-6f91edf96d3f3272945809c04702380053bff4de
   model: MiniMaxAI/MiniMax-M3-MXFP8
   model-prefix: minimaxm3
   runner: gb200