[release] 0.7.3.post1 release checklist

[Yikun]

New feature

• Qwen3 and Qwen3 Moe support backport
  • Qwen3 [Model][0.7.3] Add support for Qwen3 model #903
  • Qwen3 Moe [Model][0.7.3] Add support for Qwen3-MoE model #915

Bug Fix

• Bug fix: v1 engine with deepseek: [BugFix][0.7.3] Fix chunked prefill bugs in engine v1 #849
• Bug fix: Qwen/Qwen2.5-VL-7B-Instruct model no revision problem: [Bugfix][0.7.3] Patch transformer utils #858 [Bugfix][0.7.3] Update transformer utils patch #918
• Qwen2.5 vl doesn't work by RLHF case [Bugfix][0.7.3] Add unpadded Qwen2.5-VL for verl scenario #928
• unused log: [Bugfix] [0.7.3] fix log output bug when calling npu_mrope #911

Doc

• [Benchmark][0.7.3] Add qwen2.5 and qwen2.5-vl to benchmarks #792
• [Doc][0.7.3] Add performance tuning docs #878
• [Doc][v0.7.3] Support XLM-RoBERTa-based and MiniCPM3 model #824
• [Doc][0.7.3] Update supported models doc #921

[Yikun]

The v0.7.3.post1 is code freezing, now will be ready in next week.

You can use

• quay.io/ascend/vllm-ascend:v0.7.3-dev
• m.daocloud.io/quay.io/ascend/vllm-ascend:v0.7.3-dev

have a try.