vllm_issue_search1

"Skip to content\nNavigation Menu\nPlatform\nSolutions\nResources\nOpen Source\nEnterprise\nPricing\nSign in\nSign up\nvllm-project\n/\nvllm\nPublic\n Sponsor\nNotifications\nFork 19.4k\n Star 86.3k\nCode\nIssues\n2k\nPull requests\n3.8k\nDiscussions\nActions\nProjects\nSecurity and quality\n54\nInsights\nIssues\nAssigned to me\nCreated by me\nMentioned\nRecent activity\nViews\nProjects\nMilestones\nLabels\nFeedback\nPreview\nCollapse sidebar\n[RFC]: Changes in vLLM Model Development\n#42770 · WoosukKwon opened \n20\n[Roadmap] Rust Frontend Feature Parity\n#44280 · BugenZhao opened \n16\n[Roadmap] vLLM Roadmap Q3 2026\n#48168 · simon-mo opened \n1\nIssues\nSearch Issues\nis:issue\n \ndisaggregated\n \nmemory\nNew issue\nSearch results\nOpen\n34\n (34)\nClosed\n59\n (59)\nAuthor\nLabels\nProjects\nMilestones\nAssignees\nTypes\n[Feature]: Support fast weight updates from disk\nfeature request\nNew feature or request\nStatus: Open.\n#48644 \nIn vllm-project/vllm;\n· SumanthRH opened \n[Roadmap] 2026 Q3 vLLM × RL\nStatus: Open.\n#48314 \nIn vllm-project/vllm;\n· aoshen02 opened \n2\ncomments\n[RFC][Frontend/CLI]: Offline prefix-cache workload analyzer\nStatus: Open.\n#47993 \nIn vllm-project/vllm;\n· raravind007 opened \n1\n7\ncomments\nRuntimeError: shape mismatch during KV cache init with EP + DP on MoE model\nStatus: Open.\n#47722 \nIn vllm-project/vllm;\n· elvircrn opened \n1\n3\ncomments\n[RFC]: Streaming Derender for Disaggregated Serving\nRFC\nStatus: Open.\n#47161 \nIn vllm-project/vllm;\n· hickeyma opened \n[RFC]: Reduce multimodal payload in token-in/token-out by deferring pixel_values preprocessing to /generate\nRFC\nStatus: Open.\n#46722 \nIn vllm-project/vllm;\n· hyeongyun0916 opened \n14\ncomments\n[Bug] DeepSeekV4-Flash produces incorrect output with inline system messages after PR #46025 when preserved in-place\nbug\nSomething isn't working\nStatus: Open.\n#46710 \nIn vllm-project/vllm;\n· wqh17101 opened \n1\n16\ncomments\n[RFC]: Heterogeneous TP<->DP parallelism across the P/D disaggregation boundary (MoRIIO)\nrocm\nRelated to AMD ROCm\nStatus: Open.\n#46107 \nIn vllm-project/vllm;\n· edwinlim0919 opened \n4\ncomments\n[RFC]: Kernel-agnostic constant-stride (layer-major) KV cache layout for KV connectors\nStatus: Open.\n#45997 \nIn vllm-project/vllm;\n· xiaguan opened \n2\ncomments\n[Feature][ROCm] Support WideTP and WideEP P/D topologies on AMD GPUs\nrocm\nRelated to AMD ROCm\nStatus: Open.\n#45751 \nIn vllm-project/vllm;\n· chaeminlim-mb opened \n4\ncomments\n[Bug][ROCm][P/D] MoRI/MoRIIO READ-mode hangs and partial completions on AMD GPU P/D runs\nrocm\nRelated to AMD ROCm\nStatus: Open.\n#45749 \nIn vllm-project/vllm;\n· chaeminlim-mb opened \n7\ncomments\n[RFC]: [Roadmap] Mooncake Store Connector Feature Enrichment Roadmap\nRFC\nStatus: Open.\n#45036 \nIn vllm-project/vllm;\n· LCAIZJ opened \n3\ncomments\n[Feature]: Could you help with implement the KV cache pruning feature on vLLM?\nfeature request\nNew feature or request\nrocm\nRelated to AMD ROCm\nStatus: Open.\n#44300 \nIn vllm-project/vllm;\n· sankexin opened \n5\ncomments\n[Bug] MooncakeConnector: KV cache data corruption under concurrent PD transfers (batch_transfer_sync_write race)\nStatus: Open.\n#44238 \nIn vllm-project/vllm;\n· stmatengss opened \n[Bug]: Qwen3-VL-2B-Instruct Geo3K accuracy score lower than SGLang with deterministic sampling\nbug\nSomething isn't working\nStatus: Closed (completed).\n#43602 \nIn vllm-project/vllm;\n· by andakai was closed \n[Feature]: [RFC] Bottleneck-aware expansion for PD disaggregation\nfeature request\nNew feature or request\nStatus: Open.\n#43470 \nIn vllm-project/vllm;\n· zsn6493 opened \n2\ncomments\n[Bug]: ExampleConnector fails to inject KV cache on CPU\nbug\nSomething isn't working\nStatus: Closed (completed).\n#43059 \nIn vllm-project/vllm;\n· by revit13 was closed \n1\n1\ncomment\nPrevious1234Next\nFooter\n© 2026 GitHub, Inc.\nFooter navigation\nTerms\nPrivacy\nSecurity\nStatus\nCommunity\nDocs\nContact\nManage cookies\nDo not share my personal information"