{"slug": "vllm-merges-embedding-loader-fix-after-4-2-mb-test-allocated-3-gib", "title": "vLLM merges embedding loader fix after 4.2 MB test allocated 3 GiB", "summary": "A vLLM pull request merged into main on November 16th, 2026 moves the archive size check ahead of torch.load after contributor Leizhen Zhang showed a 4,175,968-byte base64 embedding payload allocated 3,072 MiB in 'safe_load_prompt_embeds' before the shape check rejected it. A second variant using a one-row view passed the 2 GiB check while retaining the full 3 GiB storage. The fix sums the declared uncompressed size inside the archive and rejects oversized payloads before loading begins.", "body_md": "# vLLM merges embedding loader fix after 4.2 MB test allocated 3 GiB\n\n**In a vLLM contributor's loader test, a 4,175,968-byte reproducer allocated 3,072 MiB before the shape check rejected it, and a one-row view passed with 3 GiB of storage attached. The fix merged into main on November 16th, 2026 checks archive size before torch.load.**\n\n        By [RuntimeWire Staff](https://runtimewire.com/author/runtimewire-staff)\n        · Published \n\nContributor Leizhen Zhang tested vLLM, an open-source inference and serving engine, with a 4,175,968-byte base64 embedding payload that allocated 3,072 MiB in 'safe_load_prompt_embeds' before the shape check rejected it. A second version using a one-row view passed the 2 GiB check while keeping the full 3 GiB storage attached. The merged pull request moves the check to the front of the loading path: the fix on vLLM's main branch sums the declared uncompressed size inside the archive and rejects oversized payloads before 'torch.load' runs.…", "url": "https://wpnews.pro/news/vllm-merges-embedding-loader-fix-after-4-2-mb-test-allocated-3-gib", "canonical_source": "https://runtimewire.com/article/vllm-embedding-payload-3gib-preload-check", "published_at": "2026-11-16 09:54:33+00:00", "updated_at": "2026-11-16 10:21:59.212106+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-infrastructure", "large-language-models", "ai-safety", "mlops"], "entities": ["vLLM", "Leizhen Zhang", "torch.load", "safe_load_prompt_embeds"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/vllm-merges-embedding-loader-fix-after-4-2-mb-test-allocated-3-gib", "markdown": "https://wpnews.pro/news/vllm-merges-embedding-loader-fix-after-4-2-mb-test-allocated-3-gib.md", "text": "https://wpnews.pro/news/vllm-merges-embedding-loader-fix-after-4-2-mb-test-allocated-3-gib.txt", "jsonld": "https://wpnews.pro/news/vllm-merges-embedding-loader-fix-after-4-2-mb-test-allocated-3-gib.jsonld"}}