MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 21:40:08 +07:00
61fbe0b865 [Mini-VLLM] Remove unreferenced definitions found by static scan
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 21:06:31 +07:00
5aa2ac64f8 [Mini-VLLM] Remove LoRA, plugins, speech-to-text and scale-out surfaces
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 19:31:15 +07:00
5f3c7bd8ab [Mini-VLLM] Remove the ROCm device path, debug routes and arch-gated tests
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 17:11:33 +07:00
2066d3143f [Mini-VLLM] Cut non-OpenAI surfaces, dead kernels, Mistral stack
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 09:27:27 +07:00
bd70a61c20 [Mini-VLLM] Remove orphan subsystems: snapshot, CPU workers, dead decoders
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 02:36:11 +07:00
2b544b7ba0 [Mini-VLLM] Cut dead subsystems across serving, connectors, parsers
MOVZX pushed to patches at iHore/vLLM 2026-10-19 01:09:17 +07:00
49d708e158 Merge branch 'main' into patches
ff01a6b14f [Bugfix][Frontend] Stop beam search on generation config EOS tokens (#62072)
fe8f2140ee [Doc] Document /inference/v1/generate request/response fields and DP rank header (#62305)
e80b6c642f [ROCm][DSv4] Overlap the unfused shared expert with AITER MegaMoE (#61819)
371745ebd6 [Doc][RL] Document routed experts return for MoE routing replay (R3) (#62302)
Compare 12 commits »
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 01:06:42 +07:00
54a396586e [Mini-VLLM] Repair test suite after SM86 cut
MOVZX pushed to main at iHore/vLLM 2026-10-19 00:53:11 +07:00
ff01a6b14f [Bugfix][Frontend] Stop beam search on generation config EOS tokens (#62072)
fe8f2140ee [Doc] Document /inference/v1/generate request/response fields and DP rank header (#62305)
e80b6c642f [ROCm][DSv4] Overlap the unfused shared expert with AITER MegaMoE (#61819)
371745ebd6 [Doc][RL] Document routed experts return for MoE routing replay (R3) (#62302)
072c0886d5 [Build] Keep plain CUDA arch when its suffixed variant is also requested (#56549)
Compare 525 commits »
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 00:43:53 +07:00
0152af1336 [Mini-VLLM] Remove upstream CI, docs, and lint surfaces
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 00:40:03 +07:00
e3af324d8a [Mini-VLLM] Drop upstream model recipes directory
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 00:33:26 +07:00
8dfaffb700 [Bugfix] Restore xpu_communicator for CUDA graph capture
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 00:20:57 +07:00
81074e94a3 [Mini-vLLM] Cut dead subsystems: models, entrypoints, backends, utils, tests
MOVZX pushed to sm86 at iHore/vLLM 2026-10-19 00:19:38 +07:00
6a95f62e2f [Bugfix] Load checkpoint KV-cache scales in the Qwen3.5 family
MOVZX pushed to patches at iHore/vLLM 2026-10-19 00:19:20 +07:00
6cf86cd95f [Bugfix] Load checkpoint KV-cache scales in the Qwen3.5 family
MOVZX pushed to sm86 at iHore/vLLM 2026-10-18 23:31:36 +07:00
f36850e389 [Mini-VLLM] Restore FlashInfer backend for FP8 KV & only probe quantization overrides that exist
MOVZX pushed to sm86 at iHore/vLLM 2026-10-18 23:19:39 +07:00
7470a93875 [Mini-VLLM] Restore FlashInfer backend for FP8 KV on SM86
MOVZX pushed to sm86 at iHore/vLLM 2026-10-18 23:17:54 +07:00
c66baf3199 [BugFix] Only probe quantization overrides that exist in the tree
MOVZX pushed to sm86 at iHore/vLLM 2026-10-18 19:23:22 +07:00
e61dcb9143 [Mini-vLLM] SM86-only cut: models, backends, Rust, tests
MOVZX created branch sm86 in iHore/vLLM 2026-10-18 19:23:22 +07:00