mirror of
https://gitee.com/trusted-list/sglang.git
synced 2026-10-28 16:45:11 +08:00
Commit Graph
Select branches
Hide Pull Requests
F4_support
UX-NPU
VDN-h3
ac8452/1/0
add-command-a-plus-cookbook
add-reasoning-usage-test
add_qwen3_5_ci
add_tls
add_use_seperate_warmup_in_bench_serving
adding-pin-logic-docker
agentx-mamba-envelope-transfer-20260823
agentx-nixl-contiguous-run-xfer-20260823
agentx-nixl-fewer-run-descriptors-20260823
agentx-nixl-token-budget-pipeline-20260823
agentx-post-launch-transfer-drain-20260823
agentx-upstream/gdn-compact-layer-loop-20260824
agentx/decode-assignment-ack-fresh-20260828
agentx/decode-ready-reserve-20260828
agentx/decode-router-hint-burst-20260828
agentx/verify-mha-metadata-main-20260828
alexnails/bazel-with-generate-deps
alexnails/ci-install-rust-release
alexnails/cpu-scheduler-perf-tier-ab
alexnails/cpu-scheduler-pipeline-overlap
alexnails/decoupling-scheduler-decode-fix
alexnails/dp-attn-perf-fix
alexnails/gemma-hicache-1
alexnails/gemma4-mtp-chain-selection
alexnails/gemma4-text-only
alexnails/gemma4-xpu-frozen-kv-mtp
alexnails/gemma4-xpu-greedy-logits
alexnails/gemma4-xpu-greedy-lut
alexnails/gemma4-xpu-intermediate-prefill-gemm
alexnails/gemma4-xpu-ple-add-scale
alexnails/gemma4-xpu-prefill-gemm
alexnails/gemma4-xpu-prefill-gemm-narrow
alexnails/gemma4-xpu-prefill-lut-standalone
alexnails/gemma4-xpu-qkv-rope-fma
alexnails/grpc
alexnails/grpc-phase1-fixes
alexnails/grpc-phase1/openai-stream-hook
alexnails/grpc-phase1/tests
alexnails/host-extend-bookkeeping
alexnails/inplatform-dispatch-key
alexnails/inplatform-graph-factories
alexnails/inplatform-quant-config
alexnails/kernels-1
alexnails/lint-ban-pickle
alexnails/macos-local-dev
alexnails/mlx-hybrid-batching
alexnails/model-cache-sync-migration
alexnails/moe-partial-block-optin
alexnails/platform-device-ops
alexnails/port-scheduler-radix-perf
alexnails/proto-msgpack-codec
alexnails/proto-msgpack-codegen-fixes
alexnails/rust-vlm-tile-preproc
alexnails/small-model-scheduler-perf
alexnails/tokenizer-2
alexnails/tokenizer-3
alexnails/xgrammar-parallel-fill
alexnails/xpu-bf16-gelu-mul-lut
alexnails/xpu-kernel-capabilities
amd-ci/mi355x-split-matrix-by-node-footprint
amd-diffusion-aiter-fix
amd-mi355x-disagg-nightly
amd/add-1gpu-small-recent-nv
amd/add-glm51-mxfp4-nightly-clean
amd/add-mimo-v25-pro-nightly-tests
amd/aiter-ci
amd/ci-gpu-hang-probe
amd/ci-vram-leak-selfheal
amd/deepseek_v4
amd/deepseek_v4_0430
amd/deepseek_v4_0520
amd/deepseek_v4_migrate-main_0508
amd/deepseek_v4_rebase_main_0509
amd/dpsk_v4
amd/dsv4-rebase
amd/expand-1gpu-cov-jul13
amd/expand-1gpu-cov-jul15
amd/expand-1gpu-pr-coverage
amd/fix-flashinfer-benchmarks
amd/gfx1250-fp8-swiglu-fusion
amd/glm53-flash-mi30x-nightly-accuracy
amd/kimik26-bisect
amd/kimik26-knowngood
amd/mi455x-gfx1250-gpt-oss-ci
amd/mm-gen-nested-b1-harness
amd/mm-gen-nested-b2-envskip
amd/mm-gen-unit-test
amd/nightly-glm53
amd/probe-kernel-doable
amd/register-1gpu-tier1-batch
amd/register-dsa-meta-mamba-views
amd/register-jit-benchmarks
amd/register-jit-benchmarks-batch2
amd/register-jit-benchmarks-batch3
amd/register-jit-kernel-amd
amd/register-triton-data-movement-tests
amd/rocm720-pin-triton-wheel
amd/scout-aiter-kernel-warmup
amd/test-5n-mi355x-di
amd/test-pr31368
amd/test-pr31727
amd/verify-gap-doable
amd/verify-mi35x-4gpu
amd_benchmark
amd_diffusion
amd_fix_aiter_scaled_downcast
amd_fix_dpsk_0511
amd_fix_request_timeout
amd_fix_stage_c_0715
amd_glm5.3_flash
amd_m25_long
amd_m3_moe_0723
amd_march
amd_mori
amd_pr_25266
amd_rocm10
amd_rocm10_gfx1250_triton_orginal
amd_test_26662
amd_test_29158_0625
amd_test_29181
amd_test_mi355_0710
amd_test_pr28932
amd_test_pr29677
amd_test_pr32754
amd_test_pr_21655
amd_upgrade_pr_test
ascend_moe_optimizations
auto-detect-environment
auto/lmsys-blog-cards
b12x-align-b4d6c75
b12x-chat-template-fix
b12x-ci-overlay-tweaks
b12x-vision
backup-pr26517
backup/cheng/test/attention-runner-combination
backup/mmgen-official-gt-alignment-20260425
backup/pr-30478-pre-rebase
backup/reland-pr26134
backup/sglang-miles-muse-spark-open-20261019
bailing-v3-day0
batch-7-amd-ci-audit
bbuf/hunyuan-v3-hpcops-router-gemm
beam
beam-miles
beam-sm103-moe-gemv
beam-stack-03-srt
beam-stack-06-cutile-router
beam-stack-deps-dynamo
beam-stack-deps-tileiras
bench-openhands-b300-sweep
bench-openhands-gb300-sweep-beam
bench/e4-main-combined
bench/e4-stack-combined
bench/e5-main-combined
bench/e6-main-combined
bhe/grok2-piecewise
bhe/radix_linear_attn
bingxche/amd-reusable-stage
bingxche/ci-gfx1250
bingxche/fix-tbo-forward-metadata-proxy
bingxche/image2lmsysorg
bked-default
bot/bump-docs-version-0.5.17-31225095238-1
bot/bump-docs-version-0.5.19-33912440427-1
bot/bump-docs-version-0.5.20-35322363581-1
bot/bump-docs-version-0.5.22-38824883155-1
bot/bump-flashinfer-version-0.6.8-6f8d
bot/bump-kernel-version-0.3.21.post1-5f39
bot/bump-kernel-version-to-sglang-0.4.2-3936
bot/bump-kernel-version-to-sglang-0.4.2.post1-3eff
bot/bump-kernel-version-to-sglang-0.4.2.post2-c64d
bot/bump-kernel-version-to-sglang-0.4.6-3c47
bot/lmsys-blog-cards
bot/update-est-time-20260418
bot/update-est-time-20260420
bot/update-est-time-20260504
bot/update-est-time-20260511
bot/update-est-time-20260514
bot/update-est-time-20260601
bot/update-est-time-20260608
bot/update-est-time-20260615
bot/update-est-time-20260622
bot/update-est-time-20260629
bot/update-est-time-20260706
bot/update-est-time-20260713
bot/update-est-time-20260720
bot/update-est-time-20260727
bot/update-est-time-20260803
bot/update-est-time-20260810
bot/update-est-time-20260817
bot/update-est-time-20260824
bot/update-est-time-20260831
bot/update-est-time-20260914
bot/update-est-time-20260921
bot/update-est-time-20260928
bot/update-est-time-20261005
bot/update-est-time-20261012
bot/update-est-time-20261019
bot/weekly-update-est-time
brayden/cutedsl-fused-parallel-qknorm
brayden/deepgemm-torch-2.13
brayden/fix-cutedsl-mqa
brayden/sm120-stash
brayden/splitk-tgv-bf16
brayden/support-nemotron-eplb
build-commit-sgl-kernel
byron/cherrypick-23734-onto-miles
byron/compile-deepep-timeout-fix
byron/refactor-pcg-mm-2
byron/um-pd-layout
byron/um-pd-matching-dcp
byron/um-pd-unequal-tp
cctry-add-load-counters
ch-wan-patch-1
ch-wan-patch-2
ch-wan-patch-3
chang/tokenizer
cheng/ci/in-process-group-v2
cheng/config-namespace-reland
cheng/debug-mode-batch-scoped-fault
cheng/debug/dspark-low-acceptance
cheng/debug/inkling-fa4-validation
cheng/dev/attn-dp-size-compat
cheng/dev/dp-attention-prefill-cp
cheng/dev/pplx
cheng/dev/unified-fused-drafts-base
cheng/doc/hicache-storage-config-todo
cheng/ffn-checkpoint-compat
cheng/fix-one-batch-chunked-prefill
cheng/gc-unified
cheng/gpt-oss-sinks-dtype
cheng/hot-fix/atomic-request-state-init
cheng/hot-fix/cutlass-moe-checked-reduction
cheng/hot-fix/cutlass-moe-complete-maps
cheng/hot-fix/decoder-tail-global-counter
cheng/hot-fix/dsv4-silu-kernels-empty-rows
cheng/hot-fix/ep-dispatcher-follow-layer-runner-base
cheng/hot-fix/fp8-moe-published-routing-policy
cheng/hot-fix/gguf-inactive-route-scratch
cheng/hot-fix/hunyuan-sink-weight-layout
cheng/hot-fix/lora-alignment-input-domain
cheng/hot-fix/marlin-inactive-route-consumers
cheng/hot-fix/moe-lora-gated-virtual-projections
cheng/hot-fix/moe-temporary-workspace
cheng/hot-fix/native-fp8-reloads
cheng/hot-fix/sarvam-dense-mlp-reduction
cheng/hot-fix/stack-ending-attention-sliced-rows
cheng/hot-fix/unified-joint-reclaim
cheng/hot-fix/unified-kv-hip-fused-kv-write
cheng/hot-fix/unified-kv-int8-ckpt-fail-loud
cheng/hot-fix/unified-kv-tri-conv-slice-axis
cheng/hot-fix/unified-kv-tri-host-stack
cheng/hot-fix/unified-kv-tri-state-hicache-events
cheng/hot-fix/unified-mamba-compaction
cheng/kv-loc-debt
cheng/refactor/aiter-moe-runner
cheng/refactor/attention-input-from-boundary
cheng/refactor/attention-projection-parallel-groups
cheng/refactor/attention-sink-loader-layouts
cheng/refactor/attn-01-rename-static-buffers
cheng/refactor/attn-02-eager-init-buffers
cheng/refactor/attn-03-compute-metadata-extraction
cheng/refactor/attn-04-cpu-route-eager
cheng/refactor/attn-05-deprecate-use-bound
cheng/refactor/attn-finish-plan
cheng/refactor/attn-step05-residual
cheng/refactor/attn-step08
cheng/refactor/attn-step10
cheng/refactor/auxiliary-loader-frozen-layouts
cheng/refactor/boundary-per-edge
cheng/refactor/cache-chain-reads-own-placement
cheng/refactor/cp-configs-declared
cheng/refactor/cp-moves-binding
cheng/refactor/custom-projection-parallel-groups
cheng/refactor/cutlass-moe-route-preparation
cheng/refactor/dbrx-expert-loader-layout
cheng/refactor/declared-a2a-boundary
cheng/refactor/declared-dense-fully-dp
cheng/refactor/declared-dsa-cp
cheng/refactor/declared-input-scattered
cheng/refactor/declared-mhc
cheng/refactor/declared-mhc-input-scattered
cheng/refactor/declared-moe-boundary
cheng/refactor/declared-plain-tp-boundary
cheng/refactor/declared-row-queries
cheng/refactor/declared-sp-region
cheng/refactor/decouple-forward-mode
cheng/refactor/deepep-moe
cheng/refactor/deepseek-moe-output
cheng/refactor/deepseek-v4-lagged-layer-loop
cheng/refactor/deepseek-v4-layer-loop
cheng/refactor/deepseek-v4-moe-transport
cheng/refactor/deepseek-v4-stage-boundaries
cheng/refactor/dots-fused-gate-loader-layouts
cheng/refactor/dp-pad
cheng/refactor/dp-return-selection
cheng/refactor/dp-tbo-gather
cheng/refactor/dp-tp-group-helper
cheng/refactor/draft-scope-topology
cheng/refactor/drop-dead-context-helpers
cheng/refactor/encoder-keeps-its-attention-layout
cheng/refactor/ffn-exit-fusion-entries
cheng/refactor/ffn-input-completes-owed
cheng/refactor/final-norm-skip-empty
cheng/refactor/fope-coefficient-loader-layouts
cheng/refactor/forward-metadata-staleness
cheng/refactor/fp8-block-layout-defaults
cheng/refactor/gemma-shared-vocab-layouts
cheng/refactor/gguf-moe-vector-kernels
cheng/refactor/gpt-oss-expert-loader-layouts
cheng/refactor/gptq-act-order-layout-diagnostics
cheng/refactor/group-build-drops-model-config
cheng/refactor/inkling-linear-loader-layout
cheng/refactor/inkling-moe-loader-layout
cheng/refactor/inkling-stage-boundaries
cheng/refactor/is-draft-extend-v2-rename
cheng/refactor/kimi-attention-parallel-groups
cheng/refactor/kimi-mlp-parallel-groups
cheng/refactor/kv-calibration-loader-layouts
cheng/refactor/layer-boundary-completion
cheng/refactor/layer-stack-branch-connections
cheng/refactor/layer-stack-resolved-connections
cheng/refactor/linear-attention-parallel-groups
cheng/refactor/linear-constructor-parallel-groups
cheng/refactor/linear-owned-groups
cheng/refactor/llama-inkling-mlp-parallel-groups
cheng/refactor/lora-routing-prewarm
cheng/refactor/marlin-lora-launch-paths
cheng/refactor/mhc-by-kind
cheng/refactor/mhc-residual-ops
cheng/refactor/mimo-fused-qkv-loader-layouts
cheng/refactor/model-kv-calibration-layouts
cheng/refactor/moe-cleanup
cheng/refactor/moe-cp-gathered-rows
cheng/refactor/moe-placement-in-runner-config
cheng/refactor/move-expert-mask-to-dispatcher
cheng/refactor/nemotron-h-stage-boundaries
cheng/refactor/parallel-guard-coverage
cheng/refactor/quant-weight-layout-validation
cheng/refactor/quantized-reload-scale-layouts
cheng/refactor/qwen-vision-wrapper-parallel-groups
cheng/refactor/reject-attn-tp-mismatch-layers
cheng/refactor/remaining-vision-parallel-groups
cheng/refactor/repeated-batched-linear-parallel-groups
cheng/refactor/replicated-attention-linear-groups
cheng/refactor/req-len
cheng/refactor/retire-the-group-aliases
cheng/refactor/scheduler-reads-own-placement
cheng/refactor/sharded-loader-parallel-groups
cheng/refactor/stage-communicators
cheng/refactor/stage-reads-updates
cheng/refactor/unified-kv-addressing
cheng/refactor/unified-kv-arena-and-ids
cheng/refactor/unified-kv-chain-allocator
cheng/refactor/unified-kv-declare-plan
cheng/refactor/unified-kv-draft-views
cheng/refactor/unified-kv-drop-facades
cheng/refactor/unified-kv-layout-move
cheng/refactor/unified-kv-pins
cheng/refactor/unified-kv-pricing-and-budget
cheng/refactor/unified-kv-requests
cheng/refactor/unified-kv-router
cheng/refactor/unified-kv-safety-net
cheng/refactor/unified-kv-state-access
cheng/refactor/unified-kv-storage-consumers
cheng/refactor/unified-kv-transfer
cheng/refactor/unified-kv-tree-components
cheng/refactor/unified-kv-typed-views
cheng/refactor/unify-attn-backend-cuda-graph-capture-replay
cheng/refactor/vision-parallel-groups
cheng/refactor/zaya-cca-parallel-group
cheng/scheduler-debug-mode
cherry-pick/1fb85053e7-to-v0.5.19-21323390
cherry-pick/378aea1385-to-v0.5.16-f64b7684
cherry-pick/79ea30d1f-to-v0.5.12-4fac644b
cherry-pick/c7e2c08d14-to-v0.5.18-650e8374
ci-check-moe-topk-sigmoid
ci-light-import
ci-maintain
ci-monitor/bump-est-time-20260506-233123
ci-nvidia-smi-timeout
ci-probe-rate-limit
ci-uv-venv-migration-fixes
ci/add-5090-test-stage
ci/add-hybrid-dp-ep-tp-mtp-to-nightly
ci/amd-miles-mi30x-nightly
ci/amd-stage-b-shard-14to8
ci/auto-retry-scheduled-failures
ci/deepep-egg-purge
ci/detect-consolidated-pt-format-mismatch
ci/diffusion-b200-pip-fa4
ci/disable-gb200-stage
ci/dsv4-pd-disagg-aime25-b200
ci/dsv4-pd-disagg-nixl-b200
ci/e2e-memory-capacity-thresholds
ci/eagle-dp-attn-nan-test
ci/enable-deepgemm-warmup-8gpu-h200
ci/enable-weka-docker-cache-seeding
ci/failure-tracking-infra
ci/fix-blackwell-pip-race
ci/fix_gfx1250_attn
ci/flaky-test-reporting
ci/gsm8k-answer-extraction
ci/gt-gate-skip-non-images
ci/mi355x-spur-heredoc-backticks
ci/migrate-5090-tests-to-registered
ci/migrate-quant-tests
ci/nightly-1-gpu-5090-job
ci/nightly-traces-s3
ci/nvidia-lib-selfheal
ci/pr-30762-umbp-e2e
ci/precache-hf-models
ci/precision-test-dispatch
ci/shm-teardown-purge
ci/verify-mi355x-disagg-43944-43945-43977-43985
ci/weight-checksum-validation
claude/drop-qwen-new-tag-1a6681
claude/git-bisect-debugging-cmsljv
claude/lint-hook-python39
claude/paddleocr-support-optimization-bfaf52
claude/vlm-warmup-image-size
cleanup-5090-gpu-memory
cleanup-cutlass-fp8-supported
clearing_console_output
codex-group-norm-silu-artifacts
codex/amd-nixl-ucx-autoselect-exp-20260916
codex/amd-rocm720-nightly-mi300-1gpu
codex/archive-miles-admission-source-20260905
codex/backfill-image-formats
codex/bcg-output-management
codex/cfg-vae-decode-group
codex/checkpoint-quant-metadata-no-reflection
codex/clarify-gpu-residency-log
codex/communicator-cp-ownership
codex/component-quantized-repos
codex/component-residency-strategy-compat
codex/cosmos3-lossless-perf-20260602
codex/cosmos3-video-gt-test-20260602
codex/cpu-offload-components-clean
codex/diagnose-wan-sdpa-drift
codex/diffusion-5090-perf-pool
codex/diffusion-cfg-text-batch-20260523
codex/diffusion-ci-gt-sync-files
codex/diffusion-day0-variants
codex/diffusion-platform-perf-baselines
codex/diffusion-preload-residency-planner
codex/diffusion-probe-fixes-20260621
codex/diffusion-profile-stage-names-20260523
codex/diffusion-stage-deployment
codex/direct-gpu-zero-copy
codex/disable-mi355x-di-ci
codex/dit-static-weight-mappings
codex/draft/glm5-3-datafree-fp4fp8
codex/draft/glm5-3-datafree-runtime
codex/dsv4-aime25-hard-ci
codex/dsv41-dpa-megamoe
codex/dsv41-indexer-plan
codex/engine-sampling-defaults-chat
codex/epd-grpc-registration-failure
codex/epd-health-probe-drain
codex/epd-preflight-request-validation
codex/epd-receiver-startup-consensus
codex/epd-receiver-timeout-cleanup
codex/epd-scheduler-cancel-cleanup
codex/explicit-native-only-components
codex/fix-ascii-decode-prompt
codex/fix-cu134-deepgemm-elfutils
codex/fix-deepgemm-batch-invariance
codex/fix-dense-prefill-cp
codex/fix-diffusion-final-failure-summary
codex/fix-diffusion-gt-generation
codex/fix-fa3-varlen-out
codex/fix-glm53-dsv4-pool-test
codex/fix-glm53-eagle-kpool-padding
codex/fix-hybrid-kv-release-order
codex/fix-hybrid-linear-eagle-padding
codex/fix-kda-cutedsl-rocm
codex/fix-pd-abort-terminal-semantics
codex/fix-stale-cargo-workspace-cache
codex/flashmla-mega-build-apis
codex/flux2-import-ci-main-1023
codex/forbid-distributed-transformer-fallback-v2
codex/generic-prequant-checkpoint-admission
codex/glm4-lite-shared-expert-tp-flag
codex/glm47flash-fa3-out-guard
codex/glm47flash-topk-shared-jit
codex/glm5-3-flash-h20-pd-mtp
codex/glm5-3-flash-megamoe
codex/glm5-3-flash-performance-optimization
codex/glm5-3-flash-sota-20260905
codex/glm5-3-w4a8-community
codex/glm5.3-ep_main
codex/glm53-cp-decode-attn-tp
codex/glm53-flash-dcp-base
codex/glm53-flash-dcp-mtp
codex/glm53-flash-main-baseline-20260907
codex/glm53-flash-optimize-20260908
codex/glm53-ingraph-verify-metadata
codex/glm53-kpool-deterministic
codex/glm53-lora-dense-tp
codex/glm53-verify-deepgemm-out-of-graph
codex/gptoss-bf16-flashinfer-support
codex/h3-tp-boundary-replication
codex/h3-tp-ulysses-ipc
codex/hunyuan3d-native-image-encoders
codex/ideogram4-nvfp4-ckpt
codex/ideogram4-pr27279-gt-e8dfe898ad
codex/image-benchmark-seeded-rng
codex/ipc-pool-handle-cache-default
codex/kimi-grid-metadata-cpu
codex/kimi-k27-dp-encoder
codex/kimi-k27-vlm-perf-integration
codex/kimi-moonvit-fused-rope
codex/klein-first-request-diagnosis
codex/lingbot-camera-mod-cache
codex/lingbot-full-runtime-session-refactor
codex/ltx-hq-ci-gt-refresh
codex/ltx-official-gt-workflow-20260506
codex/ltx2-from-862f-clean-main
codex/ltx23-lossless-perturb-mask-20260505
codex/ltx23-origin-main-20pct-20260505
codex/main-diffusion-test-suite
codex/mamba-track-interval-1024-20260523
codex/minimax-h3-adaln-cache
codex/minimax-h3-cross-node-ring
codex/mistral3-native-attention-20260614
codex/mlx-shutdown-ingestion-fix
codex/mm-broadcast-error-envelope
codex/mm-consumer-reconstruct-rollback
codex/mm-owner-projector
codex/mmgen-cfg-text-batch-20260521
codex/modelopt-fp8-platform-dtype-test
codex/mopd-per-position-scoring
codex/muse-cookbook-unified-memory
codex/muse-spark-launch-defaults
codex/muse-spark-stack-04-amd
codex/muse-spark-stack-05-helion-verify
codex/nccl-tms-fix
codex/neuron-overlap-sequence-relay-20260905
codex/nightly-diffusion-token-backfill-20260918
codex/nightly-flux3-20261016
codex/nixl-storage-test-isolation-1024
codex/omni-u1-orchestrator
codex/pi05-native-siglip
codex/port-pr-24082-specv2-disagg-grammar
codex/post-response-tasks-20260503
codex/pr-21247-thresholds-20260502
codex/pr-37495-validation-20260902
codex/pr-title-model-type
codex/pr/37588-kpool-request-domains
codex/pr/37588-whole-group-refresh
codex/pr28050-artifacts
codex/pr2874-nan-probe
codex/pr37587-rebased-20260907
codex/pr37587-validation-20260907
codex/pr38632-workflow-reference-cleanup
codex/pr43182-bcg-move-proof
codex/pr43774-cp-extraction-proof
codex/pr44720-amd-nightly-gfx1151-2c2986b
codex/pr44720-amd-nightly-main-7c99ace
codex/pr44720-amd-nightly-miles-7b6993a
codex/pr44720-amd-nightly-release-miles-7c99ace
codex/pr44720-amd-nightly-validation-20261016
codex/pr45498-amd-build-only
codex/qwen-qsa-prefill-metadata
codex/qwen-qsa-prefill-ragged
codex/qwen35-profile-8k1k-20260823
codex/qwen35-profile-exact-bs32-20260823
codex/rebase-sglang-27421
codex/reject-diffusers-component-attention-v2
codex/reject-unmanaged-component-residency
codex/rocm-network-rerun
codex/rocm10-copy-weight-views-slow-jobs
codex/rocm10-moe-copy-ab
codex/rocm10-rc4-aiter-a6d2b564
codex/rocm10-release-once-cc5f490b
codex/rocm10-rocm720-daily-ci
codex/rocm1000-ga-aiter-a6d2b564
codex/rocm7-tests-every-two-days
codex/router-dynamo-rendering
codex/router-engine-abort
codex/router-hf-hub-1
codex/router-pd-failfast-groups
codex/router-power-of-n
codex/router-scoring
codex/rust-frontend-compatibility-20260915
codex/sana-wm-consistency-realtime-common
codex/sana-wm-native-refiner
codex/sglang-radix-sim-core
codex/sgld-bench-case-fixes
codex/skip-mi355x-4n-unavailable-nodes
codex/sm120-diffusion-attn
codex/split-32714-gpt-oss-flashinfer
codex/srt-dynamic-cudnn-sdpa
codex/strict-component-attention-overrides
codex/swa-memory-budget-api
codex/test-mi35x-runners
codex/test-rocm720-mi300-pull-only
codex/tighten-nv-diffusion-perf-baselines
codex/tmp-glm53-layer-split-red
codex/torch211-official-gt-gen-20260502
codex/trainium3-chunked-admission-20260905
codex/ug-bagel-interleave-runtime
codex/ug-g-runtime-proof
codex/upstream-sglang-miles-latest-passed-main
codex/vae-channels-last-3d-followup
codex/vae-channels-last-true
codex/vae-tiling-and-nvfp4-gate
codex/validation/37588-kpool-request-domains
codex/vlm-preprocessed-01-core
codex/vlm-preprocessed-03-tests
codex/vlm-preprocessed-06-qwen-copies
codex/vlm-preprocessed-07-qwen-offsets
codex/vlm-preprocessed-10-pretokenized
codex/vlm-preprocessed-11-input-list
codex/vlm-preprocessed-12-shm
codex/vlm-preprocessed-13-grpc-dispatch
codex/vlm-preprocessed-input
codex/vlm-tc-prefill-graph-replay
codex/vmm-request-error-consensus
codex/wan22-bisect-9d8426870
combine/router-admission-http2-loadaware
config-lifecycle-tightening
config/resolution-policy-stages
consolidate-ci-weight-validation
cookbook-migrate-gemma4
cookbook-migrate-glm-5.1
cookbook-migrate-kimi-k2.6
cookbook-migrate-minimax-m2.7
cookbook-migrate-nemotron3-ultra
cookbook-migrate-qwen3.5
cookbook-mimo-v2.6-b300-minimal
cookbook-skill-wave1-review-fixes
cookbook/dsv4-pro-0813-dspark
copilot/add-decode-kv-cache-offload
copilot/add-epd-disaggregation-architecture
copilot/add-new-feature
copilot/add-pipeline-parallelism-sglang
copilot/add-python-decorator-introduction
copilot/add-sglang-diffusion-support
copilot/add-specforge-training-framework
copilot/check-stars-for-issue-21846
copilot/chunked-prefill-mask-attention
copilot/concurrent-programming-interview-questions
copilot/create-presentation-script
copilot/derive-softmax-gradient
copilot/discussion-on-chunk-prefill
copilot/enable-free-thread-support
copilot/enable-speculative-decoding-small-batchsize
copilot/ep-walk-tp-communication
copilot/explain-chunk-prefill-process
copilot/explain-code-functionality
copilot/explain-conn-py-code
copilot/explain-conn-py-details
copilot/explain-decode-schedule-batch-mixin
copilot/explain-inference-parameters
copilot/explain-kv-events-code
copilot/explain-prefill-py-code
copilot/explain-sglang-chunk-prefill
copilot/fix-base-a-test-cpu-failure
copilot/fix-build-and-test-job
copilot/fix-ci-errors
copilot/fix-gloo-peer-rank-connection-issues
copilot/fix-lint-ci-error
copilot/fix-lint-job-failure
copilot/fix-lint-job-failure-again
copilot/generate-ppt-script
copilot/how-role-system-and-user-combine
copilot/implement-reduce-topk-gemm-permute
copilot/install-cu13-via-pypi
copilot/load-model-process
copilot/mem-cache-block-verification
copilot/mem-fraction-static
copilot/quantization-discussion-weight
copilot/research-model-loading-sglang
copilot/review-pr-21661
copilot/set-torch-cuda-arch-list-9-0
copilot/sglang-pd-separation
copilot/sub-pr-12948
copilot/support-kimi25-epd-encoder-only
copilot/track-flaky-ci-tests
copilot/understanding-flow-matching-models
copilot/update-make-layers-function
cublas_qiaolin
cuda-graph-lifetime-fixes
cuda-graph-lifetime-tool
cuda13_memcpy_hicache_restore
cursor/amd-agentic-perf-glm52-fp4-mtp-7284
cursor/amd-agentic-perf-kimi-k3-mxfp4-42c0
cursor/amd-agentic-perf-qwen35-mxfp4-mtp-8d29
cursor/amd-ci-cache-audit-tool-eb82
cursor/amd-ci-coverage-canvas-62ce
cursor/amd-ci-resolve-image-tag-with-ls-remote-868a
cursor/amd-ci-use-system-triton-f7b1
cursor/amd-diffusion-ipc-a2a-tests-cuda-only-8e90
cursor/amd-diffusion-mock-cuda-platform-not-rocm-8e90
cursor/amd-drop-tc-piecewise-split-op-tests-51db
cursor/amd-dsv4-compact-attention-inline-qkpv-8e90
cursor/amd-kandinsky6-fa-dispatch-skip-rocm-8e90
cursor/amd-kandinsky6-layerwise-offload-deterministic-8e90
cursor/amd-kda-decode-parity-cuda-only-8e90
cursor/amd-kda-fused-verify-qk-norm-reciprocal-8e90
cursor/amd-nightly-dsv4-flash-0731-drop-v3x-ef39
cursor/amd-quark-nvfp4-mxfp4-gsm8k-threshold-8e90
cursor/amd-rocm720-nightly-runner-fallback-6ad2
cursor/amd-umbp-hicache-standalone-ssd-rebase-8e90
cursor/bazel-migration-foundation-b440
cursor/cache-amd-ci-docker-image-f124
cursor/check-readme-file-composer-1-a466
cursor/ci-experiment-workflow-9019
cursor/clip-score-log-visibility-d45f
cursor/cookbook-dsv4-amd-fp8-experts-env-d201
cursor/dispatch-pr-25301-moriep-oom-864f
cursor/drop-miles-rocm7-image-40cc
cursor/encoder-dp-amd-debug
cursor/encoder-dp-amd-fix
cursor/experiment-mi325-parallel-1319-9491
cursor/fix-amd-ci-log-formatter-prefix-5d53
cursor/fix-amd-fused-moe-single-expert-reduce-4efe
cursor/fix-ci-failure-monitor-silent-401-fc6e
cursor/fix-ci-suite-runner-exit-hang-a882
cursor/fix-dsa-idle-spec-metadata-30d6
cursor/fix-fused-moe-unwritten-intermediate-a37e
cursor/fix-hellaswag-benchmark-source-bac9
cursor/fix-rocm-image-hipify-byproducts-b101
cursor/fix-vattn-asm-rocm10-hip-runtime-990a
cursor/flag-stale-miles-rocm-image-3d8b
cursor/gateway-crate-universe-64ce
cursor/glm52-mi355x-mxfp4-agentx-hicache-00a7
cursor/glm53-flash-mi300x-validation-748f
cursor/gpu-kv-cache-management-4941
cursor/kv-cache-management-891c
cursor/logprobs-assertion-failure-37d6
cursor/miles-nightly-suite-rename-test-44cb
cursor/move-test-moriep-small-to-mi35x-864f
cursor/never-prompt-for-git-credentials-155b
cursor/nv-downsampling-ratio-behavior-d5e9
cursor/pr-states-amd-rocm-ed27
cursor/print-k8s-node-name-4515
cursor/qwen35-moe-text-shared-expert-nightly-09ab
cursor/qwen38-flash-next-cookbook-fp8-cells-fe21
cursor/reclaim-and-seed-mi35x-hf-cache-7284
cursor/register-amd-ci-unit-tests-8f21
cursor/remove-deleted-amd-checkpoints-0857
cursor/restore-amd-ci-registrations-3896
cursor/retire-legacy-gsm8k-mmlu-amd-ci-6d7a
cursor/revert-42541-rope-fp32-cache-77ed
cursor/rocm-glm53-flash-hip-fallbacks-748f
cursor/rocm-mooncake-go-fetch-retry-27b9
cursor/rocm724-probe-ci-dispatch-dba6
cursor/setup-dev-environment-5af2
cursor/sglang-core-logic-3e29
cursor/sglang-pr-18938-review-6355
cursor/sparse-attention-kernel-tests-3e52
cursor/sparse-attention-kernel-tests-888b
cursor/test-optb-on-hardware-dba6
cursor/test-qwen35-mi355x-recipes-7487
cursor/update-rocm-nightly-ci-skill-cb60
cursor/verify-rocm720-async-ckpt-3d8b
curvezmq-shared-key
dbgA-kimi26-bf16kv
dbgB-kimi26-nospec
dcp-dsa
dcp-hicache-l3
dcp-l3-mooncake
dcw02/dflash-bounded-swa-kv-cache
dcw02/dflash-grammar-support
dcw02/dflash-penalty-fix
dcw02/draft-kv-unified-memory-pool
dcw02/fix-mamba-radix-page-align
dcw02/kimi-k3-dflash
dcw02/kimi-k3-dflash-optimization
dcw02/spec-mamba-logic
dcw02/war-barrier-worker-capability-perf
debug-qwen2vl-context-length-test
deepseek_v4
deepseek_v4_compressor
deepseek_v4_dev
deepseek_v4_mxfp4_shuffle_perf
demo/single-gpu-multi-node-serving
deprecate-get-server-info-endpoint
dev-19673
dev/double-sparsity-standalone
dev/pr33561-docker-nccl-fix
dev/rust-server
dev/saienduri/test-ci
device-pool-info-dsa
dflash2-pin-1cf2b8c
dflash2-pin-1cf2b8c-nccl
diag/nixl-run-choice-first10-20260823
diffuion-refactor-denoising
diffusion
diffusion-ci-cu13
diffusion-ci-with-qwen2509
diffusion-component-residency-manager
diffusion-first-class-routes
diffusion-fix-component-resident-manager
diffusion-gemma4-support
diffusion-general
diffusion-norm-fusion-for-zimage
diffusion-perf
diffusion-refactor-arch-config
diffusion-rl-base
diffusion-test
diffusion-tmp
docker/drop-gnuplot
docker/fix-nightly-version-stamp
docs/add-glm5-usage-guide
docs/beam-serving-cookbook
docs/dspark-guide
docs/dsv4-flash-0731-cookbook
docs/glm-5.3-flash-hopper-dsa-auto
docs/playground-spec-strip-draft-path
dsa-indexer-buffer-info
dsa-indexer-state-desc
dsa-static-page-table-width
dsv4-buffer-info
dsv4-dev
dsv4-mxfp4-moe-backend
dsv4-rebase
dsv4.1-backup-before-entrypoint-fix-20260915
dsv4.1-backup-before-linear-rebase-20260915
dsv4.1-backup-before-ordinary-update-20260915
dsv4.1-backup-before-test-replay-20260915
dsv4.1-engram
dsv41-engram-fusion
dsv41-flashmla-drop-sgl-kernel
dsv41-tp8-hopper-decode
dsv4_pp_doc
duyi-dev-260610
dynamic_tree_verify
emmett/ci-5090-base-b-light
emmett/ci-health-check-api-retries
emmett/ci-shared-host-runner-isolation
emmett/diffusion-image-preprocessing-unit-suite
emmett/remove-4-gpu-gb300-suite
enable-sm90-pool-decode
enable-uv-venv-deepgemm-cache
enable_cache_dit_in_test
eos
feat/amd-deterministic-kernel
feat/async_r3
feat/deepep-shared-expert-fusion
feat/dev_20251201
feat/diffusion-placement-group-disagg
feat/fused-weight-host-spill
feat/grpc-health
feat/ideogram4-fp8-native
feat/ling3-fp8-decode-kernels
feat/ling3-mamba-fork-split
feat/ling3-pub-parser
feat/ltx-2.5-parallel-diffusion-decode
feat/ltx-two-stage
feat/nvfp4-online-dense-linear
feat/planner-unified-memory
feat/readonly-checkpoint-mappings
feat/rerun-stage-fork-support
feat/rerun-stage-path-filter
feat/residency-apply-validate
feat/residency-planner-solver
feat/router-api-profile
feat/router-responses-api
feat/sgl-router-caller-input-ids
feat/sgl-router-chat-render-parity
feat/sgl-router-m0-scaffold
feat/sgl-router-render-parity
feat/shared-pool-planner
feat/sm120-dsa
feat/stateless_scheduler_b
feat/test-summary-job-level
feat/unified-metrics-s3
feat/warmup-prefill-tokens
feature/disable_generation_healthcheck
feature/eagle3
feature/unified-memory-flashinfer-decode
fix-19320-review
fix-amd-ci-install-quoting
fix-async-d2h-worker-context
fix-bench-server-large-bs
fix-bot-cherry-pick-push-pat
fix-corruputed-weights
fix-cp-nsa-prefill-context-parallel
fix-dcp-fa3-backend
fix-deepgemm-cache-key
fix-diffusion-ci-steps
fix-dllm-prefill-stats-none
fix-explicit-chat-template-override
fix-extra-buffer-nonmamba-validation
fix-flashinfer-fusion-preflight
fix-flattened-bucket
fix-fsdp-amd
fix-host-mha-fp4-guard
fix-jit-gb200
fix-kda-kernel-args
fix-llama-4-nightly
fix-lora-batch-validation-none-count
fix-lora-eviction-test-flakiness
fix-mcp-validation
fix-mistral-eagle-use-dsa
fix-native-mla-budget
fix-nvfp4-dispatch-dummy-weights
fix-nvfp4-sm120-oom
fix-qwen3-next-gdn
fix-run-with-timeout-exception-handling
fix-streaming-parallel-auto
fix-svd
fix-swa-index-mapping-leak
fix-unified-loadback-multipin-nightly0820
fix/add-job-timeout-pr-test
fix/amd-ci-perf-bounds-and-dispatcher-test
fix/amd-mmgen-unit-suite
fix/auto-format-main-branch
fix/bailing-moe-fp8-kv-b-proj-weight-transpose
fix/bench-mmmu-multi-image
fix/bench-serving-gsp-args
fix/ci-cleanup-stale-ipc
fix/ci-env-source-cleanup
fix/ci-port-offset-for-network-host
fix/ci-runner-stability-fixes
fix/ci-stale-sglang-cleanup-v2
fix/cudnn-import-order
fix/disable-jit-deepgemm-in-ci
fix/disable-transformers-test-amd
fix/disagg-bootstrap-port-conflict
fix/disagg-bootstrap-port-conflict-v2
fix/disagg-mori-abort-notification
fix/draft_weight_checker
fix/eagle-dp-attn-large-crash
fix/eagle-filter-merge-accept-length
fix/eagle3-mla-attn-backend
fix/empty-logprobs-streaming
fix/eplb-dp-attention-token-mask
fix/flashinfer-allreduce-fusion-probe
fix/flashinfer-autotune-symm-mem-deadlock
fix/glm51-fp8-tp8-dp8-b200-oom
fix/glm52-qk-rope-head-dim
fix/gpu-cleanup-between-tests
fix/grammar-backend-disabled-message
fix/gsp-from-args-missing-attr
fix/health-check-disagg-idle
fix/hf-runner-lora-adapter-cleanup
fix/hicache-3fs-batch-set-postprocess
fix/hicache-dsv4-none-crash
fix/lora-update-test-flaky
fix/megamoe-shared-fusion-quant
fix/megamoe-waterfill
fix/mem-fraction-static-eagle-draft-kv-cache-wip-snapshot-2026-05-08
fix/mi35x-deepseek-v32-regression
fix/mi35x-remove-data2-hardcoded-paths
fix/mini-lb-flush-cache-forward-timeout
fix/modelopt-auto-map-resolution
fix/modelopt-offline-quantized-detection
fix/mooncake-nsa-l3-index-buffer
fix/mtp_moe_capture
fix/multimem-ag-skip-node-probe-single-node
fix/nvila-siglip-vision-model-prefix
fix/parallel-state-uncalled-tests
fix/pause-generation-writer-lock-deadlock-main
fix/pdl-cuda-graph-capture
fix/port-alloc-collisions
fix/qwen-image-edit-warmup-prompt
fix/qwen35-mori-expert-routing
fix/qwen35-pp-mtp-prefill-validation
fix/remove-dead-prewarm-mhc-token-count-buckets
fix/request-log-watermark-gate
fix/rerun-failed-ci-cancelled-workflows
fix/residency-calibration-probes
fix/running-req-piling-logprob
fix/rust-server-hf-tokenizer-cache
fix/sync-patch-logprob-fix
fix/test-srt-backend-smaller-model
fix/tilelang-tvm-ffi-isinstance-compat
fix/tokenizer-recv-fairness
fix/torch-cuda-version-mismatch
fix/transformers-v5-easy-fixes
fix/unified-cache-storage-recovery
fix/verify-mask-full-bound
fix/wrapper-translator-test-fixtures
fix_benchmark
fix_diffusion_1gpu_ci
fix_for_cmb
fix_glm_test
fix_grouped_topk_warp_shfl
fix_hicache_with_spec
fix_mamba_leak
fix_nightly
fix_nightly_test
fix_npu_diffusion_failure_analyzer
fix_pd_warmup_route_dp
fix_qwen_mtp_eplb
fix_v1_loads
fix_wan_channels_last_3d
flashinfer-megamoe-autotune-decode-workspace
flashinfer-swa-ragged-window
flux-klein-qwen3-mask-threshold-rebase
fp8-blockwise-sm120-tile-dispatch
fused-qk-rmsnorm-rope-gate
fusion/attn_prep_norm_mrope_cache_fuse
fusion/fuse_qk_norm_rope_cache
fusion/fused_qk_norm_mrope_cache_attn_prep
fusion/moe_topk_gating_fused
fusion/qk_norm_rope_fuse
fzyzcjy-patch-1
fzyzcjy-patch-2
fzyzcjy-patch-3
fzyzcjy-patch-4
gemma4-triton-attn-tuning
github_workflow_yaml
glm
glm-5.3-flash-opt
glm5-gb-image
glm5.2-skip-layer-indexer-weight-check-main
glm53-reasoning-parser-auto-default
glm_interleave
grpc-phase1/python-bridge
grpc-rerank-serialization
gt-consistency-check
hanming/debug-buffer-kl
hanming/eagle-beta-sync
hanming/extra-buffer-opt
hanming/extra-buffer-opt-backup
hanming/fix-extra-buffer-debug
hanming/gdn-pc-lps-decode
hanming/hicache-tombstone-backup
hanming/out-cache-loc-swa
hanming/prefill-delayer-all-branch-timeout
hanming/sdv2-fix-output
hanming/spec-v2-kl
hanming/unified-hicache-opt
hccl_automatic_sizing
healthy_worker_count
hicache-decl-b
hicache-decl-d
hicache-fast-file-backend
hicache/backup-backlog-shedding
hicache/backup-drop-visibility
hicache/bounded-storage-calls
hicache/flush-wedged-storage
hicache/recoverable-storage-sync
hicache/timeout-policy-rank-safe
hisparse-fix
hisparse-prefetch-indexshare
hisparse-prefetch-indexshare-io-only
hisparse_quest
idhanani/cherry-pick-24234-v0.5.11
idhanani/disagg-decode-radix-cache
idhanani/dyn-1986-tier-aware-pin
idhanani/dyn-optional-deps
idhanani/dyn-swa-gsm8k-chat
idhanani/fix-26974-transfer-timing
idhanani/fix-26976-chunked-tier-split
idhanani/forward-pass-metrics
idhanani/g2plus-main-review-20260524
idhanani/kvcr-direct-linker
idhanani/llm-83-sglang-emit-kv-events-for-l1l2-tier-transitions
idhanani/mx-mixed-tp-v1
idhanani/session-hicache
idhanani/sgl-bootstrap-room-u64
idhanani/sgl-hicache-server-info
idhanani/sgl-native-http-lifecycle
idhanani/sgl-native-pd-hidden-states
idhanani/sgl-native-pd-logprobs
idhanani/sgl-native-pd-sampling
idhanani/sgl-native-rejected-beam-lifecycle
idhanani/sgl-native-session-fencing
idhanani/sgl-native-session-routing
idhanani/sgl-opaque-unary-lifecycle-poc
idhanani/sglang-27574-hint-manager-rebuild
idhanani/sglang-27574-session-cache-rpcs
idhanani/sglang-27574-session-priority-admission
idhanani/sglang-conditional-aggregation-main
idhanani/sglang-omitted-output-limit
idhanani/sglang-storage-hints
ifmn/eagle-dp-attn
ifmn/npu/afd
ifmn/npu/glm-5-optim
incomplete-tokenizer-cache-fallback
indexer-topk-shape
inkling-support
int8_diffusion_online_quantization
intel_dev
ishan/add-dg-logging
ishan/dnm-fi
ishan/fix-dp
ishan/fix-minimax25-deepep
ishan/fix-nixl-hicache
ishan/radix-cache-review-fixes
ishan/retention-duration-priority
ishan/some-good-info
ishan/subagent-eviction-patterns
jh-ut-upstream-3
jiaguo/restore-cu129-matrix
jiajun/anthropic-error-sse-helpers
jiajun/chat-session-return-indexer-topk
jiajun/dsv41-bind-next-plain-ref
jiajun/dsv41-h200-rl
jiajun/filter-nonpersistent-weight-buffers
jiajun/fix-inkling-rocm-import
jiajun/indexer-topk-num-layers
jiajun/kimi-k2-raw-tool-call-id
jiajun/mimo-v2-swa-window-size
jiajun/mooncake-output-store
jialin/sync-hicache-buffer-regressions
jialin/sync-prefill-busy-accounting
jialin/sync-singleton-pd-polls
jialino/rust-tree-core-full
jiminator/ple-shared-host-tables
jit-launch-on-torch-current-stream
joesun/diffusion-unify-common-env-func
k3-kvv-conformance
k3-stream-fork-order
kan/integrate-dynamo-mm
kan/kimi-k3-mm-preprocessor
kan/processor-fixtures-hf
kan/rust-input-ids-store
kan/rust-mm-mimo
kan/rust-server-hf-cache-fix
kan/rust-server-pd
kan/rust-tm-dev
kan/rust-tm-rust-mm
kan/rust-tm-ttft-breakdown
kan/server-info-openai-env
kan/sgl-router-adopt-engine-defaults
kan/sgl-router-benchmark
kan/sgl-router-kwargs-no-forward
kan/sgl-router-pd-version-group
kan/sglang-mm-on-rust-server
kan/unify_shmem
kernel-register-amd-batch5
kernel-register-amd-ltx2
kernel/dual_fused_norm
kimi-k3-deferred-release-35049-35360
kimi-k3-hybrid-pool-import-fix
kimi-k3-zmq-max-sockets-mitigation
kimi-k3-zmq-socket-limit-fix
kimi-k3-zmq-socket-limit-fix-main
kp/DEBUG-MLA-prefill-cp
kp/add-memory-profile-instruments-flashinfer
kp/distributed-arg-max-greedy-spec
kp/fix-MHA-cp-prefix-len-calculation
kp/fix-dpsk-v32-tool-call
kp/gemma4-qkj-proj-refactor
kp/moe-finalize-fuse-with-ar-norm
kp/optimize-pack-fp8-quantize
kp/refactor-dcp-attn-comm-group
kp/rerun-ut-write-perm
kp/streaming-reasoning-parser-fix
kpham/dcp-pool-page-from-allocator
kpham/dcp-triton-extend-rank-uniform-prefix
kpham/draft-expert-location-gate
kpham/dsa-dcp-extend-gather-kv
kpham/dsa-dcp-hicache
kpham/dsv41-encoder-replay-bcg
kpham/fix-dsa-kpool-niah
kpham/fix-fake-receiver-ensure-abort
kpham/flashinfer-det-chunk-align
kpham/pd-indexshare-unfused-topk
kpham/port-36507-loading-quant
kpham/port-36507-serving-lifecycle
kpham/port-36507-spec-decode-perf
kpham/pr44623-main-merge
kpham/pr44623-tailgraph-off
kpham/speculative-dcp-size
kpham/tokenspeed-mla-nope-prefill
kv-cache-logging-temp
lianmin/dispatch-output-combine-input
lianmin/external-model-capabilities
lianmin/output-processor
lianmin/rust-frontend-extension-hooks
linyu/auto-tune
logging-worker-sglang-rotuer
lora-ci-diag-tp-overlap-cand
lora-ci-diag-tp-overlap-main
lora-ci-exp-hicache-qwen3
lora-ci-exp-update-concurrency
lora-ci-groups
lora-perf-optimize-2
lora-stack-00-shared-hooks
lora-stack-01-route-primitives
lora-stack-02-dense-kernels
lora-stack-03-dense-plans
lora-stack-04-dense-tables
lora-stack-05-dense-runner
lora-stack-06-dense-graphs
lora-stack-07-mla-correction
lora-stack-08-inkling-sink
lora-stack-09-moe-plans
lora-stack-10-moe-routing
lora-stack-11-moe-tables
lora-stack-12-moe-stages
lora-stack-13-moe-providers
lora-stack-14-moe-runner
lora-stack-15-cutedsl-schedule
lora-stack-16-cutedsl-bf16
lora-stack-17-triton-fp8
lora-stack-18-cutedsl-fp8
lora-stack-19-fp8-tables
lora-stack-20-nvfp4
lsyin/attn-backend-stamp
lsyin/bcg-dsv41-candidate-table-capture-assert
lsyin/bench-engine-steps
lsyin/concurrency-limit-attribution
lsyin/custom-ar-v2-float-only
lsyin/dedup-head-slice
lsyin/deep-gemm-post1
lsyin/draft-extend-input-staging
lsyin/ds-recipe-backend
lsyin/ds-recipe-rust-parse
lsyin/dsa-ragged-verify-graph
lsyin/dspark-compact-dsa-syncfree
lsyin/dspark-eager-draft-embed
lsyin/dspark-run-draft-proposal
lsyin/dspark-simulated-bonus-bf16
lsyin/dsv4-seq-lens-tests
lsyin/dsv41-captured-full-topk-per-row
lsyin/dsv41-prefill-graph-capture-stub
lsyin/dsv41-sm90-prefill-graph-tail
lsyin/eagle-draft-dtype-auto
lsyin/eagle-verify-fence
lsyin/engram-a2a-decode
lsyin/expand-ruff-pyflakes
lsyin/gc-debug-dev
lsyin/hybrid-allocator-sides
lsyin/hybrid-attn-sync-gate
lsyin/killall-sh-pattern-sync
lsyin/megamoe-chunk-cap
lsyin/page-tail-write
lsyin/plan-stream-poc
lsyin/ragged-layout-access
lsyin/re-introduce-tp-grammar
lsyin/rerun-test-merged-pr
lsyin/sched-pr01
lsyin/sched-pr02
lsyin/sched-pr03
lsyin/sched-pr04
lsyin/sched-pr05
lsyin/sched-pr06
lsyin/sched-pr07
lsyin/sched-pr08
lsyin/sched-pr09
lsyin/sched-pr10
lsyin/sched-pr11
lsyin/sched-pr12
lsyin/sched-pr13
lsyin/sched-pr14
lsyin/sched-pr15
lsyin/sched-pr16
lsyin/sched-pr17
lsyin/sched-pr18
lsyin/sched-rewrite
lsyin/spec-width-optional
lsyin/stash-dedup-fence
lsyin/swa-sizing-policy
lsyin/tree-mask-protocol-cleanup
lsyin/unify-draft-runner
lsyin/urc-test-long
lsyin/verify-extend-unify
lsyin/verify-metadata-in-graph
lsyin/verify-tier-resolve
ltx-2.3-two-stage-ti2v
ltx-2.3-two-stage-ti2v-clean
ltx-batched-unify-all-paths
ltx23-official-lora-merge-align
ltx23-tp-psnr20-latest-20260508
ltx23-transition-lora-support
lufang/independent-cache-window-metadata
lzy17/amd-ci-pd-keep-dmabuf
m3-mxfp8-dense-block-convert
m31-hopper-dspark
m31-nightly-e2e
main
main-ci-test
main-dev
measure-prescheduler-overhead
michael/amd-aot-kernel-cache
michael/amd-ci-openai-2.8
michael/amd-diffusion-retry-deadline
michael/amd-disagg-hip-dmabuf
michael/amd-disagg-unfused-kv
michael/amd-docker-build-only
michael/amd-gsm8k-skip-mistral-fp8
michael/amd-multimodal-max-parallel
michael/amd-pr-test-extra-tier
michael/amd-pr-test-nightly-large-models
michael/amd-pr-test-parallel-shards
michael/amd-pr-test-skip-and-tags
michael/amd-register-portable-unit-tests
michael/amd-skip-ltx25-decoder
michael/amd-vae-blend-test-reference
michael/cosmos-dreams-decode-queue-drain
michael/crash-dump-expect-exit
michael/diag-gsm8k-mistral
michael/diag-gsm8k-mixtral-r4
michael/dsv4-bcg-swa-ring-mock
michael/dsv4-hip-bcg-tests
michael/dsv4-v41-k-shfl-mask
michael/dsv41-bcg-hip-graph-width
michael/flux3-determinism-probe
michael/flux3-disagg-amd-diag
michael/flux3-exact-vae-rocm-determinism
michael/fp8-moe-aiter-py310
michael/miles-8gpu-split
michael/miles-gpu2-evict-probe
michael/miles-nightly-partition
michael/minimax-m31-mxfp4-validate
michael/mmgen-unit-rocm-skips
michael/moss-image-rocm-skip
michael/mxfp8-online-scalingtype
michael/nomas-rocm-skips
michael/qsa-prefill-relative-mi300-lds
michael/stage-b-mi35x-all
michael/stage-b-mi35x-all2
michael/usp-ring-py310
michael/wan-cutedsl-lazy-import
michaezh/test-pr-27730-rocm720
mick-dynamic-batching
mick/codex-tmp-lingbot-block5-wanvae-20260602
mick/codex-tmp-lingbot-ci-partition-append-order-20260602
mick/codex-tmp-lingbot-wanvae-batched-decode-20260602
mick/codex-tmp-lingbot-wanvae-batched-main-20260602
mick/codex-tmp-lingbot-wanvae-block4-main-20260603
mick/codex/causal-kv-profile-20260605-r2
mick/codex/diffusion-bench-framework
mick/codex/lingbot-causal-dmd-refactor-validate-d90e6f4
mick/codex/lingbot-control-signal-validate-554b486
mick/codex/lingbot-latent-spec-validate-1654379
mick/codex/lingbot-latent-spec-validate-71e3b33
mick/codex/lingbot-raw-frame-rename-validate-b18c4ad
mick/codex/ltx23-nonpr-perf-20260503
mick/codex/omni-u1-orchestrator
mick/codex/unipc-stack-candidate-20260603
mick/diffusion-auto-residency
mick/diffusion-bench-comfyui
mick/diffusion-conditioning-cache-parallel-digest
mick/diffusion-cookbook-entry
mick/diffusion-fasth3-8step-v2
mick/ltx23-hq-precision-align
mick/ltx23-official-lora-merge-align
mick/ltx23-tp-psnr20-20260506
mick/mick/mmgen-multi-output-dedup
mick/mmgen-stage-a
mick/tmp/image-edit-uploadfile-fix-20260602
mick/tmp/lingbot-local-final-proj-gather-20260602
mick/tmp/lingbot-msgspec-fix-20260602
mick/tmp/lingbot-skip-proj-after-msgspec-20260602
mick/tmp/lingbot-skip-proj-verify-20260602
mick/tmp/lingbot-uniform-usp-fastpath-20260602
mick/wan-ffn-residual-chain
mick/wan-fused-qkv-gemm
mick/wan-s2v-rebased-20260328
mick/wan-text-kv-once-per-prompt
mick/worktree-ltx2-hq-pipeline
migrate-rotary-embedding
mistral4-support
mmangkad/ci-5090-1-base-b-server
mmangkad/ci-5090-2-base-b-light
mmangkad/ci-hf-hub-rate-limit
mmangkad/dsa-kpool-skip-unused-mqa-schedule
mmangkad/dsa-mtp-share-draft-metadata
mmangkad/dsa-tilelang-bf16-split-kv-decode
mmangkad/fix-bench-serving-base-url
mmangkad/flashinfer-prefill-autotune-default
mmangkad/glm53-flash-kda-prefill-bcg
mmangkad/glm53-mhc-cross-layer-handoff
mmangkad/input-logprob-no-sync
mmangkad/kda-flashinfer-prefill-cu-seqlens-sync
mmangkad/kda-replayssm-commit-prepared-step-indices
mmangkad/kda-replayssm-fold-tiling
mmangkad/move-1gpu-large-tests-to-5090
mmangkad/nextn-draft-prefill-graph
mmangkad/remove-4-gpu-gb300-suite
mmangkad/upgrade-cuda-pytorch-2.15
moonep
msa-megakernel-route
mtp/nongraph-metadata-fix
mtp/page-align-alloc
multimodal-bench-serving-openai-compatible
multimodal-gen-tokenizer-fix
muse-spark
nightly-ci
nightly/kimi-k25-mtp
nightly_qiaolin
npu-device-mixin
npu-skip-non-python-changes
nvidia/qwen35-agentx-gb300
nvidia/qwen35-agentx-rubin
optimize-mla-kv-buffer-get
oss-sync-538ec60fd4
oss-sync-5bf04645f2
oss-sync-5cc91ba332
oss-sync-63008b6421
oss-sync-7c10a8cdcd
oss-sync-8bb92640f5
oss-sync-91e9fdc5e2
oss-sync-a288e8ae43
oss-sync-ad931fecce
oss-sync-af1e56b9be
oss-sync-b34e09bb18
oss-sync-b6d441ee64
oss-sync-c9197bfb88
oss-sync-nixl-test-tmpdir
oss/l2-transfer-consolidation
otel-2/n
page-allocator-sync
patch-docker-merge-reworkd
patch-docker-path-fix
perf-accuracy-improvement
perf-regression-dashboard
perf/eagle-fused-topk-topp-renorm
perf/tiny-n-gemm-reduce-scatter
performance-tracking-ci-failure-monitor
phase1-eager-kernel-optimization
phase1-on-main
piecewisemempool
pin-fa4-4.0.0b16
port/dsv4-flash-encoder
pr-12263
pr-13592
pr-13728
pr-14226
pr-16287
pr-17247
pr-26980
pr-33108
pr-34855-gt
pr/31902
pr23119-fix
processor-mm-qwen-vl
qiaolin/adaptive-spec-cuda-graph-buffer-reuse
qiaolin/draft-topk1-argmax-fastpath
qiaolin/qwen-next-demo
qiaolin/qwen-next-kv-allocation
qiaolin/qwen-next-relay
qiaolin/qwen-next-sampling
qiaolin/qwen-next-snapshot
qiaolin/qwen-next-state-commit
qiaolin/qwen-next-tracking
qiaolin/spec-step4-fuse-draft-decode-create-flashmla
qiaolin_commit_in_draft_extend_graph
qiaolin_fused_commit_indices
qiaolin_fused_verify_prep
qiaolin_mrope_nosync
qiaolin_replayssm_graph_commit
qiaolin_spec_mrope_opt
qimcis-ltx2-i2v-support
quant-refactor/import-shim-test
quant-refactor/string-spec-registry
quant-refactor/unify-scheme-abcs
qwen-vae-bf16-default
qwen38
qwen38_miles
qwen4-main-squashed
qwen_coder
qwen_optimize
radix-tree-n/1
rainj-me/rust-api-server-tower-layer
rainj-me/rust-benchmark
rainj-me/rust-cp-apis-migration
rainj-me/rust-fanout-control-reqs
rainj-me/rust-server-abis
rainj-me/rust-server-api-http
rainj-me/rust-server-cp-api
rainj-me/rust-server-pd-lb
rainj-me/rust-server-py-foundations
rainj-me/rust-server-py-integration
rainj-me/rust-server-python-part-1
rainj-me/rust-server-python-part-2
rainj-me/rust-server-refactor-openai-apis
rainj-me/rust-server-runtime
rainj-me/rust-server-tests
rainj-me/rust-server-tm-ingress
rainj-me/rust-server-tm-sampling
rainj-me/rust-server-tok-detok
rainj-me/rust-server-tokenizer
rebase/pr-30478-dcp-runtime-context
refactor/dsv41-multimodal-layout
refactor/release-whl-kernel-composite-actions
reland-pr26134
release-test-xpu
release/v0.5.10
release/v0.5.10rc0
release/v0.5.11
release/v0.5.12
release/v0.5.13
release/v0.5.14
release/v0.5.15
release/v0.5.16
release/v0.5.17
release/v0.5.18
release/v0.5.19
release/v0.5.20
release/v0.5.21
release/v0.5.22
release/v0.5.23
release/v0.5.7
release/v0.5.8
remove-stale-prefill-cudagraph-gqa-gate
reorganize-test-registered-structure
replay-audit-hooks
research/linear-attn-fp16-state
resize-condition-image
return_routed_expert
return_routed_expert_yizhang
revert-14717-feat/layernorm_scale_shift_kernel
revert-15829-feat/mamba2-radix-overlap
revert-20560-youngrok/fix-hicache-write-through-idle
revert-21458-fused_qk_norm_mrope_decode
revert-21631-refactor_write_back_kernel
revert-23943-yhyang201/dsv4-h200-pro-single-node
revert-25576-update-cutlass-dsl-cu13-4.5.1
revert-26235-eagle-topk1-softmax-skip
revert-26905-amd_change_hf_path_0601
revert-27529-fangyuan/fix_c128_kv_dtype_mismatch
revert-29775-yamy1234/dsv4-flashmla-sparse-prefill-default-on
revert-30947-glm52/mtp-split-1-topk1
revert-31001-do-not-merge
revert-31107-topk-glm
revert-33661-bcg-mla-all-archs
revert-34715-bmm65536-fallback
revert-40475-diffusion_register_optm_2
revert-44841-codex/comfyui-worker-patch-20261017
revert-hy3-preview-reenable-nemotron
revert-pr-19140
revert-pr26134
revert-rmsnorm-dynamo-check
revert-tf-4.57.1
rockdu-dev-mm-v2-miles/01-shared-enums
rockdu-dev-mm-v2-miles/02-processor-config
rockdu-dev-mm-v2-miles/03-shared-mixin
rockdu-dev-mm-v2-miles/04-serving-shared
rockdu-dev-mm-v2-miles/05-token-space-framework
rockdu-dev-mm-v2-miles/06-sglang-wrapper
rockdu-dev-mm-v2-miles/07-qwen-strategy
rockdu-dev-mm-v2-miles/08-parity-tests
rockdu-dev-mm-v2-miles/09-partial-expansion
rockdu-dev-mm-v2/02-processor-config
rockdu-dev-mm-v2/03-shared-mixin
rockdu-dev-mm-v2/04-serving-shared
rockdu-dev-mm-v2/05-token-space-framework
rockdu-dev-mm-v2/06-sglang-wrapper
rockdu-dev-mm-v2/07-qwen-strategy
rockdu-dev-mm-v2/08-parity-tests
rockdu-dev-mm-v2/09-partial-expansion
rockdu-dev-mm-v2/sglang-miles
rockdu-dev-mm-v2/sglang-miles-timelens
rockdu-dev-mm/02-processor-config
rockdu-dev-mm/03-shared-mixin
rockdu-dev-mm/04-token-space-core
rockdu-dev-mm/05-qwen-strategy
rockdu-dev-mm/06-startup-and-parity
rockdu/rollout-latent-step-indices
rockdu/rollout-plugin-sglang-miles-next
rockdu/rollout-plugin-sglang-miles-v0.5.22
rocm10-gfx1250
rocm101_1250
rocm_rccl_232
rocm_release_for_testing
router-3a-1-tree-prefix-view
router-3a-2-locality-metrics
router-3b-locality-dashboard
router-4d-outcome-attribution
router-fallback-handler
router/policy-reorg-1-state-move
router/policy-reorg-2-load-view
router/policy-reorg-3-core
router/policy-reorg-4-affinity
router/policy-reorg-5-cache
router/policy-reorg-6-config
router/policy-reorg-7-resolver
router/policy-reorg-8-switch
router/policy-reorg-9-remove
rust-mm-deepseek-v41
rust-mm-glm53-flash
rust-server-chat-parity-followups
rust-server-tool-choice-parity
sahithi/gb200-8k1k-configs
sahithi/gb200-accuracy-evals
sampling-top-k-first
save_mem_linear_mtp
seq-lens-datatype
server-args-cleanup
session-routing-dev
sgl-bot
sgl-compile-distributed-export-runtime
sgl-compile-platform-export
sgl-router-kv-comments
sgl-router-reorg-admitted-fallback
sgl-router-reorg-affinity-savings
sgl-router/upstream-lb-2-engine-load
sgl-router/upstream-lb-3-decision-metrics
sgl-router/upstream-lb-4-queue-gate
sgl-router/upstream-lb-5-saturation-pin
sgl-router/upstream-lb-6-attribution
sgl-router/upstream-lb-7-measure-spill
sgl-router/upstream-lb-8-spill-overlap
sgl-router/upstream-lb-9-queue-headroom
sglang-miles
sglang-miles-20260327
sglang-miles-diffusion-temp
sglang-miles-flux3-image
sglang-miles-glm53-lora
sglang-miles-glm53next
sglang-miles-h3
sglang-miles-inkling
sglang-miles-k3
sglang-miles-muse-spark-open
sglang-miles-qwen38next
sglang-miles-v0.5.10-final
sglang-miles-v0.5.12
sglang-miles-v0.5.12-final
sglang-miles-v0.5.13-final
sglang-miles-v0.5.14-final
sglang-miles-v0.5.15
sglang-miles-v0.5.15-final
sglang-miles-v0.5.16
sglang-miles-v0.5.16-final
sglang-miles-v0.5.17
sglang-miles-v0.5.18
sglang-miles-v0.5.18-final
sglang-miles-v0.5.19
sglang-miles-v0.5.19-final
sglang-miles-v0.5.20
sglang-miles-v0.5.20-final
sglang-miles-v0.5.21
sglang-miles-v0.5.21-final
sglang-miles-v0.5.22
sglang-miles-v0.5.8.post1
sglang-miles-v0.5.9
sglang-miles-v0.5.9-final
sglang-miles_weight_checker
sglang-miles_weight_loader_threshold
sglang-processor-chat-fixes
sglang-slime
sglang-slime-patched
shiyang/forwad_decouple
shiyang/load-snapshot-shm
shiyang/logits/fuse-allgather-buffer
shiyang/mem_check
shiyang/mem_v2/decouple
shiyang/mem_v2/first
shiyang/optimize-get-load-piggyback
shiyang/req_to_token
simolin/client-cancel-499
slime-patched-260303
smg-mm-11/n
spec/drop-draft-global-server-args-restore
spec_hidden
split/pr24408-01-contracts
split/pr24408-02-server-args
split/pr24408-03-batch-invariant-ops
split/pr24408-04-layernorm-batch-tests
split/pr24408-05-tp-invariant-ops
split/pr24408-06-tp-invariant-tests
split/pr24408-07-collective-wiring
split/pr24408-08-dp-attention-padding
split/pr24408-09-qwen3-moe-model
split/pr24408-12-wiring-cli-tests
split/pr24408-13-wiring-contract-tests
split/pr24408-14-wiring-runtime-tests
split/pr24408-15-wiring-reduction-tests
split/pr24408-16-wiring-cuda-tests
stack/ngram-hidden-states-test
step-3.5-flash
step5-preview-support
stress-test-debugging
support_qwen3_guard
support_usp_for_wan_vae
swa-pool-pp-start-layer
sync-pr-test-from-main
temp-docker-image
test-ci-empty-pr-2
test-gb200-runners
test-mm-amd-ci
test/ci-disagg-tp-check
test/mooncake-arm64-wheel-validation
test/pr33561-ab-both-markers-off
test/rerun-stage-8gpu-h200
test/slash-command-handler-tests
test_new_x86_node
test_qiaolin
tokenspeed_mla_prefill
tom/adhoc-sum-pad-exp-th3
tom/alloc-page-aligned/op1
tom/alloc-page-aligned/op2
tom/alloc-page-aligned/op21
tom/alloc-page-aligned/op22
tom/alloc-page-aligned/op24
tom/alloc-page-aligned/op25
tom/alloc-page-aligned/op25a
tom/alloc-page-aligned/op26
tom/alloc-page-aligned/op27
tom/alloc-page-aligned/op28
tom/alloc-page-aligned/op29a
tom/alloc-page-aligned/op29b
tom/alloc-page-aligned/op29c
tom/alloc-page-aligned/op29d
tom/alloc-page-aligned/op30
tom/alloc-page-aligned/op31
tom/alloc-page-aligned/op32
tom/alloc-page-aligned/op33
tom/alloc-page-aligned/op5
tom/alloc-page-aligned/op51-non-split
tom/alloc-page-aligned/op52
tom/alloc-page-aligned/op53
tom/alloc-page-aligned/op54
tom/alloc-page-aligned/op6
tom/alloc-page-aligned/op7
tom/alloc-page-aligned/temp-ci
tom/cache-unfinished-req-use-committed-len
tom/disable_pd_perturb_ci
tom/e2e_qiaolin_fill_ids
tom/extract-scheduler-batches
tom/fix-d1-current-gdn-lora-cli
tom/fix-dllm-abort-waiting-queue
tom/fix-eagle-chunked-next-prompt-token
tom/fix-moe-virtual-experts-gated-lora
tom/fix-preempt-stale-extend-input-len
tom/fix-vocab-boundary-off-by-one
tom/gemini-modify-component-review
tom/kv_canary
tom/kv_canary_revert_reversed
tom/lora-mq-a
tom/lora-mq-b
tom/lora_0604_opt_moe
tom/lora_qkv_b_opt
tom/merge_input_output_ids
tom/miles-snapshot-rope-compat-1019
tom/pr_24928_base
tom/pr_chain/tom_refactor_202605a/primary/mech_preflight/convert-self-x-to-locals
tom/probe-lora-bug-25690
tom/probe/20260607T131809Z-a39c428
tom/probe/20260821T001953Z-dsv3-same-ckpt
tom/probe/20260821T002009Z-dsv3-drop-refit
tom/probe/20260821T235224Z-spans-dsv3-dpattn
tom/refactor-miles-repo-sglang/ac8770
tom/refactor-miles-repo-sglang/deliver-1
tom/refactor-miles-repo-sglang/deliver-2
tom/refactor-miles-repo-sglang/deliver-4
tom/refactor-miles-repo-sglang/deliver-5
tom/refactor-miles-repo-sglang/lora-targets
tom/refactor-miles-repo-sglang/op1-3
tom/refactor-miles-repo-sglang/op1-9
tom/refactor-miles-repo-sglang/op13-116
tom/refactor-miles-repo-sglang/op21-2
tom/refactor-miles-repo-sglang/op22-2
tom/refactor-miles-repo-sglang/op26-2
tom/refactor-miles-repo-sglang/op40-1
tom/refactor-miles-repo-sglang/op43-123
tom/refactor-miles-repo-sglang/op47-62
tom/refactor_shared_array
tom/remove-full-untruncated-fill-ids
tom/rename-fill-len-extend-fill-len
tom/req-extend-range
tom/req-fill-len-none
tom/req_pool_idx_and_cache_and_kv/1
tom/req_pool_idx_and_cache_and_kv/10
tom/req_pool_idx_and_cache_and_kv/11
tom/req_pool_idx_and_cache_and_kv/12
tom/req_pool_idx_and_cache_and_kv/13
tom/req_pool_idx_and_cache_and_kv/14
tom/req_pool_idx_and_cache_and_kv/15
tom/req_pool_idx_and_cache_and_kv/16
tom/req_pool_idx_and_cache_and_kv/17
tom/req_pool_idx_and_cache_and_kv/18
tom/req_pool_idx_and_cache_and_kv/21
tom/req_pool_idx_and_cache_and_kv/22
tom/req_pool_idx_and_cache_and_kv/23
tom/req_pool_idx_and_cache_and_kv/24
tom/req_pool_idx_and_cache_and_kv/25
tom/req_pool_idx_and_cache_and_kv/26
tom/req_pool_idx_and_cache_and_kv/27
tom/req_pool_idx_and_cache_and_kv/28
tom/req_pool_idx_and_cache_and_kv/29
tom/req_pool_idx_and_cache_and_kv/3
tom/req_pool_idx_and_cache_and_kv/30
tom/req_pool_idx_and_cache_and_kv/31
tom/req_pool_idx_and_cache_and_kv/32
tom/req_pool_idx_and_cache_and_kv/33
tom/req_pool_idx_and_cache_and_kv/34
tom/req_pool_idx_and_cache_and_kv/35
tom/req_pool_idx_and_cache_and_kv/36
tom/req_pool_idx_and_cache_and_kv/37
tom/req_pool_idx_and_cache_and_kv/38
tom/req_pool_idx_and_cache_and_kv/39
tom/req_pool_idx_and_cache_and_kv/4
tom/req_pool_idx_and_cache_and_kv/5
tom/req_pool_idx_and_cache_and_kv/6
tom/req_pool_idx_and_cache_and_kv/7
tom/req_pool_idx_and_cache_and_kv/8
tom/req_pool_idx_and_cache_and_kv/9
tom/req_pool_idx_and_cache_and_kv/temp_ci
tom/req_pool_idx_step1_dataclass
tom/req_pool_idx_step2_streaming_move
tom/req_pool_idx_step3_eliminate_flags
tom/revert-25690-cutedsl
tom/revert-pr10414
tom/rm_cur_last_batch
tom/scripted_runtime_extracted
tom/scripted_runtime_extracted_chain
tom/sglang/d1-c21-dsv4-reasoning-compat
tom/stateless_scheduler_b
tom/stateless_scheduler_b_temp_run
tom/temp_ci
tom/temp_revert_27965
tom/trainer-ft-split
tom_refactor_202605a/primary/mech_model_runner
tom_refactor_202605a/primary/mech_preflight
tom_refactor_202605a/primary/mech_scheduler
tom_refactor_202605a/primary/mech_tokenizer_manager
tom_refactor_202605a/primary/nonmech_model_runner
tom_refactor_202605a/primary/nonmech_scheduler
tom_refactor_202605a/primary/nonmech_tokenizer_manager
tool_call_benchmark
top-k-first-opt
transformers-v5
trim-nemotron35-cookbook-flags
trtllm-mha-full-overlap
trtllm-mla-fused-graph-metadata
trtllm-sparse-mla-dcp
unified-envelope-buffer-info
update-transformers-v5
update_perf_dashboard
util-queue-timeline-chart
v0.5.19
vincentzed/dcp-shard-indexer
vlm-post-capture-fixed-reserve
vz/qwen3-omni
wan-overlay-only-20260328
wan-s2v-main-20260330
wan-s2v-rebased-20260328
wduan/fix_pause_generation
whl-metadata-fix
worktree-agent-a5ee4203
worktree-agent-ab2b777b
worktree-pure-nibbling-hickey
xeonnightly
xiaomi-mimo-v2-flash-0.5.5
xiezhq-schedule-new
xiezhq-schedule-simplify
xiezhq-session
xiezhq/42701-merge-main
xinyuan/dflash-overlap-radix-war-fix
xinyuan/glm47-39136-review-fixes
xinyuan/glm53flash-sota-perf
xinyuan/kimi-k3-parser-auto-upstream
xinyuan/kimi-k3-reasoning-parser
xinyuan/kimi-k3-tool-constraints
xinyuan/kimi-k3-tool-parser
xinyuan/nightly-precision-baseline-refresh
xinyuan/plan-stream-handoff
xpu/kernel-from-main
xpu/qwen3-ccl-arc-allreduce
xpu/qwen3-eagle-draft-attn-fallback
xpu/qwen3-eagle-target-only-sampling
xpu/qwen3-fp8-w8a16-dense
xpu/qwen3-moe-tuning-xpu
xpu/qwen3-qwen3-fused-qknorm-rope
xpu/qwen3-rope-native-fallback
xsun/amd_diffusion
yamy/mixed2048-fusion-piecewise-cg-bia
yanbo/feat/router-pd-loadmonitor
yanbo/k3-stream-conformance
yanbo/k3-thinking-param
yanbo/k3-tool-call-conformance
yanbo/pd-k3
yanbo/pd-k3-rebase-0919
yanbo/pr_34820
yanbo/router-dp-aware
yanbo/router-pd-inhouse
yanbo/router_test
yanbo/router_test-r1-owned-ancestor
yangmin/pr26245-fullchat-terminal-strict-rr
yihao/draft-attn-registry
yizhang/decode_store_only_mamba_radix_cache
yizhang/mamba_prefix_cache_support_large_page_size
yizhang/tp-dp-invariant
yizhang/vlm_fix_radix_cache_for_qwen3vl
yonghao/cp-sharded-kv-layout
yonghao/ump-shared-capacity
yonghao/unified-hicache-review-followup
yueming/colocated-param-bytes
yueming/colocated-param-bytes-pull
yueming/dsa-pagetable-tms
yueming/fix-lora-unit-test-fakes
yueming/fix-uninit-ue8m0-scale-padding
yueming/fix-uninit-ue8m0-scale-padding-main
yueming/nextn-init-fix
yueming/p2p-registry-fixes
yueming/qwen3-router-dtype
yueming/rank-info-record
yueming/rl-lora-streamed-updates
yueming/rl-model-side-weight-update-fixes
yueming/rl-pull-weights
yueming/rl-weight-checker-exemptions
yueming/weight-checker-skip-sincos-tables-main
yusheng/abort-rid-prefix
yusheng/cherrypick-36389-partial-sglang-miles
yusheng/cherrypick-36389-sglang-miles
yusheng/kimi-moe-lora-residual
yusheng/lora-dp-attention
yusheng/lora-fp8-col-shard
yusheng/lora-full-prefill-cg-gate
yusheng/lora-load-from-distributed
yusheng/lora-moe-buffer-dims
yusheng/lora-moe-cg-dp-attention
yusheng/lora-qwen35-gdn-in-proj-ba
yusheng/lora-rowparallel-probe
yusheng/lora-shard-spec-table
yusheng/lora-upsert
yusheng/lora-weight-update-quant-hook
yuwei/pcg
yz/amd_diffusion
zaya-pr1-short-conv
zaya-pr2-cuda-graph
zaya-pr3-model
zhichen/gdn-norm-device-nemotron-topk-layer-id
zhichen/qwen3_5-parameter-mapper
zqx/hicache-file-storage-path-followup
gateway-v0.1.5
gateway-v0.1.6
gateway-v0.1.7
gateway-v0.1.8
gateway-v0.1.9
gateway-v0.2.0
gateway-v0.2.1
gateway-v0.2.2
gateway-v0.2.3
gateway-v0.2.4
gateway-v0.3.0
gateway-v0.3.1
v0.1.10
v0.1.11
v0.1.12
v0.1.13
v0.1.14
v0.1.15
v0.1.16
v0.1.17
v0.1.18
v0.1.19
v0.1.20
v0.1.21
v0.1.22
v0.1.23
v0.1.24
v0.1.3
v0.1.5
v0.1.6
v0.1.7
v0.1.8
v0.1.9
v0.2.0
v0.2.10
v0.2.11
v0.2.12
v0.2.13
v0.2.14
v0.2.14.post1
v0.2.14.post2
v0.2.15
v0.2.5
v0.2.6
v0.2.7
v0.2.8
v0.2.9
v0.2.9.post1
v0.3.0
v0.3.1.post1
v0.3.1.post2
v0.3.1.post3
v0.3.2
v0.3.3
v0.3.3.post1
v0.3.4
v0.3.4.post1
v0.3.4.post2
v0.3.5
v0.3.5.post1
v0.3.5.post2
v0.3.6
v0.3.6.post1
v0.3.6.post2
v0.3.6.post3
v0.4.0
v0.4.0.post1
v0.4.0.post2
v0.4.1
v0.4.1.post1
v0.4.1.post2
v0.4.1.post3
v0.4.1.post4
v0.4.1.post5
v0.4.1.post6
v0.4.1.post7
v0.4.10
v0.4.10.post1
v0.4.10.post2
v0.4.2
v0.4.2.post1
v0.4.2.post2
v0.4.2.post3
v0.4.2.post4
v0.4.3
v0.4.3.post1
v0.4.3.post2
v0.4.3.post3
v0.4.3.post4
v0.4.4
v0.4.4.post1
v0.4.4.post2
v0.4.4.post3
v0.4.4.post4
v0.4.5
v0.4.5.post1
v0.4.5.post2
v0.4.5.post3
v0.4.6
v0.4.6.post1
v0.4.6.post2
v0.4.6.post3
v0.4.6.post4
v0.4.6.post5
v0.4.7
v0.4.7.post1
v0.4.8
v0.4.8.post1
v0.4.9
v0.4.9.post1
v0.4.9.post2
v0.4.9.post3
v0.4.9.post4
v0.4.9.post5
v0.4.9.post6
v0.5.0rc0
v0.5.0rc1
v0.5.0rc2
v0.5.1
v0.5.1.post1
v0.5.1.post2
v0.5.1.post3
v0.5.10
v0.5.10.post1
v0.5.10rc0
v0.5.11
v0.5.12
v0.5.12.post1
v0.5.13
v0.5.13.post1
v0.5.14
v0.5.15
v0.5.15.post1
v0.5.16
v0.5.17
v0.5.18
v0.5.19
v0.5.2
v0.5.20
v0.5.21
v0.5.22
v0.5.2rc0
v0.5.2rc1
v0.5.2rc2
v0.5.3
v0.5.3.post1
v0.5.3.post2
v0.5.3.post3
v0.5.3rc0
v0.5.3rc1
v0.5.3rc2
v0.5.4
v0.5.4.post1
v0.5.4.post2
v0.5.4.post3
v0.5.5
v0.5.5.post1
v0.5.5.post2
v0.5.5.post3
v0.5.6
v0.5.6.post1
v0.5.6.post2
v0.5.7
v0.5.8
v0.5.8.post1
v0.5.9
Select branches
Hide Pull Requests
F4_support
UX-NPU
VDN-h3
ac8452/1/0
add-command-a-plus-cookbook
add-reasoning-usage-test
add_qwen3_5_ci
add_tls
add_use_seperate_warmup_in_bench_serving
adding-pin-logic-docker
agentx-mamba-envelope-transfer-20260823
agentx-nixl-contiguous-run-xfer-20260823
agentx-nixl-fewer-run-descriptors-20260823
agentx-nixl-token-budget-pipeline-20260823
agentx-post-launch-transfer-drain-20260823
agentx-upstream/gdn-compact-layer-loop-20260824
agentx/decode-assignment-ack-fresh-20260828
agentx/decode-ready-reserve-20260828
agentx/decode-router-hint-burst-20260828
agentx/verify-mha-metadata-main-20260828
alexnails/bazel-with-generate-deps
alexnails/ci-install-rust-release
alexnails/cpu-scheduler-perf-tier-ab
alexnails/cpu-scheduler-pipeline-overlap
alexnails/decoupling-scheduler-decode-fix
alexnails/dp-attn-perf-fix
alexnails/gemma-hicache-1
alexnails/gemma4-mtp-chain-selection
alexnails/gemma4-text-only
alexnails/gemma4-xpu-frozen-kv-mtp
alexnails/gemma4-xpu-greedy-logits
alexnails/gemma4-xpu-greedy-lut
alexnails/gemma4-xpu-intermediate-prefill-gemm
alexnails/gemma4-xpu-ple-add-scale
alexnails/gemma4-xpu-prefill-gemm
alexnails/gemma4-xpu-prefill-gemm-narrow
alexnails/gemma4-xpu-prefill-lut-standalone
alexnails/gemma4-xpu-qkv-rope-fma
alexnails/grpc
alexnails/grpc-phase1-fixes
alexnails/grpc-phase1/openai-stream-hook
alexnails/grpc-phase1/tests
alexnails/host-extend-bookkeeping
alexnails/inplatform-dispatch-key
alexnails/inplatform-graph-factories
alexnails/inplatform-quant-config
alexnails/kernels-1
alexnails/lint-ban-pickle
alexnails/macos-local-dev
alexnails/mlx-hybrid-batching
alexnails/model-cache-sync-migration
alexnails/moe-partial-block-optin
alexnails/platform-device-ops
alexnails/port-scheduler-radix-perf
alexnails/proto-msgpack-codec
alexnails/proto-msgpack-codegen-fixes
alexnails/rust-vlm-tile-preproc
alexnails/small-model-scheduler-perf
alexnails/tokenizer-2
alexnails/tokenizer-3
alexnails/xgrammar-parallel-fill
alexnails/xpu-bf16-gelu-mul-lut
alexnails/xpu-kernel-capabilities
amd-ci/mi355x-split-matrix-by-node-footprint
amd-diffusion-aiter-fix
amd-mi355x-disagg-nightly
amd/add-1gpu-small-recent-nv
amd/add-glm51-mxfp4-nightly-clean
amd/add-mimo-v25-pro-nightly-tests
amd/aiter-ci
amd/ci-gpu-hang-probe
amd/ci-vram-leak-selfheal
amd/deepseek_v4
amd/deepseek_v4_0430
amd/deepseek_v4_0520
amd/deepseek_v4_migrate-main_0508
amd/deepseek_v4_rebase_main_0509
amd/dpsk_v4
amd/dsv4-rebase
amd/expand-1gpu-cov-jul13
amd/expand-1gpu-cov-jul15
amd/expand-1gpu-pr-coverage
amd/fix-flashinfer-benchmarks
amd/gfx1250-fp8-swiglu-fusion
amd/glm53-flash-mi30x-nightly-accuracy
amd/kimik26-bisect
amd/kimik26-knowngood
amd/mi455x-gfx1250-gpt-oss-ci
amd/mm-gen-nested-b1-harness
amd/mm-gen-nested-b2-envskip
amd/mm-gen-unit-test
amd/nightly-glm53
amd/probe-kernel-doable
amd/register-1gpu-tier1-batch
amd/register-dsa-meta-mamba-views
amd/register-jit-benchmarks
amd/register-jit-benchmarks-batch2
amd/register-jit-benchmarks-batch3
amd/register-jit-kernel-amd
amd/register-triton-data-movement-tests
amd/rocm720-pin-triton-wheel
amd/scout-aiter-kernel-warmup
amd/test-5n-mi355x-di
amd/test-pr31368
amd/test-pr31727
amd/verify-gap-doable
amd/verify-mi35x-4gpu
amd_benchmark
amd_diffusion
amd_fix_aiter_scaled_downcast
amd_fix_dpsk_0511
amd_fix_request_timeout
amd_fix_stage_c_0715
amd_glm5.3_flash
amd_m25_long
amd_m3_moe_0723
amd_march
amd_mori
amd_pr_25266
amd_rocm10
amd_rocm10_gfx1250_triton_orginal
amd_test_26662
amd_test_29158_0625
amd_test_29181
amd_test_mi355_0710
amd_test_pr28932
amd_test_pr29677
amd_test_pr32754
amd_test_pr_21655
amd_upgrade_pr_test
ascend_moe_optimizations
auto-detect-environment
auto/lmsys-blog-cards
b12x-align-b4d6c75
b12x-chat-template-fix
b12x-ci-overlay-tweaks
b12x-vision
backup-pr26517
backup/cheng/test/attention-runner-combination
backup/mmgen-official-gt-alignment-20260425
backup/pr-30478-pre-rebase
backup/reland-pr26134
backup/sglang-miles-muse-spark-open-20261019
bailing-v3-day0
batch-7-amd-ci-audit
bbuf/hunyuan-v3-hpcops-router-gemm
beam
beam-miles
beam-sm103-moe-gemv
beam-stack-03-srt
beam-stack-06-cutile-router
beam-stack-deps-dynamo
beam-stack-deps-tileiras
bench-openhands-b300-sweep
bench-openhands-gb300-sweep-beam
bench/e4-main-combined
bench/e4-stack-combined
bench/e5-main-combined
bench/e6-main-combined
bhe/grok2-piecewise
bhe/radix_linear_attn
bingxche/amd-reusable-stage
bingxche/ci-gfx1250
bingxche/fix-tbo-forward-metadata-proxy
bingxche/image2lmsysorg
bked-default
bot/bump-docs-version-0.5.17-31225095238-1
bot/bump-docs-version-0.5.19-33912440427-1
bot/bump-docs-version-0.5.20-35322363581-1
bot/bump-docs-version-0.5.22-38824883155-1
bot/bump-flashinfer-version-0.6.8-6f8d
bot/bump-kernel-version-0.3.21.post1-5f39
bot/bump-kernel-version-to-sglang-0.4.2-3936
bot/bump-kernel-version-to-sglang-0.4.2.post1-3eff
bot/bump-kernel-version-to-sglang-0.4.2.post2-c64d
bot/bump-kernel-version-to-sglang-0.4.6-3c47
bot/lmsys-blog-cards
bot/update-est-time-20260418
bot/update-est-time-20260420
bot/update-est-time-20260504
bot/update-est-time-20260511
bot/update-est-time-20260514
bot/update-est-time-20260601
bot/update-est-time-20260608
bot/update-est-time-20260615
bot/update-est-time-20260622
bot/update-est-time-20260629
bot/update-est-time-20260706
bot/update-est-time-20260713
bot/update-est-time-20260720
bot/update-est-time-20260727
bot/update-est-time-20260803
bot/update-est-time-20260810
bot/update-est-time-20260817
bot/update-est-time-20260824
bot/update-est-time-20260831
bot/update-est-time-20260914
bot/update-est-time-20260921
bot/update-est-time-20260928
bot/update-est-time-20261005
bot/update-est-time-20261012
bot/update-est-time-20261019
bot/weekly-update-est-time
brayden/cutedsl-fused-parallel-qknorm
brayden/deepgemm-torch-2.13
brayden/fix-cutedsl-mqa
brayden/sm120-stash
brayden/splitk-tgv-bf16
brayden/support-nemotron-eplb
build-commit-sgl-kernel
byron/cherrypick-23734-onto-miles
byron/compile-deepep-timeout-fix
byron/refactor-pcg-mm-2
byron/um-pd-layout
byron/um-pd-matching-dcp
byron/um-pd-unequal-tp
cctry-add-load-counters
ch-wan-patch-1
ch-wan-patch-2
ch-wan-patch-3
chang/tokenizer
cheng/ci/in-process-group-v2
cheng/config-namespace-reland
cheng/debug-mode-batch-scoped-fault
cheng/debug/dspark-low-acceptance
cheng/debug/inkling-fa4-validation
cheng/dev/attn-dp-size-compat
cheng/dev/dp-attention-prefill-cp
cheng/dev/pplx
cheng/dev/unified-fused-drafts-base
cheng/doc/hicache-storage-config-todo
cheng/ffn-checkpoint-compat
cheng/fix-one-batch-chunked-prefill
cheng/gc-unified
cheng/gpt-oss-sinks-dtype
cheng/hot-fix/atomic-request-state-init
cheng/hot-fix/cutlass-moe-checked-reduction
cheng/hot-fix/cutlass-moe-complete-maps
cheng/hot-fix/decoder-tail-global-counter
cheng/hot-fix/dsv4-silu-kernels-empty-rows
cheng/hot-fix/ep-dispatcher-follow-layer-runner-base
cheng/hot-fix/fp8-moe-published-routing-policy
cheng/hot-fix/gguf-inactive-route-scratch
cheng/hot-fix/hunyuan-sink-weight-layout
cheng/hot-fix/lora-alignment-input-domain
cheng/hot-fix/marlin-inactive-route-consumers
cheng/hot-fix/moe-lora-gated-virtual-projections
cheng/hot-fix/moe-temporary-workspace
cheng/hot-fix/native-fp8-reloads
cheng/hot-fix/sarvam-dense-mlp-reduction
cheng/hot-fix/stack-ending-attention-sliced-rows
cheng/hot-fix/unified-joint-reclaim
cheng/hot-fix/unified-kv-hip-fused-kv-write
cheng/hot-fix/unified-kv-int8-ckpt-fail-loud
cheng/hot-fix/unified-kv-tri-conv-slice-axis
cheng/hot-fix/unified-kv-tri-host-stack
cheng/hot-fix/unified-kv-tri-state-hicache-events
cheng/hot-fix/unified-mamba-compaction
cheng/kv-loc-debt
cheng/refactor/aiter-moe-runner
cheng/refactor/attention-input-from-boundary
cheng/refactor/attention-projection-parallel-groups
cheng/refactor/attention-sink-loader-layouts
cheng/refactor/attn-01-rename-static-buffers
cheng/refactor/attn-02-eager-init-buffers
cheng/refactor/attn-03-compute-metadata-extraction
cheng/refactor/attn-04-cpu-route-eager
cheng/refactor/attn-05-deprecate-use-bound
cheng/refactor/attn-finish-plan
cheng/refactor/attn-step05-residual
cheng/refactor/attn-step08
cheng/refactor/attn-step10
cheng/refactor/auxiliary-loader-frozen-layouts
cheng/refactor/boundary-per-edge
cheng/refactor/cache-chain-reads-own-placement
cheng/refactor/cp-configs-declared
cheng/refactor/cp-moves-binding
cheng/refactor/custom-projection-parallel-groups
cheng/refactor/cutlass-moe-route-preparation
cheng/refactor/dbrx-expert-loader-layout
cheng/refactor/declared-a2a-boundary
cheng/refactor/declared-dense-fully-dp
cheng/refactor/declared-dsa-cp
cheng/refactor/declared-input-scattered
cheng/refactor/declared-mhc
cheng/refactor/declared-mhc-input-scattered
cheng/refactor/declared-moe-boundary
cheng/refactor/declared-plain-tp-boundary
cheng/refactor/declared-row-queries
cheng/refactor/declared-sp-region
cheng/refactor/decouple-forward-mode
cheng/refactor/deepep-moe
cheng/refactor/deepseek-moe-output
cheng/refactor/deepseek-v4-lagged-layer-loop
cheng/refactor/deepseek-v4-layer-loop
cheng/refactor/deepseek-v4-moe-transport
cheng/refactor/deepseek-v4-stage-boundaries
cheng/refactor/dots-fused-gate-loader-layouts
cheng/refactor/dp-pad
cheng/refactor/dp-return-selection
cheng/refactor/dp-tbo-gather
cheng/refactor/dp-tp-group-helper
cheng/refactor/draft-scope-topology
cheng/refactor/drop-dead-context-helpers
cheng/refactor/encoder-keeps-its-attention-layout
cheng/refactor/ffn-exit-fusion-entries
cheng/refactor/ffn-input-completes-owed
cheng/refactor/final-norm-skip-empty
cheng/refactor/fope-coefficient-loader-layouts
cheng/refactor/forward-metadata-staleness
cheng/refactor/fp8-block-layout-defaults
cheng/refactor/gemma-shared-vocab-layouts
cheng/refactor/gguf-moe-vector-kernels
cheng/refactor/gpt-oss-expert-loader-layouts
cheng/refactor/gptq-act-order-layout-diagnostics
cheng/refactor/group-build-drops-model-config
cheng/refactor/inkling-linear-loader-layout
cheng/refactor/inkling-moe-loader-layout
cheng/refactor/inkling-stage-boundaries
cheng/refactor/is-draft-extend-v2-rename
cheng/refactor/kimi-attention-parallel-groups
cheng/refactor/kimi-mlp-parallel-groups
cheng/refactor/kv-calibration-loader-layouts
cheng/refactor/layer-boundary-completion
cheng/refactor/layer-stack-branch-connections
cheng/refactor/layer-stack-resolved-connections
cheng/refactor/linear-attention-parallel-groups
cheng/refactor/linear-constructor-parallel-groups
cheng/refactor/linear-owned-groups
cheng/refactor/llama-inkling-mlp-parallel-groups
cheng/refactor/lora-routing-prewarm
cheng/refactor/marlin-lora-launch-paths
cheng/refactor/mhc-by-kind
cheng/refactor/mhc-residual-ops
cheng/refactor/mimo-fused-qkv-loader-layouts
cheng/refactor/model-kv-calibration-layouts
cheng/refactor/moe-cleanup
cheng/refactor/moe-cp-gathered-rows
cheng/refactor/moe-placement-in-runner-config
cheng/refactor/move-expert-mask-to-dispatcher
cheng/refactor/nemotron-h-stage-boundaries
cheng/refactor/parallel-guard-coverage
cheng/refactor/quant-weight-layout-validation
cheng/refactor/quantized-reload-scale-layouts
cheng/refactor/qwen-vision-wrapper-parallel-groups
cheng/refactor/reject-attn-tp-mismatch-layers
cheng/refactor/remaining-vision-parallel-groups
cheng/refactor/repeated-batched-linear-parallel-groups
cheng/refactor/replicated-attention-linear-groups
cheng/refactor/req-len
cheng/refactor/retire-the-group-aliases
cheng/refactor/scheduler-reads-own-placement
cheng/refactor/sharded-loader-parallel-groups
cheng/refactor/stage-communicators
cheng/refactor/stage-reads-updates
cheng/refactor/unified-kv-addressing
cheng/refactor/unified-kv-arena-and-ids
cheng/refactor/unified-kv-chain-allocator
cheng/refactor/unified-kv-declare-plan
cheng/refactor/unified-kv-draft-views
cheng/refactor/unified-kv-drop-facades
cheng/refactor/unified-kv-layout-move
cheng/refactor/unified-kv-pins
cheng/refactor/unified-kv-pricing-and-budget
cheng/refactor/unified-kv-requests
cheng/refactor/unified-kv-router
cheng/refactor/unified-kv-safety-net
cheng/refactor/unified-kv-state-access
cheng/refactor/unified-kv-storage-consumers
cheng/refactor/unified-kv-transfer
cheng/refactor/unified-kv-tree-components
cheng/refactor/unified-kv-typed-views
cheng/refactor/unify-attn-backend-cuda-graph-capture-replay
cheng/refactor/vision-parallel-groups
cheng/refactor/zaya-cca-parallel-group
cheng/scheduler-debug-mode
cherry-pick/1fb85053e7-to-v0.5.19-21323390
cherry-pick/378aea1385-to-v0.5.16-f64b7684
cherry-pick/79ea30d1f-to-v0.5.12-4fac644b
cherry-pick/c7e2c08d14-to-v0.5.18-650e8374
ci-check-moe-topk-sigmoid
ci-light-import
ci-maintain
ci-monitor/bump-est-time-20260506-233123
ci-nvidia-smi-timeout
ci-probe-rate-limit
ci-uv-venv-migration-fixes
ci/add-5090-test-stage
ci/add-hybrid-dp-ep-tp-mtp-to-nightly
ci/amd-miles-mi30x-nightly
ci/amd-stage-b-shard-14to8
ci/auto-retry-scheduled-failures
ci/deepep-egg-purge
ci/detect-consolidated-pt-format-mismatch
ci/diffusion-b200-pip-fa4
ci/disable-gb200-stage
ci/dsv4-pd-disagg-aime25-b200
ci/dsv4-pd-disagg-nixl-b200
ci/e2e-memory-capacity-thresholds
ci/eagle-dp-attn-nan-test
ci/enable-deepgemm-warmup-8gpu-h200
ci/enable-weka-docker-cache-seeding
ci/failure-tracking-infra
ci/fix-blackwell-pip-race
ci/fix_gfx1250_attn
ci/flaky-test-reporting
ci/gsm8k-answer-extraction
ci/gt-gate-skip-non-images
ci/mi355x-spur-heredoc-backticks
ci/migrate-5090-tests-to-registered
ci/migrate-quant-tests
ci/nightly-1-gpu-5090-job
ci/nightly-traces-s3
ci/nvidia-lib-selfheal
ci/pr-30762-umbp-e2e
ci/precache-hf-models
ci/precision-test-dispatch
ci/shm-teardown-purge
ci/verify-mi355x-disagg-43944-43945-43977-43985
ci/weight-checksum-validation
claude/drop-qwen-new-tag-1a6681
claude/git-bisect-debugging-cmsljv
claude/lint-hook-python39
claude/paddleocr-support-optimization-bfaf52
claude/vlm-warmup-image-size
cleanup-5090-gpu-memory
cleanup-cutlass-fp8-supported
clearing_console_output
codex-group-norm-silu-artifacts
codex/amd-nixl-ucx-autoselect-exp-20260916
codex/amd-rocm720-nightly-mi300-1gpu
codex/archive-miles-admission-source-20260905
codex/backfill-image-formats
codex/bcg-output-management
codex/cfg-vae-decode-group
codex/checkpoint-quant-metadata-no-reflection
codex/clarify-gpu-residency-log
codex/communicator-cp-ownership
codex/component-quantized-repos
codex/component-residency-strategy-compat
codex/cosmos3-lossless-perf-20260602
codex/cosmos3-video-gt-test-20260602
codex/cpu-offload-components-clean
codex/diagnose-wan-sdpa-drift
codex/diffusion-5090-perf-pool
codex/diffusion-cfg-text-batch-20260523
codex/diffusion-ci-gt-sync-files
codex/diffusion-day0-variants
codex/diffusion-platform-perf-baselines
codex/diffusion-preload-residency-planner
codex/diffusion-probe-fixes-20260621
codex/diffusion-profile-stage-names-20260523
codex/diffusion-stage-deployment
codex/direct-gpu-zero-copy
codex/disable-mi355x-di-ci
codex/dit-static-weight-mappings
codex/draft/glm5-3-datafree-fp4fp8
codex/draft/glm5-3-datafree-runtime
codex/dsv4-aime25-hard-ci
codex/dsv41-dpa-megamoe
codex/dsv41-indexer-plan
codex/engine-sampling-defaults-chat
codex/epd-grpc-registration-failure
codex/epd-health-probe-drain
codex/epd-preflight-request-validation
codex/epd-receiver-startup-consensus
codex/epd-receiver-timeout-cleanup
codex/epd-scheduler-cancel-cleanup
codex/explicit-native-only-components
codex/fix-ascii-decode-prompt
codex/fix-cu134-deepgemm-elfutils
codex/fix-deepgemm-batch-invariance
codex/fix-dense-prefill-cp
codex/fix-diffusion-final-failure-summary
codex/fix-diffusion-gt-generation
codex/fix-fa3-varlen-out
codex/fix-glm53-dsv4-pool-test
codex/fix-glm53-eagle-kpool-padding
codex/fix-hybrid-kv-release-order
codex/fix-hybrid-linear-eagle-padding
codex/fix-kda-cutedsl-rocm
codex/fix-pd-abort-terminal-semantics
codex/fix-stale-cargo-workspace-cache
codex/flashmla-mega-build-apis
codex/flux2-import-ci-main-1023
codex/forbid-distributed-transformer-fallback-v2
codex/generic-prequant-checkpoint-admission
codex/glm4-lite-shared-expert-tp-flag
codex/glm47flash-fa3-out-guard
codex/glm47flash-topk-shared-jit
codex/glm5-3-flash-h20-pd-mtp
codex/glm5-3-flash-megamoe
codex/glm5-3-flash-performance-optimization
codex/glm5-3-flash-sota-20260905
codex/glm5-3-w4a8-community
codex/glm5.3-ep_main
codex/glm53-cp-decode-attn-tp
codex/glm53-flash-dcp-base
codex/glm53-flash-dcp-mtp
codex/glm53-flash-main-baseline-20260907
codex/glm53-flash-optimize-20260908
codex/glm53-ingraph-verify-metadata
codex/glm53-kpool-deterministic
codex/glm53-lora-dense-tp
codex/glm53-verify-deepgemm-out-of-graph
codex/gptoss-bf16-flashinfer-support
codex/h3-tp-boundary-replication
codex/h3-tp-ulysses-ipc
codex/hunyuan3d-native-image-encoders
codex/ideogram4-nvfp4-ckpt
codex/ideogram4-pr27279-gt-e8dfe898ad
codex/image-benchmark-seeded-rng
codex/ipc-pool-handle-cache-default
codex/kimi-grid-metadata-cpu
codex/kimi-k27-dp-encoder
codex/kimi-k27-vlm-perf-integration
codex/kimi-moonvit-fused-rope
codex/klein-first-request-diagnosis
codex/lingbot-camera-mod-cache
codex/lingbot-full-runtime-session-refactor
codex/ltx-hq-ci-gt-refresh
codex/ltx-official-gt-workflow-20260506
codex/ltx2-from-862f-clean-main
codex/ltx23-lossless-perturb-mask-20260505
codex/ltx23-origin-main-20pct-20260505
codex/main-diffusion-test-suite
codex/mamba-track-interval-1024-20260523
codex/minimax-h3-adaln-cache
codex/minimax-h3-cross-node-ring
codex/mistral3-native-attention-20260614
codex/mlx-shutdown-ingestion-fix
codex/mm-broadcast-error-envelope
codex/mm-consumer-reconstruct-rollback
codex/mm-owner-projector
codex/mmgen-cfg-text-batch-20260521
codex/modelopt-fp8-platform-dtype-test
codex/mopd-per-position-scoring
codex/muse-cookbook-unified-memory
codex/muse-spark-launch-defaults
codex/muse-spark-stack-04-amd
codex/muse-spark-stack-05-helion-verify
codex/nccl-tms-fix
codex/neuron-overlap-sequence-relay-20260905
codex/nightly-diffusion-token-backfill-20260918
codex/nightly-flux3-20261016
codex/nixl-storage-test-isolation-1024
codex/omni-u1-orchestrator
codex/pi05-native-siglip
codex/port-pr-24082-specv2-disagg-grammar
codex/post-response-tasks-20260503
codex/pr-21247-thresholds-20260502
codex/pr-37495-validation-20260902
codex/pr-title-model-type
codex/pr/37588-kpool-request-domains
codex/pr/37588-whole-group-refresh
codex/pr28050-artifacts
codex/pr2874-nan-probe
codex/pr37587-rebased-20260907
codex/pr37587-validation-20260907
codex/pr38632-workflow-reference-cleanup
codex/pr43182-bcg-move-proof
codex/pr43774-cp-extraction-proof
codex/pr44720-amd-nightly-gfx1151-2c2986b
codex/pr44720-amd-nightly-main-7c99ace
codex/pr44720-amd-nightly-miles-7b6993a
codex/pr44720-amd-nightly-release-miles-7c99ace
codex/pr44720-amd-nightly-validation-20261016
codex/pr45498-amd-build-only
codex/qwen-qsa-prefill-metadata
codex/qwen-qsa-prefill-ragged
codex/qwen35-profile-8k1k-20260823
codex/qwen35-profile-exact-bs32-20260823
codex/rebase-sglang-27421
codex/reject-diffusers-component-attention-v2
codex/reject-unmanaged-component-residency
codex/rocm-network-rerun
codex/rocm10-copy-weight-views-slow-jobs
codex/rocm10-moe-copy-ab
codex/rocm10-rc4-aiter-a6d2b564
codex/rocm10-release-once-cc5f490b
codex/rocm10-rocm720-daily-ci
codex/rocm1000-ga-aiter-a6d2b564
codex/rocm7-tests-every-two-days
codex/router-dynamo-rendering
codex/router-engine-abort
codex/router-hf-hub-1
codex/router-pd-failfast-groups
codex/router-power-of-n
codex/router-scoring
codex/rust-frontend-compatibility-20260915
codex/sana-wm-consistency-realtime-common
codex/sana-wm-native-refiner
codex/sglang-radix-sim-core
codex/sgld-bench-case-fixes
codex/skip-mi355x-4n-unavailable-nodes
codex/sm120-diffusion-attn
codex/split-32714-gpt-oss-flashinfer
codex/srt-dynamic-cudnn-sdpa
codex/strict-component-attention-overrides
codex/swa-memory-budget-api
codex/test-mi35x-runners
codex/test-rocm720-mi300-pull-only
codex/tighten-nv-diffusion-perf-baselines
codex/tmp-glm53-layer-split-red
codex/torch211-official-gt-gen-20260502
codex/trainium3-chunked-admission-20260905
codex/ug-bagel-interleave-runtime
codex/ug-g-runtime-proof
codex/upstream-sglang-miles-latest-passed-main
codex/vae-channels-last-3d-followup
codex/vae-channels-last-true
codex/vae-tiling-and-nvfp4-gate
codex/validation/37588-kpool-request-domains
codex/vlm-preprocessed-01-core
codex/vlm-preprocessed-03-tests
codex/vlm-preprocessed-06-qwen-copies
codex/vlm-preprocessed-07-qwen-offsets
codex/vlm-preprocessed-10-pretokenized
codex/vlm-preprocessed-11-input-list
codex/vlm-preprocessed-12-shm
codex/vlm-preprocessed-13-grpc-dispatch
codex/vlm-preprocessed-input
codex/vlm-tc-prefill-graph-replay
codex/vmm-request-error-consensus
codex/wan22-bisect-9d8426870
combine/router-admission-http2-loadaware
config-lifecycle-tightening
config/resolution-policy-stages
consolidate-ci-weight-validation
cookbook-migrate-gemma4
cookbook-migrate-glm-5.1
cookbook-migrate-kimi-k2.6
cookbook-migrate-minimax-m2.7
cookbook-migrate-nemotron3-ultra
cookbook-migrate-qwen3.5
cookbook-mimo-v2.6-b300-minimal
cookbook-skill-wave1-review-fixes
cookbook/dsv4-pro-0813-dspark
copilot/add-decode-kv-cache-offload
copilot/add-epd-disaggregation-architecture
copilot/add-new-feature
copilot/add-pipeline-parallelism-sglang
copilot/add-python-decorator-introduction
copilot/add-sglang-diffusion-support
copilot/add-specforge-training-framework
copilot/check-stars-for-issue-21846
copilot/chunked-prefill-mask-attention
copilot/concurrent-programming-interview-questions
copilot/create-presentation-script
copilot/derive-softmax-gradient
copilot/discussion-on-chunk-prefill
copilot/enable-free-thread-support
copilot/enable-speculative-decoding-small-batchsize
copilot/ep-walk-tp-communication
copilot/explain-chunk-prefill-process
copilot/explain-code-functionality
copilot/explain-conn-py-code
copilot/explain-conn-py-details
copilot/explain-decode-schedule-batch-mixin
copilot/explain-inference-parameters
copilot/explain-kv-events-code
copilot/explain-prefill-py-code
copilot/explain-sglang-chunk-prefill
copilot/fix-base-a-test-cpu-failure
copilot/fix-build-and-test-job
copilot/fix-ci-errors
copilot/fix-gloo-peer-rank-connection-issues
copilot/fix-lint-ci-error
copilot/fix-lint-job-failure
copilot/fix-lint-job-failure-again
copilot/generate-ppt-script
copilot/how-role-system-and-user-combine
copilot/implement-reduce-topk-gemm-permute
copilot/install-cu13-via-pypi
copilot/load-model-process
copilot/mem-cache-block-verification
copilot/mem-fraction-static
copilot/quantization-discussion-weight
copilot/research-model-loading-sglang
copilot/review-pr-21661
copilot/set-torch-cuda-arch-list-9-0
copilot/sglang-pd-separation
copilot/sub-pr-12948
copilot/support-kimi25-epd-encoder-only
copilot/track-flaky-ci-tests
copilot/understanding-flow-matching-models
copilot/update-make-layers-function
cublas_qiaolin
cuda-graph-lifetime-fixes
cuda-graph-lifetime-tool
cuda13_memcpy_hicache_restore
cursor/amd-agentic-perf-glm52-fp4-mtp-7284
cursor/amd-agentic-perf-kimi-k3-mxfp4-42c0
cursor/amd-agentic-perf-qwen35-mxfp4-mtp-8d29
cursor/amd-ci-cache-audit-tool-eb82
cursor/amd-ci-coverage-canvas-62ce
cursor/amd-ci-resolve-image-tag-with-ls-remote-868a
cursor/amd-ci-use-system-triton-f7b1
cursor/amd-diffusion-ipc-a2a-tests-cuda-only-8e90
cursor/amd-diffusion-mock-cuda-platform-not-rocm-8e90
cursor/amd-drop-tc-piecewise-split-op-tests-51db
cursor/amd-dsv4-compact-attention-inline-qkpv-8e90
cursor/amd-kandinsky6-fa-dispatch-skip-rocm-8e90
cursor/amd-kandinsky6-layerwise-offload-deterministic-8e90
cursor/amd-kda-decode-parity-cuda-only-8e90
cursor/amd-kda-fused-verify-qk-norm-reciprocal-8e90
cursor/amd-nightly-dsv4-flash-0731-drop-v3x-ef39
cursor/amd-quark-nvfp4-mxfp4-gsm8k-threshold-8e90
cursor/amd-rocm720-nightly-runner-fallback-6ad2
cursor/amd-umbp-hicache-standalone-ssd-rebase-8e90
cursor/bazel-migration-foundation-b440
cursor/cache-amd-ci-docker-image-f124
cursor/check-readme-file-composer-1-a466
cursor/ci-experiment-workflow-9019
cursor/clip-score-log-visibility-d45f
cursor/cookbook-dsv4-amd-fp8-experts-env-d201
cursor/dispatch-pr-25301-moriep-oom-864f
cursor/drop-miles-rocm7-image-40cc
cursor/encoder-dp-amd-debug
cursor/encoder-dp-amd-fix
cursor/experiment-mi325-parallel-1319-9491
cursor/fix-amd-ci-log-formatter-prefix-5d53
cursor/fix-amd-fused-moe-single-expert-reduce-4efe
cursor/fix-ci-failure-monitor-silent-401-fc6e
cursor/fix-ci-suite-runner-exit-hang-a882
cursor/fix-dsa-idle-spec-metadata-30d6
cursor/fix-fused-moe-unwritten-intermediate-a37e
cursor/fix-hellaswag-benchmark-source-bac9
cursor/fix-rocm-image-hipify-byproducts-b101
cursor/fix-vattn-asm-rocm10-hip-runtime-990a
cursor/flag-stale-miles-rocm-image-3d8b
cursor/gateway-crate-universe-64ce
cursor/glm52-mi355x-mxfp4-agentx-hicache-00a7
cursor/glm53-flash-mi300x-validation-748f
cursor/gpu-kv-cache-management-4941
cursor/kv-cache-management-891c
cursor/logprobs-assertion-failure-37d6
cursor/miles-nightly-suite-rename-test-44cb
cursor/move-test-moriep-small-to-mi35x-864f
cursor/never-prompt-for-git-credentials-155b
cursor/nv-downsampling-ratio-behavior-d5e9
cursor/pr-states-amd-rocm-ed27
cursor/print-k8s-node-name-4515
cursor/qwen35-moe-text-shared-expert-nightly-09ab
cursor/qwen38-flash-next-cookbook-fp8-cells-fe21
cursor/reclaim-and-seed-mi35x-hf-cache-7284
cursor/register-amd-ci-unit-tests-8f21
cursor/remove-deleted-amd-checkpoints-0857
cursor/restore-amd-ci-registrations-3896
cursor/retire-legacy-gsm8k-mmlu-amd-ci-6d7a
cursor/revert-42541-rope-fp32-cache-77ed
cursor/rocm-glm53-flash-hip-fallbacks-748f
cursor/rocm-mooncake-go-fetch-retry-27b9
cursor/rocm724-probe-ci-dispatch-dba6
cursor/setup-dev-environment-5af2
cursor/sglang-core-logic-3e29
cursor/sglang-pr-18938-review-6355
cursor/sparse-attention-kernel-tests-3e52
cursor/sparse-attention-kernel-tests-888b
cursor/test-optb-on-hardware-dba6
cursor/test-qwen35-mi355x-recipes-7487
cursor/update-rocm-nightly-ci-skill-cb60
cursor/verify-rocm720-async-ckpt-3d8b
curvezmq-shared-key
dbgA-kimi26-bf16kv
dbgB-kimi26-nospec
dcp-dsa
dcp-hicache-l3
dcp-l3-mooncake
dcw02/dflash-bounded-swa-kv-cache
dcw02/dflash-grammar-support
dcw02/dflash-penalty-fix
dcw02/draft-kv-unified-memory-pool
dcw02/fix-mamba-radix-page-align
dcw02/kimi-k3-dflash
dcw02/kimi-k3-dflash-optimization
dcw02/spec-mamba-logic
dcw02/war-barrier-worker-capability-perf
debug-qwen2vl-context-length-test
deepseek_v4
deepseek_v4_compressor
deepseek_v4_dev
deepseek_v4_mxfp4_shuffle_perf
demo/single-gpu-multi-node-serving
deprecate-get-server-info-endpoint
dev-19673
dev/double-sparsity-standalone
dev/pr33561-docker-nccl-fix
dev/rust-server
dev/saienduri/test-ci
device-pool-info-dsa
dflash2-pin-1cf2b8c
dflash2-pin-1cf2b8c-nccl
diag/nixl-run-choice-first10-20260823
diffuion-refactor-denoising
diffusion
diffusion-ci-cu13
diffusion-ci-with-qwen2509
diffusion-component-residency-manager
diffusion-first-class-routes
diffusion-fix-component-resident-manager
diffusion-gemma4-support
diffusion-general
diffusion-norm-fusion-for-zimage
diffusion-perf
diffusion-refactor-arch-config
diffusion-rl-base
diffusion-test
diffusion-tmp
docker/drop-gnuplot
docker/fix-nightly-version-stamp
docs/add-glm5-usage-guide
docs/beam-serving-cookbook
docs/dspark-guide
docs/dsv4-flash-0731-cookbook
docs/glm-5.3-flash-hopper-dsa-auto
docs/playground-spec-strip-draft-path
dsa-indexer-buffer-info
dsa-indexer-state-desc
dsa-static-page-table-width
dsv4-buffer-info
dsv4-dev
dsv4-mxfp4-moe-backend
dsv4-rebase
dsv4.1-backup-before-entrypoint-fix-20260915
dsv4.1-backup-before-linear-rebase-20260915
dsv4.1-backup-before-ordinary-update-20260915
dsv4.1-backup-before-test-replay-20260915
dsv4.1-engram
dsv41-engram-fusion
dsv41-flashmla-drop-sgl-kernel
dsv41-tp8-hopper-decode
dsv4_pp_doc
duyi-dev-260610
dynamic_tree_verify
emmett/ci-5090-base-b-light
emmett/ci-health-check-api-retries
emmett/ci-shared-host-runner-isolation
emmett/diffusion-image-preprocessing-unit-suite
emmett/remove-4-gpu-gb300-suite
enable-sm90-pool-decode
enable-uv-venv-deepgemm-cache
enable_cache_dit_in_test
eos
feat/amd-deterministic-kernel
feat/async_r3
feat/deepep-shared-expert-fusion
feat/dev_20251201
feat/diffusion-placement-group-disagg
feat/fused-weight-host-spill
feat/grpc-health
feat/ideogram4-fp8-native
feat/ling3-fp8-decode-kernels
feat/ling3-mamba-fork-split
feat/ling3-pub-parser
feat/ltx-2.5-parallel-diffusion-decode
feat/ltx-two-stage
feat/nvfp4-online-dense-linear
feat/planner-unified-memory
feat/readonly-checkpoint-mappings
feat/rerun-stage-fork-support
feat/rerun-stage-path-filter
feat/residency-apply-validate
feat/residency-planner-solver
feat/router-api-profile
feat/router-responses-api
feat/sgl-router-caller-input-ids
feat/sgl-router-chat-render-parity
feat/sgl-router-m0-scaffold
feat/sgl-router-render-parity
feat/shared-pool-planner
feat/sm120-dsa
feat/stateless_scheduler_b
feat/test-summary-job-level
feat/unified-metrics-s3
feat/warmup-prefill-tokens
feature/disable_generation_healthcheck
feature/eagle3
feature/unified-memory-flashinfer-decode
fix-19320-review
fix-amd-ci-install-quoting
fix-async-d2h-worker-context
fix-bench-server-large-bs
fix-bot-cherry-pick-push-pat
fix-corruputed-weights
fix-cp-nsa-prefill-context-parallel
fix-dcp-fa3-backend
fix-deepgemm-cache-key
fix-diffusion-ci-steps
fix-dllm-prefill-stats-none
fix-explicit-chat-template-override
fix-extra-buffer-nonmamba-validation
fix-flashinfer-fusion-preflight
fix-flattened-bucket
fix-fsdp-amd
fix-host-mha-fp4-guard
fix-jit-gb200
fix-kda-kernel-args
fix-llama-4-nightly
fix-lora-batch-validation-none-count
fix-lora-eviction-test-flakiness
fix-mcp-validation
fix-mistral-eagle-use-dsa
fix-native-mla-budget
fix-nvfp4-dispatch-dummy-weights
fix-nvfp4-sm120-oom
fix-qwen3-next-gdn
fix-run-with-timeout-exception-handling
fix-streaming-parallel-auto
fix-svd
fix-swa-index-mapping-leak
fix-unified-loadback-multipin-nightly0820
fix/add-job-timeout-pr-test
fix/amd-ci-perf-bounds-and-dispatcher-test
fix/amd-mmgen-unit-suite
fix/auto-format-main-branch
fix/bailing-moe-fp8-kv-b-proj-weight-transpose
fix/bench-mmmu-multi-image
fix/bench-serving-gsp-args
fix/ci-cleanup-stale-ipc
fix/ci-env-source-cleanup
fix/ci-port-offset-for-network-host
fix/ci-runner-stability-fixes
fix/ci-stale-sglang-cleanup-v2
fix/cudnn-import-order
fix/disable-jit-deepgemm-in-ci
fix/disable-transformers-test-amd
fix/disagg-bootstrap-port-conflict
fix/disagg-bootstrap-port-conflict-v2
fix/disagg-mori-abort-notification
fix/draft_weight_checker
fix/eagle-dp-attn-large-crash
fix/eagle-filter-merge-accept-length
fix/eagle3-mla-attn-backend
fix/empty-logprobs-streaming
fix/eplb-dp-attention-token-mask
fix/flashinfer-allreduce-fusion-probe
fix/flashinfer-autotune-symm-mem-deadlock
fix/glm51-fp8-tp8-dp8-b200-oom
fix/glm52-qk-rope-head-dim
fix/gpu-cleanup-between-tests
fix/grammar-backend-disabled-message
fix/gsp-from-args-missing-attr
fix/health-check-disagg-idle
fix/hf-runner-lora-adapter-cleanup
fix/hicache-3fs-batch-set-postprocess
fix/hicache-dsv4-none-crash
fix/lora-update-test-flaky
fix/megamoe-shared-fusion-quant
fix/megamoe-waterfill
fix/mem-fraction-static-eagle-draft-kv-cache-wip-snapshot-2026-05-08
fix/mi35x-deepseek-v32-regression
fix/mi35x-remove-data2-hardcoded-paths
fix/mini-lb-flush-cache-forward-timeout
fix/modelopt-auto-map-resolution
fix/modelopt-offline-quantized-detection
fix/mooncake-nsa-l3-index-buffer
fix/mtp_moe_capture
fix/multimem-ag-skip-node-probe-single-node
fix/nvila-siglip-vision-model-prefix
fix/parallel-state-uncalled-tests
fix/pause-generation-writer-lock-deadlock-main
fix/pdl-cuda-graph-capture
fix/port-alloc-collisions
fix/qwen-image-edit-warmup-prompt
fix/qwen35-mori-expert-routing
fix/qwen35-pp-mtp-prefill-validation
fix/remove-dead-prewarm-mhc-token-count-buckets
fix/request-log-watermark-gate
fix/rerun-failed-ci-cancelled-workflows
fix/residency-calibration-probes
fix/running-req-piling-logprob
fix/rust-server-hf-tokenizer-cache
fix/sync-patch-logprob-fix
fix/test-srt-backend-smaller-model
fix/tilelang-tvm-ffi-isinstance-compat
fix/tokenizer-recv-fairness
fix/torch-cuda-version-mismatch
fix/transformers-v5-easy-fixes
fix/unified-cache-storage-recovery
fix/verify-mask-full-bound
fix/wrapper-translator-test-fixtures
fix_benchmark
fix_diffusion_1gpu_ci
fix_for_cmb
fix_glm_test
fix_grouped_topk_warp_shfl
fix_hicache_with_spec
fix_mamba_leak
fix_nightly
fix_nightly_test
fix_npu_diffusion_failure_analyzer
fix_pd_warmup_route_dp
fix_qwen_mtp_eplb
fix_v1_loads
fix_wan_channels_last_3d
flashinfer-megamoe-autotune-decode-workspace
flashinfer-swa-ragged-window
flux-klein-qwen3-mask-threshold-rebase
fp8-blockwise-sm120-tile-dispatch
fused-qk-rmsnorm-rope-gate
fusion/attn_prep_norm_mrope_cache_fuse
fusion/fuse_qk_norm_rope_cache
fusion/fused_qk_norm_mrope_cache_attn_prep
fusion/moe_topk_gating_fused
fusion/qk_norm_rope_fuse
fzyzcjy-patch-1
fzyzcjy-patch-2
fzyzcjy-patch-3
fzyzcjy-patch-4
gemma4-triton-attn-tuning
github_workflow_yaml
glm
glm-5.3-flash-opt
glm5-gb-image
glm5.2-skip-layer-indexer-weight-check-main
glm53-reasoning-parser-auto-default
glm_interleave
grpc-phase1/python-bridge
grpc-rerank-serialization
gt-consistency-check
hanming/debug-buffer-kl
hanming/eagle-beta-sync
hanming/extra-buffer-opt
hanming/extra-buffer-opt-backup
hanming/fix-extra-buffer-debug
hanming/gdn-pc-lps-decode
hanming/hicache-tombstone-backup
hanming/out-cache-loc-swa
hanming/prefill-delayer-all-branch-timeout
hanming/sdv2-fix-output
hanming/spec-v2-kl
hanming/unified-hicache-opt
hccl_automatic_sizing
healthy_worker_count
hicache-decl-b
hicache-decl-d
hicache-fast-file-backend
hicache/backup-backlog-shedding
hicache/backup-drop-visibility
hicache/bounded-storage-calls
hicache/flush-wedged-storage
hicache/recoverable-storage-sync
hicache/timeout-policy-rank-safe
hisparse-fix
hisparse-prefetch-indexshare
hisparse-prefetch-indexshare-io-only
hisparse_quest
idhanani/cherry-pick-24234-v0.5.11
idhanani/disagg-decode-radix-cache
idhanani/dyn-1986-tier-aware-pin
idhanani/dyn-optional-deps
idhanani/dyn-swa-gsm8k-chat
idhanani/fix-26974-transfer-timing
idhanani/fix-26976-chunked-tier-split
idhanani/forward-pass-metrics
idhanani/g2plus-main-review-20260524
idhanani/kvcr-direct-linker
idhanani/llm-83-sglang-emit-kv-events-for-l1l2-tier-transitions
idhanani/mx-mixed-tp-v1
idhanani/session-hicache
idhanani/sgl-bootstrap-room-u64
idhanani/sgl-hicache-server-info
idhanani/sgl-native-http-lifecycle
idhanani/sgl-native-pd-hidden-states
idhanani/sgl-native-pd-logprobs
idhanani/sgl-native-pd-sampling
idhanani/sgl-native-rejected-beam-lifecycle
idhanani/sgl-native-session-fencing
idhanani/sgl-native-session-routing
idhanani/sgl-opaque-unary-lifecycle-poc
idhanani/sglang-27574-hint-manager-rebuild
idhanani/sglang-27574-session-cache-rpcs
idhanani/sglang-27574-session-priority-admission
idhanani/sglang-conditional-aggregation-main
idhanani/sglang-omitted-output-limit
idhanani/sglang-storage-hints
ifmn/eagle-dp-attn
ifmn/npu/afd
ifmn/npu/glm-5-optim
incomplete-tokenizer-cache-fallback
indexer-topk-shape
inkling-support
int8_diffusion_online_quantization
intel_dev
ishan/add-dg-logging
ishan/dnm-fi
ishan/fix-dp
ishan/fix-minimax25-deepep
ishan/fix-nixl-hicache
ishan/radix-cache-review-fixes
ishan/retention-duration-priority
ishan/some-good-info
ishan/subagent-eviction-patterns
jh-ut-upstream-3
jiaguo/restore-cu129-matrix
jiajun/anthropic-error-sse-helpers
jiajun/chat-session-return-indexer-topk
jiajun/dsv41-bind-next-plain-ref
jiajun/dsv41-h200-rl
jiajun/filter-nonpersistent-weight-buffers
jiajun/fix-inkling-rocm-import
jiajun/indexer-topk-num-layers
jiajun/kimi-k2-raw-tool-call-id
jiajun/mimo-v2-swa-window-size
jiajun/mooncake-output-store
jialin/sync-hicache-buffer-regressions
jialin/sync-prefill-busy-accounting
jialin/sync-singleton-pd-polls
jialino/rust-tree-core-full
jiminator/ple-shared-host-tables
jit-launch-on-torch-current-stream
joesun/diffusion-unify-common-env-func
k3-kvv-conformance
k3-stream-fork-order
kan/integrate-dynamo-mm
kan/kimi-k3-mm-preprocessor
kan/processor-fixtures-hf
kan/rust-input-ids-store
kan/rust-mm-mimo
kan/rust-server-hf-cache-fix
kan/rust-server-pd
kan/rust-tm-dev
kan/rust-tm-rust-mm
kan/rust-tm-ttft-breakdown
kan/server-info-openai-env
kan/sgl-router-adopt-engine-defaults
kan/sgl-router-benchmark
kan/sgl-router-kwargs-no-forward
kan/sgl-router-pd-version-group
kan/sglang-mm-on-rust-server
kan/unify_shmem
kernel-register-amd-batch5
kernel-register-amd-ltx2
kernel/dual_fused_norm
kimi-k3-deferred-release-35049-35360
kimi-k3-hybrid-pool-import-fix
kimi-k3-zmq-max-sockets-mitigation
kimi-k3-zmq-socket-limit-fix
kimi-k3-zmq-socket-limit-fix-main
kp/DEBUG-MLA-prefill-cp
kp/add-memory-profile-instruments-flashinfer
kp/distributed-arg-max-greedy-spec
kp/fix-MHA-cp-prefix-len-calculation
kp/fix-dpsk-v32-tool-call
kp/gemma4-qkj-proj-refactor
kp/moe-finalize-fuse-with-ar-norm
kp/optimize-pack-fp8-quantize
kp/refactor-dcp-attn-comm-group
kp/rerun-ut-write-perm
kp/streaming-reasoning-parser-fix
kpham/dcp-pool-page-from-allocator
kpham/dcp-triton-extend-rank-uniform-prefix
kpham/draft-expert-location-gate
kpham/dsa-dcp-extend-gather-kv
kpham/dsa-dcp-hicache
kpham/dsv41-encoder-replay-bcg
kpham/fix-dsa-kpool-niah
kpham/fix-fake-receiver-ensure-abort
kpham/flashinfer-det-chunk-align
kpham/pd-indexshare-unfused-topk
kpham/port-36507-loading-quant
kpham/port-36507-serving-lifecycle
kpham/port-36507-spec-decode-perf
kpham/pr44623-main-merge
kpham/pr44623-tailgraph-off
kpham/speculative-dcp-size
kpham/tokenspeed-mla-nope-prefill
kv-cache-logging-temp
lianmin/dispatch-output-combine-input
lianmin/external-model-capabilities
lianmin/output-processor
lianmin/rust-frontend-extension-hooks
linyu/auto-tune
logging-worker-sglang-rotuer
lora-ci-diag-tp-overlap-cand
lora-ci-diag-tp-overlap-main
lora-ci-exp-hicache-qwen3
lora-ci-exp-update-concurrency
lora-ci-groups
lora-perf-optimize-2
lora-stack-00-shared-hooks
lora-stack-01-route-primitives
lora-stack-02-dense-kernels
lora-stack-03-dense-plans
lora-stack-04-dense-tables
lora-stack-05-dense-runner
lora-stack-06-dense-graphs
lora-stack-07-mla-correction
lora-stack-08-inkling-sink
lora-stack-09-moe-plans
lora-stack-10-moe-routing
lora-stack-11-moe-tables
lora-stack-12-moe-stages
lora-stack-13-moe-providers
lora-stack-14-moe-runner
lora-stack-15-cutedsl-schedule
lora-stack-16-cutedsl-bf16
lora-stack-17-triton-fp8
lora-stack-18-cutedsl-fp8
lora-stack-19-fp8-tables
lora-stack-20-nvfp4
lsyin/attn-backend-stamp
lsyin/bcg-dsv41-candidate-table-capture-assert
lsyin/bench-engine-steps
lsyin/concurrency-limit-attribution
lsyin/custom-ar-v2-float-only
lsyin/dedup-head-slice
lsyin/deep-gemm-post1
lsyin/draft-extend-input-staging
lsyin/ds-recipe-backend
lsyin/ds-recipe-rust-parse
lsyin/dsa-ragged-verify-graph
lsyin/dspark-compact-dsa-syncfree
lsyin/dspark-eager-draft-embed
lsyin/dspark-run-draft-proposal
lsyin/dspark-simulated-bonus-bf16
lsyin/dsv4-seq-lens-tests
lsyin/dsv41-captured-full-topk-per-row
lsyin/dsv41-prefill-graph-capture-stub
lsyin/dsv41-sm90-prefill-graph-tail
lsyin/eagle-draft-dtype-auto
lsyin/eagle-verify-fence
lsyin/engram-a2a-decode
lsyin/expand-ruff-pyflakes
lsyin/gc-debug-dev
lsyin/hybrid-allocator-sides
lsyin/hybrid-attn-sync-gate
lsyin/killall-sh-pattern-sync
lsyin/megamoe-chunk-cap
lsyin/page-tail-write
lsyin/plan-stream-poc
lsyin/ragged-layout-access
lsyin/re-introduce-tp-grammar
lsyin/rerun-test-merged-pr
lsyin/sched-pr01
lsyin/sched-pr02
lsyin/sched-pr03
lsyin/sched-pr04
lsyin/sched-pr05
lsyin/sched-pr06
lsyin/sched-pr07
lsyin/sched-pr08
lsyin/sched-pr09
lsyin/sched-pr10
lsyin/sched-pr11
lsyin/sched-pr12
lsyin/sched-pr13
lsyin/sched-pr14
lsyin/sched-pr15
lsyin/sched-pr16
lsyin/sched-pr17
lsyin/sched-pr18
lsyin/sched-rewrite
lsyin/spec-width-optional
lsyin/stash-dedup-fence
lsyin/swa-sizing-policy
lsyin/tree-mask-protocol-cleanup
lsyin/unify-draft-runner
lsyin/urc-test-long
lsyin/verify-extend-unify
lsyin/verify-metadata-in-graph
lsyin/verify-tier-resolve
ltx-2.3-two-stage-ti2v
ltx-2.3-two-stage-ti2v-clean
ltx-batched-unify-all-paths
ltx23-official-lora-merge-align
ltx23-tp-psnr20-latest-20260508
ltx23-transition-lora-support
lufang/independent-cache-window-metadata
lzy17/amd-ci-pd-keep-dmabuf
m3-mxfp8-dense-block-convert
m31-hopper-dspark
m31-nightly-e2e
main
main-ci-test
main-dev
measure-prescheduler-overhead
michael/amd-aot-kernel-cache
michael/amd-ci-openai-2.8
michael/amd-diffusion-retry-deadline
michael/amd-disagg-hip-dmabuf
michael/amd-disagg-unfused-kv
michael/amd-docker-build-only
michael/amd-gsm8k-skip-mistral-fp8
michael/amd-multimodal-max-parallel
michael/amd-pr-test-extra-tier
michael/amd-pr-test-nightly-large-models
michael/amd-pr-test-parallel-shards
michael/amd-pr-test-skip-and-tags
michael/amd-register-portable-unit-tests
michael/amd-skip-ltx25-decoder
michael/amd-vae-blend-test-reference
michael/cosmos-dreams-decode-queue-drain
michael/crash-dump-expect-exit
michael/diag-gsm8k-mistral
michael/diag-gsm8k-mixtral-r4
michael/dsv4-bcg-swa-ring-mock
michael/dsv4-hip-bcg-tests
michael/dsv4-v41-k-shfl-mask
michael/dsv41-bcg-hip-graph-width
michael/flux3-determinism-probe
michael/flux3-disagg-amd-diag
michael/flux3-exact-vae-rocm-determinism
michael/fp8-moe-aiter-py310
michael/miles-8gpu-split
michael/miles-gpu2-evict-probe
michael/miles-nightly-partition
michael/minimax-m31-mxfp4-validate
michael/mmgen-unit-rocm-skips
michael/moss-image-rocm-skip
michael/mxfp8-online-scalingtype
michael/nomas-rocm-skips
michael/qsa-prefill-relative-mi300-lds
michael/stage-b-mi35x-all
michael/stage-b-mi35x-all2
michael/usp-ring-py310
michael/wan-cutedsl-lazy-import
michaezh/test-pr-27730-rocm720
mick-dynamic-batching
mick/codex-tmp-lingbot-block5-wanvae-20260602
mick/codex-tmp-lingbot-ci-partition-append-order-20260602
mick/codex-tmp-lingbot-wanvae-batched-decode-20260602
mick/codex-tmp-lingbot-wanvae-batched-main-20260602
mick/codex-tmp-lingbot-wanvae-block4-main-20260603
mick/codex/causal-kv-profile-20260605-r2
mick/codex/diffusion-bench-framework
mick/codex/lingbot-causal-dmd-refactor-validate-d90e6f4
mick/codex/lingbot-control-signal-validate-554b486
mick/codex/lingbot-latent-spec-validate-1654379
mick/codex/lingbot-latent-spec-validate-71e3b33
mick/codex/lingbot-raw-frame-rename-validate-b18c4ad
mick/codex/ltx23-nonpr-perf-20260503
mick/codex/omni-u1-orchestrator
mick/codex/unipc-stack-candidate-20260603
mick/diffusion-auto-residency
mick/diffusion-bench-comfyui
mick/diffusion-conditioning-cache-parallel-digest
mick/diffusion-cookbook-entry
mick/diffusion-fasth3-8step-v2
mick/ltx23-hq-precision-align
mick/ltx23-official-lora-merge-align
mick/ltx23-tp-psnr20-20260506
mick/mick/mmgen-multi-output-dedup
mick/mmgen-stage-a
mick/tmp/image-edit-uploadfile-fix-20260602
mick/tmp/lingbot-local-final-proj-gather-20260602
mick/tmp/lingbot-msgspec-fix-20260602
mick/tmp/lingbot-skip-proj-after-msgspec-20260602
mick/tmp/lingbot-skip-proj-verify-20260602
mick/tmp/lingbot-uniform-usp-fastpath-20260602
mick/wan-ffn-residual-chain
mick/wan-fused-qkv-gemm
mick/wan-s2v-rebased-20260328
mick/wan-text-kv-once-per-prompt
mick/worktree-ltx2-hq-pipeline
migrate-rotary-embedding
mistral4-support
mmangkad/ci-5090-1-base-b-server
mmangkad/ci-5090-2-base-b-light
mmangkad/ci-hf-hub-rate-limit
mmangkad/dsa-kpool-skip-unused-mqa-schedule
mmangkad/dsa-mtp-share-draft-metadata
mmangkad/dsa-tilelang-bf16-split-kv-decode
mmangkad/fix-bench-serving-base-url
mmangkad/flashinfer-prefill-autotune-default
mmangkad/glm53-flash-kda-prefill-bcg
mmangkad/glm53-mhc-cross-layer-handoff
mmangkad/input-logprob-no-sync
mmangkad/kda-flashinfer-prefill-cu-seqlens-sync
mmangkad/kda-replayssm-commit-prepared-step-indices
mmangkad/kda-replayssm-fold-tiling
mmangkad/move-1gpu-large-tests-to-5090
mmangkad/nextn-draft-prefill-graph
mmangkad/remove-4-gpu-gb300-suite
mmangkad/upgrade-cuda-pytorch-2.15
moonep
msa-megakernel-route
mtp/nongraph-metadata-fix
mtp/page-align-alloc
multimodal-bench-serving-openai-compatible
multimodal-gen-tokenizer-fix
muse-spark
nightly-ci
nightly/kimi-k25-mtp
nightly_qiaolin
npu-device-mixin
npu-skip-non-python-changes
nvidia/qwen35-agentx-gb300
nvidia/qwen35-agentx-rubin
optimize-mla-kv-buffer-get
oss-sync-538ec60fd4
oss-sync-5bf04645f2
oss-sync-5cc91ba332
oss-sync-63008b6421
oss-sync-7c10a8cdcd
oss-sync-8bb92640f5
oss-sync-91e9fdc5e2
oss-sync-a288e8ae43
oss-sync-ad931fecce
oss-sync-af1e56b9be
oss-sync-b34e09bb18
oss-sync-b6d441ee64
oss-sync-c9197bfb88
oss-sync-nixl-test-tmpdir
oss/l2-transfer-consolidation
otel-2/n
page-allocator-sync
patch-docker-merge-reworkd
patch-docker-path-fix
perf-accuracy-improvement
perf-regression-dashboard
perf/eagle-fused-topk-topp-renorm
perf/tiny-n-gemm-reduce-scatter
performance-tracking-ci-failure-monitor
phase1-eager-kernel-optimization
phase1-on-main
piecewisemempool
pin-fa4-4.0.0b16
port/dsv4-flash-encoder
pr-12263
pr-13592
pr-13728
pr-14226
pr-16287
pr-17247
pr-26980
pr-33108
pr-34855-gt
pr/31902
pr23119-fix
processor-mm-qwen-vl
qiaolin/adaptive-spec-cuda-graph-buffer-reuse
qiaolin/draft-topk1-argmax-fastpath
qiaolin/qwen-next-demo
qiaolin/qwen-next-kv-allocation
qiaolin/qwen-next-relay
qiaolin/qwen-next-sampling
qiaolin/qwen-next-snapshot
qiaolin/qwen-next-state-commit
qiaolin/qwen-next-tracking
qiaolin/spec-step4-fuse-draft-decode-create-flashmla
qiaolin_commit_in_draft_extend_graph
qiaolin_fused_commit_indices
qiaolin_fused_verify_prep
qiaolin_mrope_nosync
qiaolin_replayssm_graph_commit
qiaolin_spec_mrope_opt
qimcis-ltx2-i2v-support
quant-refactor/import-shim-test
quant-refactor/string-spec-registry
quant-refactor/unify-scheme-abcs
qwen-vae-bf16-default
qwen38
qwen38_miles
qwen4-main-squashed
qwen_coder
qwen_optimize
radix-tree-n/1
rainj-me/rust-api-server-tower-layer
rainj-me/rust-benchmark
rainj-me/rust-cp-apis-migration
rainj-me/rust-fanout-control-reqs
rainj-me/rust-server-abis
rainj-me/rust-server-api-http
rainj-me/rust-server-cp-api
rainj-me/rust-server-pd-lb
rainj-me/rust-server-py-foundations
rainj-me/rust-server-py-integration
rainj-me/rust-server-python-part-1
rainj-me/rust-server-python-part-2
rainj-me/rust-server-refactor-openai-apis
rainj-me/rust-server-runtime
rainj-me/rust-server-tests
rainj-me/rust-server-tm-ingress
rainj-me/rust-server-tm-sampling
rainj-me/rust-server-tok-detok
rainj-me/rust-server-tokenizer
rebase/pr-30478-dcp-runtime-context
refactor/dsv41-multimodal-layout
refactor/release-whl-kernel-composite-actions
reland-pr26134
release-test-xpu
release/v0.5.10
release/v0.5.10rc0
release/v0.5.11
release/v0.5.12
release/v0.5.13
release/v0.5.14
release/v0.5.15
release/v0.5.16
release/v0.5.17
release/v0.5.18
release/v0.5.19
release/v0.5.20
release/v0.5.21
release/v0.5.22
release/v0.5.23
release/v0.5.7
release/v0.5.8
remove-stale-prefill-cudagraph-gqa-gate
reorganize-test-registered-structure
replay-audit-hooks
research/linear-attn-fp16-state
resize-condition-image
return_routed_expert
return_routed_expert_yizhang
revert-14717-feat/layernorm_scale_shift_kernel
revert-15829-feat/mamba2-radix-overlap
revert-20560-youngrok/fix-hicache-write-through-idle
revert-21458-fused_qk_norm_mrope_decode
revert-21631-refactor_write_back_kernel
revert-23943-yhyang201/dsv4-h200-pro-single-node
revert-25576-update-cutlass-dsl-cu13-4.5.1
revert-26235-eagle-topk1-softmax-skip
revert-26905-amd_change_hf_path_0601
revert-27529-fangyuan/fix_c128_kv_dtype_mismatch
revert-29775-yamy1234/dsv4-flashmla-sparse-prefill-default-on
revert-30947-glm52/mtp-split-1-topk1
revert-31001-do-not-merge
revert-31107-topk-glm
revert-33661-bcg-mla-all-archs
revert-34715-bmm65536-fallback
revert-40475-diffusion_register_optm_2
revert-44841-codex/comfyui-worker-patch-20261017
revert-hy3-preview-reenable-nemotron
revert-pr-19140
revert-pr26134
revert-rmsnorm-dynamo-check
revert-tf-4.57.1
rockdu-dev-mm-v2-miles/01-shared-enums
rockdu-dev-mm-v2-miles/02-processor-config
rockdu-dev-mm-v2-miles/03-shared-mixin
rockdu-dev-mm-v2-miles/04-serving-shared
rockdu-dev-mm-v2-miles/05-token-space-framework
rockdu-dev-mm-v2-miles/06-sglang-wrapper
rockdu-dev-mm-v2-miles/07-qwen-strategy
rockdu-dev-mm-v2-miles/08-parity-tests
rockdu-dev-mm-v2-miles/09-partial-expansion
rockdu-dev-mm-v2/02-processor-config
rockdu-dev-mm-v2/03-shared-mixin
rockdu-dev-mm-v2/04-serving-shared
rockdu-dev-mm-v2/05-token-space-framework
rockdu-dev-mm-v2/06-sglang-wrapper
rockdu-dev-mm-v2/07-qwen-strategy
rockdu-dev-mm-v2/08-parity-tests
rockdu-dev-mm-v2/09-partial-expansion
rockdu-dev-mm-v2/sglang-miles
rockdu-dev-mm-v2/sglang-miles-timelens
rockdu-dev-mm/02-processor-config
rockdu-dev-mm/03-shared-mixin
rockdu-dev-mm/04-token-space-core
rockdu-dev-mm/05-qwen-strategy
rockdu-dev-mm/06-startup-and-parity
rockdu/rollout-latent-step-indices
rockdu/rollout-plugin-sglang-miles-next
rockdu/rollout-plugin-sglang-miles-v0.5.22
rocm10-gfx1250
rocm101_1250
rocm_rccl_232
rocm_release_for_testing
router-3a-1-tree-prefix-view
router-3a-2-locality-metrics
router-3b-locality-dashboard
router-4d-outcome-attribution
router-fallback-handler
router/policy-reorg-1-state-move
router/policy-reorg-2-load-view
router/policy-reorg-3-core
router/policy-reorg-4-affinity
router/policy-reorg-5-cache
router/policy-reorg-6-config
router/policy-reorg-7-resolver
router/policy-reorg-8-switch
router/policy-reorg-9-remove
rust-mm-deepseek-v41
rust-mm-glm53-flash
rust-server-chat-parity-followups
rust-server-tool-choice-parity
sahithi/gb200-8k1k-configs
sahithi/gb200-accuracy-evals
sampling-top-k-first
save_mem_linear_mtp
seq-lens-datatype
server-args-cleanup
session-routing-dev
sgl-bot
sgl-compile-distributed-export-runtime
sgl-compile-platform-export
sgl-router-kv-comments
sgl-router-reorg-admitted-fallback
sgl-router-reorg-affinity-savings
sgl-router/upstream-lb-2-engine-load
sgl-router/upstream-lb-3-decision-metrics
sgl-router/upstream-lb-4-queue-gate
sgl-router/upstream-lb-5-saturation-pin
sgl-router/upstream-lb-6-attribution
sgl-router/upstream-lb-7-measure-spill
sgl-router/upstream-lb-8-spill-overlap
sgl-router/upstream-lb-9-queue-headroom
sglang-miles
sglang-miles-20260327
sglang-miles-diffusion-temp
sglang-miles-flux3-image
sglang-miles-glm53-lora
sglang-miles-glm53next
sglang-miles-h3
sglang-miles-inkling
sglang-miles-k3
sglang-miles-muse-spark-open
sglang-miles-qwen38next
sglang-miles-v0.5.10-final
sglang-miles-v0.5.12
sglang-miles-v0.5.12-final
sglang-miles-v0.5.13-final
sglang-miles-v0.5.14-final
sglang-miles-v0.5.15
sglang-miles-v0.5.15-final
sglang-miles-v0.5.16
sglang-miles-v0.5.16-final
sglang-miles-v0.5.17
sglang-miles-v0.5.18
sglang-miles-v0.5.18-final
sglang-miles-v0.5.19
sglang-miles-v0.5.19-final
sglang-miles-v0.5.20
sglang-miles-v0.5.20-final
sglang-miles-v0.5.21
sglang-miles-v0.5.21-final
sglang-miles-v0.5.22
sglang-miles-v0.5.8.post1
sglang-miles-v0.5.9
sglang-miles-v0.5.9-final
sglang-miles_weight_checker
sglang-miles_weight_loader_threshold
sglang-processor-chat-fixes
sglang-slime
sglang-slime-patched
shiyang/forwad_decouple
shiyang/load-snapshot-shm
shiyang/logits/fuse-allgather-buffer
shiyang/mem_check
shiyang/mem_v2/decouple
shiyang/mem_v2/first
shiyang/optimize-get-load-piggyback
shiyang/req_to_token
simolin/client-cancel-499
slime-patched-260303
smg-mm-11/n
spec/drop-draft-global-server-args-restore
spec_hidden
split/pr24408-01-contracts
split/pr24408-02-server-args
split/pr24408-03-batch-invariant-ops
split/pr24408-04-layernorm-batch-tests
split/pr24408-05-tp-invariant-ops
split/pr24408-06-tp-invariant-tests
split/pr24408-07-collective-wiring
split/pr24408-08-dp-attention-padding
split/pr24408-09-qwen3-moe-model
split/pr24408-12-wiring-cli-tests
split/pr24408-13-wiring-contract-tests
split/pr24408-14-wiring-runtime-tests
split/pr24408-15-wiring-reduction-tests
split/pr24408-16-wiring-cuda-tests
stack/ngram-hidden-states-test
step-3.5-flash
step5-preview-support
stress-test-debugging
support_qwen3_guard
support_usp_for_wan_vae
swa-pool-pp-start-layer
sync-pr-test-from-main
temp-docker-image
test-ci-empty-pr-2
test-gb200-runners
test-mm-amd-ci
test/ci-disagg-tp-check
test/mooncake-arm64-wheel-validation
test/pr33561-ab-both-markers-off
test/rerun-stage-8gpu-h200
test/slash-command-handler-tests
test_new_x86_node
test_qiaolin
tokenspeed_mla_prefill
tom/adhoc-sum-pad-exp-th3
tom/alloc-page-aligned/op1
tom/alloc-page-aligned/op2
tom/alloc-page-aligned/op21
tom/alloc-page-aligned/op22
tom/alloc-page-aligned/op24
tom/alloc-page-aligned/op25
tom/alloc-page-aligned/op25a
tom/alloc-page-aligned/op26
tom/alloc-page-aligned/op27
tom/alloc-page-aligned/op28
tom/alloc-page-aligned/op29a
tom/alloc-page-aligned/op29b
tom/alloc-page-aligned/op29c
tom/alloc-page-aligned/op29d
tom/alloc-page-aligned/op30
tom/alloc-page-aligned/op31
tom/alloc-page-aligned/op32
tom/alloc-page-aligned/op33
tom/alloc-page-aligned/op5
tom/alloc-page-aligned/op51-non-split
tom/alloc-page-aligned/op52
tom/alloc-page-aligned/op53
tom/alloc-page-aligned/op54
tom/alloc-page-aligned/op6
tom/alloc-page-aligned/op7
tom/alloc-page-aligned/temp-ci
tom/cache-unfinished-req-use-committed-len
tom/disable_pd_perturb_ci
tom/e2e_qiaolin_fill_ids
tom/extract-scheduler-batches
tom/fix-d1-current-gdn-lora-cli
tom/fix-dllm-abort-waiting-queue
tom/fix-eagle-chunked-next-prompt-token
tom/fix-moe-virtual-experts-gated-lora
tom/fix-preempt-stale-extend-input-len
tom/fix-vocab-boundary-off-by-one
tom/gemini-modify-component-review
tom/kv_canary
tom/kv_canary_revert_reversed
tom/lora-mq-a
tom/lora-mq-b
tom/lora_0604_opt_moe
tom/lora_qkv_b_opt
tom/merge_input_output_ids
tom/miles-snapshot-rope-compat-1019
tom/pr_24928_base
tom/pr_chain/tom_refactor_202605a/primary/mech_preflight/convert-self-x-to-locals
tom/probe-lora-bug-25690
tom/probe/20260607T131809Z-a39c428
tom/probe/20260821T001953Z-dsv3-same-ckpt
tom/probe/20260821T002009Z-dsv3-drop-refit
tom/probe/20260821T235224Z-spans-dsv3-dpattn
tom/refactor-miles-repo-sglang/ac8770
tom/refactor-miles-repo-sglang/deliver-1
tom/refactor-miles-repo-sglang/deliver-2
tom/refactor-miles-repo-sglang/deliver-4
tom/refactor-miles-repo-sglang/deliver-5
tom/refactor-miles-repo-sglang/lora-targets
tom/refactor-miles-repo-sglang/op1-3
tom/refactor-miles-repo-sglang/op1-9
tom/refactor-miles-repo-sglang/op13-116
tom/refactor-miles-repo-sglang/op21-2
tom/refactor-miles-repo-sglang/op22-2
tom/refactor-miles-repo-sglang/op26-2
tom/refactor-miles-repo-sglang/op40-1
tom/refactor-miles-repo-sglang/op43-123
tom/refactor-miles-repo-sglang/op47-62
tom/refactor_shared_array
tom/remove-full-untruncated-fill-ids
tom/rename-fill-len-extend-fill-len
tom/req-extend-range
tom/req-fill-len-none
tom/req_pool_idx_and_cache_and_kv/1
tom/req_pool_idx_and_cache_and_kv/10
tom/req_pool_idx_and_cache_and_kv/11
tom/req_pool_idx_and_cache_and_kv/12
tom/req_pool_idx_and_cache_and_kv/13
tom/req_pool_idx_and_cache_and_kv/14
tom/req_pool_idx_and_cache_and_kv/15
tom/req_pool_idx_and_cache_and_kv/16
tom/req_pool_idx_and_cache_and_kv/17
tom/req_pool_idx_and_cache_and_kv/18
tom/req_pool_idx_and_cache_and_kv/21
tom/req_pool_idx_and_cache_and_kv/22
tom/req_pool_idx_and_cache_and_kv/23
tom/req_pool_idx_and_cache_and_kv/24
tom/req_pool_idx_and_cache_and_kv/25
tom/req_pool_idx_and_cache_and_kv/26
tom/req_pool_idx_and_cache_and_kv/27
tom/req_pool_idx_and_cache_and_kv/28
tom/req_pool_idx_and_cache_and_kv/29
tom/req_pool_idx_and_cache_and_kv/3
tom/req_pool_idx_and_cache_and_kv/30
tom/req_pool_idx_and_cache_and_kv/31
tom/req_pool_idx_and_cache_and_kv/32
tom/req_pool_idx_and_cache_and_kv/33
tom/req_pool_idx_and_cache_and_kv/34
tom/req_pool_idx_and_cache_and_kv/35
tom/req_pool_idx_and_cache_and_kv/36
tom/req_pool_idx_and_cache_and_kv/37
tom/req_pool_idx_and_cache_and_kv/38
tom/req_pool_idx_and_cache_and_kv/39
tom/req_pool_idx_and_cache_and_kv/4
tom/req_pool_idx_and_cache_and_kv/5
tom/req_pool_idx_and_cache_and_kv/6
tom/req_pool_idx_and_cache_and_kv/7
tom/req_pool_idx_and_cache_and_kv/8
tom/req_pool_idx_and_cache_and_kv/9
tom/req_pool_idx_and_cache_and_kv/temp_ci
tom/req_pool_idx_step1_dataclass
tom/req_pool_idx_step2_streaming_move
tom/req_pool_idx_step3_eliminate_flags
tom/revert-25690-cutedsl
tom/revert-pr10414
tom/rm_cur_last_batch
tom/scripted_runtime_extracted
tom/scripted_runtime_extracted_chain
tom/sglang/d1-c21-dsv4-reasoning-compat
tom/stateless_scheduler_b
tom/stateless_scheduler_b_temp_run
tom/temp_ci
tom/temp_revert_27965
tom/trainer-ft-split
tom_refactor_202605a/primary/mech_model_runner
tom_refactor_202605a/primary/mech_preflight
tom_refactor_202605a/primary/mech_scheduler
tom_refactor_202605a/primary/mech_tokenizer_manager
tom_refactor_202605a/primary/nonmech_model_runner
tom_refactor_202605a/primary/nonmech_scheduler
tom_refactor_202605a/primary/nonmech_tokenizer_manager
tool_call_benchmark
top-k-first-opt
transformers-v5
trim-nemotron35-cookbook-flags
trtllm-mha-full-overlap
trtllm-mla-fused-graph-metadata
trtllm-sparse-mla-dcp
unified-envelope-buffer-info
update-transformers-v5
update_perf_dashboard
util-queue-timeline-chart
v0.5.19
vincentzed/dcp-shard-indexer
vlm-post-capture-fixed-reserve
vz/qwen3-omni
wan-overlay-only-20260328
wan-s2v-main-20260330
wan-s2v-rebased-20260328
wduan/fix_pause_generation
whl-metadata-fix
worktree-agent-a5ee4203
worktree-agent-ab2b777b
worktree-pure-nibbling-hickey
xeonnightly
xiaomi-mimo-v2-flash-0.5.5
xiezhq-schedule-new
xiezhq-schedule-simplify
xiezhq-session
xiezhq/42701-merge-main
xinyuan/dflash-overlap-radix-war-fix
xinyuan/glm47-39136-review-fixes
xinyuan/glm53flash-sota-perf
xinyuan/kimi-k3-parser-auto-upstream
xinyuan/kimi-k3-reasoning-parser
xinyuan/kimi-k3-tool-constraints
xinyuan/kimi-k3-tool-parser
xinyuan/nightly-precision-baseline-refresh
xinyuan/plan-stream-handoff
xpu/kernel-from-main
xpu/qwen3-ccl-arc-allreduce
xpu/qwen3-eagle-draft-attn-fallback
xpu/qwen3-eagle-target-only-sampling
xpu/qwen3-fp8-w8a16-dense
xpu/qwen3-moe-tuning-xpu
xpu/qwen3-qwen3-fused-qknorm-rope
xpu/qwen3-rope-native-fallback
xsun/amd_diffusion
yamy/mixed2048-fusion-piecewise-cg-bia
yanbo/feat/router-pd-loadmonitor
yanbo/k3-stream-conformance
yanbo/k3-thinking-param
yanbo/k3-tool-call-conformance
yanbo/pd-k3
yanbo/pd-k3-rebase-0919
yanbo/pr_34820
yanbo/router-dp-aware
yanbo/router-pd-inhouse
yanbo/router_test
yanbo/router_test-r1-owned-ancestor
yangmin/pr26245-fullchat-terminal-strict-rr
yihao/draft-attn-registry
yizhang/decode_store_only_mamba_radix_cache
yizhang/mamba_prefix_cache_support_large_page_size
yizhang/tp-dp-invariant
yizhang/vlm_fix_radix_cache_for_qwen3vl
yonghao/cp-sharded-kv-layout
yonghao/ump-shared-capacity
yonghao/unified-hicache-review-followup
yueming/colocated-param-bytes
yueming/colocated-param-bytes-pull
yueming/dsa-pagetable-tms
yueming/fix-lora-unit-test-fakes
yueming/fix-uninit-ue8m0-scale-padding
yueming/fix-uninit-ue8m0-scale-padding-main
yueming/nextn-init-fix
yueming/p2p-registry-fixes
yueming/qwen3-router-dtype
yueming/rank-info-record
yueming/rl-lora-streamed-updates
yueming/rl-model-side-weight-update-fixes
yueming/rl-pull-weights
yueming/rl-weight-checker-exemptions
yueming/weight-checker-skip-sincos-tables-main
yusheng/abort-rid-prefix
yusheng/cherrypick-36389-partial-sglang-miles
yusheng/cherrypick-36389-sglang-miles
yusheng/kimi-moe-lora-residual
yusheng/lora-dp-attention
yusheng/lora-fp8-col-shard
yusheng/lora-full-prefill-cg-gate
yusheng/lora-load-from-distributed
yusheng/lora-moe-buffer-dims
yusheng/lora-moe-cg-dp-attention
yusheng/lora-qwen35-gdn-in-proj-ba
yusheng/lora-rowparallel-probe
yusheng/lora-shard-spec-table
yusheng/lora-upsert
yusheng/lora-weight-update-quant-hook
yuwei/pcg
yz/amd_diffusion
zaya-pr1-short-conv
zaya-pr2-cuda-graph
zaya-pr3-model
zhichen/gdn-norm-device-nemotron-topk-layer-id
zhichen/qwen3_5-parameter-mapper
zqx/hicache-file-storage-path-followup
gateway-v0.1.5
gateway-v0.1.6
gateway-v0.1.7
gateway-v0.1.8
gateway-v0.1.9
gateway-v0.2.0
gateway-v0.2.1
gateway-v0.2.2
gateway-v0.2.3
gateway-v0.2.4
gateway-v0.3.0
gateway-v0.3.1
v0.1.10
v0.1.11
v0.1.12
v0.1.13
v0.1.14
v0.1.15
v0.1.16
v0.1.17
v0.1.18
v0.1.19
v0.1.20
v0.1.21
v0.1.22
v0.1.23
v0.1.24
v0.1.3
v0.1.5
v0.1.6
v0.1.7
v0.1.8
v0.1.9
v0.2.0
v0.2.10
v0.2.11
v0.2.12
v0.2.13
v0.2.14
v0.2.14.post1
v0.2.14.post2
v0.2.15
v0.2.5
v0.2.6
v0.2.7
v0.2.8
v0.2.9
v0.2.9.post1
v0.3.0
v0.3.1.post1
v0.3.1.post2
v0.3.1.post3
v0.3.2
v0.3.3
v0.3.3.post1
v0.3.4
v0.3.4.post1
v0.3.4.post2
v0.3.5
v0.3.5.post1
v0.3.5.post2
v0.3.6
v0.3.6.post1
v0.3.6.post2
v0.3.6.post3
v0.4.0
v0.4.0.post1
v0.4.0.post2
v0.4.1
v0.4.1.post1
v0.4.1.post2
v0.4.1.post3
v0.4.1.post4
v0.4.1.post5
v0.4.1.post6
v0.4.1.post7
v0.4.10
v0.4.10.post1
v0.4.10.post2
v0.4.2
v0.4.2.post1
v0.4.2.post2
v0.4.2.post3
v0.4.2.post4
v0.4.3
v0.4.3.post1
v0.4.3.post2
v0.4.3.post3
v0.4.3.post4
v0.4.4
v0.4.4.post1
v0.4.4.post2
v0.4.4.post3
v0.4.4.post4
v0.4.5
v0.4.5.post1
v0.4.5.post2
v0.4.5.post3
v0.4.6
v0.4.6.post1
v0.4.6.post2
v0.4.6.post3
v0.4.6.post4
v0.4.6.post5
v0.4.7
v0.4.7.post1
v0.4.8
v0.4.8.post1
v0.4.9
v0.4.9.post1
v0.4.9.post2
v0.4.9.post3
v0.4.9.post4
v0.4.9.post5
v0.4.9.post6
v0.5.0rc0
v0.5.0rc1
v0.5.0rc2
v0.5.1
v0.5.1.post1
v0.5.1.post2
v0.5.1.post3
v0.5.10
v0.5.10.post1
v0.5.10rc0
v0.5.11
v0.5.12
v0.5.12.post1
v0.5.13
v0.5.13.post1
v0.5.14
v0.5.15
v0.5.15.post1
v0.5.16
v0.5.17
v0.5.18
v0.5.19
v0.5.2
v0.5.20
v0.5.21
v0.5.22
v0.5.2rc0
v0.5.2rc1
v0.5.2rc2
v0.5.3
v0.5.3.post1
v0.5.3.post2
v0.5.3.post3
v0.5.3rc0
v0.5.3rc1
v0.5.3rc2
v0.5.4
v0.5.4.post1
v0.5.4.post2
v0.5.4.post3
v0.5.5
v0.5.5.post1
v0.5.5.post2
v0.5.5.post3
v0.5.6
v0.5.6.post1
v0.5.6.post2
v0.5.7
v0.5.8
v0.5.8.post1
v0.5.9
-
7c98507040
Remove unnecessary comments in fp8_utils.py
xpu/qwen3-fp8-w8a16-dense
Alex Nails
2026-10-28 00:59:41 -07:00 -
2d3ef161ff
[Fix] [XPU] quantization: Honor weight_scale strides in W8A16 dequant, guard fused-op layouts
Alex Nails
2026-10-28 00:51:00 -07:00 -
f313c035e8
[Perf] [XPU] quantization: Run block-FP8 linears weight-only (W8A16) on Xe2
Yihao Wang
2026-10-27 09:10:15 +00:00 -
168154d20c
[Fix] [XPU] benchmark: Run the fused MoE Triton tuner on XPU
xpu/qwen3-moe-tuning-xpu
Yihao Wang
2026-10-27 06:11:12 +00:00 -
49657e2b5a
[Perf] [XPU] qwen3: Fuse q/k RMSNorm and RoPE on XPU
xpu/qwen3-qwen3-fused-qknorm-rope
Yihao Wang
2026-10-27 02:26:52 +00:00 -
520b855d09
[Perf] [XPU] distributed: Default oneCCL to its Arc all-reduce on Arc GPUs
xpu/qwen3-ccl-arc-allreduce
Yihao Wang
2026-10-27 02:26:09 +00:00 -
c536c44202
[Fix] [XPU] spec: Sample EAGLE verification on XPU instead of forcing argmax
xpu/qwen3-eagle-target-only-sampling
Yihao Wang
2026-10-27 05:44:49 +00:00 -
6e61eca8ee
[Fix] [XPU] spec: Give EAGLE drafts triton attention when intel_xpu lacks their head size
xpu/qwen3-eagle-draft-attn-fallback
Yihao Wang
2026-10-27 05:15:06 +00:00 -
e7fbf03f74
[Fix] [XPU] rope: Fall back to native RoPE for head sizes the XPU kernel lacks
xpu/qwen3-rope-native-fallback
Yihao Wang
2026-10-27 04:09:19 +00:00 -
6aeefd6986
test: supply state tree slots in remaining MLX component fixtures
cheng/refactor/unified-kv-tree-components
Cheng Wan
2026-10-28 00:57:53 -07:00 -
33d1b50773
Migrate the sharded eviction fixture to prefix-tree slots
Cheng Wan
2026-10-27 23:22:21 -07:00 -
74f6259256
Migrate the LLaDA image encoder to prefix-tree slots
Cheng Wan
2026-10-27 22:38:23 -07:00 -
4fe46104f9
Complete resolved geometry in the state cleanup test fixture
Cheng Wan
2026-10-27 22:16:23 -07:00 -
ddfb3e24d1
unified-kv-tree-components(native-slot-families,non_mechanical_provable): Build the native tree core's drivers by slot family
Cheng Wan
2026-10-27 15:05:23 -07:00 -
deb15daf6e
unified-kv-tree-components(family-names,non_mechanical_provable): Name the prefix tree's slots by family in the cache and both tree cores
Cheng Wan
2026-10-27 14:50:32 -07:00 -
9cabb5e2d1
unified-kv-tree-components(tracking-slots,non_mechanical_provable): Take and give back a request's state tracking slots through the KV storage
Cheng Wan
2026-10-27 12:56:09 -07:00 -
caa0e25caa
unified-kv-tree-components(tree-slot-numbers,non_mechanical_provable): Carry each tree slot's window and checkpoint interval from the plan
Cheng Wan
2026-10-27 12:33:24 -07:00 -
b30898206a
unified-kv-tree-components(state-effects,non_mechanical_provable): Reach the prefix tree's recurrent state through the KV storage
Cheng Wan
2026-10-27 08:05:36 -07:00 -
349ec7bec4
unified-kv-tree-components(tree-slot-families,non_mechanical_provable): Build the prefix tree's components by slot family from the KV storage
Cheng Wan
2026-10-27 05:37:54 -07:00 -
b4a89881f2
unified-kv-tree-components(tree-effects,non_mechanical_provable): Route the prefix tree's frees and window rebinds through the KV storage
Cheng Wan
2026-10-27 02:31:19 -07:00 -
91df012641
unified-kv-tree-components(advance-window,non_mechanical_provable): Retire a request's window rows through its KV storage
Cheng Wan
2026-10-27 01:26:52 -07:00 -
3db4e8572d
unified-kv-tree-components(tree-slots,non_mechanical_provable): Build the prefix tree's components from the KV storage
Cheng Wan
2026-10-27 01:23:06 -07:00 -
1086d4ad14
fix: initialize FlashInfer test routing through its pool
cheng/refactor/unified-kv-typed-views
Cheng Wan
2026-10-28 00:24:41 -07:00 -
8e6d38426d
Migrate new speculative decoding fixtures to runner identifiers
Cheng Wan
2026-10-27 23:22:10 -07:00 -
2924c73b31
Preserve masked window planning through FlashInfer forward
Cheng Wan
2026-10-27 22:48:46 -07:00 -
b334474081
Recognize MLA views when configuring HiCache storage
Cheng Wan
2026-10-27 22:39:48 -07:00 -
502fa981bb
Mask out-of-window ancestors in speculative verification trees
Cheng Wan
2026-10-27 22:02:29 -07:00 -
99d00cfc6f
Fix routed MLA writes to select the layer storage locations
Cheng Wan
2026-10-27 22:02:13 -07:00 -
c029cdbf7c
Fix fused draft HiCache validation after view binding
Cheng Wan
2026-10-27 22:01:56 -07:00 -
b962a7d31e
unified-kv-typed-views(id-modules,non_mechanical_provable): Keep a runner's KV id translation beside the id spaces
Cheng Wan
2026-10-27 20:14:21 -07:00 -
8315cb88ed
unified-kv-typed-views(drop-spec-declarations,non_mechanical_provable): Build every speculative forward without per-backend combination lists
Cheng Wan
2026-10-27 18:59:10 -07:00 -
93143fb813
unified-kv-typed-views(trtllm-window-draft-extend,non_mechanical_provable): A trtllm_mha draft extend reads a window layer through the plan's window table
Cheng Wan
2026-10-27 18:58:47 -07:00 -
a0bfe9b373
unified-kv-typed-views(flashinfer-window-read,non_mechanical_provable): the verify kit's FlashInfer DFLASH window is per query
Cheng Wan
2026-10-27 18:43:51 -07:00 -
5c968d90e6
unified-kv-typed-views(flashinfer-window-read,non_mechanical_provable): FlashInfer's kernel over a window read, checked against positions
Cheng Wan
2026-10-27 18:18:32 -07:00 -
0b5b89cdaa
unified-kv-typed-views(flashinfer-window-read,non_mechanical_provable): one window read for every FlashInfer mode
Cheng Wan
2026-10-27 18:16:12 -07:00 -
26c900cd57
unified-kv-typed-views(callers-kv-layers,non_mechanical_provable): Reach per-layer KV through the layer router in the remaining callers
Cheng Wan
2026-10-27 17:57:07 -07:00 -
3ec1e82ac0
unified-kv-typed-views(other-backends-kv-layers,non_mechanical_provable): Read and write the remaining attention backends' per-layer KV through the layer router
Cheng Wan
2026-10-27 17:57:06 -07:00 -
6c20a42d0b
unified-kv-typed-views(aiter-kv-layers,non_mechanical_provable): Read and write AITER's per-layer KV through the pool's layer router
Cheng Wan
2026-10-27 17:57:05 -07:00 -
4f157a97cc
unified-kv-typed-views(models-mla-kv-layers,non_mechanical_provable): Read and write the remaining models' per-layer KV through the layer router
Cheng Wan
2026-10-27 16:59:20 -07:00 -
c2d92ef8ac
unified-kv-typed-views(mla-kv-layers,non_mechanical_provable): Read and write the MLA backends' per-layer KV through the layer router
Cheng Wan
2026-10-27 16:59:20 -07:00 -
5c02b627c0
unified-kv-typed-views(trtllm-torch-kv-layers,non_mechanical_provable): Read and write TRT-LLM MHA's and the torch backends' per-layer KV through the layer router
Cheng Wan
2026-10-27 16:51:36 -07:00 -
4d54fd3aae
unified-kv-typed-views(models-kv-layers,non_mechanical_provable): Read and write models' per-layer KV through the active pool's layer router
Cheng Wan
2026-10-27 16:51:36 -07:00 -
dc06fd3c88
unified-kv-typed-views(flashinfer-kv-layers,non_mechanical_provable): Read and write FlashInfer's per-layer KV through the runner's layer router
Cheng Wan
2026-10-27 16:42:35 -07:00 -
7adc3601e0
unified-kv-typed-views(fa3-kv-layers,non_mechanical_provable): Read and write FlashAttention's per-layer KV through the runner's layer router
Cheng Wan
2026-10-27 16:42:35 -07:00 -
e427a2d815
unified-kv-typed-views(triton-kv-layers,non_mechanical_provable): Read and write Triton's per-layer KV through the runner's layer router
Cheng Wan
2026-10-27 16:41:09 -07:00 -
af62163fe3
unified-kv-typed-views(kv-layer-access,non_mechanical_provable): Read and write a layer's KV through its router's KVLayer
Cheng Wan
2026-10-27 16:40:56 -07:00 -
dbef5a6a2e
unified-kv-typed-views(draft-views,non_mechanical_provable): Read a fused draft's KV and state through typed views of its host entries
Cheng Wan
2026-10-27 12:35:17 -07:00 -
3343cccbf5
unified-kv-typed-views(state-view,non_mechanical_provable): Read, clear and copy recurrent state through a typed view of the arena
Cheng Wan
2026-10-27 12:32:49 -07:00 -
81ebfe212d
unified-kv-typed-views(mla-view,non_mechanical_provable): Read and store MLA latent rows through a typed view of the arena
Cheng Wan
2026-10-27 12:30:32 -07:00 -
1624edd242
unified-kv-typed-views(mha-view,non_mechanical_provable): Read and store MHA KV through a typed view of the arena
Cheng Wan
2026-10-27 12:19:32 -07:00 -
a558f73489
style: format static KV storage transfer boundary
cheng/refactor/unified-kv-storage-consumers
Cheng Wan
2026-10-28 00:53:58 -07:00 -
675d5486c3
unified-kv-static-transfer(pins-and-sizing,non_mechanical_provable): Register transfer pins and size HiCache alike for both placements
Cheng Wan
2026-10-26 22:01:51 -07:00 -
c5a5fda900
unified-kv-static-transfer(static-pd-surface,non_mechanical_provable): Register static pools with PD peers through the storage interface
Cheng Wan
2026-10-26 22:01:50 -07:00 -
6a385b1f5c
unified-kv-observers(beam-rows,non_mechanical_provable): Take and give back beam member rows and their ids through the KV storage
Cheng Wan
2026-10-26 22:04:36 -07:00 -
45bc0d6112
unified-kv-observers(usage,non_mechanical_provable): Read the pool observers' numbers from the KV storage's per-sub-pool usage
Cheng Wan
2026-10-26 22:04:29 -07:00 -
a5d88ad44c
unified-kv-lifecycle(lifecycle-through-storage,non_mechanical_provable): Take a request's rows and KV ids through the KV storage
cheng/refactor/unified-kv-requests
Cheng Wan
2026-10-26 19:58:59 -07:00 -
b4b37bbdcd
unified-kv-requests(request-table,non_mechanical_provable): Hold every unified plan's request rows in one RequestTable
Cheng Wan
2026-10-26 19:03:44 -07:00 -
c51c0b8c00
unified-kv-id-spaces(unit-ids,non_mechanical_provable): Keep a sub-pool's unit names in its own record
cheng/refactor/unified-kv-arena-and-ids
Cheng Wan
2026-10-26 18:33:59 -07:00 -
4dc90e1090
unified-kv-id-spaces(bands,non_mechanical_provable): Place each unified sub-pool's units in a band of the arena
Cheng Wan
2026-10-26 18:26:24 -07:00 -
0258081e8c
unified-kv-id-spaces(id-space-home,non_mechanical_provable): Define the id space in the KV allocation package
Cheng Wan
2026-10-26 18:21:49 -07:00 -
ff6df073a4
unified-kv-memory-and-ids(storage-result,non_mechanical_provable): Return the unified KV storage from the builder
Cheng Wan
2026-10-26 17:45:12 -07:00 -
e78544d4f4
unified-kv-memory-and-ids(arena,non_mechanical_provable): Hold the unified placement's bytes in a KVArena of bands
Cheng Wan
2026-10-26 17:27:45 -07:00 -
266e581a62
fix: use resolved decode graph configuration for speculative capabilities
cheng/refactor/unified-kv-draft-views
Cheng Wan
2026-10-28 00:52:54 -07:00 -
400086a358
unified-kv-draft-views(drop-draft-backend-lists,non_mechanical_provable): Fuse a draft whatever attention backend reads it
Cheng Wan
2026-10-26 16:53:49 -07:00 -
5120120a9b
unified-kv-draft-views(routed-draft,non_mechanical_provable): Route a fused draft's layers to its host sub-pools' views
Cheng Wan
2026-10-26 16:53:20 -07:00 -
568669aebe
fix: retain one current-role transfer pin across PD switches
cheng/refactor/unified-kv-pins
Cheng Wan
2026-10-28 00:51:30 -07:00 -
8db0f341d2
unified-kv-pins(host-ids-from-storage,non_mechanical_provable): Map host transfer ids through the KV storage instead of the device views
Cheng Wan
2026-10-26 16:46:33 -07:00 -
391da7416b
unified-kv-pins(pin-sources,non_mechanical_provable): Keep unified pages in place through pin sources registered on the KV storage
Cheng Wan
2026-10-26 16:46:08 -07:00 -
af9e5981a7
unified-kv-drop-facades(combination-ratchet,non_mechanical_provable): Fail a new unified pool class named for a combination of KV kinds
cheng/refactor/unified-kv-drop-facades
Cheng Wan
2026-10-26 16:06:51 -07:00 -
1cc372255b
unified-kv-drop-facades(routed-kv-cache,non_mechanical_provable): Route a unified pool's KV through one RoutedKVCache and delete the per-combination facades
Cheng Wan
2026-10-26 16:05:12 -07:00 -
506129af97
unified-kv-drop-facades(consumer-questions,non_mechanical_provable): Ask the allocator and the router what the unified KV facades answered
Cheng Wan
2026-10-26 16:05:04 -07:00 -
13042505d3
unified-kv-pins(hip-fused-window-write,non_mechanical_provable): Leave a unified window's KV writes to the pool on ROCm
cheng/hot-fix/unified-kv-hip-fused-kv-write
Cheng Wan
2026-10-26 16:35:12 -07:00 -
e1644d157b
unified-kv-transfer(transfer-compare,non_mechanical_provable): Add a script comparing what PD peers and HiCache see of a unified pool
cheng/refactor/unified-kv-transfer
Cheng Wan
2026-10-26 14:14:12 -07:00 -
5fc472243b
unified-kv-transfer(pd-from-plan,non_mechanical_provable): Register a unified pool with PD peers from its plan
Cheng Wan
2026-10-26 14:14:00 -07:00 -
070d3844d2
unified-kv-transfer(hicache-from-plan,non_mechanical_provable): Build a unified pool's HiCache host stack from its plan
Cheng Wan
2026-10-26 14:13:38 -07:00 -
053f96c727
unified-kv-transfer(storage-names,non_mechanical_provable): Give each unified sub-pool its external names and a transfer surface
Cheng Wan
2026-10-26 14:13:03 -07:00 -
c512039b81
unified-kv-transfer(tri-host-stack,non_mechanical_provable): Build the full+SWA+state HiCache stack like the other hybrid stacks
cheng/hot-fix/unified-kv-tri-host-stack
Cheng Wan
2026-10-26 13:35:29 -07:00 -
2faf40c84a
fix: bind EAGLE buffers through KV layer routes
cheng/refactor/unified-kv-state-access
Cheng Wan
2026-10-27 23:32:37 -07:00 -
71c5f0b7e2
Update GDN graph fixtures for state views and identifier spaces
Cheng Wan
2026-10-27 23:12:04 -07:00 -
238b2fd8c7
Route GDN prefill graph state through the KV layer view
Cheng Wan
2026-10-27 23:06:47 -07:00 -
abc20183b2
unified-kv-outside-branches(kind-facts,non_mechanical_provable): Read combined kind checks outside the pools as plan facts
Cheng Wan
2026-10-26 12:27:02 -07:00 -
3943b6df97
unified-kv-outside-branches(pool-predicates,non_mechanical_provable): Ask the KV pool's router what PD and HiCache dispatch on
Cheng Wan
2026-10-26 12:26:51 -07:00 -
ca2831e222
unified-kv-state-access(state-built-by-builder,non_mechanical_provable): Build the unified state storage in the pool builder
Cheng Wan
2026-10-26 12:03:09 -07:00 -
2aadb602a5
unified-kv-state-access(state-acquire,non_mechanical_provable): Bind a request's row and state slots through one acquire
Cheng Wan
2026-10-26 10:38:01 -07:00 -
273f607416
unified-kv-state-access(state-readers,non_mechanical_provable): Reach recurrent state through the KV pool's router and the token allocator
Cheng Wan
2026-10-26 10:11:27 -07:00 -
0f0f1a76ca
unified-kv-state-access(state-storage,non_mechanical_provable): Hold a table's recurrent state in one StateStorage
Cheng Wan
2026-10-26 10:06:26 -07:00 -
c669c71e48
unified-kv-state-access(state-routes,non_mechanical_provable): Read recurrent state through the KV pool's layer router
Cheng Wan
2026-10-26 09:53:10 -07:00 -
83e2b5a85e
unified-kv-state-access(clear-slots-no-sync,non_mechanical_provable): Clear unified state slots without a host synchronization
Cheng Wan
2026-10-26 08:27:56 -07:00 -
007434e24a
unified-kv-state-access(state-attrs,non_mechanical_provable): Read state pool attributes without getattr
Cheng Wan
2026-10-26 08:20:51 -07:00 -
317f6eb73f
unified-kv-state-access(state-id-space,non_mechanical_provable): Map state slot ids through the allocator's id space
Cheng Wan
2026-10-26 08:05:55 -07:00 -
6d0c45528f
unified-kv-router(drop-chain-window-translate,non_mechanical_provable): Remove ChainAllocator.translate_loc_from_full_to_swa
cheng/refactor/unified-kv-router
Cheng Wan
2026-10-26 07:09:48 -07:00 -
f447a3f6b5
unified-kv-router(backend-routing,non_mechanical_provable): Route attention backends' window and layout checks through the layer router
Cheng Wan
2026-10-26 07:09:26 -07:00 -
b71814774b
unified-kv-router(router-and-views,non_mechanical_provable): Route each layer to the sub-pool view that holds its KV
Cheng Wan
2026-10-26 07:08:53 -07:00 -
9f0af30395
unified-kv-addressing(drop-target-allowlists,non_mechanical_provable): Stop gating the unified target's attention backend by speculative algorithm
cheng/refactor/unified-kv-addressing
Cheng Wan
2026-10-26 05:31:46 -07:00 -
13d111c864
unified-kv-addressing(call-sites,non_mechanical_provable): Map window ids through id spaces at every call site
Cheng Wan
2026-10-26 05:26:31 -07:00 -
9429cd5324
unified-kv-addressing(id-spaces,non_mechanical_provable): Build the KV index translator from the pools' id spaces
Cheng Wan
2026-10-26 05:26:18 -07:00 -
ed46981e6d
unified-kv-request-lifecycle(chain-budget,non_mechanical_provable): Admit every unified chain through one prefill budget that owns its state-slot gate
cheng/refactor/unified-kv-pricing-and-budget
Cheng Wan
2026-10-26 05:11:00 -07:00 -
3ca3103467
unified-kv-planner(snapshot-memo,non_mechanical_provable): Keep the pricing snapshot between decisions and price on resolved bands
Cheng Wan
2026-10-26 05:20:31 -07:00 -
6f60f12534
unified-kv-planner(pricing-property,non_mechanical_provable): Check that each capacity price is what the allocation then does
Cheng Wan
2026-10-26 04:05:10 -07:00 -
e5afb50cd9
unified-kv-planner(snapshot-pricing,non_mechanical_provable): Price chain capacity on a host-only snapshot and take through one path
Cheng Wan
2026-10-26 04:03:33 -07:00