Repository navigation
Basic attention kernel that supports cached KV + (multi-)prompts - #24
Merged
Merged
Conversation
suquark
commented
Apr 3, 2023
WoosukKwon
reviewed
Apr 3, 2023
WoosukKwon
left a comment
WoosukKwon
left a comment
Collaborator
There was a problem hiding this comment.
Thanks a lot for your effort. Left minor comments.
Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Contributor
Author
|
@WoosukKwon any comments? |
hongxiayang
pushed a commit
to hongxiayang/vllm
that referenced
this pull request
Feb 13, 2024
slyalin
pushed a commit
to slyalin/vllm
that referenced
this pull request
Apr 4, 2024
Fix for multi-query attention
z103cb
referenced
this pull request
in z103cb/opendatahub_vllm
May 9, 2024
`format.sh` now has mypy checks after pulling in upstream changes. This PR makes the mypy suggested modifications to our code. --------- Signed-off-by: Travis Johnson <tsjohnso@us.ibm.com>
z103cb
referenced
this pull request
in z103cb/opendatahub_vllm
May 9, 2024
`format.sh` now has mypy checks after pulling in upstream changes. This PR makes the mypy suggested modifications to our code. --------- Signed-off-by: Travis Johnson <tsjohnso@us.ibm.com>
z103cb
referenced
this pull request
in opendatahub-io/vllm
May 9, 2024
`format.sh` now has mypy checks after pulling in upstream changes. This PR makes the mypy suggested modifications to our code. --------- Signed-off-by: Travis Johnson <tsjohnso@us.ibm.com>
dtrifiro
pushed a commit
to dtrifiro/vllm
that referenced
this pull request
May 21, 2024
bump ubi base image tag
Closed
njhill
pushed a commit
to njhill/vllm
that referenced
this pull request
Nov 6, 2024
…trings Stop String Plumbing
1 task done
1 task done
1 task done
1 task done
heheda12345
added a commit
to heheda12345/vllm
that referenced
this pull request
Sep 29, 2025
Signed-off-by: Chen Zhang <zhangch99@outlook.com>
1 task done
yma11
added a commit
to yma11/vllm
that referenced
this pull request
Nov 14, 2025
* fix vision attn for whisper/llama4 Signed-off-by: Yan Ma <yan.ma@intel.com> * address comments Signed-off-by: Yan Ma <yan.ma@intel.com> --------- Signed-off-by: Yan Ma <yan.ma@intel.com>
dik654
pushed a commit
to dik654/vllm-for-study
that referenced
this pull request
Nov 18, 2025
New Industry Use Cases (vllm-project#21-30): - vllm-project#21 Game Development: AI game testing + balance tuning - vllm-project#22 Construction: Vision AI safety inspection - vllm-project#23 Agriculture/Smart Farm: Crop monitoring + pest detection - vllm-project#24 Government/Public: Document automation + citizen services - vllm-project#25 Energy/Utilities: Grid monitoring + anomaly detection - vllm-project#26 Environment/Sustainability: Carbon tracking + ESG reporting - vllm-project#27 Fashion/Apparel: Trend analysis + inventory optimization - vllm-project#28 Sports/Fitness: Performance analytics + tactical analysis - vllm-project#29 Automotive/Mobility: Autonomous driving simulation - vllm-project#30 Space/Aerospace: Satellite image analysis Advanced Architecture Patterns: 1. Event-Driven Pattern: Webhook → Event Bus → Agent triggers 2. Streaming Pattern: Large dataset processing with chunking 3. Batch Processing Pattern: Celery-based parallel processing 4. Circuit Breaker Pattern: Fault tolerance + auto recovery 5. CQRS + Event Sourcing: Command/Query separation 6. Saga Pattern: Distributed transaction management Guide now covers: - 30+ industry-specific MCP implementations - 6 production-ready architecture patterns - Real-world scalability solutions - Enterprise integration strategies - Total: 8,672 lines (from 7,249)
1 task done
1 task done
amd-xavierwang
pushed a commit
to amd-xavierwang/vllm
that referenced
this pull request
Apr 8, 2026
Asymmetric AWQ models (zero_points=True) previously fell back to the slower ConchLinearKernel. This adds zero-point support to the skinny GEMM int4 kernels so asymmetric models route through the fast path. A `bool HAS_ZERO_POINTS` template parameter (default false) is added to the existing kernel templates. When false, the compiler eliminates all zero-point code, producing identical codegen to the current path. When true, per-group zero-point subtraction is inserted after dequantization using the algebraic identity: (nibble - zp) = (nibble - 8) - (zp - 8) The existing dequant (which always produces nibble-8) is unchanged; pre-adjusted (zp-8) values are subtracted afterward.
jxmorris12
added a commit
to jxmorris12/vllm
that referenced
this pull request
Apr 21, 2026
starpit
added a commit
to starpit/vllm
that referenced
this pull request
May 26, 2026
prefill + skip-logging eprintlns - codegen.rs: wire I::RopeAppend in dispatch_instruction_to_render so prefill canonicals containing fused_qkv_rope_cache reach render_fused_qkv_rope_cache and emit role bodies. Without this arm the dispatch fell through to `_ => Ok(None)` which silently dropped emit_for_canonical / inventory::submit, while build_mega_tape kept emitting — explaining the "build_mega_tape says emitted but cu-emit inventory empty" mismatch. - cuda_emit/render.rs: gate_up_activate_mul prefill rewritten to the WGMMA_N + N_MICRO microtile loop pattern matching the gemm and gemm_add prefill renders, so the inner mma uses cols ∈ [16, 256] step 16 (TK 2.0 wgmma::base specialization range from third_party/thunderkittens/include/ops/group/mma/base/base.cuh:28-43). - ferrite-forward-macro/src/codegen.rs: emit_slice_render_with_loop_expansion now eprintln!s the canonical / instruction kind on each render-skip return so silent drops are visible in cu-emit logs. Pod two-pass build green at 4m 57s under the 5-min cicc cap; serve + curl E2E on Llama-3.2-1B-Instruct returns coherent generation (\" Paris. The Eiffel Tower is located in Paris...\") for the 5 canonicals currently emitted (m_1_sk_128/192, m_8_sk_128, m_64_sk_128/512). M=512 / M=4096 prefill canonicals are tracked separately under tasks vllm-project#23/vllm-project#24. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
1 task done
heliubj18
added a commit
to heliubj18/vllm
that referenced
this pull request
Aug 8, 2026
401 of the step's 402 cases need no device, but test_scheduler.py::test_async_scheduling_pp_allows_rescheduling_with_output_placeholders builds a ParallelConfig with pipeline_parallel_size=2. It never touches a second card - the step allocates 0.0 GiB across its whole run - but on a CUDA image ParallelConfig validates world size against the visible GPU count and rejects 2. Upstream never sees this: the step depends on image-build-cpu, where that check is inert. So the marking upstream gives a step describes the image it runs on, not what its cases ask for, and _gpus_for() had no way to say otherwise. Adds docker.devices_by_step, a per-key override of the device count that indexes into the existing gpus_by_devices map. Only v1-others-cpu uses it; the generated pipeline differs from before by exactly one line, and the other CPU-marked steps still get "device=0". Two cards costs nothing when the step allocates no VRAM, and leaving it at one meant a guaranteed failure every run - which is expensive now that fail_fast is back on. In vllm-project#24 it cancelled Basic Models Test and Torch Stable ABI Audit, both of which pass (3m19s in vllm-project#17). I had written this case off as "not worth handling" without accounting for that. Co-authored-by: Claude Opus 5
wenwzhan
pushed a commit
to wenwzhan/vllm-kimi-fz
that referenced
this pull request
Aug 12, 2026
…cheduling [Kimi-K3][ROCm] Do not run DSpark under both data parallelism and async scheduling
Chloe-orz
pushed a commit
to Chloe-orz/vllm-maintomain
that referenced
this pull request
Nov 5, 2026
Merge vllm-data-comm/vllm main at 928b990 while retaining the local gate fix b40d5cf as the first parent. Preserve all existing commits, original SHAs and ancestry without rebasing, reordering or squashing either history. Bring in PR vllm-project#54: remove the long-request exclusive prefill batch rule, its model-specific threshold argument and the associated upstream test and documentation updates. Keep the incoming resume-first, FCFS picker within its token budget and native request-slot limit. The scheduler auto-merges without textual conflicts. Retain the local explicit num_running_reqs argument to is_gated and compute free_seq from len(self.running). Waiting requests continue to affect routing scores but cannot reserve native slots before admission. Preserve shared edge slots across DP routes, the in-flight resume exemption, per-DP pending caps and the current Host Staging token budget. Relative to the incoming main, only the existing four-file gate-fix patch remains. Pair with vllm-ascend old-main-1026 fast-forwarded to 14602594f485da18702e57d8c10dade8c700b734. That upstream history includes our graph-placeholder storage fix through PR vllm-project#24, the cloud DSV4 Hash routing repair in PR vllm-project#26 and the dead FlashComm v1 sharding removal in PR vllm-project#27. Keep the runner-owned placeholder and both slice-based uses. Validation: source/diff review only. No tests, lint, pre-commit, CI, syntax checks, startup or benchmarks were run, as requested. NPU validation is pending. Backup branches were created before either update. Signed-off-by: zzyccccccccc <zzyccccccccc@noreply.gitcode.com>
cxywhite
added a commit
to cxywhite/vllm
that referenced
this pull request
Nov 8, 2026
…ucible config - vllm submodule -> 6e1011be4 (tau scalar EMA sum unit + TTFT-margin force-admit + WT_SIMPLIFY_PREDICT predicted_remaining). - test_evaluation_scheme4_rr_compare.sh: pass --activation-predict to decode so scheme4 (greedy) tracks predictive pressure too (admission-only A/B, pitfall vllm-project#24). - experiment_log.md: o2000/o1600/o1200/o800 sweep, TTFT-margin finding (o2000 margin=250 +24%, o1600 margin=100 +4.3%), 复现约定 + 已证伪路径 + result index. - winning_config.md: reproducible config snapshot (model/PD/gpu_util/dataset/theta/margin). - admission_ab_base: qwen_o800/o1200/o1600/o2000 (output-mean scaled), burstgpt_real (real trace); traces sanitized sk- -> sk_ for secret scanning.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR implements a basic and not highly optimized kernel to support cached KV with multiple import prompts.