FlagGems
CLAUDE.md1.2k starsChanged 4 months agoApache-2.0View on GitHub
1 other project uses a version of this.
Trending
More talk lately.about the whole repository
Most recently taken up by
- NVIDIA/TensorRT-LLMword for word, first seen Aug 17, 2026
Who's talking about flagos-ai/FlagGems
- linkedin/Liger-Kernel“[RFC] Code Organization for New Backend Support”GitHub · Dec 4, 2025 · 4 reactions
- gpu-mode/resource-stream“Docs: Can we add FlagGems under the session of "Triton Kernels / Examples"”GitHub · Dec 22, 2025 · 1 reactions
- flagos-ai/Torch-FL“[AI] MetaX SDPA cannot compile FlagGems 5.4.0 attention kernel”GitHub · Nov 10, 2026
- flagos-ai/Torch-FL“[Hygon] FP32 addmm: four CPU-reference discrepancies across DAS boxing and FlagGems”GitHub · Nov 10, 2026
- flagos-ai/Torch-FL“[AI][DCU] Track FlagGems public global aminmax empty-input error”GitHub · Nov 9, 2026
The biggest 5 of 46 mentions linking to the repository that our crawl found. Each line is their own headline.
See all 46
Talked about on GitHub (46 times).
On GitHub · 46
- Dec 4, 2025[RFC] Code Organization for New Backend Support — linkedin/Liger-Kernel4 reactions · 3 comments
- Dec 22, 2025Docs: Can we add FlagGems under the session of "Triton Kernels / Examples" — gpu-mode/resource-stream1 reaction
- Nov 10, 2026[AI] MetaX SDPA cannot compile FlagGems 5.4.0 attention kernel — flagos-ai/Torch-FL0 reactions
- Nov 10, 2026[Hygon] FP32 addmm: four CPU-reference discrepancies across DAS boxing and FlagGems — flagos-ai/Torch-FL0 reactions
- Nov 9, 2026[AI][DCU] Track FlagGems public global aminmax empty-input error — flagos-ai/Torch-FL0 reactions
41 more on GitHub
- Nov 9, 2026[AI][DCU] Track FlagGems public global aminmax NaN propagation — flagos-ai/Torch-FL0 reactions
- Nov 9, 2026[AI][DCU] Track FlagGems _to_copy fractional-to-bool conversion — flagos-ai/Torch-FL0 reactions
- Nov 9, 2026[BUG][DCU] Advanced indexing rejects valid non-contiguous row and column index views — flagos-ai/Torch-FL0 reactions
- Nov 4, 2026Add linear-5379: reshape non-contiguous linear input instead of view — FLAG-OP/flag-fixes0 reactions · 4 comments
- Nov 2, 2026[KMCompiler] Rewrite the operator using loop vectorization — flagos-ai/FlagGems-vllm0 reactions
- Oct 29, 2026Collect original group_mm H800 records with native per-group CUDA composite — zhaojs12345/ops-baseline0 reactions
- Oct 27, 2026Add clean H800 baseline for mkldnn_rnn_layer — zhaojs12345/ops-baseline0 reactions
- Oct 27, 2026[QC][MThreads] Optimize W8A16 FP8 RMSNorm for real shapes — July-h5kf3/FlagGems0 reactions
- Oct 26, 2026[AI][BUG][DCU] Track FlagGems round rejecting strided tensors — flagos-ai/Torch-FL0 reactions
- Oct 26, 2026[AI][BUG][DCU] Track FlagGems remainder errors for infinite divisors — flagos-ai/Torch-FL0 reactions
- Oct 26, 2026[AI][BUG][DCU] Track FlagGems clamp/clamp_min losing NaN propagation — flagos-ai/Torch-FL0 reactions
- Oct 26, 2026[AI][BUG][DCU] Track FlagGems rand/rand_like returning 1.0 in BF16/FP16 — flagos-ai/Torch-FL0 reactions · 2 comments
- Oct 26, 2026[Bug Fix] Guard attn_res on TLE load capability — flagos-ai/FlagGems-vllm0 reactions
- Oct 22, 2026[NVIDIA] Add guarded MiniMax TP8 W8A8 MoE profiles and mixed-policy fusion — flagos-ai/FlagGems-vllm0 reactions
- Oct 20, 2026[KMCompiler][Ascend] perf-opt for fused_experts_impl — flagos-ai/FlagGems-vllm0 reactions
- Oct 20, 2026Collect and replay explicit native CPU and host baselines — zhaojs12345/ops-baseline0 reactions
- Oct 19, 2026[AI][DCU] Track native_layer_norm CPU-parity regression and boxing control for FlagGems #7247 — flagos-ai/Torch-FL0 reactions
- Oct 18, 2026[AI][Performance][DCU] R5 same-device DAS/FlagOS/boxing survey: 29 cases and classified follow-ups — flagos-ai/Torch-FL0 reactions
- Oct 18, 2026[AI][Performance][DCU] Track FlagGems repeat_interleave output_size readback and R5 integration — flagos-ai/Torch-FL0 reactions
- Oct 18, 2026[AI][Performance][DCU] Track FlagGems Hygon int64 sort multistage latency on R5 — flagos-ai/Torch-FL0 reactions
- Oct 15, 2026Explicit complex dtype becomes float32 in zeros on the default DCU backend — flagos-ai/Torch-FL0 reactions · 1 comment
- Oct 14, 2026Route Ascend GDN decode through the FlagGems operator — iteratorlee/sglang-plugin-FL0 reactions
- Oct 12, 2026docs(wheel-info): give each package a source-tarball download URL — flagos-ai/build-infra0 reactions
- Oct 10, 2026[AI][DCU] Track batched cdist failure in the FlagGems _euclidean_dist route — flagos-ai/Torch-FL0 reactions
- Oct 10, 2026[AI][DCU] Track FlagGems exp float64 precision loss on the default route — flagos-ai/Torch-FL0 reactions
- Oct 10, 2026[AI][DCU] Track FlagGems amin dtype rejection on the default route — flagos-ai/Torch-FL0 reactions
- Sep 27, 2026[OP] Route Hygon FP32 scaled_dot_product_attention to a supported fused backend — flagos-ai/Torch-FL0 reactions · 1 comment
- Sep 27, 2026[MThreads][TLE] Add persistent GEMM pipeline and SPMC synchronization support — flagos-ai/FlagTree0 reactions · 10 comments
- Sep 24, 2026Fix FlagOS attention layouts and packed sequence boundaries — flagos-ai/TransformerEngine-FL0 reactions
- Sep 21, 2026[KMCompiler][Ascend][test] perf-opt for pack_seq_triton/unpack_seq_triton on ascend — flagos-ai/FlagGems-vllm0 reactions
- Sep 18, 2026Fix amin_ accuracy test to exercise the inplace Gems operator — flagos-ai/FlagGems-Experimental0 reactions
- Sep 18, 2026Fix add_relu benchmark to measure the fused Gems operator — flagos-ai/FlagGems-Experimentalsays it tested it0 reactions
- Sep 16, 2026fix(benchmark): sync backend functional batch norm target from FlagGems #6352 — flagos-ai/FlagGems-Experimentalsays it tested it0 reactions · 2 comments
- Sep 16, 2026fix(benchmark): sync explicit batch norm targets from FlagGems #6338 — flagos-ai/FlagGems-Experimentalsays it tested it0 reactions
- Sep 2, 2026Arm: integrate Qwen packed W4A8 and GDN on CPU — flagos-ai/vllm-plugin-FL0 reactions · 3 comments
- Aug 26, 2026[XPU] tl.store 在向量计算之后写入标量时,实际写入的是过期的寄存器值 — flagos-ai/FlagTree0 reactions
- Aug 18, 2026[Ascend] flagos 融合算子精度污染: prefer=flagos 输出乱码(prefer=vendor 正常) — flagos-ai/sglang-plugin-FL0 reactions
- Aug 7, 2026[KMCompiler] perf-opt grouped_topk for DeepSeek-v3.2 — flagos-ai/FlagGems-vllm0 reactions · 1 comment
- Jul 7, 2026fix: add in-place input comparison in test_unsqueeze_ — XDYuanzhuLee/FlagGems0 reactions
- Jun 24, 2026This project is not maintained anymore 😢 — github/command0 reactions · 1 comment
- Mar 10, 2026算子开发流程 — siliconflow/FlagGems0 reactions
Other versions1 project uses a version of this.
Word for word · 1
- NVIDIA/TensorRT-LLM★ 15k100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repositoryOne other file it gives its agents.
- AGENTS.mdAGENTS.md
The file
We found no rules to list in this file: it reads as prose or reference rather than instructions.
Files
In ./CLAUDE.md
@AGENTS.md
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

