## What's Changed * Update news entry for cache-dit release by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/395 * misc: Update .gitignore to keep git from user local settings by @KarhouTam in https://github.com/xlite-dev/LeetCUDA/pull/400 * fix: remove redundant CP_ASYNC_WAIT_GROUP by @YangFan0918 in https://github.com/xlite-dev/LeetCUDA/pull/401 * Change device retrieval method to use torch.device because Triton 3.0… by @clveryang in https://github.com/xlite-dev/LeetCUDA/pull/402 * feat: add lmyybh's zhihu blogs by @lmyybh in https://github.com/xlite-dev/LeetCUDA/pull/403 * fix: update mat_transpose_f32_row2col2d_kernel by @AgainstEntropy in https://github.com/xlite-dev/LeetCUDA/pull/404 * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/406 * misc: automated submodule update by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/407 * chore: update blogs indexes by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/408 ## New Contributors * @KarhouTam made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/400 * @YangFan0918 made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/401 * @clveryang made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/402 * @lmyybh made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/403 * @AgainstEntropy made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/404 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.16...v3.0.17

## What's Changed * misc: automated submodule update by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/369 * softmax.cu spell error by @napleon-liu in https://github.com/xlite-dev/LeetCUDA/pull/371 * feat: taylorseer zhihu blog by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/373 * cache-dit: Qwen-Image 1.5x↑ blog by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/374 * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/375 * fix: bracket error in mat_transpose.cu by @wxwxwwxxx in https://github.com/xlite-dev/LeetCUDA/pull/376 * fix: add tail-case handling for elementwise_add_f16x8_pack_kernel to … by @lhycms in https://github.com/xlite-dev/LeetCUDA/pull/380 * feat: add Wan2.2-MoE 2.4x speedup zhihu blog by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/383 * feat: Qwen-Image-Lightning & HunyuanImage-2.1 SpeedUp blog by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/387 ## New Contributors * @napleon-liu made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/371 * @wxwxwwxxx made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/376 * @lhycms made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/380 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.14...v3.0.15

## What's Changed * pin HelloGitHub by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/342 * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/343 * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/344 * fix comments by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/345 * fix comments by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/346 * fix all2all example by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/347 * fix softmax.cu by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/348 * fix CONTRIBUTE.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/349 * feat: add torch.compile blogs by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/350 * feat: triton layer-norm by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/351 * fix flash-attn comments by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/354 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.12...v3.0.13

## What's Changed * feat: add cute hgemv implement by @kitecats in https://github.com/xlite-dev/LeetCUDA/pull/331 * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/333 * feat: add a cute bank-free mat transpose vectorize impelment by @kitecats in https://github.com/xlite-dev/LeetCUDA/pull/334 * bugfix: fix layernorm & rmsnorm f16 overflow by @hebangwen in https://github.com/xlite-dev/LeetCUDA/pull/335 * Bugfix: fix a compilation error by @lixiaoquan in https://github.com/xlite-dev/LeetCUDA/pull/336 ## New Contributors * @hebangwen made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/335 * @lixiaoquan made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/336 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.10...v3.0.11

## What's Changed * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/322 * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/323 * Fix: missing source by @botbw in https://github.com/xlite-dev/LeetCUDA/pull/325 * Use 128-bit data loading by @kitecats in https://github.com/xlite-dev/LeetCUDA/pull/326 * Create FUNDING.yml by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/327 * Add open-collective badge by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/328 * Update open-collective contributors badge by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/329 ## New Contributors * @kitecats made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/326 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.9...v3.0.10

## What's Changed * feat: add some torch.distributed examples by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/313 * feat: add some torch.distributed examples by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/315 * feat: add a naive CuTe flash-attn by @botbw in https://github.com/xlite-dev/LeetCUDA/pull/314 * fix(kernels): correct typo in LayerNorm kernel at line 73 110 346 443 by @nxdxml in https://github.com/xlite-dev/LeetCUDA/pull/317 * misc: manually update submodules by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/318 * chore: add naive cute flash-attn index by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/319 * add triton merge_attn_states zhihu blog by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/320 ## New Contributors * @nxdxml made their first contribution in https://github.com/xlite-dev/LeetCUDA/pull/317 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.8...v3.0.9

## What's Changed * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/308 * misc: add triton vector add zhihu blog by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/310 * Update README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/311 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.7...v3.0.8

## What's Changed * Update mat-transpose/README.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/300 * feat: add triton fused-softmax by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/301 * misc: add pre-commit & format by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/302 * misc: add developer guide by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/303 * misc: add developer guide by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/304 * misc: fix typo by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/305 * Update CONTRIBUTE.md by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/306 * feat: update pre-commit max-length=80 by @DefTruth in https://github.com/xlite-dev/LeetCUDA/pull/307 **Full Changelog**: https://github.com/xlite-dev/LeetCUDA/compare/v3.0.6...v3.0.7