biondizzle 40dd56eac2 KV-1: Fix shared memory corruption in block_reduce ...

block_reduce_sum/max write to smem[0..n_warps-1] but we passed &s_amax
(single float). For 128 threads / 4 warps, this wrote 4 floats starting
at &s_amax, corrupting adjacent shared variables (s_inv_rms, s_vals).

Fix: use s_scratch[8] array (4 for sum, 4 for max) with proper sizing.

2026-06-02 09:49:12 +00:00

..

cache

E1: Wire LayerCacheHandle gather methods + CUDA gather kernels

2026-05-30 21:09:21 +00:00

kernels

KV-1: Fix shared memory corruption in block_reduce

2026-06-02 09:49:12 +00:00

layers

wip: SE fused SwiGLU deinterleave fix

2026-06-02 08:41:00 +00:00

loader

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00

model

P0 COMPLETE: Eliminate ALL .item() CPU-GPU syncs from NVFP4 activation path

2026-06-01 21:05:03 +00:00

ops

fix: rope_cuda path — kernels/cuda not ops/cuda

2026-06-02 09:06:36 +00:00

reference

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00

__init__.py

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00