mahorozte e9e661bd59 CUDA: remove unnecessary warp reduce in FA (ggml/1032) ...

* kqmax_new_j in every thread within warp is same after operate at line 199,this reduce can be omit

* same problem in vec32

---------

Co-authored-by: ZhaoXiaoYu <zhao.xiaoyu@zte.com.cn>

2024-12-03 20:04:49 +02:00

..

template-instances

…

vendors

Add some minimal optimizations for CDNA (#10498)

2024-11-27 17:10:08 +01:00

acc.cu

…

acc.cuh

…

arange.cu

…

arange.cuh

…

argmax.cu

cuda : optimize argmax (#10441)

2024-11-21 18:18:50 +01:00

argmax.cuh

ggml/ex: calculate accuracy in graph, adapt MNIST (ggml/980)

2024-10-03 21:17:26 +03:00

argsort.cu

…

argsort.cuh

…

binbcast.cu

…

binbcast.cuh

…

clamp.cu

…

clamp.cuh

…

CMakeLists.txt

cmake : enable warnings in llama (#10474)

2024-11-26 14:18:08 +02:00

common.cuh

Add some minimal optimizations for CDNA (#10498)

2024-11-27 17:10:08 +01:00

concat.cu

…

concat.cuh

…

conv-transpose-1d.cu

…

conv-transpose-1d.cuh

…

convert.cu

…

convert.cuh

…

count-equal.cu

ggml: fix zero division in ‘dne’ calculation in CUDA COUNT_EQUAL operator when ‘ne’ is small (#10213)

2024-11-09 08:35:46 +01:00

count-equal.cuh

ggml/ex: calculate accuracy in graph, adapt MNIST (ggml/980)

2024-10-03 21:17:26 +03:00

cpy.cu

…

cpy.cuh

increase cuda_cpy block size (ggml/996)

2024-10-26 10:33:56 +03:00

cross-entropy-loss.cu

…

cross-entropy-loss.cuh

…

dequantize.cuh

…

diagmask.cu

…

diagmask.cuh

…

fattn-common.cuh

ggml : build backends as libraries (#10256)

2024-11-14 18:04:35 +01:00

fattn-tile-f16.cu

ggml : build backends as libraries (#10256)

2024-11-14 18:04:35 +01:00

fattn-tile-f16.cuh

…

fattn-tile-f32.cu

ggml : build backends as libraries (#10256)

2024-11-14 18:04:35 +01:00

fattn-tile-f32.cuh

…

fattn-vec-f16.cuh

CUDA: remove unnecessary warp reduce in FA (ggml/1032)

2024-12-03 20:04:49 +02:00

fattn-vec-f32.cuh

CUDA: remove unnecessary warp reduce in FA (ggml/1032)

2024-12-03 20:04:49 +02:00

fattn-wmma-f16.cuh

ggml : build backends as libraries (#10256)

2024-11-14 18:04:35 +01:00

fattn.cu

metal : optimize FA kernels (#10171)

2024-11-08 13:47:22 +02:00

fattn.cuh

…

getrows.cu

…

getrows.cuh

…

ggml-cuda.cu

Add some minimal optimizations for CDNA (#10498)

2024-11-27 17:10:08 +01:00

im2col.cu

CUDA: fix 1D im2col, add tests (ggml/993)

2024-10-23 16:50:02 +03:00

im2col.cuh

…

mma.cuh

…

mmq.cu

Add some minimal optimizations for CDNA (#10498)

2024-11-27 17:10:08 +01:00

mmq.cuh

Add some minimal optimizations for CDNA (#10498)

2024-11-27 17:10:08 +01:00

mmv.cu

CUDA: remove DMMV, consolidate F16 mult mat vec (#10318)

2024-11-17 09:09:55 +01:00

mmv.cuh

CUDA: remove DMMV, consolidate F16 mult mat vec (#10318)

2024-11-17 09:09:55 +01:00

mmvq.cu

Add some minimal optimizations for CDNA (#10498)

2024-11-27 17:10:08 +01:00

mmvq.cuh

…

norm.cu

…

norm.cuh

…

opt-step-adamw.cu

ggml: new optimization interface (ggml/988)

2024-11-17 08:30:29 +02:00

opt-step-adamw.cuh

…

out-prod.cu

…

out-prod.cuh

…

pad.cu

…

pad.cuh

…

pool2d.cu

…

pool2d.cuh

…

quantize.cu

cuda : optimize argmax (#10441)

2024-11-21 18:18:50 +01:00

quantize.cuh

…

rope.cu

…

rope.cuh

…

scale.cu

…

scale.cuh

…

softmax.cu

…

softmax.cuh

…

sum.cu

ggml : build backends as libraries (#10256)

2024-11-14 18:04:35 +01:00

sum.cuh

…

sumrows.cu

…

sumrows.cuh

…

tsembd.cu

…

tsembd.cuh

…

unary.cu

…

unary.cuh

…

upscale.cu

…

upscale.cuh

…

vecdotq.cuh

…

wkv6.cu

Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (#10133)

2024-11-07 15:19:10 +08:00

wkv6.cuh

Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (#10133)

2024-11-07 15:19:10 +08:00