Johannes Gäßler 9b596417af CUDA: quantized KV support for FA vec (#7527) ...

* CUDA: quantized KV support for FA vec

* try CI fix

* fix commented-out kernel variants

* add q8_0 q4_0 tests

* fix nwarps > batch size

* split fattn compile via extern templates

* fix flake8

* fix metal tests

* fix cmake

* make generate_cu_files.py executable

* add autogenerated .cu files

* fix AMD

* error if type_v != FP16 and not flash_attn

* remove obsolete code

2024-06-01 08:44:14 +02:00

..

template-instances

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

acc.cu

…

acc.cuh

…

arange.cu

…

arange.cuh

…

argsort.cu

ggml : mul_mat_id use the same tensor for all the experts (#6387)

2024-04-03 16:07:05 +03:00

argsort.cuh

…

binbcast.cu

ggml : group all experts in a single ggml_mul_mat_id (#6505)

2024-04-18 15:18:48 +02:00

binbcast.cuh

…

clamp.cu

Introduction of CUDA Graphs to LLama.cpp (#6766)

2024-05-08 22:55:49 +02:00

clamp.cuh

…

common.cuh

update HIP_UMA #7399 (#7414)

2024-05-28 01:40:47 +02:00

concat.cu

cuda : non-cont concat support (#7610)

2024-05-29 15:38:26 +03:00

concat.cuh

…

convert.cu

ggml : drop support for QK_K=64 (#7473)

2024-05-23 10:00:21 +03:00

convert.cuh

llama : add Command R Plus support (#6491)

2024-04-09 11:16:13 +03:00

cpy.cu

Introduction of CUDA Graphs to LLama.cpp (#6766)

2024-05-08 22:55:49 +02:00

cpy.cuh

Introduction of CUDA Graphs to LLama.cpp (#6766)

2024-05-08 22:55:49 +02:00

dequantize.cuh

llama : add Command R Plus support (#6491)

2024-04-09 11:16:13 +03:00

diagmask.cu

…

diagmask.cuh

…

dmmv.cu

ggml : drop support for QK_K=64 (#7473)

2024-05-23 10:00:21 +03:00

dmmv.cuh

sync : ggml (#6351)

2024-03-29 17:45:46 +02:00

fattn-common.cuh

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

fattn-tile-f16.cu

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

fattn-tile-f16.cuh

CUDA: faster large batch FA without tensor cores (#7314)

2024-05-17 18:54:52 +02:00

fattn-tile-f32.cu

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

fattn-tile-f32.cuh

CUDA: faster large batch FA without tensor cores (#7314)

2024-05-17 18:54:52 +02:00

fattn-vec-f16.cuh

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

fattn-vec-f32.cuh

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

fattn-wmma-f16.cuh

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

fattn.cu

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

fattn.cuh

ggml : add Flash Attention (#5021)

2024-04-30 12:16:08 +03:00

getrows.cu

…

getrows.cuh

…

im2col.cu

…

im2col.cuh

…

mmq.cu

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00

mmq.cuh

…

mmvq.cu

Introduction of CUDA Graphs to LLama.cpp (#6766)

2024-05-08 22:55:49 +02:00

mmvq.cuh

…

norm.cu

ggml : fix YARN + add tests + add asserts (#7617)

2024-05-29 20:17:31 +03:00

norm.cuh

…

pad.cu

…

pad.cuh

…

pool2d.cu

…

pool2d.cuh

…

quantize.cu

llama : add Command R Plus support (#6491)

2024-04-09 11:16:13 +03:00

quantize.cuh

llama : add Command R Plus support (#6491)

2024-04-09 11:16:13 +03:00

rope.cu

ggml : fix YARN + add tests + add asserts (#7617)

2024-05-29 20:17:31 +03:00

rope.cuh

…

scale.cu

Introduction of CUDA Graphs to LLama.cpp (#6766)

2024-05-08 22:55:49 +02:00

scale.cuh

…

softmax.cu

CUDA: deduplicate FlashAttention code (#7352)

2024-05-18 12:36:25 +02:00

softmax.cuh

…

sumrows.cu

…

sumrows.cuh

…

tsembd.cu

…

tsembd.cuh

…

unary.cu

feat: implemented sigmoid function (ggml/806)

2024-05-11 15:38:34 +03:00

unary.cuh

feat: implemented sigmoid function (ggml/806)

2024-05-11 15:38:34 +03:00

upscale.cu

ggml : add ggml_upscale_ext (ggml/814)

2024-05-15 13:23:33 +03:00

upscale.cuh

…

vecdotq.cuh

CUDA: quantized KV support for FA vec (#7527)

2024-06-01 08:44:14 +02:00