public_git/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-10-03 13:03:07 +02:00
Code Issues Packages Projects Releases Wiki Activity
Files
fbdbbc7847e030166531e3b3edf30faac2e693a4
whisper.cpp/ggml/src/ggml-cuda
T
History
uvos 94d6e24e4c CUDA/HIP: improve access patterns in im2col (llama/28013)
2026-09-23 20:46:47 +03:00
..
template-instances
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
vendors
HIP: Enable AllReduce for ROCm (llama/27825)
2026-09-23 20:46:47 +03:00
acc.cu
…
acc.cuh
…
add-id.cu
…
add-id.cuh
…
allreduce.cu
HIP: Enable AllReduce for ROCm (llama/27825)
2026-09-23 20:46:47 +03:00
allreduce.cuh
HIP: Enable AllReduce for ROCm (llama/27825)
2026-09-23 20:46:47 +03:00
arange.cu
…
arange.cuh
…
argmax.cu
ggml : use WARP_SIZE/2 for argmax reduction offset (llama/18092)
2025-12-18 08:20:56 +02:00
argmax.cuh
…
argsort.cu
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
argsort.cuh
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
binbcast.cu
ggml : address integer overflows in binary ops CUDA implementation (llama/24706)
2026-06-26 16:03:57 +03:00
binbcast.cuh
CUDA: fuse muls (llama/21665)
2026-04-30 11:29:05 +03:00
clamp.cu
…
clamp.cuh
…
CMakeLists.txt
CUDA: replace GGML_FA_ALL_QUANTS with GGML_FA_QUANTS, more control over what is compiled (llama/28079)
2026-09-14 20:45:06 +03:00
col2im-1d.cu
Ggml/cuda col2im 1d (llama/24417)
2026-06-19 12:53:43 +03:00
col2im-1d.cuh
Ggml/cuda col2im 1d (llama/24417)
2026-06-19 12:53:43 +03:00
common.cuh
ggml-cuda: fallback to F32 on device without BF16 hardware acceleration (llama/28846)
2026-09-14 20:45:06 +03:00
concat.cu
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
concat.cuh
…
conv2d-dw.cu
…
conv2d-dw.cuh
…
conv2d-transpose.cu
CUDA & CPU: support F32 kernel type for CONV_TRANSPOSE_2D (llama/17094)
2026-03-29 15:04:36 +03:00
conv2d-transpose.cuh
CUDA & CPU: support F32 kernel type for CONV_TRANSPOSE_2D (llama/17094)
2026-03-29 15:04:36 +03:00
conv2d.cu
Test support for alternative conv layout (llama/25617)
2026-08-04 13:37:47 +03:00
conv2d.cuh
CUDA: add conv2d (llama/15635)
2025-09-20 13:42:44 +03:00
conv-transpose-1d.cu
ggml-cuda: optimize conv_transpose_1d indexing (llama/25310)
2026-07-10 13:06:42 +03:00
conv-transpose-1d.cuh
…
convert.cu
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
convert.cuh
CUDA: fix BF16 FA compilation (llama/20865)
2026-03-29 15:04:36 +03:00
count-equal.cu
…
count-equal.cuh
…
cp-async.cuh
…
cpy-utils.cuh
cuda : support non-contiguous i32 to i32 copy (llama/17326)
2025-12-12 17:53:06 +02:00
cpy.cu
cuda : enable i16 and i32 for DUP (llama/28897)
2026-09-23 20:46:47 +03:00
cpy.cuh
cuda : remove legacy copy-op pointer indirection code (llama/16485)
2025-10-15 09:29:17 +03:00
cross-entropy-loss.cu
…
cross-entropy-loss.cuh
…
cumsum.cu
sampling : add support for backend sampling (llama/17004)
2026-01-14 09:11:59 +02:00
cumsum.cuh
Add support for CUMSUM and TRI for CUDA. (llama/17584)
2025-12-12 17:53:17 +02:00
dequantize.cuh
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
diag.cu
Add DIAG for CUDA (llama/17873)
2025-12-12 17:53:23 +02:00
diag.cuh
Add DIAG for CUDA (llama/17873)
2025-12-12 17:53:23 +02:00
diagmask.cu
…
diagmask.cuh
…
dsv4-hc.cu
qwen4exp: add hc ops (llama/28901)
2026-09-23 20:46:47 +03:00
dsv4-hc.cuh
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
fattn-common.cuh
CUDA/HIP: Flash Attention tuning (gfx1201) (llama/28102)
2026-09-14 20:45:06 +03:00
fattn-mma-f16.cuh
HIP: fattn-mma: use fp32 accumulation on MFMA devices (llama/28576)
2026-09-23 20:46:47 +03:00
fattn-swizzle.cuh
CUDA: XOR swizzle flash attn K,V smem fp16 tiles (llama/25635)
2026-09-04 13:39:37 +03:00
fattn-tile.cu
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
fattn-tile.cuh
CUDA + ggml: add sparse-fa for DSV4/GLM (llama/27970)
2026-09-04 13:39:40 +03:00
fattn-vec.cuh
cuda: fixes races in mmid and mmf (llama/28475)
2026-09-14 20:45:06 +03:00
fattn.cu
CUDA/HIP: Flash Attention tuning (gfx1201) (llama/28102)
2026-09-14 20:45:06 +03:00
fattn.cuh
cuda: reserve space for quantize kv-cache at startup (llama/23907)
2026-06-08 14:36:36 +03:00
fill.cu
ggml : allow fill node alloc inplace (llama/17870)
2025-12-12 17:53:22 +02:00
fill.cuh
cuda : add FILL op support (llama/17851)
2025-12-12 17:53:22 +02:00
fwht.cu
CUDA: missing PDL sync for FWHT, better fallback (llama/23690)
2026-05-29 09:47:30 +03:00
fwht.cuh
CUDA: missing PDL sync for FWHT, better fallback (llama/23690)
2026-05-29 09:47:30 +03:00
gated_delta_net.cu
Remove redundant CUDA copies after gated_delta_net. (llama/23940)
2026-07-10 13:06:42 +03:00
gated_delta_net.cuh
Remove redundant CUDA copies after gated_delta_net. (llama/23940)
2026-07-10 13:06:42 +03:00
getrows.cu
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
getrows.cuh
…
ggml-cuda.cu
qwen4exp: add hc ops (llama/28901)
2026-09-23 20:46:47 +03:00
gla.cu
…
gla.cuh
…
im2col.cu
CUDA/HIP: improve access patterns in im2col (llama/28013)
2026-09-23 20:46:47 +03:00
im2col.cuh
ggml: add ops for WAN video model (cuda && cpu) (llama/15669)
2025-09-20 13:42:49 +03:00
lightning-indexer.cu
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
lightning-indexer.cuh
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
mean.cu
cuda: support row-contiguous SUM_ROWS (llama/26308)
2026-09-23 20:46:47 +03:00
mean.cuh
…
mma.cuh
HIP: RDNA3 mma FA, faster AMD transpose, tune AMD (llama/22880)
2026-05-25 12:26:07 +03:00
mmf.cu
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
mmf.cuh
cuda: fixes races in mmid and mmf (llama/28475)
2026-09-14 20:45:06 +03:00
mmid.cu
cuda: fixes races in mmid and mmf (llama/28475)
2026-09-14 20:45:06 +03:00
mmid.cuh
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
mmq-config-ampere.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-blackwell.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-cdna.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-gcn.cuh
ggml-cuda: hip add specific config table for AMD GCN (llama/27841)
2026-09-14 20:45:06 +03:00
mmq-config-pascal-dp4a.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-pascal-older.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-rdna2.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-rdna3-5.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-rdna3.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-config-rdna4.cuh
Revert "CUDA: size routed MoE MMQ N-tiles from typical expert width on RDNA3 (#24546)" (llama/28551)
2026-09-14 20:45:06 +03:00
mmq-load-tiles.cuh
hip : optimize Q2_0 dot-product path for gfx1201 (llama/26753)
2026-09-04 13:39:34 +03:00
mmq-vec-dot.cuh
ggml-cuda : remove unused vars (llama/28235)
2026-09-04 13:39:40 +03:00
mmq.cu
HIP: broaden MoE ncols_opt tile heuristic on RDNA3.5 architecture (llama/28935)
2026-09-23 20:46:47 +03:00
mmq.cuh
ggml-cuda: hip add specific config table for AMD GCN (llama/27841)
2026-09-14 20:45:06 +03:00
mmvf.cu
ggml: add SWIGLU_CLAMP (llama/27930)
2026-09-04 13:39:35 +03:00
mmvf.cuh
CUDA: use mmvq for mul-mat-id for small batch sizes (llama/18958)
2026-02-08 09:29:10 +02:00
mmvq.cu
CUDA: branchless Q4_K/Q5_K unpack to speed up mmvq, L2 prefetch on DGX Spark (llama/26705)
2026-09-14 20:45:06 +03:00
mmvq.cuh
CUDA: route batch>=4 quantized matmul to MMQ on AMD MFMA hardware (llama/23227)
2026-05-29 09:47:30 +03:00
moe-weighted-reduction.cu
cuda: fuse MoE weighted expert reduction (llama/25952)
2026-09-04 13:39:39 +03:00
moe-weighted-reduction.cuh
cuda: fuse MoE weighted expert reduction (llama/25952)
2026-09-04 13:39:39 +03:00
norm.cu
CUDA: Fix data-races when reusing SMEM in block_reduce (llama/26385)
2026-08-04 13:37:47 +03:00
norm.cuh
CUDA: fuse adds, fuse add with rms norm (llama/15631)
2025-09-20 13:42:44 +03:00
opt-step-adamw.cu
…
opt-step-adamw.cuh
…
opt-step-sgd.cu
…
opt-step-sgd.cuh
…
out-prod.cu
ggml-cuda: provide static workspace for cuBLAS handles (llama/26574)
2026-08-21 19:23:26 +03:00
out-prod.cuh
…
pad_reflect_1d.cu
musa: fix build warnings (llama/15611)
2025-09-29 15:18:10 +03:00
pad_reflect_1d.cuh
cuda : add Pad Reflect 1D support (llama/14659)
2025-09-20 13:42:39 +03:00
pad.cu
cuda : extend GGML_OP_PAD to work with non-cont src0 (llama/19429)
2026-02-15 21:44:37 +02:00
pad.cuh
…
pool1d.cu
cuda : add POOL_1D support (llama/27573)
2026-08-25 15:28:56 +03:00
pool1d.cuh
cuda : add POOL_1D support (llama/27573)
2026-08-25 15:28:56 +03:00
pool2d.cu
…
pool2d.cuh
…
quantize.cu
cuda: fix warnings for unused variable/function (llama/26688)
2026-08-07 21:59:49 +03:00
quantize.cuh
sync : ggml (#3962)
2026-07-31 09:11:28 +02:00
reduce_rows.cuh
cuda: support row-contiguous SUM_ROWS (llama/26308)
2026-09-23 20:46:47 +03:00
roll.cu
…
roll.cuh
…
rope.cu
ggml: add ggml_rope_set_offset (+ metal support) (llama/27120)
2026-08-21 19:23:26 +03:00
rope.cuh
CUDA: fuse rms_norm + mul + rope (+ view + set_rows) (llama/26767)
2026-08-14 22:16:06 +03:00
scale.cu
Programmatic Dependent Launch (PDL) for more performance on newer NVIDIA GPUs (Hopper+) (llama/22522)
2026-05-25 12:26:07 +03:00
scale.cuh
…
set-rows.cu
cuda : add support for f16->f16 GGML_OP_SET_ROWS (llama/25367)
2026-07-10 13:06:42 +03:00
set-rows.cuh
…
set.cu
cuda: add SET operation support (llama/16804)
2025-11-09 23:38:03 +02:00
set.cuh
cuda: add SET operation support (llama/16804)
2025-11-09 23:38:03 +02:00
snake.cu
cuda: fuse snake activation (mul, sin, sqr, mul, add) (llama/22667)
2026-05-14 21:26:48 +03:00
snake.cuh
cuda: fuse snake activation (mul, sin, sqr, mul, add) (llama/22667)
2026-05-14 21:26:48 +03:00
softcap.cu
Programmatic Dependent Launch (PDL) for more performance on newer NVIDIA GPUs (Hopper+) (llama/22522)
2026-05-25 12:26:07 +03:00
softcap.cuh
…
softmax.cu
CUDA: Fix data-races when reusing SMEM in block_reduce (llama/26385)
2026-08-04 13:37:47 +03:00
softmax.cuh
…
solve_tri.cu
ggml-cuda: provide static workspace for cuBLAS handles (llama/26574)
2026-08-21 19:23:26 +03:00
solve_tri.cuh
SOLVE_TRI CUDA kernel for small matrices (llama/17457)
2025-12-12 17:53:11 +02:00
ssm-conv.cu
Avoid PDL race conditions by disabling __restrict__ when PDL is used (llama/24030)
2026-06-08 14:36:36 +03:00
ssm-conv.cuh
CUDA: fuse SSM_CONV + ADD(bias) + SILU (llama/22478)
2026-05-01 13:07:35 +03:00
ssm-scan.cu
ggml-cuda: provide static workspace for cuBLAS handles (llama/26574)
2026-08-21 19:23:26 +03:00
ssm-scan.cuh
…
sum.cu
…
sum.cuh
…
sumrows.cu
cuda: support row-contiguous SUM_ROWS (llama/26308)
2026-09-23 20:46:47 +03:00
sumrows.cuh
…
top-k.cu
ROCm: add radix TOP_K for long rows (llama/27466)
2026-09-04 13:39:36 +03:00
top-k.cuh
sampling : add support for backend sampling (llama/17004)
2026-01-14 09:11:59 +02:00
topk-moe.cu
CUDA: extend MOE fusion to specdec, earlier MOE glu fusion and topk-router fusion were restricted to 1 token (llama/27621)
2026-09-04 13:39:36 +03:00
topk-moe.cuh
CUDA: extend MOE fusion to specdec, earlier MOE glu fusion and topk-router fusion were restricted to 1 token (llama/27621)
2026-09-04 13:39:36 +03:00
tri.cu
Add support for CUMSUM and TRI for CUDA. (llama/17584)
2025-12-12 17:53:17 +02:00
tri.cuh
Add support for CUMSUM and TRI for CUDA. (llama/17584)
2025-12-12 17:53:17 +02:00
tsembd.cu
ggml : fix padding in timestep embedding kernels (llama/15932)
2025-09-20 13:45:30 +03:00
tsembd.cuh
…
unary.cu
ggml: add SWIGLU_CLAMP (llama/27930)
2026-09-04 13:39:35 +03:00
unary.cuh
ggml: add SWIGLU_CLAMP (llama/27930)
2026-09-04 13:39:35 +03:00
upscale.cu
model: LFM2-VL fixes (llama/17577)
2025-12-12 17:53:14 +02:00
upscale.cuh
…
vecdotq.cuh
CUDA: branchless Q4_K/Q5_K unpack to speed up mmvq, L2 prefetch on DGX Spark (llama/26705)
2026-09-14 20:45:06 +03:00
wkv.cu
cuda : add warp-per-row wkv7 kernel for single-token decode (llama/26111)
2026-08-14 22:16:06 +03:00
wkv.cuh
…