Georgi Gerganov
|
a876861455
metal : update support condition for im2col + fix warning (#0)
|
1 سال پیش |
Johannes Gäßler
|
202084d31d
tests: add gradient tests for all backends (ggml/932)
|
1 سال پیش |
Salvatore Mesoraca
|
efe6a83e30
ggml : fix cont with transposed tensors when one dimension is 1 (ggml/934)
|
1 سال پیش |
compilade
|
9bc6db28d0
ggml-quants : ternary packing for TriLMs and BitNet b1.58 (#8151)
|
1 سال پیش |
Georgi Gerganov
|
231cff5f6f
sync : ggml
|
1 سال پیش |
Georgi Gerganov
|
fc18425b6a
ggml : add SSM Metal kernels (#8546)
|
1 سال پیش |
slaren
|
0c41e03ceb
metal : gemma2 flash attention support (#9159)
|
1 سال پیش |
Johannes Gäßler
|
e11bd856d5
CPU/CUDA: Gemma 2 FlashAttention support (#8542)
|
1 سال پیش |
zhentaoyu
|
4f8d19ff17
[SYCL] Fix SYCL `im2col` and `convert` Overflow with Large Dims (#9052)
|
1 سال پیش |
Molly Sophia
|
2d5dd7bb3f
ggml : add epsilon as a parameter for group_norm (#8818)
|
1 سال پیش |
0cc4m
|
064cdc265f
vulkan : fix Qantized Mat-Vec Mul on AMD GPUs for ncols < 64 (#8855)
|
1 سال پیش |
Mengqing Cao
|
e09a800f9a
cann: Fix ggml_cann_im2col for 1D im2col (#8819)
|
1 سال پیش |
slaren
|
7a11eb3a26
cuda : fix dmmv cols requirement to 2*GGML_CUDA_DMMV_X (#8800)
|
1 سال پیش |
slaren
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 سال پیش |
slaren
|
87e397d00b
ggml : fix quant dot product with odd number of blocks (#8549)
|
1 سال پیش |
hipudding
|
1bdd8ae19f
[CANN] Add Ascend NPU backend (#6035)
|
1 سال پیش |
Georgi Gerganov
|
6847d54c4f
tests : fix whitespace (#0)
|
1 سال پیش |
John Balis
|
fde13b3bb9
feat: cuda implementation for `ggml_conv_transpose_1d` (ggml/854)
|
1 سال پیش |
slaren
|
0e0590adab
cuda : update supports_op for matrix multiplication (#8245)
|
1 سال پیش |
Georgi Gerganov
|
f3f65429c4
llama : reorganize source code + improve CMake (#8006)
|
1 سال پیش |
slaren
|
b6b9a8e606
fix CI failures (#8066)
|
1 سال پیش |
Calvin Laurenson
|
43b35e38ba
Add support for sqrt on CUDA (#7953)
|
1 سال پیش |
Georgi Gerganov
|
a9cae48003
tests : add non-cont unary tests (#7857)
|
1 سال پیش |
Georgi Gerganov
|
2b3389677a
ggml : refactor rope norm/neox (#7634)
|
1 سال پیش |
Johannes Gäßler
|
e141ce624a
Fix FlashAttention debug test, FP32 assert (#7684)
|
1 سال پیش |
Johannes Gäßler
|
9b596417af
CUDA: quantized KV support for FA vec (#7527)
|
1 سال پیش |
Georgi Gerganov
|
fb76ec31a9
ggml : fix YARN + add tests + add asserts (#7617)
|
1 سال پیش |
Georgi Gerganov
|
cce3dcffc5
cuda : non-cont concat support (#7610)
|
1 سال پیش |
Georgi Gerganov
|
0548a4187f
ggml : generalize GGML_OP_CONCAT (#7563)
|
1 سال پیش |
Georgi Gerganov
|
3e5faa8503
cuda : fix rope + add tests (#7452)
|
1 سال پیش |