 Georgi Gerganov
|
e92d53b29e
sampling : optimize samplers by reusing bucket sort (#15665)
|
4 months ago |
|
Georgi Gerganov
|
0d161f021a
server : enable /slots by default and make it secure (#15630)
|
4 months ago |
|
Georgi Gerganov
|
4efd5a8316
metal : fix checks for available FA kernels (#15700)
|
4 months ago |
 Diego Devesa
|
274966226f
llama : fix fattn reserve call n_seqs parameter (#15699)
|
4 months ago |
|
Diego Devesa
|
9777032dcc
llama : separate compute buffer reserve from fattn check (#15696)
|
4 months ago |
 Sigbjørn Skjæret
|
7d3c9f2b21
ci : explicitly set fa off or on (#15692)
|
4 months ago |
 Jeff Bolz
|
bbbf5ecccb
vulkan: handle large sizes for get_rows (#15686)
|
4 months ago |
|
Jeff Bolz
|
c37052ab4d
vulkan: mul_mat_id coopmat2 optimizations (#15546)
|
4 months ago |
 Daniel Bevenius
|
5c16b9c87d
vulkan : remove unused portability_enumeration_ext variable (#15679)
|
4 months ago |
|
Jeff Bolz
|
b97c9edc59
vulkan: Allow fallback to sysmem memory when vidmem is full (#15649)
|
4 months ago |
|
Jeff Bolz
|
94e82c7ead
vulkan: clamp matmul and FA results to the max finite value (#15652)
|
4 months ago |
 Charles Xu
|
4d74393bcc
ggml: update kleidiai to v1.13.0 (#15663)
|
4 months ago |
|
Diego Devesa
|
dd892555b0
Update build.md to remove MSVC arm64 notes (#15684)
|
4 months ago |
 Johannes Gäßler
|
e81b8e4b7f
llama: use FA + max. GPU layers by default (#15434)
|
4 months ago |
|
Johannes Gäßler
|
38ad381f9f
CUDA: use FP32 arithmetic for conv2d (#15683)
|
4 months ago |
|
Jeff Bolz
|
696fccf354
vulkan: Skip syncing for prealloc_y when it is reused (#15544)
|
4 months ago |
 Chenguang Li
|
ef476916bb
CANN: FIx compiler warnings (#15661)
|
4 months ago |
 Sergey Alirzaev
|
d82f6aa34a
server : removed obsolete doc (#15670)
|
4 months ago |
|
Johannes Gäßler
|
3d16b29c3b
scripts: strip "AMD Instinct" from GPU name (#15668)
|
4 months ago |
 ExtReMLapin
|
792b44f2ed
server : add documentation for `parallel_tool_calls` param (#15647)
|
4 months ago |
 Aman Gupta
|
81017865ee
CUDA: fix bug in rms_norm fusion (#15660)
|
4 months ago |
 Piotr Wilkin (ilintar)
|
60e5eee31f
chat : Seed OSS thinking + tool call support (#15552)
|
4 months ago |
|
Aman Gupta
|
009b709d6e
CUDA: fuse adds, fuse add with rms norm (#15631)
|
4 months ago |
 Gabe Goodhart
|
e8d99dd0b6
nvidia nemotron nano v2 (nemotronh) (#15507)
|
4 months ago |
|
Gabe Goodhart
|
a8bca68f72
fix: Compute the full sum in llama-eval-callback, not just the sum of printed values (#15637)
|
4 months ago |
 mnehete32
|
c97dc09391
CUDA: add conv2d (#15635)
|
4 months ago |
 Aaron Teo
|
6c442f42ff
ggml-cpu: fix invalid hsum build in debug s390x (#15634)
|
4 months ago |
 compilade
|
73804145ab
ggml : fix SSM_SCAN for n_groups > 1 (#15625)
|
4 months ago |
|
Georgi Gerganov
|
c8d0d14e77
kv-cache : fix find_slot to not search for continuous slot (#15638)
|
4 months ago |
|
Sigbjørn Skjæret
|
84ab83cc0b
model : jina-embeddings-v3 support (#13693)
|
4 months ago |
