Ali Tariq
|
648ebcdb73
ci : Added CI with RISC-V RVV1.0 Hardware (#14439)
|
5 месяцев назад |
Sigbjørn Skjæret
|
07aa869a91
ci : add more python requirements to copilot-setup-steps (#15289)
|
5 месяцев назад |
Georgi Gerganov
|
00f35d509e
ggml : repack block_iq4_nlx8 (#14904)
|
5 месяцев назад |
Oliver Simons
|
6028bf7435
CUDA: Optimize `reduce_rows_f32` kernel, leading up to 25x perf improvement on kernel-level and 10% perf increase for Gemma3n (#15132)
|
5 месяцев назад |
Sigbjørn Skjæret
|
bc5182272c
ci : add copilot-setup-steps.yml (#15214)
|
5 месяцев назад |
Tak-RS
|
e71d48e326
ggml-rpc: chunk send()/recv() to avoid EINVAL for very large tensors over RPC (macOS & others) (#15188)
|
5 месяцев назад |
uvos
|
b0493156fa
HIP: disable sync warp shuffel operators from clr amd_warp_sync_functions.h (#15273)
|
5 месяцев назад |
Romain Biessy
|
f4586ee598
sycl: Fix and disable more configurations of mul_mat (#15151)
|
5 месяцев назад |
rmatif
|
60a7658810
opencl: allow mixed f16/f32 `add` (#15140)
|
5 месяцев назад |
Aman Gupta
|
efe3a90996
CUDA cmake: add `-lineinfo` for easier debug (#15260)
|
5 месяцев назад |
Chenguang Li
|
bbd57b7eaf
CANN: GGML_OP_CPY optimization (#15070)
|
5 месяцев назад |
R0CKSTAR
|
25ff6f7659
musa: fix failures in test-backend-ops for mul_mat_id op (#15236)
|
5 месяцев назад |
hipudding
|
be48528b06
CANN: Add broadcast for softmax and FA (#15208)
|
5 месяцев назад |
rainred
|
cf9e5648a7
mtmd : Fix MinicpmV model converter and clip to avoid using hardcode. (#14750)
|
5 месяцев назад |
Xuan-Son Nguyen
|
fba5c0d680
chat : hotfix gpt-oss jinja raising an exception (#15243)
|
5 месяцев назад |
Xuan-Son Nguyen
|
53d0a12658
server : allow specifying reasoning_format in HTTP request (#15238)
|
5 месяцев назад |
Zagaj
|
27093afe78
readme : update infra list (#15234)
|
5 месяцев назад |
Georgi Gerganov
|
228f724d9c
kv-cache : fix seq_rm with seq_id == -1 (#15226)
|
5 месяцев назад |
Daniel Bevenius
|
cd3069dfcb
kv-cache : log (debug) all streams in find_slot (#15176)
|
5 месяцев назад |
Sigbjørn Skjæret
|
50e81bdf5d
convert : fix merge conflicts (#15229)
|
5 месяцев назад |
Daniel Bevenius
|
1ebbaddff2
perplexity : update comments/error msg to use decode [no ci] (#15227)
|
5 месяцев назад |
Julien Denize
|
a3a7874272
convert : improve Mistral models integration (#14737)
|
5 месяцев назад |
Charles Xu
|
002cb1bb33
kleidiai: fix unsigned overflow bug (#15150)
|
5 месяцев назад |
David Zhao
|
79c1160b07
cuda: refactored ssm_scan and use CUB (#13291)
|
5 месяцев назад |
Aman Gupta
|
34c9d765bf
CUDA: add attention sinks for tile and wmma (#15178)
|
5 месяцев назад |
compilade
|
e54d41befc
gguf-py : add Numpy MXFP4 de/quantization support (#15111)
|
5 месяцев назад |
Johannes Gäßler
|
4850b52aed
server-bench: external OAI servers, sqlite (#15179)
|
5 месяцев назад |
AN Long
|
cd6983d56d
ggml : fix field name when new ggml_backend (#14944)
|
5 месяцев назад |
Olivier Chafik
|
6c7e9a5440
vendor: sync minja (#15161)
|
5 месяцев назад |
Johannes Gäßler
|
1425f587a8
CUDA: attention sinks for mma FlashAttention (#15157)
|
5 месяцев назад |