 Gaurav Garg
|
b1b132efcb
cuda : enable CUDA Graph on CUDA Toolkit < 12.x (#12394)
|
преди 10 месеца |
 Guus Waals
|
01e8f2138b
ggml-vulkan: remove unused find_program(glslc) (#12416)
|
преди 10 месеца |
 Jeff Bolz
|
484a8ab513
vulkan: Add N/2 and N/4 optimized paths in coopmat2 shader (#12312)
|
преди 10 месеца |
 Daniele
|
cf2270e4d3
vulkan: subgroup size tuning (#12087)
|
преди 10 месеца |
|
Jeff Bolz
|
f07690c930
vulkan: use fp32 in coopmat2 q4_k dequant function (#12309)
|
преди 10 месеца |
|
Jeff Bolz
|
891c63956d
vulkan: Pad N dimension of B matrix for coopmat2 perf, to avoid bounds checking (#12273)
|
преди 10 месеца |
|
Jeff Bolz
|
2f21123c1d
vulkan: Adjust coopmat2 tile sizes and selection heuristic (#12258)
|
преди 10 месеца |
 Christian Kastner
|
374101fd74
cmake : enable building llama.cpp using system libggml (#12321)
|
преди 10 месеца |
 Akarshan Biswas
|
b3c9a65673
SYCL: set extras only on GGML_TYPE_Q4_0 (#12366)
|
преди 10 месеца |
 Sigbjørn Skjæret
|
8ba95dca20
llama : fix OLMo-2-0325-32B-Instruct K-norm size (#12400)
|
преди 10 месеца |
 Georgi Gerganov
|
dc079cfdff
context : fix init of n_outputs (#12397)
|
преди 10 месеца |
 Daniel Bevenius
|
7b61bcc87c
ci : add --symlinks to xcframework zip command (#12409)
|
преди 10 месеца |
 marcoStocchi
|
f4c3dd5daa
llama-tts : add '-o' option (#12398)
|
преди 10 месеца |
 aubreyli
|
3d35d87b41
SYCL: Delete redundant plus sign and space (#12391)
|
преди 10 месеца |
 fairydreaming
|
b19bd064c0
SYCL : support non-contiguous tensors in binary ops (add, sub, etc) (#12399)
|
преди 10 месеца |
 Chenguang Li
|
92a391327e
[CANN]MUL_MAT optimization (#12382)
|
преди 10 месеца |
 Eric Curtin
|
9f2250ba72
Add CLI arg to llama-run to adjust the number of threads used (#12370)
|
преди 10 месеца |
|
Sigbjørn Skjæret
|
774973b8f3
main : add -sysf / --system-prompt-file (#12249) (#12250)
|
преди 10 месеца |
|
fairydreaming
|
8fcb563613
Load all MoE experts during warmup (#11571)
|
преди 10 месеца |
 Victor
|
add2a3aa5a
server: fix "--grammar-file" parameter (#12285)
|
преди 10 месеца |
|
Georgi Gerganov
|
c522ce4143
graph : simplify attn input build for unified KV cache (#12381)
|
преди 10 месеца |
|
Georgi Gerganov
|
081bee8c64
hparams : add SWA rope parameters (#12374)
|
преди 10 месеца |
|
Georgi Gerganov
|
84d5475541
llama : fix Gemma3 SWA KV cache shift (#12373)
|
преди 10 месеца |
 Xuan-Son Nguyen
|
be7c303410
arg : no n_predict = -2 for examples except for main and infill (#12364)
|
преди 10 месеца |
|
Georgi Gerganov
|
e0dbec0bc6
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
|
преди 10 месеца |
 Ishaan Gandhi
|
2048b5913d
server : fix crash when using verbose output with input tokens that are not in printable range (#12178) (#12338)
|
преди 10 месеца |
 Oscar Barenys
|
f08f4b3187
Update build.yml for Windows Vulkan builder to use Vulkan 1.4.304 SDK for VK_NV_cooperative_matrix2 support (#12301)
|
преди 10 месеца |
|
Daniel Bevenius
|
80a02aa858
llama.swiftui : fix xcframework dir in README [no ci] (#12353)
|
преди 10 месеца |
 Alberto Cabrera Pérez
|
363f8c5d67
sycl : variable sg_size support for mmvq kernels (#12336)
|
преди 10 месеца |
 uvos
|
34c961b181
CUDA/HIP: Fix fattn-vec-* when device warp size is not 32 (#12315)
|
преди 10 месеца |
