 Aaron Teo
|
c7f3169cd5
ggml-cpu : disable GGML_NNPA by default due to instability (#14880)
|
6 mesi fa |
 Gabe Goodhart
|
793c0d7f46
metal: SSM_SCAN performance (#14743)
|
6 mesi fa |
 lhez
|
ce111d39d6
opencl: add fused `rms_norm_mul` (#14841)
|
6 mesi fa |
 wooksong
|
e7fecba934
docs : update HOWTO‑add‑model.md for ModelBase and new model classes (#14874)
|
6 mesi fa |
 Oliver Simons
|
e2b7621e7c
ggml : remove invalid portPos specifiers from dot files (#14838)
|
6 mesi fa |
 Georgi Gerganov
|
c1dbea752a
context : restore preemptive sched reset when LLAMA_SET_ROWS=0 (#14870)
|
6 mesi fa |
 kiwi
|
749e0d27f0
mtmd : fix 32-bit narrowing issue in export-lora and mtmd clip (#14503)
|
6 mesi fa |
 Chris Rohlf
|
64bf1c3744
rpc : check for null buffers in get/set/copy tensor endpoints (#14868)
|
6 mesi fa |
 Diego Devesa
|
c12bbde372
sched : fix multiple evaluations of the same graph with pipeline parallelism (#14855)
|
6 mesi fa |
 R0CKSTAR
|
3f4fc97f1d
musa: upgrade musa sdk to rc4.2.0 (#14498)
|
6 mesi fa |
|
Georgi Gerganov
|
2df255da3c
sync : ggml
|
6 mesi fa |
 Kai Pastor
|
60f816a79d
cmake : fix usage issues (ggml/1257)
|
6 mesi fa |
 Daniel Bevenius
|
5592f278b6
ggml-cpu : remove stdlib include from repack.cpp (ggml/1276)
|
6 mesi fa |
|
Georgi Gerganov
|
e4868d16d2
context : perform output reorder lazily upon access after sync (#14853)
|
6 mesi fa |
 Xuan-Son Nguyen
|
820de57d4f
chat : fix kimi-k2 chat template (#14852)
|
6 mesi fa |
 Alberto Cabrera Pérez
|
cb4a63aad6
sycl: fixed semantics of block offset calculation (#14814)
|
6 mesi fa |
 yummy
|
86f5623d90
llama : fix MiniCPM inference after Granite Four changes (#14850)
|
6 mesi fa |
 Pouya
|
39cffdf188
docs: add libcurl-dev install hint for Linux distros (#14801)
|
6 mesi fa |
|
Georgi Gerganov
|
065908cb09
metal : fix fusion across different encoders (#14849)
|
6 mesi fa |
 Donghyeon Jeong
|
4ec6291a24
sycl: fix undefined variable in work group size check (#14843)
|
6 mesi fa |
 jacekpoplawski
|
a12363bbf0
convert : text-only support for GLM-4.1V-9B-Thinking (#14823)
|
6 mesi fa |
 Johannes Gäßler
|
a86f52b285
CUDA: fix overflow in FA, tune performance (#14840)
|
6 mesi fa |
|
Johannes Gäßler
|
b284197df4
CUDA: fix compilation with GGML_CUDA_F16 (#14837)
|
6 mesi fa |
 Sigbjørn Skjæret
|
221c0e0c58
ci : correct label refactor->refactoring (#14832)
|
6 mesi fa |
|
Johannes Gäßler
|
07a19e27a2
CUDA: fix quantized KV cache + multiple sequences (#14822)
|
6 mesi fa |
|
Georgi Gerganov
|
18f3b5ff9e
tests : add non-cont K,V FA tests
|
6 mesi fa |
 l3utterfly
|
7233358d29
memory : handle saving/loading null layers in recurrent memory (#14675)
|
6 mesi fa |
 lixing-star
|
6c88b3bb25
ggml: fix loongarch quantize_row_q8_1 error (#14827)
|
6 mesi fa |
 chen fan
|
14c28dfc50
CANN: weight format to NZ for Ascend310P3 (#14407)
|
6 mesi fa |
 Aman Gupta
|
8c988fa41d
CUDA: add fused rms norm (#14800)
|
6 mesi fa |
