furyhawk/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-07-20 09:45:54 +00:00
Code Issues Packages Projects Releases Wiki Activity
10,069 Commits 618 Branches 6,869 Tags

Default Branch

master

178a6c4493 · opencl: Support broadcast for Adreno MUL_MAT and honor view_offs for Adreno Q8_0 MUL_MAT for llama-server multi-stream (#25910) · Updated 2026-07-20 05:48:57 +00:00
 ZIP  TAR.GZ

Branches

gg/ggml-op-offload-fix

9f364c765a · llama : dsv4 graph fixes · Updated 2026-07-17 15:43:19 +00:00
6
2
 ZIP  TAR.GZ
0cc4m/vulkan-mul-mm-refactor

bdccb7da38 · use minimal shmem size 8 instead of 1 to workaround cm2 compiler bug · Updated 2026-07-17 11:21:28 +00:00
11
12
 ZIP  TAR.GZ
gg/server-slot-similarity-trace

2c3ed6086f · server : add trace logs for slot prefix similarity computation · Updated 2026-07-17 10:01:28 +00:00
9
1
 ZIP  TAR.GZ
cross-profiler

6eac19fe3e · Add tensor name to JSON output · Updated 2026-07-16 12:35:13 +00:00
26
14
 ZIP  TAR.GZ
ci/cuda-virtual-devices

234252e03d · label virtual devices in description; add GPUx2 server CI jobs · Updated 2026-07-14 13:45:17 +00:00
142
3
 ZIP  TAR.GZ
xsn/server_cors_args

6554db6ed0 · add link to PR · Updated 2026-07-14 11:07:03 +00:00
73
5
 ZIP  TAR.GZ
gg/server-fix-lru-evict-loading

3203ab47a6 · server : prevent LRU eviction of models still loading · Updated 2026-07-10 13:49:57 +00:00
110
1
 ZIP  TAR.GZ
xsn/mtmd_ds_ocr_tiles

1f66d544cb · Merge branch 'master' into xsn/mtmd_ds_ocr_tiles · Updated 2026-07-10 13:11:44 +00:00
107
13
 ZIP  TAR.GZ
dev-metal

65dd9133ba · metal : add CONV_2D_DW (depthwise convolution) support (#21565) · Updated 2026-07-10 11:51:31 +00:00
110
4
 ZIP  TAR.GZ
xsn/cli_output

7ab9323d1d · cli: add --output option · Updated 2026-07-09 11:44:21 +00:00
129
1
 ZIP  TAR.GZ
xsn/cli_http_based

99b9a6c08a · also show model aliases · Updated 2026-07-07 22:35:05 +00:00
156
17
 ZIP  TAR.GZ
gg/spec-naming

14f254515f · spec : fix naming, spacing · Updated 2026-07-07 15:52:11 +00:00
158
1
 ZIP  TAR.GZ
gg/metal-set-rows-f16

d4a193e269 · metal : add set_rows with src0 f16 · Updated 2026-07-07 11:18:12 +00:00
189
5
 ZIP  TAR.GZ
jg/ggml-backend-async-copy

124d3d9428 · ggml: asynchronous scheduler memory copies · Updated 2026-07-05 10:47:02 +00:00
202
1
 ZIP  TAR.GZ
xsn/ggml_backend_dev_reset

ff9e7673cf · musaDeviceReset · Updated 2026-07-03 16:57:48 +00:00
203
6
 ZIP  TAR.GZ
xsn/server_sleep_free_backend_2

697372dddb · wip · Updated 2026-07-03 15:15:50 +00:00
206
4
 ZIP  TAR.GZ
0cc4m/vulkan-device-cpy-benchmark

2d9cf41192 · check once · Updated 2026-06-29 14:03:26 +00:00
231
12
 ZIP  TAR.GZ
0cc4m/server-memory-limit

e4d2e198b9 · server: add --models-memory-max parameter to allow dynamically unloading models when they exceed a memory size threshold · Updated 2026-06-29 07:38:11 +00:00
231
1
 ZIP  TAR.GZ
xsn/jinja_preseve_thinking

fcf66c1d80 · correct help message · Updated 2026-06-28 17:52:36 +00:00
237
2
 ZIP  TAR.GZ
0cc4m/vulkan-d2d-copy

0aac18be29 · use memcpy for small copies across host-visible memory · Updated 2026-06-27 10:59:58 +00:00
243
9
 ZIP  TAR.GZ