This website requires JavaScript.
0949beb5a3
fix build number for sycl release (#22283 )
2026-04-23 15:38:58 +02:00
9012c50fc8
model-conversion : fix mmproj output file name [no ci] (#22274 )
2026-04-23 15:07:38 +02:00
0dd7f915fd
cli : cleanup auto-completion code (#21745 )
2026-04-23 15:03:28 +02:00
550d684bd1
server: Enable transcriptions API for LFM2-Audio (#22000 )
2026-04-23 10:47:26 +02:00
8635e221c8
metal : fix event synchronization (#22260 )
2026-04-23 08:22:49 +03:00
930e0210d1
gitignore: add AGENTS.local.md (#22246 )
2026-04-23 08:22:24 +03:00
96c1db26c4
ggml-base: use MATH_LIBRARY variable instead of hardcoded 'm' (#22239 )
2026-04-23 08:22:08 +03:00
4ead6fd957
[SYCL] Update oneapi 2025.3.3, Seperate SYCL build, release Ubuntu 24 package. (#22078 )
2026-04-23 13:21:36 +08:00
5eaee65384
convert : Handle ModelOpt produced mixed precision model during convert to GGUF (#22247 )
2026-04-23 05:19:51 +00:00
60b68a6279
sycl : fused MoE mul_mat_vec_q for TG (#21920 )
2026-04-22 23:18:56 -06:00
b76429a69c
ggml-webgpu: add support for im2col (#22259 )
2026-04-22 23:17:41 -04:00
86db42e97f
CUDA: fuse relu + sqr (#22249 )
2026-04-23 02:28:56 +00:00
6217b49583
HIP: flip GGML_HIP_GRAPHS to default on (#22254 )
2026-04-23 02:34:31 +02:00
0d0764dfd2
[WebGPU] Implement async tensor api and event api (#22099 )
2026-04-22 10:52:01 -07:00
6da7168312
ggml-webgpu: Add fused RMS_NORM + MUL (#21983 )
2026-04-23 02:51:40 +09:00
8bccdbbff9
chat: fix parallel_tool_calls default setting based on model capabilities, add tests for parallel tool calls and structured outputs (#22217 )
2026-04-22 18:10:56 +02:00
bcb5eeb645
speculative-simple : add checkpoint support (#22227 )
2026-04-22 15:44:45 +03:00
225088ea76
sycl: Improve mul_mat_id memory efficiency and add BF16 fast path (#22119 )
2026-04-22 18:02:56 +05:30
82d3f4d3b2
mtmd: also support LLAMA_ROPE_TYPE_NONE (#22242 )
2026-04-22 12:16:29 +02:00
17f6245168
server: ignore reasoning content from transcription api (#21905 )
2026-04-22 12:10:50 +02:00
7bfe60fdf9
mtmd, llama : Update HunyuanVL vision-language model support (#22037 )
2026-04-22 17:58:43 +08:00
750579ff14
common: Refactoring sampler parameters (#20429 ) (#22233 )
2026-04-22 01:40:19 -07:00
134d6e54d4
common/chat, server: refactor, move all conversion functions to common, add tests (#20690 )
2026-04-22 10:28:45 +02:00
ca7f7b7b94
ggml-webgpu(shader): support conv2d kernels. (#21964 )
2026-04-21 23:18:57 -04:00
0dedb9ef7a
hexagon: add support for FILL op (#22198 )
2026-04-22 04:54:20 +05:30
2799d933b5
ggml-webgpu: reset CPU/GPU profiling time when freeing context (#22050 )
2026-04-22 08:05:21 +09:00
04fe84b69d
server: allow cancel loading model (#21814 )
2026-04-22 00:26:09 +02:00
5a4cd6741f
Hexagon: DAIG op (#22195 )
2026-04-21 14:16:04 -07:00
2248799a58
hexagon: fix missing v79 entry in libggml-htp.inf (#22194 )
2026-04-22 04:53:44 +08:00
72d693e4fb
spec : reset i_last when low acceptance streak occurs (#22168 )
2026-04-21 20:29:07 +02:00
98d2d2884e
mtmd: Add support for Reka Edge 2603 (#21616 )
2026-04-22 02:02:49 +08:00
84652b80cf
arg : add --spec-default (#22223 )
2026-04-21 19:52:02 +03:00
52f1096f21
openvino: driver setup, CI split, thread safety, and NPU optimizations (#21944 )
2026-04-21 23:58:34 +08:00
606fa42f5d
vendor : update cpp-httplib to 0.43.1 (#22143 )
2026-04-21 11:45:48 -03:00
7fc1c4ef78
metal : workaround macOS GPU interactivity watchdog (#22216 )
2026-04-21 17:24:55 +03:00
82209efb7e
vulkan: Support F16 OP_FILL (#22177 )
2026-04-21 11:01:56 +02:00
9998d88bc8
mtmd: correct mtmd_decode_use_mrope() (#22188 )
2026-04-21 10:53:37 +02:00
cd03ec7642
llama-ext : fix exports (#22202 )
2026-04-21 11:04:46 +03:00
4889afba5f
sync : ggml
2026-04-21 11:03:42 +03:00
041fe83d74
ggml : bump version to 0.10.0 (ggml/1463)
2026-04-21 11:02:56 +03:00
cfe9838d26
fit-params : refactor + add option to output estimated memory per device (#22171 )
2026-04-21 09:54:36 +03:00
ff6b1062af
server : fix hardcoded proxy connection timeout in router mode (#18760 ) (#22003 )
2026-04-21 06:41:14 +02:00
97895129e5
ggml-cuda: flush legacy pool on OOM and retry (#22155 )
2026-04-21 05:30:38 +08:00
86f8daacfe
mtmd: correct get_n_pos / get_decoder_pos (#22175 )
2026-04-20 23:29:19 +02:00
cf8b0dbda9
server : remove /api endpoints (#22165 )
2026-04-20 20:41:19 +03:00
fd6ae4ca1c
Tensor-parallel: Fix delayed AllReduce on Gemma-4 MoE (#22129 )
2026-04-20 21:55:39 +05:30
fb19f94c71
TP: fix 0-sized tensor slices, AllReduce fallback (#21808 )
2026-04-20 18:09:39 +02:00
7f251fdbce
ggml-cpu: Optimized x86 and generic cpu q1_0 dot (follow up) (#21636 )
2026-04-20 21:02:54 +05:00
a6cc43c286
ggml-webgpu: updated matrix-vector multiplication (#21738 )
2026-04-20 07:37:17 -07:00
a678916623
mtmd: refactor mtmd_decode_use_mrope (#22161 )
2026-04-20 14:45:11 +02:00
81df3f7cfa
fix: GLM-DSA crash in llama-tokenize when using vocab_only (#22102 )
2026-04-20 02:32:46 -05:00
de71b5f81c
server : refactor "use checkpoint" logic (#22114 )
2026-04-20 08:42:37 +03:00
788fcbc5dd
[SYCL] Fix reorder MMVQ assert on unaligned vocab sizes (#22035 )
2026-04-20 01:39:45 -04:00
9d49acb2a7
server: rename --clear-idle to --cache-idle-slots (#21741 )
2026-04-20 08:30:24 +03:00
e365e658f0
vendor : update cpp-httplib to 0.42.0 (#21781 )
2026-04-19 19:41:43 -03:00
4eac5b4509
CUDA: refactor mma data loading for AMD (#22051 )
2026-04-19 18:26:59 +02:00
d5b780a676
common/autoparser : allow space after tool call (#22073 )
2026-04-19 06:28:35 -05:00
471540ae8a
HIP: Remove unesscary NCCL_CHECK (#21914 )
2026-04-19 12:59:44 +02:00
19124078be
mtmd: add pos_0 to mtmd_image_tokens_get_decoder_pos (breaking change) (#22082 )
2026-04-19 11:57:21 +02:00
bcdcc1044f
ggml : reduce CPU overhead in meta backend (#22041 )
2026-04-19 15:18:35 +05:30
037bfe38d0
ci : install spirv-headers for vulkan-cross (#22109 )
2026-04-19 09:32:08 +02:00
8685e7b075
convert : support sentence-transformer 5.4 config files (#22087 )
2026-04-19 16:25:39 +09:00
09b4efa95f
cmake: remove CMP0194 policy to restore MSVC builds (#21934 )
2026-04-19 02:25:05 -05:00
455d8e4be8
server : speculative checkpointing (#19493 )
2026-04-19 09:24:06 +02:00
91fef95362
rpc : refactor the RPC transport (#21998 )
2026-04-19 10:21:53 +03:00
9e5647affa
server: Expose media_tag on /props endpoint. (#22028 )
2026-04-18 19:27:17 -03:00
4f02d47339
model : refactor bias tensor variable names (#22079 )
2026-04-18 20:12:00 +02:00
23b8cc4991
android : libcommon -> libllama-common (#22076 )
2026-04-18 11:19:40 +02:00
59accc8863
ggml-backend-meta: add multi-segment read support in get_tensor (#22063 )
2026-04-18 03:04:51 -05:00
83d58e02fc
ci : free disk space for rocm release (#22012 )
2026-04-18 09:37:30 +02:00
89a5474f0e
convert : fix (ignore for now) typings errors (#22002 )
2026-04-18 09:36:41 +02:00
fd1c0ec3f0
llama: fit ctx size for CPU only (#21568 )
2026-04-18 08:16:04 +02:00
45cac7ca70
ggml-webgpu: fix compiler warnings and refactor FlashAttention encoding (#21052 )
2026-04-17 09:17:11 -07:00
b94050e896
CUDA: use LRU based eviction for cuda graphs (#21611 )
2026-04-17 23:24:21 +08:00
a279d0f0f4
ci : add android arm64 build and release (#21647 )
2026-04-17 05:32:24 -04:00
268d61e178
mtmd: add missing struct tag (#22023 )
2026-04-17 01:48:33 -07:00
6990e2f1f7
libs : rename libcommon -> libllama-common (#21936 )
2026-04-17 11:11:46 +03:00
fcc7508759
model : Gemma4 model type detection (#22027 )
2026-04-17 16:07:11 +08:00
5e6c0e18b6
opencl: refactor q8_0 set_tensor and mul_mat host side dispatch for Adreno (#21938 )
2026-04-16 22:28:33 -07:00
30dce2cf29
cli : use get_media_marker (#22017 )
2026-04-17 00:12:31 +02:00
089dd41fe3
cmake: use glob to collect src/models sources (#22005 )
2026-04-16 23:25:16 +02:00
85dde8dc4a
hexagon: optimize HMX matmul operations (#21071 )
2026-04-17 04:48:34 +08:00
4fbdabdc61
model: using single llm_build per arch (#21970 )
2026-04-16 21:10:22 +02:00
e45dbdece8
opencl: add q5_K gemm and gemv kernels for Adreno (#21595 )
2026-04-16 12:08:33 -07:00
4adac43f6f
server: tests: fetch random media marker via /apply-template (#21962 ) (#21980 )
2026-04-16 19:46:21 +02:00
9db77a020c
model : refactor QKV into common build_qkv and create_tensor_qkv helpers (#21245 )
2026-04-16 23:41:34 +08:00
f772f6e434
model : support NVFP4 tensors for Gemma4 (#21971 )
2026-04-16 16:51:47 +02:00
b572d1ecd6
codeowners: add team member comments (#21714 )
2026-04-16 12:13:11 +02:00
03b3d07798
Convert: Fix NemotronH Config Parsing (#21664 )
2026-04-16 10:11:45 +00:00
3f7c29d318
ggml: add graph_reused (#21764 )
2026-04-16 17:21:28 +08:00
ae2d34899e
metal: Implement ROLL op (#21946 )
2026-04-16 03:54:37 -05:00
1e796eb41f
ggml-cpu: add 128-bit RVV implementation for Quantization Vector Dot (#20633 )
2026-04-16 13:15:15 +05:00
5637536517
ggml : implemented simd_gemm kernel for riscv vector extension (#20627 )
2026-04-16 13:14:26 +05:00
90fb96a7b3
devops : added spirv-headers to nix (#21965 )
2026-04-16 08:12:52 +00:00
82677a6ede
ggml-webgpu: compute pass batching and removing profiling overhead (#21873 )
2026-04-16 01:12:19 -07:00
8612ed18b7
ci : Use ggml-org/ccache-action on RISC-V as well (#21632 )
2026-04-16 10:11:25 +02:00
b1be68e8ca
[SYCL] Fix Q8_0 reorder: garbage on 2nd prompt + crash on full VRAM (#21638 )
2026-04-16 01:34:05 -04:00
408225bb1a
server: use random media marker (#21962 )
2026-04-15 23:52:22 +02:00
b3d758750a
vulkan: optimize im2col (#21713 )
2026-04-15 19:04:51 +02:00
7e72b38bc1
cuda: Q1_0 initial backend (#21629 )
2026-04-15 09:38:38 -07:00
Sigbjørn Skjæret
Daniel Bevenius
Matthias Straka
Tarek Dakhran
Georgi Gerganov
Neo Zhang Jianyu
ynankani
abotsis
Chen Yuan
Anav Prasad
uvos
Nikhil Jain
Masashi Yoshimura
Piotr Wilkin (ilintar)
Akarshan Biswas
Xuan-Son Nguyen
manayang
Ethan Turner
Aparna M P
Shreya Jain
Mengsheng Wu
Paul Dubs
Kwa Jie Hao
Zijun Yu
Alessandro de Oliveira Faria (A.K.A.CABELO)
Jeff Bolz
xris99
leonardHONG
Gaurav Garg
Johannes Gäßler
pl752
neha-ha
SamareshSingh
Katostrofik
Yes You Can Have Your Own
Aldehir Rojas
Dowon
texasich
Sascha Rogmann
Radoslav Gerganov
Cetarthoriphros
Reese Levine
Aman Gupta
Yuri Khrustalev
65a
Eric Zhang
lhez
nullname
shaofeiqi
Pascal
PikaPikachu
Ruben Ortlam
Kusha Gharahi
rehan-10xengineer
Yuannan
Ludovic Henry
Pasha Khosravi