llama.cpp-xrip

wmantly/llama.cpp-xrip

0 Fork 0

e86f5fd1e9 cmake: WASI doesn't support spawning (#30751) Mendy Berger 2026-10-27 10:14:03 -04:00
27b7a8a009 llama : build the k-pool sparse selection independently of the kq_mask shape (#31243) Aman GuptaandGeorgi Gerganov 2026-10-27 21:57:45 +08:00
7ecb57f7ad rpc : catch std::length_error in recv_msg (#31204) Katara 2026-10-27 14:04:55 +01:00
8542e94350 tests : speed up tensor data generation and model saving (#31145) Georgi Gerganov 2026-10-27 14:44:41 +02:00
922f53ed33 vulkan: fix CI llvmpipe by allowing dim 3 broadcast in MUL_MAT (#31239) Pascal 2026-10-27 13:39:03 +01:00
0c8c59ef3d mtmd-cli: Reduced scope of batch pointer (#31155) (#31165) Sam Malayek 2026-10-27 07:56:03 -04:00
3ab37aedfe ui: fix system message row stretching to full prompt height in WebKit (#31246) Aleksander Grygier 2026-10-27 12:19:02 +01:00
85cc432e2c server: add router device memory margin parameter for dynamic unloading (#21231)
Ruben Ortlam Xuan Son Nguyen Pascal
2026-10-27 12:14:42 +01:00
74a1fd3d03 common : --model or -hf is required (#31242) Julien Chaumond 2026-10-27 11:55:48 +01:00
7eed6b7685 model: Add MiniCPM-OCR-Jev (#30980) tc-mbandXuan Son Nguyen 2026-10-27 18:49:12 +08:00
276378515c metal : add GGML_OP_TOP_K_MASK (#31230) Georgi Gerganov 2026-10-27 12:13:23 +02:00
d78f40d9d3 glm5-next : QSA without kq_mask (+ DSA) (#31009) Alexey Kopytko 2026-10-27 18:54:36 +09:00
fae9ac561f ci : use ccache depend mode for ROCm release builds (#31231) Sreeram 2026-10-27 01:24:00 -07:00
a8276b70cc ci : force python 3.12 on ubuntu 26.04 (#31222) Sigbjørn Skjæret 2026-10-27 09:09:44 +01:00
de5122d879 ggml : relax scale to non-contiguous views (#31146) Georgi GerganovandPascal 2026-10-27 07:55:31 +02:00
6666a7330c tests : fix multi-seq split replay and partial-state cases for trained models (#31216) Georgi Gerganov 2026-10-27 07:54:51 +02:00
7a2e5aba90 llama : remove obsolete n_pad from memory instances (#31200) Georgi Gerganov 2026-10-27 07:53:44 +02:00
0d5ca60098 ui: list org-less models first in the model selector (#31215) Aleksander Grygier 2026-10-27 06:52:54 +01:00
d930fe3c26 opencl: refine mul mat id -1 index (#31170) lhez 2026-10-26 22:52:07 -07:00
a2b3e121ce opencl: initial support for q3_k (#31172) lhezandHongqiang Wang 2026-10-26 22:51:22 -07:00
823df7d0fb ui: Highlight reasoning toggle on hover and align it with model selector (#31212) Aleksander Grygier 2026-10-27 06:50:15 +01:00
1a97e387d3 ui: add conversation compaction (#27401) PascalandAleksander Grygier 2026-10-27 06:49:27 +01:00
bc262f33cc ui: Open model details in a drawer without leaving the page (#31176) Aleksander Grygier 2026-10-27 06:46:25 +01:00
f5892ca5ad ui: use shared confirmation dialog for conversation deletion (#31187) Aleksander Grygier 2026-10-27 06:45:53 +01:00
7aca0d2ea8 hexagon: add GGML_HEXAGON_VTCM_SIZE (#31221) Ehsan Bateni 2026-10-26 19:56:01 -04:00
961c376a2d llama : enable moe amax rescaling for laguna (#31147) begcdn 2026-10-27 06:04:19 +08:00
4244ad7acd ci : add windows behavior check (#31191) Sigbjørn Skjæret 2026-10-26 23:01:04 +01:00
63b4b303bd opencl: write KV-cache views in set_tensor so a state restore is not dropped (#30471) Hongqiang Wang 2026-10-26 11:39:22 -07:00
caa8191df1 hexagon: MUL_MAT_ID row skipping optimizations and improvements (#31169) Max Krasnyansky 2026-10-26 08:21:14 -07:00
a4699400e3 llama : move no_alloc out of llama_hparams (#31141) Georgi Gerganov 2026-10-26 16:21:25 +02:00
4b851fbee8 models : replace ggml_tensor** out-params with output structs (#31142) Georgi Gerganov 2026-10-26 16:19:47 +02:00
4d71b42fd8 ggml-cpu : rewrite concat as one row kernel for all types (#31179) Pascal 2026-10-26 15:18:45 +01:00
70c5f26324 vulkan: rms norm optimization using subgroup reductions (#29882) virajwad 2026-10-26 07:14:31 -07:00
5cd62ffee9 sycl: template the plain MMVQ launch triples (#31013) Titaniumtown 2026-10-26 07:13:16 -07:00
38645c2ad9 sycl: template the per-quant concat impls (#31008) Titaniumtown 2026-10-26 07:12:56 -07:00
02269b4c00 sycl: add set_tensor_2d for faster tensor-split loading (#30358) Captain-TrippsandJon Staples 2026-10-26 09:12:25 -05:00
5320c87ff0 sycl : bound reorder temp memory to 32 MiB (#31062) KyleHagy 2026-10-26 07:11:57 -07:00
74b104c9ba common: pin model resolution on real HF repo layouts (#30548) Pascal 2026-10-26 14:18:29 +01:00
c6fb3b5be2 mtmd: add mtmd_helper_get_capabilities (#31183) Xuan-Son Nguyen 2026-10-26 12:45:35 +01:00
8b28c23943 ggml-cpu : fix unused variables in arm arch (#31180) Sigbjørn Skjæret 2026-10-26 11:38:55 +01:00
fd8f8451a2 tests : refactor test-llama-archs to not link internal libllama symbols (#30880) Georgi Gerganov 2026-10-26 12:16:02 +02:00
9125a2ebe8 ci : fix race in webgpu memset (#31162) Evan Huus 2026-10-26 05:02:12 -04:00
9b8446b914 metal : stop overflow from happening in large matrix output offsets (#30072) feni6andfeni6 2026-10-26 02:13:39 -06:00
dc9e3bed42 ui: UX improvements (#31030) Aleksander Grygier 2026-10-26 09:13:08 +01:00
69b3cc406e ui: Sidebar resizing improvements (#31119) Aleksander Grygier 2026-10-26 09:07:25 +01:00
224d94c02b ci : bump server timeout during startup (#31174) Georgi Gerganov 2026-10-26 09:59:15 +02:00
48c36110ab ci : drop the macos-latest-x64 job (#31175) Georgi Gerganov 2026-10-26 09:56:19 +02:00
a074215352 deepseekv4.1: follow-up review fixes (#31137) Aman GuptaandGeorgi Gerganov 2026-10-26 15:52:02 +08:00
d854d98bae cuda : use std::string instead of VLA for archName (#30463) Adrien Gallouët 2026-10-26 00:08:22 +01:00
48c9f08b7b common, grammar: add recursive depth limit, make it a requirement (#31153)
Xuan-Son Nguyen Alde Rojas Pascal
2026-10-25 23:53:08 +01:00
6050b467c0 tests : adjust default threads for test-llama-context (#31157) Georgi Gerganov 2026-10-25 19:46:44 +02:00
2280f0fd16 qwen: merge tensors on convert (#31028) Aman Gupta 2026-10-25 21:23:22 +08:00
1a0b56cec7 ci : disable test-llama-context with fa off in the webgpu (#30977) Masashi Yoshimura 2026-10-25 20:53:55 +09:00
63412d8737 llama: fix exhausting CPU and MEM caused by dry sequence breaker (#31035) Trunk Trick 2026-10-25 19:31:09 +08:00
5c965fd8ca llama-quant : fix 4D tensor quantization using ggml_nrows() and name-based expert detection (#30820) Anthony Wang 2026-10-25 19:28:38 +08:00
f1168ac417 grammar: limit recursion depth to prevent stack overflow (#29495) (#29755) Hrushikesh Ramilla 2026-10-25 16:40:10 +05:30
fa3645feb5 llama : respect no_alloc in all memory caches (#31087) Georgi Gerganov 2026-10-25 11:56:04 +02:00
7790a9231d tools : reject truncated GGUF tensor data in merge (#31044) ShreyaandShreya Agrawal 2026-10-25 14:55:10 +05:30
c49edf0963 CUDA: fuse FFN gate/up matmuls and GLU in MMQ with interleaved gate/up rows (#29948) anujj 2026-10-25 14:54:38 +05:30
a55c16e4ee ggml-cuda: fuse qwen MRoPE + set-rows (#31025) Aman Gupta 2026-10-25 17:15:45 +08:00
e321bdad7e vulkan: Add a larger tile size for conv3d (coopmat2 only) (#29920) Jeff Bolz 2026-10-25 04:13:48 -05:00
459a1ba549 vulkan : one warp tile per subgroup in the small matmul tiles (#30106) Ian McKellar 2026-10-25 02:13:14 -07:00
f0fb5984b0 hexagon: generic dims collapse/reshape for unary, binary and matmul ops (#31071) Max Krasnyansky 2026-10-25 02:11:48 -07:00
9a4580b61c opencl: add topk and lightning indexer (#31026) lhez 2026-10-25 02:11:15 -07:00
b083c384f7 opencl: add initial q2_k support (#30999) lhezandHongqiang Wang 2026-10-25 02:09:56 -07:00
344be0a600 ci : update CPU job assignments (#31136) Georgi Gerganov 2026-10-25 10:35:11 +02:00
b3fa69955e chat : reject JSON nested more than 128 levels before walking it (#30242) Yuri KhrustalevandXuan Son Nguyen 2026-10-25 01:46:35 -04:00
74bca39500 model: DSV4.1 flash (#31086) Aman GuptaandDaniel Han 2026-10-25 09:43:51 +08:00
d327bf1f07 jinja : fix select and selectattr argument handling (#31059) Joseph Patricio 2026-10-24 13:54:46 -07:00
78a8ae14be hexagon: KV-head team split for multicore HMX flash attention (#31055) Ehsan Bateni 2026-10-24 15:19:32 -04:00
f3269ef96a server: refactor server entrypoint (#31113) Xuan-Son Nguyen 2026-10-24 20:04:28 +02:00
3f03b17724 metal : remove mul_mm_id amax rescaling (#30924) Georgi Gerganov 2026-10-24 18:40:19 +03:00
9a0fde9967 rpc : add bounds check in rdma_recv (#30399) Katara 2026-10-24 16:34:46 +01:00
e168e041ec ui: Models Manager as Route + UI improvements (#30988) Aleksander Grygier 2026-10-24 17:34:11 +02:00
ea1d7d66f3 fix(ui): align hover card and context popup styling with dropdown menu (#30978) Aleksander Grygier 2026-10-24 17:33:27 +02:00
71255fefc0 ci : pass per-runner results dir to self-hosted jobs on T4 runners (#31100) Georgi Gerganov 2026-10-24 17:17:04 +03:00
73adf89e38 minimax-m3 : forward quant scales to the graph (fixes #31063) (#31097) Satyam Thakurandorigin-labs-ai 2026-10-24 19:39:02 +05:30
df9beac715 tests : fix test-llama-context for large n_batch (#31083) Alexey Kopytko 2026-10-24 20:14:58 +09:00
4d32b29aaa ci : exclude test-speculative-rejection-sampling from macos-latest-arm64 (#31088) Georgi Gerganov 2026-10-24 14:14:31 +03:00
3032209758 llama : deny -sm tensor for arches with moe amax rescaling (#31081) Georgi Gerganov 2026-10-24 13:48:32 +03:00
701fd25b3a ggml : allow MUL_MAT_ID -1 index to skip the computation (#26631) Xuan-Son Nguyenandyomaytk 2026-10-24 12:15:36 +02:00
3642238cc1 tests : add speculative rejection draft sampling test (#30213) Daniel Bevenius 2026-10-24 11:47:49 +02:00
6b9ba4f2e8 ci : fix data race between runners sharing the model cache (#31046) Georgi Gerganov 2026-10-24 09:21:24 +03:00
8e5a58df87 vulkan: ggml src1 precision support (#30656) Ruben Ortlam 2026-10-24 06:05:05 +02:00
1ddffe9f70 HIP: add byte permute fallback when __builtin_amdgcn_perm is unavailable (#30915) Arseniy Obolenskiy 2026-10-24 06:04:05 +02:00
7894e8b924 mtmd: move build_inp_temporal_merge to base class (#31065) Xuan-Son Nguyen 2026-10-24 03:38:25 +02:00
edc353842c mtmd : lfm2 resize matches torchvision (#30943) Yuri KhrustalevandXuan Son Nguyen 2026-10-23 20:21:06 -04:00
9ae73e68f5 mtmd : add Soprano TTS support (#28607) Hans 2026-10-24 07:09:50 +08:00
0ae0e19766 hexagon: add SGN op (#30346) kurquharandMax Krasnyansky 2026-10-23 14:58:44 -07:00
21ec307146 common : allow concurrent downloads across processes (#29312) Adrien Gallouët 2026-10-23 23:52:56 +02:00
9145ddb955 hip: fix bf16 WMMA operand vector type for gfx11/gfx12 (#30885) WalterCool 2026-10-23 18:31:28 -03:00
1a24df1639 ggm-cuda: Enforce DP4A MMQ tile height multiple of warp size (#30837) thelittlefireman 2026-10-23 21:37:20 +02:00
9ff0d7e90f CUDA: fix build after chunked CUDA graphs (#30804) (#31053) SreeramandJohannes Gäßler 2026-10-23 11:47:20 -07:00
06230cb877 allows the usage of 4-10 gpus for stepfun and laguna in tensor parallel (#24554) Christian Loehnert 2026-10-23 11:56:38 -06:00
5c31142682 qwen4exp : build the QSA selection without the kq_mask (#30965) Alexey Kopytko 2026-10-24 02:49:08 +09:00
da2ee1e7a6 vendor : update LibreSSL to 4.4.0 [no ci] (#31049) Adrien Gallouët 2026-10-23 19:12:35 +02:00
280141e6d4 CUDA: chunk graphs to amortize graph launch (#30804) Aman Gupta 2026-10-24 00:26:16 +08:00
36e36e63a5 hexagon: add CEIL op (#31006) Max Krasnyansky 2026-10-23 07:46:53 -07:00
e1e3b2c409 ui : bump packages to clear npm audit (#31038) Aleksander Grygier 2026-10-23 16:33:13 +02:00
4231b23dc2 ggml : fix time overflow in ggml_time_us (#30969) Kartik Gulia 2026-10-23 18:26:59 +05:30

1 2 3 4 5 ...

Commit Graph Select branches Hide Pull Requests master Mono Color

Commit Graph

Select branches

Hide Pull Requests

master