wylab/llama.cpp
1
0
Fork 1
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-07-16 01:15:58 +02:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • d414db02d3 vulkan: Use fewer rows for scalar FA when HS is not a multiple of 16 (#17455) b7152 Jeff Bolz 2025-11-25 00:11:27 -06:00
  • 2b4c7927ee Merge remote-tracking branch 'upstream/master' into backend-sampling Daniel Bevenius 2025-11-25 06:10:33 +01:00
  • 877566d512 llama: introduce support for model-embedded sampling parameters (#17120) b7151 Aaron Teo 2025-11-25 09:56:07 +08:00
  • 3d07caa99b vulkan: more FA details in vk_perf_logger (#17443) b7150 Jeff Bolz 2025-11-24 15:25:24 -06:00
  • 134e6940ca llama : skip output reordering for single token batches (#17466) b7149 Daniel Bevenius 2025-11-24 21:06:17 +01:00
  • a02adf4211 sampling : add assertions for contiguous tensors in async copy functions Daniel Bevenius 2025-11-24 21:00:03 +01:00
  • 883a87043a samplers : add missing cont Georgi Gerganov 2025-11-24 21:46:57 +02:00
  • 0543f928a3 HIP: WMMA-MMQ kernels for RDNA 4 (#17156) b7148 Jiacheng (Jason) Chen 2025-11-24 14:00:10 -05:00
  • b26c7069fb common : initialize backend samplers Georgi Gerganov 2025-11-24 20:25:44 +02:00
  • e2d4f0829c llama-cli : fix dangling reference to sampler config Georgi Gerganov 2025-11-24 19:51:32 +02:00
  • d0bea21a3c examples : update batched to use backend sampling Daniel Bevenius 2025-11-24 16:37:22 +01:00
  • b61de2b2df convert : allow quantizing lora again (#17453) Sigbjørn Skjæret 2025-11-24 15:50:55 +01:00
  • 25f33806d3 sampling : add debug log when backend sampler selects token Daniel Bevenius 2025-11-24 15:03:41 +01:00
  • b8372eecd9 server: split server.cpp code into server/common/task/queue (#17362) b7146 Xuan-Son Nguyen 2025-11-24 14:41:53 +01:00
  • 6ab8eacddf examples : add -kvu to batched usage example [no ci] (#17469) Daniel Bevenius 2025-11-24 14:38:45 +01:00
  • 2d50b9d8cb sync : ggml b7144 Georgi Gerganov 2025-11-24 14:28:37 +02:00
  • 697edfeead ggml : remove dirty flag from version string (ggml/1391) Daniel Bevenius 2025-11-24 12:51:50 +01:00
  • 8eb9b4769d sampling : remove redundant checks for stride and size [no ci] Daniel Bevenius 2025-11-24 13:53:29 +01:00
  • 4a90583d7d sampling : cleanup and clarify output_reserve Daniel Bevenius 2025-11-24 13:26:18 +01:00
  • dbb852b549 ggml-cpu: arm64: q4_K repack gemm and gemv implementations (i8mm) (#16739) b7142 Alberto Cabrera Pérez 2025-11-24 11:08:11 +00:00
  • 5f55c385cb ggml: add RISC-V cpu-feats (#17461) b7141 ixgbe 2025-11-24 19:07:14 +08:00
  • 72f80499ee server : headers cleanup gg/tmp Georgi Gerganov 2025-11-24 10:43:56 +02:00
  • d88ba1813c common : remove build-info.cpp from commit [no ci] Daniel Bevenius 2025-11-24 09:31:14 +01:00
  • 7816f0bb56 Merge remote-tracking branch 'upstream/master' into backend-sampling Daniel Bevenius 2025-11-24 07:44:06 +01:00
  • 50d21aa4a4 tests : cleanup test-backend-sampler.cpp Daniel Bevenius 2025-11-24 07:18:39 +01:00
  • 4902eebe33 models : Added support for RND1 Diffusion Language Model (#17433) b7140 william pan 2025-11-23 22:16:56 -08:00
  • 923ae3c619 hexagon: add support for ROPE_NEOX (#17458) b7139 Max Krasnyansky 2025-11-23 18:55:56 -08:00
  • 01ad35e6d6 CANN: Define cann_graph_update_required before macro (#17434) b7138 Raul Torres 2025-11-24 02:02:52 +00:00
  • fcb013847c ggml-hexagon: Initial Hexagon v68/v69 support (#17394) b7137 M. Mediouni 2025-11-24 01:54:49 +01:00
  • d5bc1ad110 ggml-hexagon: add hex_supported_buffer for better buffer supported check (#17212) b7136 nullname 2025-11-24 06:26:36 +08:00
  • 0c7220db56 webui: minor settings reorganization and add disable autoscroll option (#17452) Pascal 2025-11-23 18:42:00 +01:00
  • 9e273f7aa4 sampling : fix copying both sampled tokens and logits/probs from backend Daniel Bevenius 2025-11-23 13:08:08 +01:00
  • ae23d2d2c1 sampling: clarify candidate ids usage in comments Daniel Bevenius 2025-11-23 11:28:19 +01:00
  • 65500d05ab sampling : add stride variable for clarity Daniel Bevenius 2025-11-23 11:27:54 +01:00
  • 96ac5a2329 cuda : support non-contiguous i32 to i32 copy (#17326) b7134 Sigbjørn Skjæret 2025-11-23 11:13:34 +01:00
  • bc809e9c53 vulkan: Update docker image to Ubuntu 26.04 to enable glslc features (#17439) Eric Curtin 2025-11-23 09:29:36 +00:00
  • 722f9defe9 vulkan: intel mmv fix attempt 0cc4m/vulkan-intel-mmv-fix 0cc4m 2025-11-23 10:13:19 +01:00
  • 54d83bbe85 vulkan: remove a couple unnecessary switches (#17419) b7132 Jeff Bolz 2025-11-22 23:29:40 -06:00
  • 4949ac0f18 ci : switch to BoringSSL on Server workflow (#17441) b7131 Adrien Gallouët 2025-11-22 21:38:19 +01:00
  • 3f3a4fb9c3 Revive MUL_MAT_ID to perf testing (#17397) b7130 Masato Nakasaka 2025-11-22 18:55:43 +09:00
  • 8174e29b0e release: fix linting Aaron Teo 2025-11-22 13:59:24 +08:00
  • 49d4164952 release: fix duplicate libs, store symbolic links Aaron Teo 2025-11-16 19:51:53 +08:00
  • d6abfe8c84 release: add deprecation notice to release.yml fix-release-duplicate-libs-b7083-d6abfe8 Aaron Teo 2025-11-22 11:52:11 +08:00
  • 028f93ef98 HIP: RDNA4 tensor core support for MMF (#17077) b7129 yulo 2025-11-22 07:03:24 +08:00
  • 8e9ddba610 opencl: refine condition for kqv mm (#17392) b7128 lhez 2025-11-21 14:34:48 -08:00
  • 79b8cf2a75 Merge remote-tracking branch 'upstream/master' into backend-sampling Daniel Bevenius 2025-11-21 16:38:32 +01:00
  • 79bfc1c0d3 release: rm gunzip Aaron Teo 2025-11-21 22:49:48 +08:00
  • dec86a0c8c release: add .tar release Aaron Teo 2025-11-21 22:45:07 +08:00
  • 23bc779a6e model : detect GigaChat3-10-A1.8B as deepseek lite (#17420) b7127 ubergarm 2025-11-21 08:51:38 -05:00
  • 9b2439347f common, tools : refactor model loading to support backend samplers Daniel Bevenius 2025-11-21 14:26:52 +01:00
  • 61ffe41dc1 sampling : use pinned memory for backend sampling buffers Daniel Bevenius 2025-11-21 14:02:16 +01:00
  • 28175f857d cmake : add option to build and link BoringSSL (#17205) b7126 Adrien Gallouët 2025-11-21 11:46:45 +01:00
  • 9cc4080441 ci : start using OpenSSL (#17235) Adrien Gallouët 2025-11-21 11:45:00 +01:00
  • f1ffbba68e vulkan: disable async for older Intel devices (#17369) b7124 Jeff Bolz 2025-11-21 02:58:17 -06:00
  • 2370665e56 CANN: Refactor evaluate_and_capture_cann_graph (#17333) b7123 Raul Torres 2025-11-21 08:23:29 +00:00
  • c1625620f6 sampling : return early if backend sampling is disabled Daniel Bevenius 2025-11-21 08:44:25 +01:00
  • 21d31e0810 ggml-hexagon: fix swiglu failure at test-backend-ops (#17344) b7122 nullname 2025-11-21 07:45:05 +08:00
  • dd0f321941 readme : add Unsloth exporting to GGUF in tools (#17411) Daniel Han 2025-11-20 11:07:36 -08:00
  • 054a45c3d3 grammar: fix regression caused by #17381 (#17412) b7120 Xuan-Son Nguyen 2025-11-20 18:35:10 +01:00
  • 6cdda87baf ci : disable op offload in some tests sl/realloc-error-cont Georgi Gerganov 2025-11-20 17:16:50 +02:00
  • 0d28b16bdc sampling : introduce sampling_info struct Daniel Bevenius 2025-11-20 14:31:37 +01:00
  • 4c91f2633f Improved file naming & structure for UI components (#17405) Aleksander Grygier 2025-11-20 14:07:31 +01:00
  • 92c0b387a9 grammar : fix integer overflow (#17381) b7118 Piotr Wilkin (ilintar) 2025-11-20 13:47:04 +01:00
  • 2286a360ff sync : ggml b7117 Georgi Gerganov 2025-11-20 14:09:48 +02:00
  • 1d321e592b metal : fix compile on macos 11 (whisper/3533) YangLe 2025-11-20 19:54:54 +08:00
  • 196f5083ef common : more accurate sampling timing (#17382) Georgi Gerganov 2025-11-20 13:40:10 +02:00
  • 5088b435d4 convert : fix TypeError when loading base model remotely in convert_lora_to_gguf (#17385) o7si 2025-11-20 19:30:12 +08:00
  • 845f200b28 ggml : Fix transposed SOLVE_TRI result (#17323) b7113 Piotr Wilkin (ilintar) 2025-11-20 11:58:21 +01:00
  • a7784a8b1d DGX Spark: UMA support (#17368) b7112 Scott Fudally 2025-11-20 02:32:02 -08:00
  • 79bb743512 ggml : remove useless and error-prone variadic macros (#17399) b7111 Adrien Gallouët 2025-11-20 11:18:27 +01:00
  • 3ae282a06f kleidiai: fix zero-size array declaration (#17240) b7110 sudhiarm 2025-11-20 09:45:49 +00:00
  • ed4345bdd9 squash! common : fix regression caused by extra memory allocations during sampling Daniel Bevenius 2025-11-20 07:56:33 +01:00
  • 5be353ec4a ggml-cpu:add RISC-V RVV (Zvfh) optimization for FP16 vector scaling (#17314) b7109 ixgbe 2025-11-20 14:09:18 +08:00
  • 0c660e7390 Merge remote-tracking branch 'upstream/master' into backend-sampling Daniel Bevenius 2025-11-20 06:57:24 +01:00
  • 7d77f07325 vulkan: implement ADD1, ARANGE, FILL, SOFTPLUS, STEP, ROUND, CEIL, FLOOR, TRUNC (#17319) b7108 Giuseppe Scrivano 2025-11-19 17:29:45 +01:00
  • 1fa4551af0 vulkan: support larger argsort (#17313) b7107 Jeff Bolz 2025-11-19 10:25:50 -06:00
  • 2eba631b81 vulkan: Add copy_transpose shader (#17371) b7106 Jeff Bolz 2025-11-19 09:50:43 -06:00
  • 18ed4d8f96 squash! sampling : simplify backend sampling logic decode Daniel Bevenius 2025-11-19 15:10:15 +01:00
  • 99c53d6558 webui: Add a "Continue" Action for Assistant Message (#16971) Aleksander Grygier 2025-11-19 14:39:50 +01:00
  • 38f408c253 common : fix regression caused by extra memory allocations during sampling Georgi Gerganov 2025-11-19 13:43:29 +02:00
  • 07b0e7a5ac convert : use self.block_count everywhere instead of reading hparams (#17359) Sigbjørn Skjæret 2025-11-19 11:52:38 +01:00
  • d74eb61aa7 squash! sampling : simplify backend sampling logic decode Daniel Bevenius 2025-11-19 11:29:26 +01:00
  • fd7353d5eb cuda: fix rope fusion for gemma3 (#17378) b7103 Aman Gupta 2025-11-19 18:25:05 +08:00
  • 6fd4f95367 Fix too relaxed check on CUDA "fast copy" (can_be_transposed) condition (#17332) b7102 Piotr Wilkin (ilintar) 2025-11-19 10:36:33 +01:00
  • 7e98ebcc6b sampling : simplify backend sampling logic decode Daniel Bevenius 2025-11-19 09:31:33 +01:00
  • e4838046f3 llama : update worst-case graph for unified cache Georgi Gerganov 2025-11-19 09:44:04 +02:00
  • 980b7cd17e vulkan: force full subgroups for flash attention to fix intel subgroup crash (#17356) b7101 Ruben Ortlam 2025-11-19 08:46:26 +01:00
  • c49daff5ba ggml-cpu: Don't pass -mpowerpc64 when -mcpu already implies it (#17308) b7100 Jeremy Rand 2025-11-19 06:19:00 +00:00
  • 51fee29822 sampling : always populate logits for sampled probs Daniel Bevenius 2025-11-19 07:14:11 +01:00
  • 0da7e7dccc sampling : remove version from sampler chain Daniel Bevenius 2025-11-19 06:59:03 +01:00
  • 10e9780154 chat: fix int overflow, prevent size calculation in float/double (#17357) b7099 Xuan-Son Nguyen 2025-11-18 19:11:53 +01:00
  • a045492088 vocab : call reserve() for building plamo-2-translate suffix (#17343) Haiyue Wang 2025-11-19 01:58:22 +08:00
  • 1920345c3b common : Generalized XML-style tool-call parsing with streaming support (GLM 4.5/4.6 + MiniMax M2 + SeedOSS + Kimi-K2 + Qwen3-Coder + Apriel-1.5 + Xiaomi-MiMo) (#16932) b7097 hksdpc255 2025-11-19 04:54:15 +11:00
  • 26be108be8 CUDA: Optimize argsort for gpu-based token sampling Oliver Simons 2025-11-18 18:17:44 +01:00
  • 561a3e2788 ci : change the openEuler-310p image to fix release (#17361) b7096 jiahao su 2025-11-19 01:10:23 +08:00
  • 625010d42d move enum stop_type to server-task Xuan Son Nguyen 2025-11-18 17:28:21 +01:00
  • 311c1a347f sampling : ensure at most one output token per seq Daniel Bevenius 2025-11-18 16:01:54 +01:00
  • f40a2e5f11 gitignore : be more specific about ignored stuff (#17354) Georgi Gerganov 2025-11-18 16:44:53 +02:00
  • 82957a90f2 sampling : always expose sampled_ids Daniel Bevenius 2025-11-18 14:54:49 +01:00
  • ca993bad51 rm redundant includes Xuan Son Nguyen 2025-11-18 15:01:17 +01:00