upstream-archive/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-07-28 13:41:01 -05:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • cd811b7a9d kv-cache : use ggml_set_rows Georgi Gerganov 2025-06-19 19:26:47 +03:00
  • 55a1c5a5fd CUDA: add softmax broadcast (#14475) b5808 Aman Gupta 2025-07-02 20:34:24 +08:00
  • 12a81af45f CUDA: broadcasting for FlashAttention mask (#14500) Johannes Gäßler 2025-07-02 13:42:12 +02:00
  • 8875523eb3 vulkan: support softmax/FA batch and broadcast (#14449) Jeff Bolz 2025-07-01 03:32:56 -05:00
  • ec68e84c32 ggml : support bcast ggml_soft_max_ext, ggml_flash_attn_ext (#14435) Georgi Gerganov 2025-06-27 21:50:57 +03:00
  • 307e79d33d opencl : fix possible buffer overflow in dump_tensor (#14490) b5804 zhouwg 2025-07-02 20:38:10 +08:00
  • d7f5f4e578 simple-chat : fix context-exceeded condition (#14494) b5803 Georgi Gerganov 2025-07-02 14:12:07 +03:00
  • c8a4e470f6 opencl : skip empty nodes on cgraph compute (#14491) b5802 Eric Zhang 2025-07-02 19:00:04 +08:00
  • 71bef66591 cuda : graceful fallback for Mamba-1 models with weird embd size compilade/mamba2 Francis Couture-Harpin 2025-07-02 02:56:42 -04:00
  • 603e43dc91 opencl : update upscale to support align corners (#14488) b5801 lhez 2025-07-02 00:07:42 -07:00
  • 611ba4b264 ci : add OpenCL to labeler workflow (#14496) Sigbjørn Skjæret 2025-07-02 09:02:51 +02:00
  • 73de1fd170 Merge branch 'master' into compilade/mamba2 Francis Couture-Harpin 2025-07-02 02:39:04 -04:00
  • 85841e121d github : add OpenCL backend to issue templates (#14492) Eric Zhang 2025-07-02 13:41:35 +08:00
  • 68b3cd6514 ggml : Callback before abort (#14481) b5798 Björn Ganster 2025-07-02 07:19:31 +02:00
  • de56944147 ci : disable fast-math for Metal GHA CI (#14478) b5797 Georgi Gerganov 2025-07-01 18:04:08 +03:00
  • 1b2aaf28ac Add Vulkan images to docker.md (#14472) Grzegorz Grasza 2025-07-01 15:44:11 +02:00
  • 343b6e94b6 CANN: update aclnnGroupedMatmulV2 to aclnnGroupedMatmulV3 (#14411) b5795 Chenguang Li 2025-07-01 16:47:30 +08:00
  • 6a746cf9c4 vulkan: Split large mul_mat_id to fit in shared memory (#14451) b5794 Jeff Bolz 2025-07-01 03:43:08 -05:00
  • eff5e45443 add GELU_ERF (#14455) b5793 Sigbjørn Skjæret 2025-07-01 10:14:21 +02:00
  • a6a47958a1 ggml : remove trailing whitespace (#0) b5792 Georgi Gerganov 2025-07-01 11:05:48 +03:00
  • f61c05d4b1 sync : ggml Georgi Gerganov 2025-07-01 10:27:52 +03:00
  • 431b2c24f3 ggml-cpu : "align corners" for bilinear upscale/downscale (ggml/1285) Acly 2025-07-01 09:11:00 +02:00
  • 497be7c01d ggml-quants : rename best_mad to best_error (ggml/1283) Daniel Bevenius 2025-06-24 06:10:16 +02:00
  • 79b33b2317 opencl : add GEGLU, REGLU, SWIGLU (#14456) b5788 lhez 2025-07-01 00:19:16 -07:00
  • 0a5a3b5cdf Add Conv2d for CPU (#14388) b5787 Aman Gupta 2025-06-30 23:57:04 +08:00
  • 745f11fed0 memory : correctly handle failure in apply() (#14438) Georgi Gerganov 2025-06-30 18:03:03 +03:00
  • 5dd942de59 metal : disable fast-math for some cpy kernels (#14460) b5785 Georgi Gerganov 2025-06-30 17:04:05 +03:00
  • a7417f5594 ggml-cpu: sycl: Re-enable exp f16 (#14462) b5784 Romain Biessy 2025-06-30 14:52:02 +02:00
  • eb3fa2913e test-backend-ops : disable llama test (#14461) b5783 Diego Devesa 2025-06-30 03:43:15 -07:00
  • c839a2da1a cmake : Remove redundant include path in CMakeLists.txt (#14452) b5782 xiaobing318 2025-06-30 17:48:24 +08:00
  • e9b6350e61 scripts : make the shell scripts cross-platform (#14341) Vedran Miletić 2025-06-30 10:17:18 +02:00
  • caf5681fcb server : support jinja extra template kwargs (Qwen3 enable_thinking feature), from command line and from client (#13196) b5780 matteo 2025-06-29 20:02:53 +02:00
  • 83790b0e7e server : fix appearance of the chats list context menu for Safari (#14322) Renat 2025-06-29 19:29:57 +02:00
  • f47c1d7106 SYCL: disable faulty fp16 exp kernel (#14395) b5778 Akarshan Biswas 2025-06-29 21:07:58 +05:30
  • a5d1fb6212 ggml : fix unmerged GGML_FPxx_TO_FPxx refactoring (#14443) b5777 Sigbjørn Skjæret 2025-06-29 14:38:10 +02:00
  • a0535ffa0d ggml : implement REGLU/GEGLU/SWIGLU ops (#14158) Sigbjørn Skjæret 2025-06-29 11:04:10 +02:00
  • bd9c981d72 vulkan: Add fusion support for RMS_NORM+MUL (#14366) b5775 Jeff Bolz 2025-06-29 02:43:36 -05:00
  • 27208bf657 CUDA: add bf16 and f32 support to cublas_mul_mat_batched (#14361) b5774 Aman Gupta 2025-06-29 01:30:53 +08:00
  • 63a7bb3c7e vulkan: handle noncontig in the final case of ggml_vk_get_cpy_pipeline (#14378) b5773 Jeff Bolz 2025-06-28 10:36:40 -05:00
  • 00d5282c7f vulkan: lock accesses of pinned_memory vector (#14333) b5772 Jeff Bolz 2025-06-28 10:17:09 -05:00
  • 566c16fcce model : add support for ERNIE 4.5 0.3B model (#14408) b5771 Weizhao Ouyang 2025-06-28 22:08:21 +08:00
  • b25e92774e fix async_mode bug (#14432) b5770 Xinpeng Dou 2025-06-28 17:35:41 +08:00
  • 6609507a91 ci : fix windows build and release (#14431) b5769 Sigbjørn Skjæret 2025-06-28 09:57:07 +02:00
  • ceb1bf5a34 vulkan: Fix GGML_VULKAN_SHADER_DEBUG_INFO (#14427) Jeff Bolz 2025-06-27 22:35:30 -05:00
  • 72babea5de graph : make llm_graph_context destructor virtual (#14410) Georgi Gerganov 2025-06-27 21:42:02 +03:00
  • 43678060c1 recurrent : call balloc split_reset() in init_batch() (#14414) Georgi Gerganov 2025-06-27 17:55:45 +03:00
  • 8d94219a4a ggml : add ggml_set_rows (#14274) Radoslav Gerganov 2025-06-27 16:41:40 +03:00
  • 50f88fc4ca ggml : add ggml_scale_bias Xuan Son Nguyen 2025-06-27 11:21:26 +02:00
  • f667f1e624 convert : fix broken sentencepiece vocab (#14416) Sigbjørn Skjæret 2025-06-27 10:42:19 +02:00
  • dc1d109da8 mamba : fix mismatched new and delete size for llm_build_mamba Francis Couture-Harpin 2025-06-26 17:52:28 -04:00
  • 7c3f9c226f Merge branch 'master' into compilade/test-model-random Francis Couture-Harpin 2025-06-26 17:23:16 -04:00
  • 8846aace49 model : gemma3n text-only (#14400) Xuan-Son Nguyen 2025-06-26 19:34:02 +02:00
  • a01047b041 cmake: regen vulkan shaders when shaders-gen sources change (#14398) bandoti 2025-06-26 13:46:53 -03:00
  • b25346221d llama : return mistral-v7-tekken as default template only (#14390) Sigbjørn Skjæret 2025-06-26 15:01:14 +02:00
  • e8215dbb96 metal : add special-case mat-vec mul for ne00 == 4 (#14385) b5760 Georgi Gerganov 2025-06-26 15:51:19 +03:00
  • 5783ae4359 metal : batch rows copy in a single threadgroup (#14384) b5759 Georgi Gerganov 2025-06-26 15:50:15 +03:00
  • bf5bcd0b85 docs: update s390x documentation + add faq (#14389) Aaron Teo 2025-06-26 18:41:41 +08:00
  • 716301d1b0 musa: enable fp16 mma (all) and cublas on qy2 (#13842) b5757 R0CKSTAR 2025-06-26 12:11:59 +08:00
  • 60ef23d6c1 ggml-cpu: enable IBM NNPA Vector Intrinsics (#14317) b5756 Aaron Teo 2025-06-26 05:49:04 +08:00
  • b193d53069 ggml : do not output unprintable characters on GGUF load failure (#14381) b5755 Sigbjørn Skjæret 2025-06-25 23:26:51 +02:00
  • 2bf9d539dd sycl: GGML_SYCL_DISABLE_OPT on by default for all Intel Devices (#13973) b5754 Anton Mitkov 2025-06-25 17:09:55 +01:00
  • 6179578988 batch : require non-coupled batch with sequential split_equal gg/llama-high-throughput-save2 Georgi Gerganov 2025-06-25 17:20:46 +03:00
  • 5eb1a88dc0 batch : optional requirement for sequential sequence ids Georgi Gerganov 2025-06-25 17:02:38 +03:00
  • 6663128448 kv-cache : rework kv_idxs, support seq_cp Georgi Gerganov 2025-06-25 14:48:47 +03:00
  • 0bb1da5854 kv-cache : simplify set_rows logic Georgi Gerganov 2025-06-24 23:14:24 +03:00
  • 73e53dc834 opencl: ref count ggml_backend_opencl_context and refactor profiling (#14254) b5753 lhez 2025-06-24 11:46:25 -07:00
  • 165d822044 graph : support iSWA virtual sequences Georgi Gerganov 2025-06-24 20:35:16 +03:00
  • 1b74b9d73b ggml : extend support for n_seq for soft_max and fattn Georgi Gerganov 2025-06-24 20:14:22 +03:00
  • 8c68219835 kv-cache : fix non-FA path with virutal sequences Georgi Gerganov 2025-06-24 20:01:05 +03:00
  • 7c6487b22f metal : extend ggml_soft_max_ext() to support n_seq dim Georgi Gerganov 2025-06-24 20:00:40 +03:00
  • 62af464227 batch : fix check for empty sequences in memory (#14364) b5752 Georgi Gerganov 2025-06-24 18:26:30 +03:00
  • c148cf1946 cmake : use LLAMA_BUILD_NUMBER when defining LLAMA_INSTALL_VERSION (#14362) b5751 Mathieu Baudier 2025-06-24 15:05:31 +02:00
  • 401c13e3c3 cont : fix build Georgi Gerganov 2025-06-24 15:59:47 +03:00
  • 132143938f tools : tmp adjustments (TMP) Georgi Gerganov 2025-06-24 15:02:58 +03:00
  • 52b9007176 llama : add "virtual sequences" Georgi Gerganov 2025-06-23 16:29:02 +03:00
  • 1b809cee22 server : move no API key doc to /health (#14352) Nigel Bosch 2025-06-24 08:59:11 +00:00
  • 37bdfbef8c wip 3 gg/llama-high-throughput-save Georgi Gerganov 2025-06-24 11:00:05 +03:00
  • abf241045d main : honor --verbose-prompt on interactive prompts (#14350) b5749 Sigbjørn Skjæret 2025-06-24 09:31:00 +02:00
  • 901e20bbe5 jinja : Add Mistral-Small-3.2-24B-Instruct-2506.jinja (#14349) Bartowski 2025-06-24 02:17:58 -04:00
  • efc33ea60d wip 2 Georgi Gerganov 2025-06-24 07:17:44 +03:00
  • 0142961a2e CUDA/HIP: optimize mmv paths taken for HIP devices (#14324) b5747 uvos 2025-06-24 01:12:56 +02:00
  • e33de128c7 common : move string_remove_suffix from quantize and imatrix Francis Couture-Harpin 2025-06-23 16:22:27 -04:00
  • ce82bd0117 ci: add workflow for relocatable cmake package (#14346) bandoti 2025-06-23 15:30:51 -03:00
  • 118d52fefc Merge branch 'master' into compilade/imatrix-batched-chunks Francis Couture-Harpin 2025-06-23 12:54:56 -04:00
  • 0e79355075 quantize : fix dataset name loading from gguf imatrix Francis Couture-Harpin 2025-06-23 12:43:25 -04:00
  • 43cd2b3eb5 imatrix : support 3d tensors with MUL_MAT Francis Couture-Harpin 2025-06-23 11:50:54 -04:00
  • afdb669206 Merge branch 'master' into compilade/mamba2 Francis Couture-Harpin 2025-06-23 10:40:16 -04:00
  • bf2a99e3cb vulkan: update windows SDK in release.yml (#14344) b5745 Jeff Bolz 2025-06-23 08:44:48 -05:00
  • 7664390bc8 wip Georgi Gerganov 2025-06-23 16:29:02 +03:00
  • 72c6bc3f3d llama : better rwkv chat template and add missing inputs.use_jinja setting (#14336) b5744 Molly Sophia 2025-06-23 19:56:19 +08:00
  • defe2158dd CUDA: mul_mat_v support for batch sizes > 1 (#14262) b5743 Johannes Gäßler 2025-06-23 13:11:31 +02:00
  • 36f8e20d08 kv-cache : utilize ggml_set_rows broadcast Georgi Gerganov 2025-06-22 10:28:22 +03:00
  • 332f073589 cont : support non-continuous slots Georgi Gerganov 2025-06-21 16:23:31 +03:00
  • 39d0b1e8df cont : kv-cells cp/set for non-cont slots Georgi Gerganov 2025-06-21 15:26:01 +03:00
  • f875d6cb72 cont : migrate to using set of indices instead of slot head Georgi Gerganov 2025-06-21 11:57:07 +03:00
  • db2bb378b1 cont : gate the ggml_set_rows usage with env var Georgi Gerganov 2025-06-21 10:37:06 +03:00
  • 79dac3c861 kv-cache : use ggml_set_rows Georgi Gerganov 2025-06-19 19:26:47 +03:00
  • 1f647b5992 ggml : fix supports_op Radoslav Gerganov 2025-06-23 11:25:16 +03:00
  • eba97574da ggml : simplify forward_dup_f32 Radoslav Gerganov 2025-06-23 11:16:54 +03:00
  • c0cfc2f78b metal : add ggml_set_rows implementation Georgi Gerganov 2025-06-22 18:45:52 +03:00