| 2024-07-08 |
Georgi Gerganov | sync : ggml |
commit | commitdiff | tree |
| 2024-07-08 |
Georgi Gerganov | ggml : sync sycl (skip) (#0) |
commit | commitdiff | tree |
| 2024-07-08 |
Georgi Gerganov | scripts : fix sync scripts |
commit | commitdiff | tree |
| 2024-07-08 |
Daniel Bevenius | ggml : remove unnecessary UNUSED macro call (ggml/880) |
commit | commitdiff | tree |
| 2024-07-08 |
Natsu | cmake : add GGML_BUILD and GGML_SHARED macro definition... |
commit | commitdiff | tree |
| 2024-07-08 |
Ouadie EL FAROUKI | Enabled more data types for oneMKL gemm_batch (llama... |
commit | commitdiff | tree |
| 2024-07-08 |
Johannes Gäßler | CUDA: MMQ support for iq4_nl, iq4_xs (llama/8278) |
commit | commitdiff | tree |
| 2024-07-08 |
Daniele | CUDA: revert part of the RDNA1 optimizations (llama... |
commit | commitdiff | tree |
| 2024-07-08 |
Johannes Gäßler | CUDA: fix MMQ stream-k rounding if ne00 % 128 != 0... |
commit | commitdiff | tree |
| 2024-07-08 |
luoyu-intel | Fix WARP_SIZE=16 bug of Intel GPU (llama/8266) |
commit | commitdiff | tree |
| 2024-07-08 |
Neo Zhang Jianyu | rm get_work_group_size() by local cache for performance... |
commit | commitdiff | tree |
| 2024-07-08 |
Daniele | Define and optimize RDNA1 (llama/8085) |
commit | commitdiff | tree |
| 2024-07-08 |
Judd | fix typo (llama/8267) |
commit | commitdiff | tree |
| 2024-07-08 |
Clint Herron | Removes multiple newlines at the end of files that... |
commit | commitdiff | tree |
| 2024-07-08 |
slaren | cuda : update supports_op for matrix multiplication... |
commit | commitdiff | tree |
| 2024-07-08 |
luoyu-intel | Fix win build conflict of math library (llama/8230) |
commit | commitdiff | tree |
| 2024-07-08 |
luoyu-intel | Fix the sub group size of Intel (llama/8106) |
commit | commitdiff | tree |
| 2024-07-08 |
Johannes Gäßler | CUDA: refactor and optimize IQ MMVQ (llama/8215) |
commit | commitdiff | tree |
| 2024-07-08 |
zhentaoyu | Update SYCL-Rope op and Refactor (llama/8157) |
commit | commitdiff | tree |
| 2024-07-08 |
Johannes Gäßler | CUDA: fix MMQ stream-k for --split-mode row (llama... |
commit | commitdiff | tree |
| 2024-07-08 |
John Balis | feat: cuda implementation for `ggml_conv_transpose_1d... |
commit | commitdiff | tree |
| 2024-07-08 |
Georgi Gerganov | ci : disable java build |
commit | commitdiff | tree |
| 2024-07-08 |
Emmanuel Schmidbauer | server : add inference path to make OAI API compatible... |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | sync : ggml + fix sync script |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | make : disable CUDA graphs |
commit | commitdiff | tree |
| 2024-06-26 |
slaren | ggml : add GGML_CUDA_USE_GRAPHS option, restore GGML_CU... |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | make : disable CUDA mel build |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | cmake : minor fixes |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | make : fix missing -O3 |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | whisper : disable CUDA mel + fix FFMPEG |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | sync : ggml |
commit | commitdiff | tree |
| 2024-06-26 |
Georgi Gerganov | whisper : reorganize source code + improve CMake (... |
commit | commitdiff | tree |
| 2024-06-18 |
mky_coder | whisper : optimize fft() function (#2242) |
commit | commitdiff | tree |
| 2024-06-18 |
Georgi Gerganov | talk-llama : sync llama.cpp |
commit | commitdiff | tree |
| 2024-06-18 |
Georgi Gerganov | whisper : use ggml_backend_sched (#2239) |
commit | commitdiff | tree |
| 2024-06-18 |
Georgi Gerganov | fix : remove extra files |
commit | commitdiff | tree |
| 2024-06-18 |
Georgi Gerganov | scripts : sync ggml-blas |
commit | commitdiff | tree |
| 2024-06-18 |
Georgi Gerganov | build : update make / cmake |
commit | commitdiff | tree |
| 2024-06-18 |
Georgi Gerganov | sync : ggml |
commit | commitdiff | tree |
| 2024-06-18 |
slaren | move BLAS to a separate backend (cont) (llama/6210) |
commit | commitdiff | tree |
| 2024-06-18 |
0cc4m | Vulkan Shader Refactor, Memory Debugging Option (llama... |
commit | commitdiff | tree |
| 2024-06-18 |
Georgi Gerganov | scripts : stop sync whisper example from ggml |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | cmake : fix sycl build (#0) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : remove OpenCL (#0) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | sycl : sync (#0) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | cuda : enable CUDA graphs (#0) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | talk-llama : sync llama.cpp |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | cmake : fix CUDA build (#0) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | sync : ggml |
commit | commitdiff | tree |
| 2024-06-16 |
Hong Bo PENG | ggml : fix and optimize ppc64le (ggml/849) |
commit | commitdiff | tree |
| 2024-06-16 |
Daniel Bevenius | ggml : remove duplicate include of ggml-common.h (ggml... |
commit | commitdiff | tree |
| 2024-06-16 |
Meng, Hengyu | remove global variables (llama/7710) |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: faster q2_K, q3_K MMQ + int8 tensor cores (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | metal : utilize max shared memory for mul_mat_id (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
Radoslav Gerganov | rpc : fix ggml_backend_rpc_supports_buft() (llama/7918) |
commit | commitdiff | tree |
| 2024-06-16 |
slaren | move BLAS to a separate backend (llama/6210) |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: fix broken oob check for FA vec f32 kernel (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | tests : add non-cont unary tests (llama/7857) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : improve ggml_is_contiguous logic (llama/7856) |
commit | commitdiff | tree |
| 2024-06-16 |
k.h.lai | vulkan: select only one device for single gpu with... |
commit | commitdiff | tree |
| 2024-06-16 |
0cc4m | Update Vulkan RoPE implementation (llama/7818) |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: int8 tensor cores for MMQ (q4_K, q5_K, q6_K)... |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: use tensor cores for MMQ (llama/7676) |
commit | commitdiff | tree |
| 2024-06-16 |
Ben Ashbaugh | use the correct SYCL context for host USM allocations... |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: revise q8_1 data layout for mul_mat_q (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
slaren | vulkan : reuse parent extra for views (llama/7806) |
commit | commitdiff | tree |
| 2024-06-16 |
pengxin99 | fix softmax r2r result wrong issue (llama/7811) |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: refactor mmq, dmmv, mmvq (llama/7716) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : refactor rope norm/neox (llama/7634) |
commit | commitdiff | tree |
| 2024-06-16 |
agray3 | Allow number of nodes in CUDA graph to change (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : remove OpenCL (llama/7735) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : prevent builds with -ffinite-math-only (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
Radoslav Gerganov | llama : offload to RPC in addition to other backends... |
commit | commitdiff | tree |
| 2024-06-16 |
Masaya, Kato | ggml : use OpenMP as a thread pool (llama/7606) |
commit | commitdiff | tree |
| 2024-06-16 |
0cc4m | Vulkan Mixture of Experts (MoE) support (llama/7628) |
commit | commitdiff | tree |
| 2024-06-16 |
woachk | kompute : implement op_getrows_f32 (llama/6403) |
commit | commitdiff | tree |
| 2024-06-16 |
Dave Airlie | fix bug introduced in using calloc (llama/7701) |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | Fix FlashAttention debug test, FP32 assert (llama/7684) |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: fix Pascal FA, deq. KV to FP16 for batch > 8... |
commit | commitdiff | tree |
| 2024-06-16 |
Johannes Gäßler | CUDA: quantized KV support for FA vec (llama/7527) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : fix loongson compile warnings (llama/7537) |
commit | commitdiff | tree |
| 2024-06-16 |
Chris Elrod | faster avx512 exp implementation (llama/7551) |
commit | commitdiff | tree |
| 2024-06-16 |
junchao-loongson | ggml : fix loongarch build (O2 issue) (llama/7636) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | metal : remove invalid asserts (llama/7617) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | metal : add missing asserts (llama/7617) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : fix YARN + add tests + add asserts (llama/7617) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | cuda : non-cont concat support (llama/7610) |
commit | commitdiff | tree |
| 2024-06-16 |
Radoslav Gerganov | llama-bench : add support for the RPC backend (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
slaren | ggml : use atomic_flag for critical section (llama... |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | examples : adapt to new ggml_concat (ggml/0) |
commit | commitdiff | tree |
| 2024-06-16 |
zhouwg | ggml : fix typo in ggml.c (llama/7603) |
commit | commitdiff | tree |
| 2024-06-16 |
Meng, Hengyu | Align GEMM dispatch (llama/7566) |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | sycl : fix assert (llama/7563) |
commit | commitdiff | tree |
| 2024-06-16 |
k.h.lai | vulkan: properly initialize vulkan devices for LLAMA_SP... |
commit | commitdiff | tree |
| 2024-06-16 |
Radoslav Gerganov | rpc : resource management rework (llama/7562) |
commit | commitdiff | tree |
| 2024-06-16 |
Neo Zhang | fix ggml_sycl_mul_mat_id() to match the change of api... |
commit | commitdiff | tree |
| 2024-06-16 |
Georgi Gerganov | ggml : generalize GGML_OP_CONCAT (llama/7563) |
commit | commitdiff | tree |
| 2024-06-16 |
Djip007 | update HIP_UMA #7399 (llama/7414) |
commit | commitdiff | tree |
| 2024-06-16 |
agray3 | Allow multiple copy function pointers for CUDA graph... |
commit | commitdiff | tree |
| 2024-06-16 |
AidanBeltonS | Fix q_xxs using mul_mat_q (llama/7459) |
commit | commitdiff | tree |
| next |
overview / pkg / ggml / sources / whisper.cpp / shortlog