为了安全，强烈建议开启2FA双因子认证：User Settings -> Account -> Enable two-factor authentication！！！

Tags

Tags give the ability to mark specific points in history as being important

b4113

467576b6 · CMake: default to -arch=native for CUDA build (#10320) · Nov 17, 2024
b4112

eda7e1d4 · ggml : fix possible buffer use after free in sched reserve (#9930) · Nov 17, 2024
b4111

24203e9d · ggml : inttypes.h -> cinttypes (#0) · Nov 17, 2024
b4103

4e54be0e · llama/ex: remove --logdir argument (#10339) · Nov 16, 2024
b4102

db4cfd5d · llamafile : fix include path (#0) · Nov 16, 2024
b4100

bcdb7a23 · server: (web UI) Add samplers sequence customization (#10255) · Nov 16, 2024
b4098

772703c8 · vulkan: Optimize some mat-vec mul quant shaders (#10296) · Nov 16, 2024
b4096

1e58ee13 · ggml : optimize Q4_0 into Q4_0_X_Y repack (#10324) · Nov 16, 2024
b4095

89e4caaa · llama : save number of parameters and the size in llama_model (#10286) · Nov 16, 2024
b4094

74d73dc8 · Make updates to fix issues with clang-cl builds while using AVX512 flags (#10314) · Nov 15, 2024
b4093

4047be74 · scripts: update compare-llama-bench.py (#10319) · Nov 15, 2024
b4092

883d206f · ggml : fix some build issues · Nov 15, 2024
b4091

09ecbcb5 · cmake : fix ppc64 check (whisper/0) · Nov 15, 2024
b4088

18429220 · AVX BF16 and single scale quant optimizations (#10212) · Nov 15, 2024
b4087

f0204a0e · ci: build test musa with cmake (#10298) · Nov 15, 2024
b4085

9901068a · server : (web UI) add copy button for code block, fix api key (#10242) · Nov 15, 2024
b4082

5a54af4d · sycl: Use syclcompat::dp4a (#10267) · Nov 15, 2024
b4081

1607a5e5 · backend cpu: add online flow for aarch64 Q4_0 GEMV/GEMM kernels (#9921) · Nov 15, 2024
b4080

ae8de6d5 · ggml : build backends as libraries (#10256) · Nov 14, 2024
b4079

4a8ccb37 · CUDA: no -sm row for very small matrices (#10185) · Nov 14, 2024

1
…
40
41
42
43
44
45
46
47
48
…
178