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

b4174

0eb4e12b · vulkan: Fix a vulkan-shaders-gen arugment parsing error (#10484) · Nov 26, 2024
b4173

0cc63754 · Introduce llama-run (#10291) · Nov 25, 2024
b4172

50d5cecb · ci : build docker images only once daily (#10503) · Nov 25, 2024
b4171

9fd8c268 · server : add more information about error (#10455) · Nov 25, 2024
b4170

47f931c8 · server : enable cache_prompt by default (#10501) · Nov 25, 2024
b4169

106964e3 · metal : enable mat-vec kernels for bs <= 4 (#10491) · Nov 25, 2024
b4168

80acb7b4 · Rename Olmo1124 to Olmo2 (#10500) · Nov 25, 2024
b4167

10bce045 · llama : accept a list of devices to use to offload a model (#10497) · Nov 25, 2024
b4164

9ca2e677 · server : add speculative decoding support (#10455) · Nov 25, 2024
b4163

5931c1f2 · ggml : add support for dynamic loading of backends (#10469) · Nov 25, 2024
b4162

f6d12e7d · tests : fix compile warning · Nov 25, 2024
b4161

b7564411 · metal : minor code formatting · Nov 25, 2024
b4160

5a898779 · [SYCL] Fix building Win package for oneAPI 2025.0 update (#10483) · Nov 25, 2024
b4157

dc39012c · llama : fix op mul check with command-r-plus (#10476) · Nov 24, 2024
b4154

55ed008b · ggml : do not use ARM features not included in the build (#10457) · Nov 23, 2024
b4153

6dfcfef0 · ci: Update oneAPI runtime dll packaging (#10428) · Nov 22, 2024
b4151

c18610b4 · CANN: Support Ascend310P to accelerate F32 and F16 Model (#10216) · Nov 22, 2024
b4150

a5e47592 · cuda : optimize argmax (#10441) · Nov 21, 2024
b4149

1bb30bf2 · llama : handle KV shift for recurrent models (#10402) · Nov 21, 2024
b4148

87a533be · sync : ggml · Nov 21, 2024

1
…
38
39
40
41
42
43
44
45
46
…
178