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

b3873

a7ad5535 · ggml-backend : add device description to CPU backend (#9720) · Oct 03, 2024
b3872

d6fe7abf · ggml: unify backend logging mechanism (#9709) · Oct 03, 2024
b3870

841713e1 · rpc : enable vulkan (#9714) · Oct 03, 2024
b3869

56399714 · Fixed dequant precision issues in Q4_1 and Q5_1 (#9711) · Oct 03, 2024
b3868

c83ad6d0 · ggml-backend : add device and backend reg interfaces (#9707) · Oct 03, 2024
b3867

a39ab216 · llama : reduce compile time and binary size (#9712) · Oct 02, 2024
b3866

f536f4c4 · [SYCL] Initial cmake support of SYCL for AMD GPUs (#9658) · Oct 02, 2024
b3865

00b7317e · vulkan : do not use tensor->extra (#9407) · Oct 02, 2024
b3864

76b37d15 · gguf-split : improve --split and --merge logic (#9619) · Oct 02, 2024
b3863

148844fe · examples : remove benchmark (#9704) · Oct 02, 2024
b3861

f1b8c427 · sync : ggml · Oct 01, 2024
b3856

cad341d8 · metal : reduce command encoding overhead (#9698) · Oct 01, 2024
b3855

a90484c6 · llama : print correct model type for Llama 3.2 1B and 3B · Oct 01, 2024
b3853

6f1d9d71 · Fix Docker ROCM builds, use AMDGPU_TARGETS instead of GPU_TARGETS (#9641) · Sep 30, 2024
b3849

8277a817 · console : utf-8 fix for windows stdin (#9690) · Sep 30, 2024
b3848

c919d5db · ggml : define missing HWCAP flags (#9684) · Sep 29, 2024
b3847

d0b1d663 · sync : ggml · Sep 29, 2024
b3841

faac0bae · common : ensure llama_batch size does not exceed max size (#9668) · Sep 29, 2024
b3837

1b2f992c · test-backend-ops : use flops for some performance tests (#9657) · Sep 28, 2024
b3835

6102037b · vocab : refactor tokenizer to reduce init overhead (#9449) · Sep 28, 2024

1
…
48
49
50
51
52
53
54
55
56
…
178