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

b4988

3714c3ee · llama : fix incorrect Qwen2Moe ffn_moe_out graph callback (#12631) · Mar 28, 2025
b4987

b4ae5081 · metal : improve FA + improve MoE (#12612) · Mar 28, 2025
b4986

b86f6007 · vulkan: fix coopmat shader generation when cross-compiling (#12272) · Mar 28, 2025
b4985

dd373dd3 · llama: fix error on bad grammar (#12628) · Mar 28, 2025
b4984

5d016702 · server : include speculative decoding stats when timings_per_token is enabled (#12603) · Mar 28, 2025
b4982

13731766 · llamafile : ppc64le GEMV forwarding for FP32. (#12594) · Mar 28, 2025
b4981

ab6ab8f8 · rpc : send hash when tensor data is above some fixed threshold (#12496) · Mar 28, 2025
b4980

2099a9d5 · server : Support listening on a unix socket (#12613) · Mar 27, 2025
b4978

5dec47dc · opencl: add multi and vision rope, `gelu_quick` and `im2col` (#12600) · Mar 27, 2025
b4977

f125b8dc · llama : add PLM GGUF Conversion & Inference Support (#12457) · Mar 27, 2025
b4976

953c2a62 · model : restore support for T5Encoder (#12590) · Mar 27, 2025
b4974

029c693f · sync : ggml · Mar 27, 2025
b4972

df0665a4 · sync : ggml · Mar 27, 2025
b4970

c7b43ab6 · llamafile : ppc64le MMA implementation for Q4_0. (#12489) · Mar 27, 2025
b4969

24feaec0 · ggml : riscv: add 128-bit RVV support (#12530) · Mar 27, 2025
b4967

f17a3bb4 · SYCL: implement memset ggml backend buffer interface (#12580) · Mar 27, 2025
b4966

bd40678d · HIP: Add support for RDNA4 targets (#12372) · Mar 26, 2025
b4964

2447ad8a · upgrade to llguidance 0.7.10 (#12576) · Mar 26, 2025
b4963

02082f15 · clip: Fix llama-llava-clip-quantize-cli quantization error under CUDA backend (#12566) · Mar 26, 2025
b4961

5ed38b68 · ggml : fix MUL_MAT_ID repack with Q8_K (#12544) · Mar 26, 2025

1
…
11
12
13
14
15
16
17
18
19
…
178