summaryrefslogtreecommitdiff
path: root/ggml/src
AgeCommit message (Expand)Author
2025-05-13Better CPU FA performance for DeepSeek-Lite (#410)Kawrakow
2025-05-12Fix new CUDA FA on Touring (#413)Kawrakow
2025-05-12Faster DeepSeek FA on CUDA (#408)Kawrakow
2025-05-12GPU offload policy (#405)Kawrakow
2025-05-11Revert "Fix race in the CUDA DeepSeek FA kernel (#406)"Iwan Kawrakow
2025-05-11Fix race in the CUDA DeepSeek FA kernel (#406)Kawrakow
2025-05-10TG improvements for MoE models (#404)Kawrakow
2025-05-09Fix CUDA FlashMLA-3 with quantized KV cache (#400)Kawrakow
2025-05-07FlashMLA-3 for DeepSeek models on CUDA (#386)Kawrakow
2025-05-07Fix DeepSeek q8_0 cache (#391)Kawrakow
2025-05-07Fix build for Xeon Gold 6226R (#390)Kawrakow
2025-05-05Fix DeepSeek FA (#382)Kawrakow
2025-05-04CUDA: MMQ for IQ4_KS (#374)Kawrakow
2025-05-04CUDA: faster FA TG for GQA models (#370)Kawrakow
2025-05-04Another attempt to fix #367 (#371)Kawrakow
2025-05-03Trying to fix iq1_s_r4/iq1_m_r4 quantization failure (#368)Kawrakow
2025-05-02Fix FA bug on AVX2 (#364)Kawrakow
2025-04-30Fix IQK_FA_ALL_QUANTS on AVX2 (#360)Kawrakow
2025-04-29CPU FA improvements (#351)Kawrakow
2025-04-26Fix division by zero bug (#349)Kawrakow
2025-04-26Add support for Cohere2 (#341)Kawrakow
2025-04-25Fix q4_1 and q5_1 on Arm (#348)Kawrakow
2025-04-25Add ability to manually set arch flags (#347)Kawrakow
2025-04-25Fix FA on ARM (#346)Kawrakow
2025-04-24cuda: use switch in constexpr funcs (#343)Kawrakow
2025-04-21Fix termux/android build (#336)saood06
2025-04-17Better TG performance for GQA models (CPU) (#332)Kawrakow
2025-04-15Better gemm/gemv on AVX2 fr q4_0_r8 (#331)Kawrakow
2025-04-15Allow q8_0 KV cache for head size 256 (#330)Kawrakow
2025-04-13Improved IQ1_M quantization (#327)Kawrakow
2025-04-07Better iq2_xs quantization (#312)Kawrakow
2025-04-07Add copyright notices (#317)Kawrakow
2025-04-05We need to synchronize before using device to host async memcpy (#313)Kawrakow
2025-04-04Add -flax-vector-conversions for GCC on ARM (#311)Kawrakow
2025-04-03Metal: FA and FlashMLA (#310)Kawrakow
2025-04-03Fix GCC compilation errors on ARM (#309)Kawrakow
2025-04-03Metal: much faster MoE prompt processing (#307)Kawrakow
2025-04-01Fix ARM_NEON build failure due to q8_2 (#303)Kawrakow
2025-04-01Quantization improvements (2) (#302)Kawrakow
2025-04-01Fix #300 (#301)Kawrakow
2025-03-29Quantization improvements (#295)Kawrakow
2025-03-27Use bf16 instead of fp16 block scales for q8_1 (#292)Kawrakow
2025-03-25CUDA: better MoE implementation (#283)Kawrakow
2025-03-23Improve DeepSeek batched processing speed (#282)Kawrakow
2025-03-23Attempt to improve FlashMLA on the CPU (#277)Kawrakow
2025-03-22Native build ooption for CUDA when GGML_NATIVE is set (#280)Kawrakow
2025-03-22Fighting with cmake (#279)Kawrakow
2025-03-21Convert models to row-interleaved quants using the quantize tool (#272)Kawrakow
2025-03-19Fix ggml_compute_forward_dup_q (#269)Kawrakow
2025-03-19Prevent FlashMLA-1 from running on CUDA (#268)Kawrakow
generated by cgit v1.2.3 (git 2.25.1) at 2025-12-08 09:37:32 +0000