Releases: ggml-org/llama.cpp
Releases · ggml-org/llama.cpp
b6593
common : use cpp-httplib as a cURL alternative for downloads (#16185) * vendor : update httplib Signed-off-by: Adrien Gallouët <[email protected]> * common : use cpp-httplib as a cURL alternative for downloads The existing cURL implementation is intentionally left untouched to prevent any regressions and to allow for safe, side-by-side testing by toggling the `LLAMA_CURL` CMake option. Signed-off-by: Adrien Gallouët <[email protected]> * ggml : Bump to Windows 10 Signed-off-by: Adrien Gallouët <[email protected]> --------- Signed-off-by: Adrien Gallouët <[email protected]>
b6591
ggml-cpu: implement MXFP4 SIMD for s390x (#16193) * ggml-cpu: impl mxfp4 s390x Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: missing s = sumf Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: fix incorrect kval_mxfp4 type Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: rework mxfp4 Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: missing delta calc Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: fix typo Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: fix typo for vec_splats Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: expand to 2 blocks per loop Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: add unroll to boost perf Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: back to 1 block per loop to test perf Signed-off-by: Aaron Teo <[email protected]> * Revert "ggml-cpu: back to 1 block per loop to test perf" This reverts commit 1fe55724e2dc295701101bf838bdd4a512237492. Signed-off-by: Aaron Teo <[email protected]> * ggml-cpu: rm unroll from single block Signed-off-by: Aaron Teo <[email protected]> --------- Signed-off-by: Aaron Teo <[email protected]>
b6587
musa: fix build warnings (#15611) Signed-off-by: Xiaodong Ye <[email protected]>
b6586
model : add GroveMoE support (#15510) * add GroveMoE support * remove constexpr that fails on certain compilers * revert crude scalar div implementation, use cast * build_attn_inp_kv_unified -> build_attn_inp_kv * fix build_attn * re-apply ffn_exps regex changes
b6585
vendors: update miniaudio version (#16212) * vendor: update miniaudio.h Signed-off-by: Aaron Teo <[email protected]> * vendor: update miniaudio.h Signed-off-by: Aaron Teo <[email protected]> --------- Signed-off-by: Aaron Teo <[email protected]>
b6583
CUDA: add a fused top-K MoE kernel (#16130) * CUDA: add a fused top-K MoE kernel This kernel does the following: 1. softmax over the logits per token [n_experts, n_tokens] 2. argmax reduce over the top-k (n_experts_used) logits 3. write weights + ids to global memory It is intended as fusion of softmax->top-k->get_rows pipeline for MoE models * Refactor into ggml_cuda_should_use_topk_moe * Review: Use better coalescing pattern, use WARP_SIZE, store logits into registers before * Review: format + micro-optimizations * Fix bug: fix tie breakers * Add optional norm + clean-up code * Use smem for final write * Add bounds check * Use better memory pattern for writeback
b6582
model-conversion : add embedding prompt file support (#15871) This commit adds support for passing a prompt file to the model conversion targets/scripts. It also updates the logits.cpp to print out embedding information in the same format as when running the original embedding model. The motivation for this is that it allows us to pass files of different sizes when running the converted models and validating the logits. This can be particularly important when testing the sliding window functionality of models where the sequence length needs to exceed a certain number of tokens to trigger the sliding window logic.
b6580
ggml : fix loongarch lsx compilation error (#15864)
b6578
llama : add support for qwen3 reranker (#15824)
b6576
metal : relax reorder conditions (#16216)