| 2026-03-15 |
sprayandwipe | kv-cache : fix reading llama_kv_cell_ext during state... |
commit | commitdiff | tree |
| 2026-03-14 |
Michael Wand | model : wire up Qwen3.5/Qwen3.5MoE tensors for NVFP4... |
commit | commitdiff | tree |
| 2026-03-14 |
Georgi Gerganov | metal : add FA specialization for HSK = 320, HSV =... |
commit | commitdiff | tree |
| 2026-03-14 |
Georgi Gerganov | ci : move self-hosted workflows to separate files ... |
commit | commitdiff | tree |
| 2026-03-14 |
Gerard Guillemas... | docker : force Python 3.13 in Vulkan container (#20530) |
commit | commitdiff | tree |
| 2026-03-14 |
Eve | ci : try to optimize some jobs (#20521) |
commit | commitdiff | tree |
| 2026-03-14 |
Max Krasnyansky | hexagon: Q4_0 and MXFP4 repack fixes (#20527) |
commit | commitdiff | tree |
| 2026-03-14 |
Georgi Gerganov | ci : reduce webgpu tests timeout to 900s (#20538) |
commit | commitdiff | tree |
| 2026-03-14 |
Xuan-Son Nguyen | mtmd: add llama-mtmd-debug binary (#20508) |
commit | commitdiff | tree |
| 2026-03-14 |
Neo Zhang | add op gated_delta_net (#20455) |
commit | commitdiff | tree |
| 2026-03-14 |
Chedrian07 | webui: restore code preview iframe origin isolation... |
commit | commitdiff | tree |
| 2026-03-14 |
Adrien Gallouët | scripts : remove get-wikitext-103.sh (#20543) |
commit | commitdiff | tree |
| 2026-03-14 |
Adrien Gallouët | scripts : update get-hellaswag.sh and get-winogrande... |
commit | commitdiff | tree |
| 2026-03-14 |
Adrien Gallouët | ggml : add native AVX512-FP16 support for F16 operation... |
commit | commitdiff | tree |
| 2026-03-14 |
Wallentri | Use fp32 in cuBLAS V100 to avoid overflows, env variabl... |
commit | commitdiff | tree |
| 2026-03-14 |
Zijun Yu | ggml : add OpenVINO backend (#15307) |
commit | commitdiff | tree |
| 2026-03-14 |
Adrien Gallouët | vendor : update cpp-httplib to 0.37.2 (#20484) |
commit | commitdiff | tree |
| 2026-03-14 |
Rail Chabdarov | Fix data race in CUDA's "cpy" kernel (influences GGML... |
commit | commitdiff | tree |
| 2026-03-14 |
lhez | opencl: fix l2_norm (#20480) |
commit | commitdiff | tree |
| 2026-03-13 |
Adrien Gallouët | tools : enable kvu in perplexity for hellaswag, winogra... |
commit | commitdiff | tree |
| 2026-03-13 |
Georgi Gerganov | graph : remove redundant GDN state transposes (#20443) |
commit | commitdiff | tree |
| 2026-03-13 |
Piotr Wilkin... | common/parser: gracefully handle undetected tool parser... |
commit | commitdiff | tree |
| 2026-03-13 |
ZeroV0LT | llama : fix pooling assertion crash in chunked GDN... |
commit | commitdiff | tree |
| 2026-03-13 |
SoftwareRenderer | server: reset counter related to kill-switch on client... |
commit | commitdiff | tree |
| 2026-03-13 |
rehan-10xengineer | ggml-cpu: add RVV vec dot kernels for quantization... |
commit | commitdiff | tree |
| 2026-03-13 |
Adrien Gallouët | ggml : fix typo gmml (#20512) |
commit | commitdiff | tree |
| 2026-03-13 |
Daniel Bevenius | mtmd : rename mtmd_get_audio_bitrate to mtmd_get_audio_... |
commit | commitdiff | tree |
| 2026-03-13 |
Piotr Wilkin... | general: CONTRIBUTING.md - guidelines for quantization... |
commit | commitdiff | tree |
| 2026-03-13 |
Georgi Gerganov | metal : fix l2 norm scale (#20493) |
commit | commitdiff | tree |
| 2026-03-13 |
Daniel Bevenius | convert : fix/suppress pyright errors (#20442) |
commit | commitdiff | tree |
| 2026-03-12 |
Georgi Gerganov | llama : disable graph reuse with pipeline parallelism... |
commit | commitdiff | tree |
| 2026-03-12 |
Alessandro... | vendor : update cpp-httplib to 0.37.1 (#20390) |
commit | commitdiff | tree |
| 2026-03-12 |
Piotr Wilkin... | tests : use `reasoning` instead of `reasoning_budget... |
commit | commitdiff | tree |
| 2026-03-12 |
Ruben Ortlam | test-backend-ops: allow loading tests from file and... |
commit | commitdiff | tree |
| 2026-03-12 |
Daniel Bevenius | common : update completion executables list [no ci... |
commit | commitdiff | tree |
| 2026-03-12 |
Asbjørn Olling | grammar: Fix grammar root symbol check (#19761) |
commit | commitdiff | tree |
| 2026-03-12 |
ProgenyAlpha | vulkan: add GATED_DELTA_NET op support (#20334) |
commit | commitdiff | tree |
| 2026-03-12 |
Sigbjørn Skjæret | convert : better mtp check and fix return [no ci] ... |
commit | commitdiff | tree |
| 2026-03-12 |
ProgenyAlpha | vulkan: fix SSM_CONV PP scaling with large ubatch sizes... |
commit | commitdiff | tree |
| 2026-03-12 |
Pascal | New conversations now auto-select the first loaded... |
commit | commitdiff | tree |
| 2026-03-12 |
Masashi Yoshimura | ggml-virtgpu: Fix some build commands (#20341) |
commit | commitdiff | tree |
| 2026-03-12 |
Georgi Gerganov | metal : avoid divisions in bin kernel (#20426) |
commit | commitdiff | tree |
| 2026-03-12 |
Masato Nakasaka | ci: Setup self-hosted CI for Intel Linux Vulkan backend... |
commit | commitdiff | tree |
| 2026-03-12 |
Jeff Bolz | vulkan: fix l2_norm epsilon handling (#20350) |
commit | commitdiff | tree |
| 2026-03-12 |
Jeff Bolz | vulkan: fix OOB check in flash_attn_mask_opt (#20296) |
commit | commitdiff | tree |
| 2026-03-12 |
Masato Nakasaka | vulkan: Fix ErrorOutOfHostMemory on Intel GPU when... |
commit | commitdiff | tree |
| 2026-03-12 |
lhez | opencl: use larger workgroup size for get_rows (#20316) |
commit | commitdiff | tree |
| 2026-03-12 |
shaofeiqi | opencl: add cumsum op (#18981) |
commit | commitdiff | tree |
| 2026-03-12 |
uvos | hip: compile debug builds with -O2 on hip to avoid... |
commit | commitdiff | tree |
| 2026-03-12 |
Mishusha | common/parser: add GigaChatV3/3.1 models support (... |
commit | commitdiff | tree |
| 2026-03-11 |
DAN™ | model : add support for Phi4ForCausalLMV (#20168) |
commit | commitdiff | tree |
| 2026-03-11 |
Richard Davison | graph : add optional scale parameter to build_lora_mm... |
commit | commitdiff | tree |
| 2026-03-11 |
ddh0 | common : fix --n-cpu-moe, --cpu-moe for models with... |
commit | commitdiff | tree |
| 2026-03-11 |
Masashi Yoshimura | ggml-webgpu: Add supports for `GGML_OP_REPEAT` (#20230) |
commit | commitdiff | tree |
| 2026-03-11 |
Georgi Gerganov | llama : enable chunked fused GDN path (#20340) |
commit | commitdiff | tree |
| 2026-03-11 |
Sigbjørn Skjæret | llama : whitespace cleanup (#20422) |
commit | commitdiff | tree |
| 2026-03-11 |
Richard Davison | ggml : add NVFP4 quantization type support (#19769) |
commit | commitdiff | tree |
| 2026-03-11 |
Georgi Gerganov | benches : add nemotron super (#20420) |
commit | commitdiff | tree |
| 2026-03-11 |
Daniel Bevenius | llama : add support for Nemotron 3 Super (#20411) |
commit | commitdiff | tree |
| 2026-03-11 |
Georgi Gerganov | metal : fix capture_compute counter logic (#20410) |
commit | commitdiff | tree |
| 2026-03-11 |
Aman Gupta | compare-llama-bench: check remotes as well (#20406) |
commit | commitdiff | tree |
| 2026-03-11 |
Georgi Gerganov | metal : fix q5_k mul_mv register spill (#20399) |
commit | commitdiff | tree |
| 2026-03-11 |
Georgi Gerganov | metal : add env var to trigger graph capture (#20398) |
commit | commitdiff | tree |
| 2026-03-11 |
Neo Zhang | [SYCL] Update SYCL.md for binary package for Windows... |
commit | commitdiff | tree |
| 2026-03-11 |
Ruben Ortlam | ci: disable coopmat on ubuntu-24-cmake-vulkan job ... |
commit | commitdiff | tree |
| 2026-03-11 |
Aldehir Rojas | common/parser: use nlohmann::ordered_json to preserve... |
commit | commitdiff | tree |
| 2026-03-11 |
Piotr Wilkin... | common/parser: handle reasoning budget (#20297) |
commit | commitdiff | tree |
| 2026-03-11 |
uvos | ggml-cuda: gdn use shared mem for HIP (#20366) |
commit | commitdiff | tree |
| 2026-03-11 |
uvos | cuda/hip: fix loop unrolling in ssm-conv (#20369) |
commit | commitdiff | tree |
| 2026-03-11 |
Pascal | Fix agentic mcp image single model (#20339) |
commit | commitdiff | tree |
| 2026-03-11 |
Alessandro... | vendor : update cpp-httplib to 0.37.0 (#20207) |
commit | commitdiff | tree |
| 2026-03-11 |
Alessandro... | vendor : update miniaudio to 0.11.25 (#20209) |
commit | commitdiff | tree |
| 2026-03-11 |
Neo Zhang | fix op rope, add rope_back (#20293) |
commit | commitdiff | tree |
| 2026-03-11 |
Neo Zhang | fix for failed UT case: ACC, L2_NORM, UPSCALE, fused_gl... |
commit | commitdiff | tree |
| 2026-03-10 |
Vinicios Lugli | model : qwen3vl reranker text support (#20332) |
commit | commitdiff | tree |
| 2026-03-10 |
ddh0 | llama-quant : correct `n_attention_wv` usage (#20357) |
commit | commitdiff | tree |
| 2026-03-10 |
Georgi Gerganov | ggml : bump RPC version (#20330) |
commit | commitdiff | tree |
| 2026-03-10 |
Reese Levine | ggml webgpu: faster normal quant and some k-quant matri... |
commit | commitdiff | tree |
| 2026-03-10 |
Piotr Wilkin... | Reduce level of content parser warning message to avoid... |
commit | commitdiff | tree |
| 2026-03-10 |
Ray Xu | examples : fix empty items in json_schema_to_grammar... |
commit | commitdiff | tree |
| 2026-03-10 |
a3894281 | docs: update CPU backend ops to mark POOL_1D as support... |
commit | commitdiff | tree |
| 2026-03-10 |
Georgi Gerganov | models : fix assert in mamba2 (cont) (#20335) |
commit | commitdiff | tree |
| 2026-03-10 |
Georgi Gerganov | server : make 2 checkpoints near the end of the prompt... |
commit | commitdiff | tree |
| 2026-03-10 |
Sigbjørn Skjæret | common : fix incorrect uses of stoul (#20313) |
commit | commitdiff | tree |
| 2026-03-10 |
Charles Xu | kleidiai : support for concurrent sme and neon kernel... |
commit | commitdiff | tree |
| 2026-03-10 |
Taimur Ahmad | ggml-cpu: add RVV repack GEMM and GEMV for quantization... |
commit | commitdiff | tree |
| 2026-03-10 |
Julian Pscheid | metal: handle command buffer failures gracefully in... |
commit | commitdiff | tree |
| 2026-03-10 |
ddh0 | llama-quant : fail early on missing imatrix, refactor... |
commit | commitdiff | tree |
| 2026-03-09 |
Aldehir Rojas | common: consolidate PEG string parsers (#20263) |
commit | commitdiff | tree |
| 2026-03-09 |
Xuan-Son Nguyen | model: fix step3.5 n_rot (#20318) |
commit | commitdiff | tree |
| 2026-03-09 |
Xuan-Son Nguyen | llama: dynamic head_dim and n_rot for SWA (#20301) |
commit | commitdiff | tree |
| 2026-03-09 |
Evan Huus | server: Parse port numbers from MCP server URLs in... |
commit | commitdiff | tree |
| 2026-03-09 |
Paul Flynn | metal : extend mul_mv_ext to BF16, Q2_K, Q3_K (#20250) |
commit | commitdiff | tree |
| 2026-03-09 |
Georgi Gerganov | server : fix checkpoints n_tokens calculation (#20287) |
commit | commitdiff | tree |
| 2026-03-09 |
Georgi Gerganov | metal : add upscale (#20284) |
commit | commitdiff | tree |
| 2026-03-09 |
Georgi Gerganov | server : warn swa-full is not supported for non-SWA... |
commit | commitdiff | tree |
| 2026-03-09 |
Georgi Gerganov | server : fix off-by-1 in server_tokens::size_up_to_pos... |
commit | commitdiff | tree |
| 2026-03-09 |
Piotr Wilkin... | common: map developer role to system (#20215) |
commit | commitdiff | tree |
| 2026-03-09 |
Georgi Gerganov | models : fix assert in mamba2 graph (#20270) |
commit | commitdiff | tree |
| 2026-03-09 |
Georgi Gerganov | server : add kill switch when server is stuck (#20277) |
commit | commitdiff | tree |
| next |
overview / pkg / ggml / sources / llama.cpp / shortlog