| 2026-05-26 |
Georgi Gerganov | ci : reduce (disable SYCL and CANN builds/releases... |
commit | commitdiff | tree |
| 2026-05-26 |
ghleg | convert : support Gemma4ForCausalLM architecture (... |
commit | commitdiff | tree |
| 2026-05-26 |
Michael Wand | models : Attach Mistral3 NVFP4 weight scales (#23629) |
commit | commitdiff | tree |
| 2026-05-26 |
Alexey Kopytko | SYCL: implement ggml_sycl_pool_vmm (#22862) |
commit | commitdiff | tree |
| 2026-05-26 |
Jeff Bolz | tests: test-backend-ops -j <N> to run tests in parallel... |
commit | commitdiff | tree |
| 2026-05-26 |
Niklas Sheth | model : add support for talkie-1930-13b (#22596) |
commit | commitdiff | tree |
| 2026-05-26 |
Masashi Yoshimura | ggml-webgpu: Add MMVQ path for Q4/Q8/Q2_K/Q4_K and... |
commit | commitdiff | tree |
| 2026-05-26 |
Nikhil Jain | [WebGPU] Check batch_compute_passes before sending... |
commit | commitdiff | tree |
| 2026-05-26 |
Johannes Gäßler | CUDA: missing PDL sync for FWHT, better fallback (... |
commit | commitdiff | tree |
| 2026-05-25 |
forforever73 | metal : add apple device id (#23566) |
commit | commitdiff | tree |
| 2026-05-25 |
Max Krasnyansky | snapdragon: bump toolchain docker to v0.7 to fix ui... |
commit | commitdiff | tree |
| 2026-05-25 |
Georgi Gerganov | ci : reduce PR jobs by matching backend paths (#23675) |
commit | commitdiff | tree |
| 2026-05-25 |
Pascal | model: tag ffn_latent as MUL_MAT to fix buft probe... |
commit | commitdiff | tree |
| 2026-05-25 |
Aman Gupta | CUDA: add fast walsh-hadamard transform (#23615) |
commit | commitdiff | tree |
| 2026-05-25 |
Pascal | ui: fix stop/continue during an agentic loop (#23356) |
commit | commitdiff | tree |
| 2026-05-25 |
Michael Wand | convert : add compressed-tensors NVFP4 support (#21095) |
commit | commitdiff | tree |
| 2026-05-25 |
Georgi Gerganov | sync : ggml |
commit | commitdiff | tree |
| 2026-05-25 |
Georgi Gerganov | ggml : bump version to 0.13.0 (ggml/1510) |
commit | commitdiff | tree |
| 2026-05-25 |
Georgi Gerganov | sync : ggml |
commit | commitdiff | tree |
| 2026-05-25 |
Georgi Gerganov | ggml : bump version to 0.12.1 (ggml/1508) |
commit | commitdiff | tree |
| 2026-05-25 |
Ori Pekelman | ggml.h: correct ggml_silu_back arg docstring (a=dy... |
commit | commitdiff | tree |
| 2026-05-25 |
Dev-X25874 | ggml-alloc: fix out-of-bounds read in ggml_dyn_tallocr_... |
commit | commitdiff | tree |
| 2026-05-25 |
Johannes Gäßler | TP: fix ggml context size calculation (#22616) |
commit | commitdiff | tree |
| 2026-05-25 |
Gilad S. | ggml: `gguf_init_from_callback` and `gguf_init_from_buf... |
commit | commitdiff | tree |
| 2026-05-25 |
Aman Gupta | server: MTP layer kv-cache should respect draft type... |
commit | commitdiff | tree |
| 2026-05-25 |
alex-spacemit | ci : update spacemit toolchain url and enhance curl... |
commit | commitdiff | tree |
| 2026-05-25 |
Sigbjørn Skjæret | ci : fix pre-tokenizer-hashes check (#23651) |
commit | commitdiff | tree |
| 2026-05-25 |
Tim Neumann | llama : document that only one on-device state can... |
commit | commitdiff | tree |
| 2026-05-25 |
Aldehir Rojas | ci : install host compiler on android-ndk build (#23630) |
commit | commitdiff | tree |
| 2026-05-25 |
Jeff Bolz | ggml : Parallelize quant LUT init (#23595) |
commit | commitdiff | tree |
| 2026-05-25 |
Saba Fallah | ui: media attachments before text (#23467) |
commit | commitdiff | tree |
| 2026-05-25 |
Alessandro... | vendor : update cpp-httplib to 0.45.1 (#23639) |
commit | commitdiff | tree |
| 2026-05-25 |
jacekpoplawski | server: fix checkpoints creation (#22929) |
commit | commitdiff | tree |
| 2026-05-25 |
fairydreaming | perplexity : fix even more integer overflows (#23623) |
commit | commitdiff | tree |
| 2026-05-25 |
Georgi Gerganov | ci : move most slim jobs to self-hosted runners (#23619) |
commit | commitdiff | tree |
| 2026-05-24 |
Georgi Gerganov | ci : update build-self-hosted.yml (#23616) |
commit | commitdiff | tree |
| 2026-05-24 |
Sigbjørn Skjæret | convert : minor fixes for numpy 2.x (#23571) |
commit | commitdiff | tree |
| 2026-05-24 |
Aldehir Rojas | cmake : fix ui build (#23592) |
commit | commitdiff | tree |
| 2026-05-24 |
Aman Gupta | server: add margin for draft model for `fit` (#23485) |
commit | commitdiff | tree |
| 2026-05-24 |
Johannes Gäßler | TP: fix entirely zero-sized slices per device (#23525) |
commit | commitdiff | tree |
| 2026-05-24 |
shaofeiqi | opencl: batch profiling to improve speed and prevent... |
commit | commitdiff | tree |
| 2026-05-24 |
Yiwei Shao | hexagon: apply repl optimization in flash attn softmax... |
commit | commitdiff | tree |
| 2026-05-24 |
Aparna M P | snapdragon: update windows toolchain to use hsdk v6... |
commit | commitdiff | tree |
| 2026-05-23 |
Aldehir Rojas | cmake/ui : refactor the build (#23352) |
commit | commitdiff | tree |
| 2026-05-23 |
Aditya Singh | requirements : bump torch to 2.11.0 (#23503) |
commit | commitdiff | tree |
| 2026-05-23 |
Michael Wand | model : add NVFP4 MTP scale tensors (#23563) |
commit | commitdiff | tree |
| 2026-05-23 |
dskwe | ggml : Check the right iface method before using the... |
commit | commitdiff | tree |
| 2026-05-23 |
Jeff Bolz | vulkan: fix windows find_package of SPIRV-Headers ... |
commit | commitdiff | tree |
| 2026-05-23 |
Shawn Gu | opencl: generalize Adreno MoE kernels on M (#23449) |
commit | commitdiff | tree |
| 2026-05-22 |
Aldehir Rojas | server: only parse empty msg if continuing an assistant... |
commit | commitdiff | tree |
| 2026-05-22 |
fairydreaming | perplexity : fix integer overflow (#23496) |
commit | commitdiff | tree |
| 2026-05-22 |
Alexey Kopytko | SYCL: improve MoE prefill throughput (#23142) |
commit | commitdiff | tree |
| 2026-05-22 |
Alexey Kopytko | sycl : Level Zero detection in ggml_sycl_init (#23097) |
commit | commitdiff | tree |
| 2026-05-22 |
karavayev | SYCL : gated_delta_net K>1 (#23174) |
commit | commitdiff | tree |
| 2026-05-22 |
Katostrofik | SYCL: add BF16 to DMMV kernel path (~4x tg speedup... |
commit | commitdiff | tree |
| 2026-05-22 |
Jesus Talavera | docs: Update documentation with Granite 4.0/4.1 (#23404) |
commit | commitdiff | tree |
| 2026-05-22 |
Sachin Sharma | ggml-zendnn : add Q8_0 quantization support (#23414) |
commit | commitdiff | tree |
| 2026-05-22 |
fairydreaming | cmake : build router app only during standalone builds... |
commit | commitdiff | tree |
| 2026-05-22 |
Kashif Rasul | vocab : fix HybridDNA tokenizer (#23466) |
commit | commitdiff | tree |
| 2026-05-22 |
Georgi Gerganov | cmake : add install() for impl libraries + fix apple... |
commit | commitdiff | tree |
| 2026-05-21 |
Johannes Gäßler | CUDA: fix PDL CC check for JIT compilation (#23471) |
commit | commitdiff | tree |
| 2026-05-21 |
Georgi Gerganov | cmake : remove STATIC from impl libraries, enable LLAMA... |
commit | commitdiff | tree |
| 2026-05-21 |
Reese Levine | Update WebGPU support and add link to blog/demo (#23483) |
commit | commitdiff | tree |
| 2026-05-21 |
Pascal | vulkan: fuse snake activation (mul, sin, sqr, mul,... |
commit | commitdiff | tree |
| 2026-05-21 |
Chen Yuan | fix(flash-attn): replace f32 with kv_type and q_type... |
commit | commitdiff | tree |
| 2026-05-21 |
Georgi Gerganov | tests : move save-load-state from examples to tests... |
commit | commitdiff | tree |
| 2026-05-21 |
ScrewTSW | server: expose prompt token counts in /slots endpoint... |
commit | commitdiff | tree |
| 2026-05-21 |
Georgi Gerganov | metal : optimize concat kernel and fix set kernel threa... |
commit | commitdiff | tree |
| 2026-05-21 |
Aman Gupta | server : free draft/MTP resources on sleep to fix VRAM... |
commit | commitdiff | tree |
| 2026-05-21 |
Pascal | server: re-inject subcommand when router spawns childre... |
commit | commitdiff | tree |
| 2026-05-21 |
Adrien Gallouët | app : add batched-bench, fit-params, quantize & perplex... |
commit | commitdiff | tree |
| 2026-05-21 |
Aman Gupta | mtp: use inp_out_ids for skipping logit computation... |
commit | commitdiff | tree |
| 2026-05-21 |
Kashif Rasul | vocab : add Carbon-3B (HybridDNATokenizer) support... |
commit | commitdiff | tree |
| 2026-05-21 |
Ruixiang Wang | doc: fix spec mtp typo (#23435) |
commit | commitdiff | tree |
| 2026-05-21 |
Aleksander... | ui: Improve Git Hooks for UI development (#23403) |
commit | commitdiff | tree |
| 2026-05-21 |
Matt Corallo | ggml : Check the right iface method before using the... |
commit | commitdiff | tree |
| 2026-05-21 |
Daniel Elliott | llama-graph: fix null-buffer crash in llm_graph_input_a... |
commit | commitdiff | tree |
| 2026-05-21 |
Todor Boinovski | hexagon: ssm-conv fix for large prompts (#23307) |
commit | commitdiff | tree |
| 2026-05-21 |
Adrien Gallouët | app : show version (#23426) |
commit | commitdiff | tree |
| 2026-05-20 |
wendadawen | mtmd, model : merge HunyuanOCR into HunyuanVL and fix... |
commit | commitdiff | tree |
| 2026-05-20 |
stduhpf | ui: Add max image size option (#22849) |
commit | commitdiff | tree |
| 2026-05-20 |
Gaurav Garg | Move to backend sampling for MTP draft path (#23287) |
commit | commitdiff | tree |
| 2026-05-20 |
lhez | opencl: refactor backend initilization (#23318) |
commit | commitdiff | tree |
| 2026-05-20 |
Georgi Gerganov | common/speculative : fix nullptr crash in get_devices_s... |
commit | commitdiff | tree |
| 2026-05-20 |
Saba Fallah | mtmd : DeepSeek-OCR image processing fixes, img_tool... |
commit | commitdiff | tree |
| 2026-05-20 |
Daniele | vulkan: optimize operations in the IM2COL shader (... |
commit | commitdiff | tree |
| 2026-05-20 |
Aleksander... | feat: Add WAV MIME type variants and improve audio... |
commit | commitdiff | tree |
| 2026-05-20 |
Max Krasnyansky | hexagon: HMX quantized matmul rework (#23368) |
commit | commitdiff | tree |
| 2026-05-20 |
Andreas Kieslinger | Programmatic Dependent Launch (PDL) for more performanc... |
commit | commitdiff | tree |
| 2026-05-20 |
Adrien Gallouët | app : introduce the llama unified executable (#23296) |
commit | commitdiff | tree |
| 2026-05-20 |
Aleksander... | refactor: Move text attachments up before the message... |
commit | commitdiff | tree |
| 2026-05-20 |
Xuan-Son Nguyen | mtmd: fit_params now take into account mmproj (#21489) |
commit | commitdiff | tree |
| 2026-05-20 |
Sigbjørn Skjæret | docker : copy conversion files (#23370) |
commit | commitdiff | tree |
| 2026-05-20 |
Aleksander... | ui: Refactor `isMobile` as reactive value in `viewport... |
commit | commitdiff | tree |
| 2026-05-20 |
Aleksander... | fix: Div wrapper no pointer events on hidden (#23390) |
commit | commitdiff | tree |
| 2026-05-20 |
Georgi Gerganov | metal : optimize pad + cpy (#23354) |
commit | commitdiff | tree |
| 2026-05-20 |
Max Krasnyansky | snapdragon: update toolchain to v0.6 (#23369) |
commit | commitdiff | tree |
| 2026-05-20 |
ravel7524 | ggml-cuda: tune RDNA3 Q6_K MMVQ nwarps (#23349) |
commit | commitdiff | tree |
| 2026-05-19 |
shaofeiqi | opencl: add MoE support for q4_k, q5_k, q6_k on Adreno... |
commit | commitdiff | tree |
| 2026-05-19 |
Aparna M P | hexagon: add MROPE and IMROPE support in HTP rope op... |
commit | commitdiff | tree |
| next |
overview / pkg / ggml / sources / llama.cpp / shortlog