]> git.djapps.eu Git - pkg/ggml/sources/whisper.cpp/shortlog
pkg/ggml/sources/whisper.cpp
2026-05-26 texasichcli : merge tokens split across UTF-8 boundaries in...
2026-05-25 Georgi Gerganovrelease : v1.8.5
2026-05-25 Georgi Gerganovbenches : update
2026-05-25 Georgi Gerganovsync : ggml
2026-05-25 Georgi Gerganovggml : bump version to 0.13.0 (ggml/1510)
2026-05-25 Johannes GäßlerTP: fix ggml context size calculation (llama/22616)
2026-05-25 Gilad Sggml: `gguf_init_from_callback` and `gguf_init_from_buf...
2026-05-25 Kaihui-AMDreadme : add AMD ROCm/HIP GPU build instructions (...
2026-05-25 Georgi Gerganovtalk-llama : sync llama.cpp
2026-05-25 Georgi Gerganovsync : ggml
2026-05-25 Georgi Gerganovggml : bump version to 0.12.1 (ggml/1508)
2026-05-25 Jeff Bolzggml : Parallelize quant LUT init (llama/23595)
2026-05-25 Johannes GäßlerTP: fix entirely zero-sized slices per device (llama...
2026-05-25 shaofeiqiopencl: batch profiling to improve speed and prevent...
2026-05-25 Yiwei Shaohexagon: apply repl optimization in flash attn softmax...
2026-05-25 dskweggml : Check the right iface method before using the...
2026-05-25 Jeff Bolzvulkan: fix windows find_package of SPIRV-Headers ...
2026-05-25 Shawn Guopencl: generalize Adreno MoE kernels on M (llama/23449)
2026-05-25 Alexey KopytkoSYCL: improve MoE prefill throughput (llama/23142)
2026-05-25 Alexey Kopytkosycl : Level Zero detection in ggml_sycl_init (llama...
2026-05-25 karavayevSYCL : gated_delta_net K>1 (llama/23174)
2026-05-25 KatostrofikSYCL: add BF16 to DMMV kernel path (~4x tg speedup...
2026-05-25 Sachin Sharmaggml-zendnn : add Q8_0 quantization support (llama...
2026-05-25 Johannes GäßlerCUDA: fix PDL CC check for JIT compilation (llama/23471)
2026-05-25 Pascalvulkan: fuse snake activation (mul, sin, sqr, mul,...
2026-05-25 Chen Yuanfix(flash-attn): replace f32 with kv_type and q_type...
2026-05-25 Georgi Gerganovmetal : optimize concat kernel and fix set kernel threa...
2026-05-25 Matt Coralloggml : Check the right iface method before using the...
2026-05-25 Todor Boinovskihexagon: ssm-conv fix for large prompts (llama/23307)
2026-05-25 lhezopencl: refactor backend initilization (llama/23318)
2026-05-25 Danielevulkan: optimize operations in the IM2COL shader (llama...
2026-05-25 Max Krasnyanskyhexagon: HMX quantized matmul rework (llama/23368)
2026-05-25 Andreas KieslingerProgrammatic Dependent Launch (PDL) for more performanc...
2026-05-25 Georgi Gerganovmetal : optimize pad + cpy (llama/23354)
2026-05-25 ravel7524ggml-cuda: tune RDNA3 Q6_K MMVQ nwarps (llama/23349)
2026-05-25 shaofeiqiopencl: add MoE support for q4_k, q5_k, q6_k on Adreno...
2026-05-25 Aparna M Phexagon: add MROPE and IMROPE support in HTP rope op...
2026-05-25 Aparna M Phexagon: enable support for NORM op (llama/23319)
2026-05-25 Reese Levineggml-webgpu : extend GDN for K>1 (llama/23299)
2026-05-25 Intel AI Get... sycl: add GGML_SYCL_USE_ASYNC_MEM_OP env toggle (llama...
2026-05-25 Radoslav Gerganovrpc : keep last_graph_uid in the device context (llama...
2026-05-25 Pranav Dhinakarhexagon: add support for TRI op (llama/22822)
2026-05-25 Pranav Dhinakarggml-hexagon: add PAD op HVX kernel (llama/23078)
2026-05-25 Intel AI Get... sycl: scalar SWAR byte-subtract in Q6_K MMVQ dot produc...
2026-05-25 Intel AI Get... sycl: route small f32 matmuls to oneMKL, bypass oneDNN...
2026-05-25 Gabe Goodhartfeat: Support d_conv=15 for ssm-conv.cu (llama/23017)
2026-05-25 Oliver SimonsCUDA: Continue directly including cuda/iterator (llama...
2026-05-25 Jan Ekströmggml-vulkan/CMakeLists: add a check for SPIRV-Headers...
2026-05-25 Pascalvulkan: add cpy bf16 -> f32 pipelines (llama/22677)
2026-05-25 Jeff Bolzvulkan: Support unaligned tensors for ROPE (llama/22637)
2026-05-25 Jeff Bolzvulkan: fuse SSM_CONV + BIAS + SILU (llama/22653)
2026-05-25 Winston Mavulkan: removed duplicate #include <memory> in headers...
2026-05-25 Ori Pekelmanggml.h: correct ggml_silu_back arg docstring (a=dy...
2026-05-25 Dev-X25874ggml-alloc: fix out-of-bounds read in ggml_dyn_tallocr_...
2026-05-25 Georgi Gerganovggml : bump version to 0.12.0 (ggml/1494)
2026-05-25 Aman Guptallama + spec: MTP Support (llama/22673)
2026-05-25 Pranav Dhinakarggml-hexagon: cpy: add contiguous fast-path in reshape...
2026-05-25 Johannes GäßlerHIP: RDNA3 mma FA, faster AMD transpose, tune AMD ...
2026-05-25 Zheyuan Chenggml-webgpu: makes the flash attn vec path subgroup...
2026-05-25 Georgi Gerganovlogs : reduce (llama/23021)
2026-05-25 alex-spacemitggml-cpu: Add IME2 Instruction Support for the SpacemiT...
2026-05-25 Ruben Ortlamvulkan: fix matmul integer pipeline selection (llama...
2026-05-25 KatostrofikSYCL: fix multi-GPU system RAM exhaustion by using...
2026-05-25 Daniel Beveniuscmake : add CMakePresets.json [no ci] (#3808)
2026-05-25 OrbisAI Securityfix: in bindings/ruby/test/jfk_reader/jfk_reader in...
2026-05-22 Pascalcommon : fix server /inference fails to decode in-memor...
2026-05-21 Daniel Beveniusci : use github ubuntu-22.04-arm runner instead of...
2026-05-19 Daniel Beveniuswhisper : set bench data for each iteration (#3812)
2026-05-18 petterreinholdtsenexamples : fix memory leak in read_audio_data (#3810)
2026-05-18 Andreas Lubbeserver : Return speaker information in JSON (#3782)
2026-05-15 Andreas Lubbeserver: add support for carry_initial_prompt (#3781)
2026-05-14 Georgi Gerganovtalk-llama : sync llama.cpp
2026-05-14 Georgi Gerganovsync : ggml
2026-05-14 Zheyuan Chenggml-webgpu: only use subgroup-matrix path when head...
2026-05-14 scutler-nvFix for issue #22974. Cast intermediate results to...
2026-05-14 shaofeiqiopencl: add q5_0 and q5_1 MoE for Adreno (llama/22985)
2026-05-14 lhezopencl: fix crash when warming up MoE on Adreno (llama...
2026-05-14 Masashi Yoshimuraflush the gpu profile timestamp before the queryset...
2026-05-14 Max Krasnyanskyhexagon: add unary tanh op (llama/22999)
2026-05-14 Sachin Sharmaggml-zendnn : adaptive fallback to CPU backend for...
2026-05-14 Trivikram Reddyhexagon: eliminate scalar VTCM loads via HVX splat...
2026-05-14 yzyyzyhhhopencl: add opt-in Adreno xmem F16xF32 GEMM for prefill...
2026-05-14 Masashi Yoshimuraggml-webgpu: Enables running gpt-oss-20b (llama/22906)
2026-05-14 Chen Yuanggml-webgpu: address precision issues for multimodal...
2026-05-14 Masato Nakasakavulkan: Fix Windows performance regression on Intel...
2026-05-14 Jeff Bolzvulkan: Check shared memory size for mmq shaders (llama...
2026-05-14 guyfischmanmetal : promote mul_mv/mul_mm batch divisors to functio...
2026-05-14 Shawn Guopencl: add q4_1 MoE for Adreno (llama/22856)
2026-05-14 CrispStrobeCUDA: handle OW > 65535 in im2col (2D and 3D) (llama...
2026-05-14 PascalGgml/cuda snake fusion hardening (llama/22912)
2026-05-14 Jeff Bolzvulkan: Support asymmetric FA in scalar/mmq/coopmat1...
2026-05-14 Oliver SimonsCUDA: directly include cuda/iterator (llama/22936)
2026-05-14 Neo ZhangAdd OP im2col_3d (llama/22903)
2026-05-14 Oliver Walshggml-virtgpu : include missing mutex header (llama...
2026-05-14 CrispStrobemetal : tighten input-position loop in kernel_conv_tran...
2026-05-14 Steve Lhommeggml: install ggml.pc in <libdir>/pkgconfig (ggml/1480)
2026-05-14 Daniel BeveniusRevert "ggml-cuda : add explicit casts to -INFINITY...
2026-05-14 Daniel BeveniusRevert "ggml-cuda : add ar_add() to avoid ambiguous...
2026-05-14 Daniel Beveniussquash! ci : update ONEAPI version to 2025.3.3-0-devel...
2026-05-14 Daniel Beveniusci : update ONEAPI version to 2025.3.3-0-devel-ubuntu24.04
next