ggerganov/llama.cpp

loader: map sparse mmap tensor ranges

json-schema-to-grammar: harden _visit_pattern against malformed regex patterns

ui: restore Ctrl+B sidebar toggle shortcut server/ui

server-context: chain the caller-provided load_progress_callback instead of discarding it server

cuda : concat implementation for quantized types ggml merge ready CUDA

ui: fake 200 for proxy DELETE req server/ui

kv-cache : fix SWA state save/load round-trip past n_swa

llama : stream MoE routed experts from disk

scripts : use HF_TOKEN when downloading UI assets

llama-batch: enable parallel sequences for partial rollback model

common : resolve non-positive --threads to the number of math cores documentation testing server

ggml-backend-meta: abort if we see a multi buffer ggml

CANN: Refactor `#ifdef` blocks to avoid unreachable code after `return` ggml Ascend NPU

ggml, server: add ggml_backend_dev_reset() for sleep mode Vulkan server ggml SYCL Apple Metal Ascend NPU OpenCL IBM zDNN Hexagon CUDA AMD ZenDNN OpenVINO WebGPU

[SYCL] support OP OPT_STEP_ADAMW, OPT_STEP_SGD documentation ggml merge ready SYCL

[SYC:] support op get_rows_back, only support fp32/fp16 documentation ggml SYCL

[SYCL] support op col2im_1d documentation ggml SYCL

Kmoren/add penalties cu backend testing

feat: add --threads-all option to llama-bench examples

vulkan: fix 32-bit integer overflow in CEIL_DIV Vulkan ggml

vulkan: for small AMD GPUs, reduce submission threshold based on CU count Vulkan ggml

common: Set optimal default thread count for ppc ( linux as well as AIX)

[SYCL] support OP cross_entropy_loss, cross_entropy_loss_back documentation ggml merge ready SYCL

common,server : fix custom preset dedup against cached models server

[UT] enhance UT to show all real unsupported backends testing

llama : clear error when MTP draft shares KV cache across backends

[SYCL] fix unsupported UT cases of CONT & CPY documentation ggml merge ready SYCL

Ensure unique node names and add org_src to track the org tensor for OpenVINO backend testing ggml

vulkan: when using transfer queue for async copies, sync on event_wait to avoid race Vulkan ggml

CUDA: Support CUDA Virtual Devices ggml CUDA