github
GitHub APIExtracts: repo, release, stars delta · metadata or bounded excerpt
Retention: D1 event history with canonical source link and deduplication metadata.
Access basis: Public API; token recommended for production rate limits.
Last ingested: 2026-08-26 · run status: success with data
- 2026-07-05ggerganov/llama.cpp b9876: b9876
<details open> ggml : fix tensor-parallel + -ncmoe crash on MoE models (#25028) Tensor parallelism (-sm tensor) combined with -ncmoe (CPU-offloaded MoE experts) aborts during warm-up on MoE models with GGML_ASSERT(ggml_is_contiguous(tensor)) in ggml-backend-meta.cpp. The faili
github:ggerganov/llama.cpp - 2026-07-05ggerganov/llama.cpp b9874: b9874
<details open> cuda : concat implementation for quantized types (#25303) * cuda : concat implementation for quantized types * chore : apply am17an clever suggestion to shorten the code --------- Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com> </details> **macOS/iOS
github:ggerganov/llama.cpp - 2026-07-04ggerganov/llama.cpp b9873: b9873
<details open> llama : add guard for K/V rotation input when buffer is unallocated (#25215) llm_graph_input_attn_kv::set_input and llm_graph_input_attn_kv_iswa::set_input call set_input_k_rot / set_input_v_rot whenever the rotation tensor pointer is non-null, but the tensor's b
github:ggerganov/llama.cpp