Search results

Misc. bug: llama.cpp-deepdeek-v4-flash aborts on launch when np > 1

Status: Open.

#9 In antirez/llama.cpp-deepseek-v4-flash;

· kstjohn1 opened on May 13, 2026

Feature Request: Disable x390 build so that docker containers actually publish

enhancementNew feature or requestNew feature or request

Status: Open.

#8 In antirez/llama.cpp-deepseek-v4-flash;

· devedse opened on May 12, 2026

ggml_cuda_op_concat crash on first inference — Blackwell sm_120a, IQ2_XXS GGUF

Status: Open.

#6 In antirez/llama.cpp-deepseek-v4-flash;

· louisamystudio opened on May 7, 2026

Potential 2x decoding speedup with MTP

Status: Open.

#5 In antirez/llama.cpp-deepseek-v4-flash;

· MirkoCovizzi opened on May 4, 2026

llama.cpp pull request

Status: Open.

#1 In antirez/llama.cpp-deepseek-v4-flash;

· dzenc opened on Apr 27, 2026