LettuceAI/app android-dev-227-1-8d71e82

Android Dev Build #227.1 (8d71e82)

on GitHub

Automated Android build artifacts for commit 8d71e829d575daf5788cc51d479a3cf98b24c340.
Workflow run: https://github.com/LettuceAI/app/actions/runs/26717152674

Changes since previous Android build android-dev-226-1-07a286a

Compare: android-dev-226-1-07a286a...8d71e82

• 8d71e82 docs(changelog): add llama.cpp mixed-offload context sizing and VRAM headroom fixes
• b57cee6 fix(chat): keep configured bubble width when header-above or message-info row is shown
• 11760d0 fix(llama): account for GPU-offloaded weights when sizing context for mixed offload
• b49e338 refactor(llama-cpp): derive compute-buffer VRAM reserve from model dims and batch
• 97d552f fix(llama-cpp): retry smaller context on OOM even when an explicit KV type is set
• 1f23950 fix(llama-cpp): reserve compute-buffer VRAM by context and stop offload from exceeding the safe layer estimate
• 1086fb9 docs(changelog): clarify per-message info shows the generating model
• 72029bd fix(chat-appearance): persist per-message model id so info shows the generating model, not the current one
• ba538db fix(chat-appearance): only show per-message info on assistant and scene messages