2 days ago
-
L-AI 1.20: f32 KV cache, per-model Precise kernels (auto-fix for models that only output end-of-text), polished memory box
3 days ago
-
L-AI 1.19: fix GPT-2-style models (TinyGPT) replying with ????? (NaN guard: F32 attention fallback); new start screen buttons
-
L-AI 1.19: polished Hugging Face model browser (search pill, tiles, For your phone, RAM-aware recommendations, chat models only)
-
L-AI 1.18: new models start with no system prompt (the model default), stale per-model prompts reset
18 days ago
-
Super-squash branch 'main' using huggingface_hub
