Commit History

fix: Vicuna template instead of ChatML
8addbf1

Z User commited on

fix: reduce num_ctx 8192->2048 to prevent OOM on 16GB
e8f464e

Z User commited on

feat: switch to Ollama (zero compilation, no OOM)
968fe67

Z User commited on

init: WHOAMISec model API
8b369b6

kepom commited on