rts-commander / model_manager.py

Commit History

perf: Apply all thread optimizations + complete documentation
8e1770a

Luigi commited on

feat: Implement cancel-on-new-request strategy (no timeouts)
fa2c1d8

Luigi commited on

perf: Non-blocking LLM architecture to prevent game lag
7e8483f

Luigi commited on

perf: Optimize LLM for HuggingFace 2 vCPU environment
03eb9aa

Luigi commited on

fix: Change chat_format from 'qwen2' to 'qwen'
6d41c63

Luigi commited on

fix: Critical fixes for model path and game loop
6efa796

Luigi commited on

fix: Shared LLM model manager and UI positioning
1e98ab1

Luigi commited on