chinese-llm
AirLLM 70B inference with single 4GB GPU
Python toolkit for Chinese LLMs, with flexible batch capacity, structured real-time visulization and automated accumulation for streaming, and explicit feedback on vendor-native parameter validation.
