Software description:
https://github.com/OswaldHe/HeteroLLM
[ICML 2026] Efficient deployment framework for fine-grained disaggregated LLM inference with Memory Processing Pipeline (MPP).
Year:
2026
https://github.com/OswaldHe/HeteroLLM
[ICML 2026] Efficient deployment framework for fine-grained disaggregated LLM inference with Memory Processing Pipeline (MPP).