Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces ModTrans, a translator that converts real-world machine learning models into a format compatible with the ASTRA-sim distributed training simulator. This enables ML researchers to simulate distributed training scenarios without requiring physical hardware or extensive knowledge of simulator-specific input formats. Experiments demonstrate that ModTrans introduces negligible overhead, making it a practical tool for bridging the gap between model development and system-level simulation.
Unlock the power of distributed training simulation for your real-world models with a new translator that eliminates the need for specialized simulator expertise.
Large-scale distributed training has been a research hot spot in machine learning systems for industry and academia in recent years. However, conducting experiments without physical machines and corresponding resources is difficult. One solution is to leverage distributed training simulators, but current ones like ASTRA-sim do not support importing real-world developed models, which poses challenges for ML researchers seeking to use them. Based on this challenge, we developed ModTrans, a translator supporting format translation from any real-world model to the ASTRA-sim simulator's input, removing the barrier between machine learning experts and machine learning system researchers. The experiment results show that ModTrans's cost is negligible.