This is the official PyTorch implementation of Enhancing Channel-Independent Time Series Forecasting via Cross-Variate Patch Embedding.
Implementation extends original Time-LLM code at https://github.com/KimMeen/Time-LLM by adding Cross-Variate Patch Embeddings, consisting of position-enhanced patch embeddings in layers/Embed.py and a router attention mechanism in layers/utils.py. We exchange the vanilla patch embedding with CVPE in models/TimeLLM_CVPE.py.