KimMeen/Time-LLM
An ICLR 2024 research paper that reprograms large language models for time series forecasting using prompt-based reprogramming.
Not currently ranked — collecting fresh signals.
star history
Time-LLM is an official implementation of a research paper that adapts pre-trained LLMs for time series forecasting without fine-tuning. The method reprograms input time series by remapping them into natural language descriptions compatible with LLM inputs, leveraging the knowledge embedded in foundation models. It evaluates across multiple benchmark time series datasets, demonstrating competitive forecasting performance by aligning time series with language model capabilities.
Frequently asked
- What is KimMeen/Time-LLM?
- An ICLR 2024 research paper that reprograms large language models for time series forecasting using prompt-based reprogramming.
- Is Time-LLM open source?
- Yes — KimMeen/Time-LLM is open source, released under the Apache-2.0 license.
- What language is Time-LLM written in?
- KimMeen/Time-LLM is primarily written in Python.
- How popular is Time-LLM?
- KimMeen/Time-LLM has 2.7k stars on GitHub.
- Where can I find Time-LLM?
- KimMeen/Time-LLM is on GitHub at https://github.com/KimMeen/Time-LLM.