Lag-Llama
Visit ToolLag-Llama is an open-source foundation model for probabilistic time series forecasting. It offers zero-shot forecasting capabilities and supports finetuning for improved performance.
At a glance
Trending
Lag-Llama is an open-source foundation model for probabilistic time series forecasting. It offers zero-shot forecasting capabilities and supports finetuning for improved performance.
Trending
About
Lag-Llama is the first open-source foundation model specifically designed for probabilistic time series forecasting. It provides robust capabilities for zero-shot forecasting on datasets of any frequency and prediction length, making it highly versatile. Users can also finetune the model on their specific datasets to achieve maximum performance, with recommendations provided for optimizing hyperparameters like context length and learning rate. The project includes scripts for replicating pretraining and finetuning experiments from the associated paper, ensuring reproducibility. Lag-Llama aims to advance the field of time series analysis by offering a powerful, adaptable foundation model.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending