timesfm

{"problem_it_solves": "Provides a pretrained model for time-series forecasting tasks, eliminating the need to train models from scratch for each forecasting use case.", "target_customer": "Data scientists, ML engineers, and developers working on time-series forecasting tasks.", "use_cases": ["Time-series forecasting", "Point forecasting", "Quantile forecasting", "Anomaly detection in time series", "Demand forecasting", "Financial forecasting"], "pricing_details": "Open source under Apache-2.0 license", "differentiator": "First pretrained time-series foundation model from Google Research, tested at ICML 2024, with support for up to 16k context length and built-in quantile forecasting head.", "traction": {"notable_metrics": "15.8k GitHub stars, 1.4k forks, 109 watchers, cited in paper at ICML 2024", "press_mentions": ["Paper: A decoder-only foundation model for time-series forecasting, ICML 2024", "TimesFM Hugging Face Collection", "Google Research blog post"]}}