TimesFM
A large-scale time-series forecasting foundation model with parameter counts in the billions range. Its scale results in relatively high computational resource requirements for both training and inference.
Timeline (last 90 days)
TimesFM originally comes from the ICML 2024 paper "A decoder-only foundation model for time-series forecasting."
The code and weights of TimesFM 2.5 and earlier versions are open-sourced under the Apache-2.0 license.
TimesFM is a decoder-only transformer whose input tokens are patches of 32 consecutive numbers, with a prediction task of the next 128 numbers.
Lumina uses Google's TimesFM for time series forecasting, outputting a 5-day probabilistic fan chart with an 80% confidence interval and a crash probability estimate.
Time-Series Foundation Models are general-purpose forecasting models pretrained on large-scale, multi-domain time-series data with cross-task generalization capabilities, with representative works including TimesFM, Chronos, and Moirai.
Google TimesFM is a foundation model dedicated to time series forecasting, capable of zero-shot prediction without requiring specialized training on specific business data.
TimesFM 3.0 is a foundation model dedicated to time series forecasting, with zero-shot forecasting capability.
Representative time-series foundation models include Google's TimesFM, Amazon's Chronos, and Salesforce's Moirai.
TimesFM draws on the pre-training + zero-shot inference paradigm of large language models, enabling direct forecasting of unseen time series after pre-training without targeted training.
TimesFM provides implementations based on both the JAX and PyTorch deep learning frameworks.
11 more timeline events
All Facts (20)
Google officially disclosed that TimesFM was pre-trained on 200B time points of data, covering multiple domains including finance, retail, energy, and healthcare.
70%UnverifiedTimesFM adopts patch-based tokenization, packing several consecutive time steps into a patch and treating it as a single token, similar to the image patch strategy used by ViT.
70%UnverifiedTimesFM originally comes from the ICML 2024 paper "A decoder-only foundation model for time-series forecasting."
50%UnverifiedTimesFM is a decoder-only transformer whose input tokens are patches of 32 consecutive numbers, with a prediction task of the next 128 numbers.
50%UnverifiedThe code and weights of TimesFM 2.5 and earlier versions are open-sourced under the Apache-2.0 license.
50%UnverifiedLumina uses Google's TimesFM for time series forecasting, outputting a 5-day probabilistic fan chart with an 80% confidence interval and a crash probability estimate.
50%UnverifiedTime-Series Foundation Models are general-purpose forecasting models pretrained on large-scale, multi-domain time-series data with cross-task generalization capabilities, with representative works including TimesFM, Chronos, and Moirai.
50%UnverifiedGoogle TimesFM is a foundation model dedicated to time series forecasting, capable of zero-shot prediction without requiring specialized training on specific business data.
50%UnverifiedTimesFM 3.0 is a foundation model dedicated to time series forecasting, with zero-shot forecasting capability.
50%UnverifiedRepresentative time-series foundation models include Google's TimesFM, Amazon's Chronos, and Salesforce's Moirai.
50%UnverifiedTimesFM draws on the pre-training + zero-shot inference paradigm of large language models, enabling direct forecasting of unseen time series after pre-training without targeted training.
50%UnverifiedTimesFM provides implementations based on both the JAX and PyTorch deep learning frameworks.
50%UnverifiedTimesFM adopts a decoder-only Transformer architecture, consistent with the design philosophy of the GPT series of models.
50%UnverifiedTimesFM's patch mechanism reduces computational complexity from O(n²) to O(p²), where p is the number of patches.
50%UnverifiedAccording to a Google Research paper, the pre-training data scale of TimesFM reaches approximately 100 billion time points.
50%UnverifiedTimesFM supports not only point forecasting but also probabilistic forecasting (outputting quantiles of the predicted distribution).
50%UnverifiedTimesFM is implemented in Python and released as open source on GitHub, providing both source code and pre-trained weights.
50%UnverifiedThe patch-based input approach is inspired by the Vision Transformer's idea of dividing images into patches.
50%UnverifiedTimesFM completed pre-training on thousands of time series datasets spanning different domains, frequencies, and patterns.
50%UnverifiedTimesFM is implemented entirely in Python and has been released as open source on GitHub.
50%