How Much Computation Does Every Forecast Need? Structured Block Skipping in Two Time-Series Foundation Models

Key points

• Time-series foundation models use the same network depth for every input, although forecasting windows differ in structure and difficulty.1
• At moderate sparsity, Chronos-Bolt with 7/11 candidate blocks and TimesFM with 16/20 blocks show little change in mean relative forecasting quality across datasets while reducing median A100 batch-1 latency by 11.2% and 13.4%.1
• More aggressive TimesFM skipping accelerates inference further but substantially harms Traffic and Weather.1
• These results establish cross-architecture structured redundancy and input-dependent opportunity, but not a deployable dynamic inference method.1

Sources (1)

• [1]How Much Computation Does Every Forecast Need? Structured Block Skipping in Two Time-Series Foundation Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 9, 08:26 AM

  We test whether whole Transformer blocks can be removed at inference time in two pretrained models: Chronos-Bolt and TimesFM 2.5.

  Time-series foundation models use the same network depth for every input, although forecasting windows differ in structure and difficulty.

Extractive summary: sentences quoted from the sources.