Contamination MapTime series / Model
Chronos-Bolt
- Organisation
- Amazon
- Family
- Chronos
- Released
- 2024-11-26 (Repository README news entry '26 Nov 2024: Chronos-Bolt models released on HuggingFace' (https://raw.githubusercontent.com/amazon-science/chronos-forecasting/main/README.md); HF repos amazon/chronos-bolt-* created 2024-11-25 (HF API createdAt))
- Checkpoints
- amazon/chronos-bolt-tiny, amazon/chronos-bolt-mini, amazon/chronos-bolt-small, amazon/chronos-bolt-base
- Weights
- Open · apache-2.0
- Links
- Model cardCode
Training data disclosure: Partial list
Level determined for 18% of benchmark cells.
- TestTest data seen0
- SeenSeen, split unclear8
- TrainTraining data seen0
- Related data seen1
- NoNot seen0
- ?Unknown40
“has been trained on nearly 100 billion time series observations”
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
Training corpora
-
Chronos training corpus (real-world datasets)
· 28 sources (complete list) · pre-training
· homepage
Evidence that Chronos-Bolt uses it
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
github.com · website · Comment by abdulfatir (Chronos first author), 22 Dec 2024 · retrieved 2026-10-15 “The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
github.com · website · Comment by abdulfatir, 22 Dec 2024 · retrieved 2026-10-15 Contents of this corpus
-
Brazilian Cities Temperature · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Brazilian Cities Temperature”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Mexico City Bikes · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Mexico City Bikes”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
NREL Solar Power Data for Integration Studies (2006) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Solar (5 Min.)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
NREL Solar Power Data for Integration Studies (2006) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Solar (Hourly)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Spanish Energy and Weather · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Spanish Energy and Weather”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Taxi (Hourly) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Taxi (Hourly)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
USHCN · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“USHCN”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Weatherbench (Daily) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Weatherbench (Daily)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Weatherbench (Hourly) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Weatherbench (Hourly)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Weatherbench (Weekly) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Weatherbench (Weekly)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Wiki Daily (100k) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Wiki Daily (100k)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Wind Farms (Daily) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Wind Farms (Daily)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Wind Farms (Hourly) · split: unknown
Pretraining-only dataset (not part of Chronos Benchmark I/II). The paper does not state explicitly that the full series were used, but describes no held-out portion for this group.
Evidence
“Wind Farms (Hourly)”
arxiv.org · paper · Appendix B, Table 3, section 'Pretraining-only' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
ElectricityLoadDiagrams20112014 (UCI) · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Electricity (15 Min.)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Electricity (ECL, 321 clients) · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Electricity (Hourly)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Electricity Weekly (Monash) · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Electricity (Weekly)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
KDD Cup 2018 · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“KDD Cup 2018”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
London Smart Meters · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“London Smart Meters”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
M4 Daily · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“M4 (Daily)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
M4 Hourly · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“M4 (Hourly)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
M4 Monthly · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“M4 (Monthly)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
M4 Weekly · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“M4 (Weekly)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Pedestrian Counts · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Pedestrian Counts”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Rideshare · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Rideshare”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Taxi (30 Min.) · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Taxi (30 Min.)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Temperature Rain · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Temperature-Rain”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Uber TLC (Daily) · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Uber TLC (Daily)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15 -
Uber TLC (Hourly) · split: train
Chronos Benchmark I (in-domain) dataset: used for training except the last H observations of each series (H is task-specific, Table 3), which are held out for testing. This is the Chronos paper's own split, not GIFT-Eval's or long-horizon's.
Evidence
“Uber TLC (Hourly)”
arxiv.org · paper · Appendix B, Table 3, section 'In-domain evaluation' · retrieved 2026-10-15 “pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
arxiv.org · paper · Appendix B, Table 3 caption · retrieved 2026-10-15
-
Brazilian Cities Temperature · split: unknown
Benchmark cells
GIFT-Eval
Unknown Jena Weather (GIFT-Eval) · training data not disclosed in enough detail
Unknown BizITObs - Application · training data not disclosed in enough detail
Unknown BizITObs - Service · training data not disclosed in enough detail
Unknown BizITObs - L2C · training data not disclosed in enough detail
Unknown Bitbrains - Fast Storage · training data not disclosed in enough detail
Unknown Bitbrains - rnd · training data not disclosed in enough detail
Unknown Restaurant (Recruit Restaurant Visitor Forecasting) · training data not disclosed in enough detail
Unknown ETT1 (GIFT-Eval) · training data not disclosed in enough detail
Unknown ETT2 (GIFT-Eval) · training data not disclosed in enough detail
Unknown Loop Seattle (LibCity) · training data not disclosed in enough detail
Unknown SZ-Taxi (LibCity) · training data not disclosed in enough detail
Unknown M_DENSE (LibCity) · training data not disclosed in enough detail
Solar (LSTNet, 137 Alabama PV plants) · related dataset NREL Solar Power Data for Integration Studies (2006) was used (via Chronos training corpus (real-world datasets))
Evidence (6)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“Solar (5 Min.)”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
“The uploaded dataset contains the aggregated version of a subset of the original dataset used by Lai et al. (2017).”
“Solar (5 Min., Hourly) contains data about solar power generation in the US in 2006. The original data has 5 minute frequency and was obtained from https://www.nrel.gov/grid/solar-power-data.html”
Unknown Hierarchical Sales (Mancuso et al., 2020) · training data not disclosed in enough detail
Unknown M4 Yearly · training data not disclosed in enough detail
Unknown M4 Quarterly · training data not disclosed in enough detail
Seen, split unclear M4 Monthly · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“M4 (Monthly)”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Seen, split unclear M4 Weekly · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“M4 (Weekly)”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Seen, split unclear M4 Daily · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“M4 (Daily)”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Seen, split unclear M4 Hourly · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“M4 (Hourly)”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Unknown Hospital · training data not disclosed in enough detail
Unknown COVID Deaths · training data not disclosed in enough detail
Unknown US Births · training data not disclosed in enough detail
Unknown Saugeen River Flow · training data not disclosed in enough detail
Seen, split unclear Temperature Rain · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“Temperature-Rain”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Seen, split unclear KDD Cup 2018 · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“KDD Cup 2018”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Unknown Car Parts · training data not disclosed in enough detail
Seen, split unclear ElectricityLoadDiagrams20112014 (UCI) · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“Electricity (15 Min.)”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI)
Unknown ETTh1 · training data not disclosed in enough detail
Unknown ETTh2 · training data not disclosed in enough detail
Unknown ETTm1 · training data not disclosed in enough detail
Unknown ETTm2 · training data not disclosed in enough detail
Seen, split unclear Electricity (ECL, 321 clients) · a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))
Evidence (4)
“The real datasets used for training Chronos-Bolt are the same as Chronos.”
“The datasets under "Pretraining only" and "In-domain evaluation" were used for pretraining whereas the datasets under "Zero-shot evaluation" (and all other datasets not listed in the table) were not used for training.”
“Electricity (Hourly)”
“pretraining-only data is only used for Chronos training; in-domain evalution data is used for training Chronos models and other task-specific baselines, except for the H observations that are held out for in-domain testing only”
Unknown Traffic (PeMS San Francisco Bay, 862 sensors) · training data not disclosed in enough detail
Unknown Weather (Jena, 2020; Autoformer) · training data not disclosed in enough detail
Unknown Exchange-Rate · training data not disclosed in enough detail
Unknown ILI (CDC influenza-like illness, national) · training data not disclosed in enough detail
Time-series anomaly detection
Unknown TSB-AD-U (TSB-AD univariate) · training data not disclosed in enough detail
Unknown TSB-AD-M (TSB-AD multivariate) · training data not disclosed in enough detail
Unknown UCR Time Series Anomaly Archive (Hexagon ML/UCR, KDD Cup 2021) · training data not disclosed in enough detail
Unknown Server Machine Dataset (SMD) · training data not disclosed in enough detail
Unknown Mars Science Laboratory (MSL) telemetry anomaly dataset · training data not disclosed in enough detail
Unknown Soil Moisture Active Passive (SMAP) telemetry anomaly dataset · training data not disclosed in enough detail
Unknown Pooled Server Metrics (PSM) · training data not disclosed in enough detail
Unknown Secure Water Treatment (SWaT) · training data not disclosed in enough detail
Unknown Yahoo S5 Labeled Anomaly Detection Dataset (Yahoo Webscope S5) · training data not disclosed in enough detail
Unknown Numenta Anomaly Benchmark (NAB) · training data not disclosed in enough detail
Time-series classification
Unknown UCR Time Series Classification Archive (2018) · training data not disclosed in enough detail
Unknown UEA Multivariate Time Series Classification Archive (2018) · training data not disclosed in enough detail
Notes
No paper. The only data statement in the model card/blog is 'nearly 100 billion time series observations' (vs ~84B for Chronos), so the synthetic/augmented part of Bolt's training data is undisclosed. Identical copies exist under the autogluon/ HF org. Other stated exclusions, not tracked here: chronos-benchmark-ii (all).
Sources
- amazon/chronos-bolt-base model card · model card · retrieved 2026-10-15
- Fast and accurate zero-shot forecasting with Chronos-Bolt and AutoGluon (AWS ML blog) · website · retrieved 2026-10-15
- chronos-forecasting issue #245: Does the traffic involved in the pretraining for the chronos-bolt version? · website · retrieved 2026-10-15
- amazon-science/chronos-forecasting README · readme · retrieved 2026-10-15