BOOM π₯ Observability Time-Series Forecasting Leaderboard
BOOM (Benchmark of Observability Metrics) is a large-scale, real-world time series dataset designed for evaluating models on forecasting tasks in complex observability environments. Consisting of around 350 million time-series data points spanning 32,887 variables, the benchmark is derived from real-world metrics collected via Datadog, a leading observability platform. It therefore captures the irregularity, structural complexity, and heavy-tailed statistics typical of production observability data.
To support research in resource-constrained settings, we also introduce BOOMLET, a smaller yet representative subset of the BOOM benchmark. It includes 32 metric queries, covering 1,627 variates and approximately 23 million observation points. It preserves key distributional properties while focusing on queries with more variates to ensure meaningful coverage.
For more information, please refer to the BOOM Dataset Card and the BOOM GitHub repository
- "headers": [
- "Type",
- "Model",
- "CRPS",
- "MASE",
- "Model Type",
- "Rank"
- "data": [
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Datadog/Toto-Open-Base-1.0" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Toto_Open_Base_1.0</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/toto.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.375,
- 0.617,
- "π’ pretrained",
- 2.369
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.428,
- 0.71,
- "π’ pretrained",
- 4.328
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-large" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_large</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.436,
- 0.72,
- "π’ pretrained",
- 4.561
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.442,
- 0.729,
- "π’ pretrained",
- 4.905
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/google/timesfm-2.0-500m-jax" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">timesfm_2_0_500m</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timesfm.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.447,
- 0.725,
- "π’ pretrained",
- 5.243
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.451,
- 0.726,
- "π’ pretrained",
- 5.576
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.455,
- 0.733,
- "π’ pretrained",
- 5.927
- [
- "π£",
- "Auto_Arima",
- 0.736,
- 0.824,
- "π£ statistical",
- 9.708
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-200M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-200M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.643,
- 0.881,
- "π’ pretrained",
- 9.809
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-50M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-50M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.649,
- 0.806,
- "π’ pretrained",
- 9.88
- [
- "π’",
- "<a target="_blank" href="https://github.com/thuml/Large-Time-Series-Model" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Timer</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timer.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.639,
- 0.796,
- "π’ pretrained",
- 9.923
- [
- "π’",
- "<a target="_blank" href="https://github.com/Keytoyze/VisionTS" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">VisionTS</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/visonts.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.673,
- 0.988,
- "π’ pretrained",
- 10.989
- [
- "π£",
- "Auto_ETS",
- 1.975,
- 0.842,
- "π£ statistical",
- 11.677
- [
- "π£",
- "Auto_Theta",
- 1.018,
- 1.123,
- "π£ statistical",
- 12.473
- [
- "π£",
- "Seasonal_Naive",
- 1,
- 1,
- "π£ statistical",
- 12.631
- [
- "metadata": null
- "headers": [
- "Type",
- "Model",
- "Application usage (CRPS)",
- "Application usage (MASE)",
- "Database (CRPS)",
- "Database (MASE)",
- "Infrastructure (CRPS)",
- "Infrastructure (MASE)",
- "Model Type",
- "Networking (CRPS)",
- "Networking (MASE)",
- "Security (CRPS)",
- "Security (MASE)"
- "data": [
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Datadog/Toto-Open-Base-1.0" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Toto_Open_Base_1.0</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/toto.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.377,
- 0.625,
- 0.33,
- 0.627,
- 0.401,
- 0.585,
- "π’ pretrained",
- 0.342,
- 0.629,
- 0.501,
- 0.63
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.425,
- 0.714,
- 0.378,
- 0.735,
- 0.464,
- 0.681,
- "π’ pretrained",
- 0.381,
- 0.716,
- 0.597,
- 0.762
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-large" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_large</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.432,
- 0.718,
- 0.383,
- 0.744,
- 0.48,
- 0.702,
- "π’ pretrained",
- 0.386,
- 0.715,
- 0.586,
- 0.749
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.441,
- 0.732,
- 0.387,
- 0.74,
- 0.481,
- 0.707,
- "π’ pretrained",
- 0.393,
- 0.735,
- 0.617,
- 0.792
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/google/timesfm-2.0-500m-jax" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">timesfm_2_0_500m</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timesfm.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.444,
- 0.728,
- 0.398,
- 0.761,
- 0.481,
- 0.694,
- "π’ pretrained",
- 0.409,
- 0.745,
- 0.574,
- 0.702
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.45,
- 0.732,
- 0.395,
- 0.747,
- 0.484,
- 0.691,
- "π’ pretrained",
- 0.409,
- 0.743,
- 0.611,
- 0.733
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.453,
- 0.737,
- 0.394,
- 0.742,
- 0.493,
- 0.706,
- "π’ pretrained",
- 0.416,
- 0.754,
- 0.621,
- 0.744
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-200M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-200M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.617,
- 0.868,
- 0.584,
- 0.928,
- 0.733,
- 0.876,
- "π’ pretrained",
- 0.615,
- 0.915,
- 0.747,
- 0.804
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-50M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-50M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.63,
- 0.826,
- 0.578,
- 0.74,
- 0.731,
- 0.806,
- "π’ pretrained",
- 0.619,
- 0.914,
- 0.772,
- 0.828
- [
- "π’",
- "<a target="_blank" href="https://github.com/thuml/Large-Time-Series-Model" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Timer</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timer.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.63,
- 0.828,
- 0.586,
- 0.75,
- 0.679,
- 0.749,
- "π’ pretrained",
- 0.608,
- 0.909,
- 0.831,
- 0.899
- [
- "π’",
- "<a target="_blank" href="https://github.com/Keytoyze/VisionTS" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">VisionTS</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/visonts.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.683,
- 1.013,
- 0.613,
- 1.061,
- 0.688,
- 0.893,
- "π’ pretrained",
- 0.615,
- 1.026,
- 0.875,
- 0.951
- [
- "π£",
- "Auto_Arima",
- 0.752,
- 0.846,
- 0.735,
- 0.866,
- 0.692,
- 0.735,
- "π£ statistical",
- 0.711,
- 0.843,
- 0.801,
- 0.963
- [
- "π£",
- "Seasonal_Naive",
- 1,
- 1,
- 1,
- 1,
- 1,
- 1,
- "π£ statistical",
- 1,
- 1,
- 1,
- 1
- [
- "π£",
- "Auto_Theta",
- 1.037,
- 1.142,
- 1.054,
- 1.266,
- 0.984,
- 1.008,
- "π£ statistical",
- 0.945,
- 1.18,
- 0.955,
- 1.013
- [
- "π£",
- "Auto_ETS",
- 1.623,
- 0.854,
- 1.394,
- 0.875,
- 4.428,
- 0.759,
- "π£ statistical",
- 1.777,
- 0.913,
- 0.838,
- 0.931
- [
- "metadata": null
- "headers": [
- "Type",
- "Model",
- "Count (CRPS)",
- "Count (MASE)",
- "Distribution (CRPS)",
- "Distribution (MASE)",
- "Gauge (CRPS)",
- "Gauge (MASE)",
- "Model Type",
- "Rate (CRPS)",
- "Rate (MASE)"
- "data": [
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Datadog/Toto-Open-Base-1.0" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Toto_Open_Base_1.0</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/toto.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.317,
- 0.687,
- 0.382,
- 0.658,
- 0.382,
- 0.583,
- "π’ pretrained",
- 0.369,
- 0.634
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.353,
- 0.795,
- 0.422,
- 0.724,
- 0.444,
- 0.686,
- "π’ pretrained",
- 0.418,
- 0.728
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-large" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_large</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.372,
- 0.813,
- 0.428,
- 0.729,
- 0.456,
- 0.7,
- "π’ pretrained",
- 0.422,
- 0.733
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.379,
- 0.808,
- 0.432,
- 0.737,
- 0.459,
- 0.704,
- "π’ pretrained",
- 0.433,
- 0.75
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.402,
- 0.88,
- 0.446,
- 0.753,
- 0.463,
- 0.696,
- "π’ pretrained",
- 0.443,
- 0.739
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.403,
- 0.883,
- 0.452,
- 0.759,
- 0.469,
- 0.706,
- "π’ pretrained",
- 0.445,
- 0.742
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/google/timesfm-2.0-500m-jax" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">timesfm_2_0_500m</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timesfm.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.403,
- 0.919,
- 0.44,
- 0.745,
- 0.466,
- 0.706,
- "π’ pretrained",
- 0.431,
- 0.726
- [
- "π’",
- "<a target="_blank" href="https://github.com/Keytoyze/VisionTS" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">VisionTS</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/visonts.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.591,
- 1.199,
- 0.672,
- 1.03,
- 0.674,
- 0.925,
- "π’ pretrained",
- 0.68,
- 1.032
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-50M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-50M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.651,
- 0.652,
- 0.604,
- 0.878,
- 0.694,
- 0.76,
- "π’ pretrained",
- 0.619,
- 0.846
- [
- "π’",
- "<a target="_blank" href="https://github.com/thuml/Large-Time-Series-Model" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Timer</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timer.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.662,
- 0.663,
- 0.608,
- 0.89,
- 0.658,
- 0.721,
- "π’ pretrained",
- 0.63,
- 0.864
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-200M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-200M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.698,
- 1.035,
- 0.596,
- 0.877,
- 0.703,
- 0.89,
- "π’ pretrained",
- 0.596,
- 0.862
- [
- "π£",
- "Auto_Arima",
- 0.767,
- 1.031,
- 0.736,
- 0.864,
- 0.701,
- 0.755,
- "π£ statistical",
- 0.777,
- 0.88
- [
- "π£",
- "Seasonal_Naive",
- 1,
- 1,
- 1,
- 1,
- 1,
- 1,
- "π£ statistical",
- 1,
- 1
- [
- "π£",
- "Auto_Theta",
- 1.132,
- 1.578,
- 0.976,
- 1.144,
- 1.015,
- 1.062,
- "π£ statistical",
- 1.034,
- 1.158
- [
- "π£",
- "Auto_ETS",
- 1.557,
- 1.074,
- 1.22,
- 0.854,
- 3.059,
- 0.774,
- "π£ statistical",
- 1.496,
- 0.907
- [
- "metadata": null
- "headers": [
- "Type",
- "Model",
- "Long (CRPS)",
- "Long (MASE)",
- "Medium (CRPS)",
- "Medium (MASE)",
- "Model Type",
- "Short (CRPS)",
- "Short (MASE)"
- "data": [
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Datadog/Toto-Open-Base-1.0" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Toto_Open_Base_1.0</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/toto.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.424,
- 0.688,
- 0.406,
- 0.657,
- "π’ pretrained",
- 0.318,
- 0.535
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.473,
- 0.78,
- 0.46,
- 0.753,
- "π’ pretrained",
- 0.37,
- 0.627
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.485,
- 0.797,
- 0.475,
- 0.769,
- "π’ pretrained",
- 0.387,
- 0.649
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-large" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_large</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.491,
- 0.799,
- 0.475,
- 0.77,
- "π’ pretrained",
- 0.369,
- 0.626
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.519,
- 0.798,
- 0.507,
- 0.782,
- "π’ pretrained",
- 0.365,
- 0.632
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/google/timesfm-2.0-500m-jax" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">timesfm_2_0_500m</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timesfm.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.522,
- 0.817,
- 0.499,
- 0.78,
- "π’ pretrained",
- 0.359,
- 0.619
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.528,
- 0.813,
- 0.508,
- 0.782,
- "π’ pretrained",
- 0.368,
- 0.637
- [
- "π’",
- "<a target="_blank" href="https://github.com/thuml/Large-Time-Series-Model" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Timer</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timer.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.661,
- 0.809,
- 0.671,
- 0.804,
- "π’ pretrained",
- 0.597,
- 0.779
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-200M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-200M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.694,
- 0.95,
- 0.692,
- 0.929,
- "π’ pretrained",
- 0.57,
- 0.794
- [
- "π’",
- "<a target="_blank" href="https://github.com/Keytoyze/VisionTS" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">VisionTS</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/visonts.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.698,
- 1.028,
- 0.694,
- 1.006,
- "π’ pretrained",
- 0.637,
- 0.943
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-50M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-50M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.724,
- 0.886,
- 0.725,
- 0.866,
- "π’ pretrained",
- 0.541,
- 0.704
- [
- "π£",
- "Auto_Arima",
- 0.807,
- 0.895,
- 0.804,
- 0.853,
- "π£ statistical",
- 0.634,
- 0.749
- [
- "π£",
- "Seasonal_Naive",
- 1,
- 1,
- 1,
- 1,
- "π£ statistical",
- 1,
- 1
- [
- "π£",
- "Auto_Theta",
- 1.296,
- 1.368,
- 1.183,
- 1.163,
- "π£ statistical",
- 0.738,
- 0.928
- [
- "π£",
- "Auto_ETS",
- 2.83,
- 0.938,
- 2.399,
- 0.885,
- "π£ statistical",
- 1.253,
- 0.739
- [
- "metadata": null
- "headers": [
- "Type",
- "Model",
- "CRPS",
- "MASE",
- "Model Type",
- "Rank"
- "data": [
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Datadog/Toto-Open-Base-1.0" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Toto_Open_Base_1.0</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/toto.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.519,
- 0.617,
- "π’ pretrained",
- 1.244
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/google/timesfm-2.0-500m-jax" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">timesfm_2_0_500m</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timesfm.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.603,
- 0.685,
- "π’ pretrained",
- 4.178
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-large" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_large</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.621,
- 0.767,
- "π’ pretrained",
- 4.322
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.63,
- 0.779,
- "π’ pretrained",
- 4.644
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Salesforce/moirai-1.1-R-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Moirai_1.1_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/moirai.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.631,
- 0.786,
- "π’ pretrained",
- 5.044
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-base" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_base</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.637,
- 0.711,
- "π’ pretrained",
- 5.578
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/amazon/chronos-bolt-small" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">chronos_bolt_small</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/chronos.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.642,
- 0.717,
- "π’ pretrained",
- 5.778
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-200M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-200M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.78,
- 0.793,
- "π’ pretrained",
- 9.478
- [
- "π’",
- "<a target="_blank" href="https://huggingface.co/Maple728/TimeMoE-50M" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">TimeMoE-50M</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/time-moe.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.788,
- 0.81,
- "π’ pretrained",
- 9.5
- [
- "π’",
- "<a target="_blank" href="https://github.com/thuml/Large-Time-Series-Model" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">Timer</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/timer.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.793,
- 0.807,
- "π’ pretrained",
- 9.844
- [
- "π£",
- "Auto_Arima",
- 0.88,
- 0.922,
- "π£ statistical",
- 10.3
- [
- "π’",
- "<a target="_blank" href="https://github.com/Keytoyze/VisionTS" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">VisionTS</a> (<a target="_blank" href="https://github.com/DataDog/toto/blob/main/boom/notebooks/visonts.ipynb" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">code</a>)",
- 0.885,
- 0.912,
- "π’ pretrained",
- 11.722
- [
- "π£",
- "Seasonal_Naive",
- 1,
- 1,
- "π£ statistical",
- 12.222
- [
- "π£",
- "Auto_ETS",
- 15.664,
- 0.969,
- "π£ statistical",
- 12.856
- [
- "π£",
- "Auto_Theta",
- 1.182,
- 1.03,
- "π£ statistical",
- 13.289
- [
- "metadata": null
BOOM is a large-scale, real-world time series dataset designed for benchmarking forecasting models in observability environments. The dataset captures the complexity and irregularity of production observability data, making it a challenging and realistic testbed for time series forecasting. BOOM consists of approximately 350 million time-series points across 32,887 variates. The dataset is split into 2,807 individual time series with one or multiple variates. For more details and dataset structure, please refer to the BOOM Dataset Card.
The evaluation procedure is inspired by Gift-Eval: We evaluate models using MASE (Mean Absolute Scaled Error) for forecast accuracy, CRPS (Continuous Ranked Probability Score) for probabilistic forecast quality, and Rankβwhich determines overall performance and is used to order models on the leaderboard.
To reproduce our results, we provide a guide in the BOOM GitHub repository that explains how to install the required dependencies and includes example notebooks demonstrating how to evaluate both foundation models and statistical baselines on BOOM.