feat: add T0Model foundation model

CHANGELOG.md

@@ -17,7 +17,8 @@ but cannot always guarantee backwards compatibility. Changes that may **break co
   - Custom PyTorch datasets and Lightning modules are easier to read, extend, and debug: samples use named fields (`past_target`, `future_covariates`, ...) instead of positional tuples, modules receive each feature as a separate tensor rather than one concatenated input, and recurrent state is returned in a structured output. Models saved with previous Darts versions continue to load for inference.
     - 🔴 Custom `TorchTrainingDataset` / `TorchInferenceDataset` implementations must return `TorchTrainingSample` / `TorchInferenceSample`.
     - 🔴 Custom module `forward()` methods must accept `PLModuleInput` and return `PLModuleOutput`.
-- 🚀🚀 Added new forecasting model `TimesFM3Model` : Google's pre-trained 330M-parameter foundation model for zero-shot forecasting. Unlike previous versions, it natively supports multivariate time series, past covariates, and future covariates, and can output deterministic or probabilistic forecasts without training. The TimesFM 3.0 pre-trained weights are non-commercial: users must accept the license with `accept_license=True` when creating the model. [#3199](https://github.com/unit8co/darts/pull/3199) by [JuanCruzC97](https://github.com/JuanCruzC97).
+- 🚀 Added new forecasting model `TimesFM3Model` : Google's pre-trained 330M-parameter foundation model for zero-shot forecasting. Unlike previous versions, it natively supports multivariate time series, past covariates, and future covariates, and can output deterministic or probabilistic forecasts without training. The TimesFM 3.0 pre-trained weights are non-commercial: users must accept the license with `accept_license=True` when creating the model. [#3199](https://github.com/unit8co/darts/pull/3199) by [JuanCruzC97](https://github.com/JuanCruzC97).
+- 🚀 Added new forecasting model `T0Model` : The Forecasting Company's open-weights ~100M-parameter foundation model (see [here](https://huggingface.co/theforecastingcompany/t0-alpha)) for zero-shot forecasting. It supports univariate, multivariate, and multiple time series as well as past and future covariates, without training, and can output deterministic or probabilistic forecasts. It can also be fine-tuned (full or partial) with `enable_finetuning`. [#3142](https://github.com/unit8co/darts/pull/3142) by [Geoffrey Négiar](https://github.com/GeoffNN), [Huikan Xiang](https://github.com/huikan-tfc) and [Lucas Meyer](https://github.com/LTMeyer).
 - `FittableAnomalyScorer.fit_from_prediction()` now returns the fitted scorer object similar to `fit()`. [#3202](https://github.com/unit8co/darts/pull/3202) by [Venish Paneliya](https://github.com/VenishPaneliya).
 - Calling `ForecastingModel.historical_forecasts()` with a `start` value that is later than what is forecastable given the supplied covariates now raises an informative exception. [#3207](https://github.com/unit8co/darts/pull/3207) by [Dennis Bader](https://github.com/dennisbader).
 - Calling `ForecastingModel.gridsearch()` with a sequence of `TimeSeries` now raises an informative exception. [#3191](https://github.com/unit8co/darts/pull/3191) by [Geovanny Basantes](https://github.com/COMPUMAX-EC).

INSTALL.md

@@ -22,6 +22,7 @@ Some models have additional dependencies that are not included in the `all` inst
 |-----------------------|-----------------------|
 | `NeuralForecastModel` | neuralforecast>=3.0.0 |
 | `TiRexModel`          | tirex-ts>=1.4.0       |
+| `T0Model`             | tfc-t0>=0.5.0         |
 
 Some optional integrations also require additional dependencies:
 
@@ -59,6 +60,7 @@ Some models have dependencies not available on conda-forge. To use them, you nee
 | Model                 | Dependencies          |
 |-----------------------|-----------------------|
 | `TiRexModel`          | tirex-ts>=1.4.0       |
+| `T0Model`             | tfc-t0>=0.5.0         |
 
 
 ## Other Information

README.md

@@ -284,6 +284,7 @@ Here's a breakdown of the forecasting models currently implemented in Darts. Our
 | [TimesFM3Model](https://unit8co.github.io/darts/generated_api/darts.models.forecasting.timesfm3_model.html#darts.models.forecasting.timesfm3_model.TimesFM3Model)                                                                                                                                | [TimesFM 1.0 paper](https://arxiv.org/abs/2310.10688), [TimesFM 3.0 model card](https://huggingface.co/google/timesfm-3.0-pytorch)                                                                                                | ✅ ✅                                                        | ✅ ✅ 🔴                                                                 | ✅ ✅                                                                    | ✅                                        |
 | [TiRexModel](https://unit8co.github.io/darts/generated_api/darts.models.forecasting.tirex_model.html#darts.models.forecasting.tirex_model.TiRexModel)                                                                                                                                            | [TiRex paper](https://arxiv.org/abs/2505.23719), [TiRex GitHub](https://github.com/NX-AI/tirex)                                                                                                                                   | ✅ ✅                                                        | 🔴 🔴 🔴                                                                 | ✅ ✅                                                                    | ✅                                        |
 | [PatchTSTFMModel](https://unit8co.github.io/darts/generated_api/darts.models.forecasting.patchtst_fm_model.html#darts.models.forecasting.patchtst_fm_model.PatchTSTFMModel)                                                                                                                      | [PatchTST-FM paper](https://arxiv.org/abs/2602.06909), [PatchTST-FM GitHub](https://github.com/ibm-granite/granite-tsfm)                                                                                                          | ✅ ✅                                                        | 🔴 🔴 🔴                                                                 | ✅ ✅                                                                    | ✅                                        |
+| [T0Model](https://unit8co.github.io/darts/generated_api/darts.models.forecasting.t0_model.html#darts.models.forecasting.t0_model.T0Model)                                                                                                                                                        | [T0 report](https://www.theforecastingcompany.com/papers/files/t0-technical-report.pdf), [tfc-t0 GitHub](https://github.com/theforecastingcompany/tfc-t0)                                                                         | ✅ ✅                                                        | ✅ ✅ 🔴                                                                 | ✅ ✅                                                                    | ✅                                        |
 | **Ensemble Models**<br/>([GlobalForecastingModel](https://unit8co.github.io/darts/userguide/covariates.html#global-forecasting-models-gfms)): Model support is dependent on ensembled forecasting models and the ensemble model itself                                                           |                                                                                                                                                                                                                                   |                                                              |                                                                          |                                                                          |                                           |
 | [NaiveEnsembleModel](https://unit8co.github.io/darts/generated_api/darts.models.forecasting.naive_ensemble_model.html#darts.models.forecasting.naive_ensemble_model.NaiveEnsembleModel)                                                                                                          |                                                                                                                                                                                                                                   | ✅ ✅                                                        | ✅ ✅ ✅                                                                 | ✅ ✅                                                                    | ✅                                        |
 | [RegressionEnsembleModel](https://unit8co.github.io/darts/generated_api/darts.models.forecasting.regression_ensemble_model.html#darts.models.forecasting.regression_ensemble_model.RegressionEnsembleModel)                                                                                      |                                                                                                                                                                                                                                   | ✅ ✅                                                        | ✅ ✅ ✅                                                                 | ✅ ✅                                                                    | ✅                                        |

darts/models/__init__.py

@@ -106,6 +106,7 @@
         SKLearnClassifierModel as SKLearnClassifierModel,
     )
     from darts.models.forecasting.sklearn_model import SKLearnModel as SKLearnModel
+    from darts.models.forecasting.t0_model import T0Model as T0Model
     from darts.models.forecasting.tcn_model import TCNModel as TCNModel
     from darts.models.forecasting.tft_model import TFTModel as TFTModel
     from darts.models.forecasting.theta import FourTheta as FourTheta
@@ -195,6 +196,7 @@
     "TimesFM2p5Model": ("darts.models.forecasting.timesfm2p5_model", "(Py)Torch"),
     "TimesFM3Model": ("darts.models.forecasting.timesfm3_model", "(Py)Torch"),
     "TiRexModel": ("darts.models.forecasting.tirex_model", "(Py)Torch and/or TiRex-TS"),
+    "T0Model": ("darts.models.forecasting.t0_model", "(Py)Torch and/or tfc-t0"),
     # --- Forecasting: NeuralForecast ---
     "NeuralForecastModel": ("darts.models.forecasting.nf_model", "NeuralForecast"),
     # --- Forecasting: Prophet ---

darts/models/components/huggingface_connector.py

@@ -43,18 +43,18 @@ def __init__(
         Parameters
         ----------
         model_name
-             The HuggingFace repository name where the model is stored, e.g., "amazon/chronos-2".
+            The HuggingFace repository name where the model is stored, e.g., "amazon/chronos-2".
         model_revision
-             The revision of the model in the HuggingFace repository. Must be a branch name, tag name, or commit hash.
-                If not provided, the default branch and the latest commit will be used.
+            The revision of the model in the HuggingFace repository. Must be a branch name, tag name, or commit hash.
+            If not provided, the default branch and the latest commit will be used.
         local_dir
             Optional local directory to load the pre-downloaded model. If specified and the directory is empty, the
             model will be downloaded from HuggingFace Hub and saved to this directory. Default is ``None``, which will
             use a cache directory managed by ``huggingface_hub`` instead.
         config_file
-             The name of the configuration file. Default is "config.json".
+            The name of the configuration file. Default is "config.json".
         model_file
-             The name of the model weight file. Default is "model.safetensors".
+            The name of the model weight file. Default is "model.safetensors".
         """
         if local_dir is not None:
             local_dir_path = Path(local_dir)

darts/models/forecasting/__init__.py

@@ -60,6 +60,7 @@
     - :class:`~darts.models.forecasting.timesfm3_model.TimesFM3Model`
     - :class:`~darts.models.forecasting.tirex_model.TiRexModel`
     - :class:`~darts.models.forecasting.patchtst_fm_model.PatchTSTFMModel`
+    - :class:`~darts.models.forecasting.t0_model.T0Model`
 Ensemble Models (`GlobalForecastingModel <https://unit8co.github.io/darts/userguide/covariates.html#global-forecasting-models-gfms>`__)
     - :class:`~darts.models.forecasting.naive_ensemble_model.NaiveEnsembleModel`
     - :class:`~darts.models.forecasting.regression_ensemble_model.RegressionEnsembleModel`