fix(t0): bump tfc-t0 to 0.5.0 for t0-beta API and gift-eval CI

README.md

@@ -103,7 +103,7 @@ Some models require specific Python versions (e.g. FlowState 3.11-3.13, TabPFN &
 - **Moirai:** `Salesforce/moirai-{1.0,1.1,2.0}-R-{small,base,large}`, `Salesforce/moirai-moe-1.0-R-*`
 - **PatchTST-FM:** `ibm-research/patchtst-fm-r1`
 - **Sundial:** `thuml/sundial-base-128m`
-- **T0:** `theforecastingcompany/t0-alpha`
+- **T0:** `theforecastingcompany/t0-alpha`, `theforecastingcompany/t0-beta` (requires `tfc-t0>=0.5.0`)
 - **TabPFN:** default `tabpfn-v2-regressor-2noar4o2.ckpt`; TabPFN-3 via `model_path="tabpfn-v3-regressor-v3_20260506_timeseries.ckpt"` (LOCAL, `TABPFN_TOKEN`); or CLIENT mode with a Prior Labs API token
 - **Tafsut:** `Tafsut-FM/tafsut-univariate-base`
 - **TiRex:** `NX-AI/TiRex`, `NX-AI/TiRex-2`

docs/changelogs/index.md

@@ -2,6 +2,7 @@
 
 Welcome to the FoundationForecast Changelog. Here you will find a list of changes, updates, and improvements for each release. Explore the details of each version below.
 
+- [v0.1.8](v0.1.8.md)
 - [v0.1.7](v0.1.7.md)
 - [v0.1.6](v0.1.6.md)
 - [v0.1.5](v0.1.5.md)

docs/changelogs/v0.1.8.md

@@ -0,0 +1,31 @@
+### Features
+
+* **T0 beta support**: The shared [`T0`](../api/models/foundation/models.md#foundationforecast.models.t0.T0) wrapper now works with [`theforecastingcompany/t0-beta`](https://huggingface.co/theforecastingcompany/t0-beta) (~256M). Register it in gift-eval `models.yaml`; GIFT-Eval CI still replicates `t0-alpha` until leaderboard results exist for beta.
+
+    ```python
+    import pandas as pd
+    from foundationforecast.models.t0 import T0
+
+    df = pd.read_csv(
+        "https://timecopilot.s3.amazonaws.com/public/data/air_passengers.csv",
+        parse_dates=["ds"],
+    )
+    model = T0(
+        repo_id="theforecastingcompany/t0-beta",
+        alias="t0-beta",
+    )
+    fcst = model.forecast(df, h=12, freq="D", quantiles=[0.1, 0.5, 0.9])
+    assert "t0-beta" in fcst.columns
+    ```
+
+### Fixes
+
+* **T0 / gift-eval CI**: Bump `tfc-t0` to `>=0.5.0` and call `T0Forecaster.predict(quantile_levels=...)` instead of the removed `quantiles=` kwarg. Fixes gift-eval Modal runs after PyPI resolved `tfc-t0` 0.5.0 alongside the t0-beta release.
+
+### Dependencies
+
+* Bumped `tfc-t0` from `>=0.2.3` to `>=0.5.0` (required for t0-beta; still supports t0-alpha).
+
+---
+
+**Full Changelog**: https://github.com/TimeCopilot/foundationforecast/compare/v0.1.7...v0.1.8

docs/model-hub.md

@@ -41,7 +41,7 @@ Below is the list of available foundation models. Click a model name for API det
 - [Moirai](api/models/foundation/models.md#foundationforecast.models.moirai.Moirai) ([arXiv:2402.02592](https://arxiv.org/abs/2402.02592))
 - [PatchTST-FM](api/models/foundation/models.md#foundationforecast.models.patchtst_fm.PatchTSTFM) ([arXiv:2602.06909](https://arxiv.org/abs/2602.06909)) — Python 3.11–3.13
 - [Sundial](api/models/foundation/models.md#foundationforecast.models.sundial.Sundial) ([arXiv:2502.00816](https://arxiv.org/pdf/2502.00816))
-- [T0](api/models/foundation/models.md#foundationforecast.models.t0.T0) ([model card](https://huggingface.co/theforecastingcompany/t0-alpha)) — Python 3.11–3.13
+- [T0](api/models/foundation/models.md#foundationforecast.models.t0.T0) ([t0-alpha](https://huggingface.co/theforecastingcompany/t0-alpha), [t0-beta](https://huggingface.co/theforecastingcompany/t0-beta); requires `tfc-t0>=0.5.0` for beta) — Python 3.11–3.13
 - [TabPFN](api/models/foundation/models.md#foundationforecast.models.tabpfn.TabPFN) ([arXiv:2501.02945](https://arxiv.org/abs/2501.02945)) — Python 3.10–3.12; TabPFN-2 (default) and TabPFN-3 via `model_path`; TabPFN-3 is LOCAL-only; weights use the [TabPFN NC license](https://ux.priorlabs.ai) (`TABPFN_TOKEN`)
 - [Tafsut](api/models/foundation/models.md#foundationforecast.models.tafsut.Tafsut) ([GitHub](https://github.com/Tafsut-FM/tafsut))
 - [TiRex / TiRex-2](api/models/foundation/models.md#foundationforecast.models.tirex.TiRex) ([arXiv:2505.23719](https://arxiv.org/abs/2505.23719), [arXiv:2607.01204](https://arxiv.org/abs/2607.01204)) — Python 3.11+

experiments/gift-eval/configs/models.yaml

@@ -250,6 +250,14 @@ models:
       repo_id: theforecastingcompany/t0-alpha
       batch_size: 32
 
+  theforecastingcompany--t0-beta:
+    class: foundationforecast.models.t0.T0
+    reference_slug: null
+    kwargs:
+      repo_id: theforecastingcompany/t0-beta
+      alias: t0-beta
+      batch_size: 16
+
   thuml--sundial-base-128m:
     class: foundationforecast.models.sundial.Sundial
     reference_slug: sundial_base_128m

experiments/gift-eval/tests/test_models_config.py

@@ -0,0 +1,10 @@
+from src.eval.jobs import load_models_config
+
+
+def test_t0_beta_registered_in_models_yaml():
+    models = load_models_config()
+    entry = models["theforecastingcompany--t0-beta"]
+    assert entry["class"] == "foundationforecast.models.t0.T0"
+    assert entry["reference_slug"] is None
+    assert entry["kwargs"]["repo_id"] == "theforecastingcompany/t0-beta"
+    assert entry["kwargs"]["alias"] == "t0-beta"

foundationforecast/models/t0.py

@@ -61,14 +61,15 @@ def __init__(
         Notes:
             **Requirements:**
 
-            - T0 requires Python 3.11 to 3.13 (via the
-              [`tfc-t0`](https://pypi.org/project/tfc-t0/) package).
+            - T0 requires Python 3.11 to 3.13 and ``tfc-t0>=0.5.0`` (required
+              for ``t0-beta``; also supports ``t0-alpha``).
 
             **Available models:**
 
             | Model ID                                                                                          | Parameters |
             | ------------------------------------------------------------------------------------------------- | ---------- |
             | [`theforecastingcompany/t0-alpha`](https://huggingface.co/theforecastingcompany/t0-alpha)         | ~102M      |
+            | [`theforecastingcompany/t0-beta`](https://huggingface.co/theforecastingcompany/t0-beta)           | ~256M      |
 
             **Resources:**
 
@@ -79,9 +80,9 @@ def __init__(
 
             - The model is loaded onto the best available device (GPU if
               available, otherwise CPU).
-            - T0 predicts 5 quantile knots (0.1, 0.25, 0.5, 0.75, 0.9); the
-              median (0.5) is used as the point forecast and other requested
-              quantiles are obtained by linear interpolation across the knots.
+            - ``t0-alpha`` predicts 5 quantile knots (0.1, 0.25, 0.5, 0.75, 0.9);
+              ``t0-beta`` predicts 21 native levels (0.01–0.99). Requested
+              quantiles are interpolated; the median (0.5) is the point forecast.
             - NaN values in the context are treated as missing observations.
             - T0 natively supports past and known-future covariates through its
               `predict` API; this integration currently exposes the univariate
@@ -194,7 +195,7 @@ def forecast(
                 out = model.predict(
                     self._to_context(batch),
                     horizon=h,
-                    quantiles=pred_quantiles,
+                    quantile_levels=pred_quantiles,
                 )
                 # shape: (batch, h, n_quantiles)
                 fcsts.append(out.quantiles.cpu().numpy())

mkdocs.yml

@@ -41,6 +41,7 @@ nav:
       - Contributing: contributing.md
   - Changelogs:
       - changelogs/index.md
+      - changelogs/v0.1.8.md
       - changelogs/v0.1.7.md
       - changelogs/v0.1.6.md
       - changelogs/v0.1.5.md

pyproject.toml

@@ -54,7 +54,7 @@ dependencies = [
   "scipy",
   "tabpfn-time-series>=1.2.0 ; python_full_version < '3.13'",
   "tafsut>=0.1.0",
-  "tfc-t0>=0.2.3 ; python_full_version >= '3.11' and python_full_version < '3.14'",
+  "tfc-t0>=0.5.0 ; python_full_version >= '3.11' and python_full_version < '3.14'",
   "timecopilot-chronos-forecasting>=0.2.2",
   "timecopilot-granite-tsfm>=0.2.1 ; python_full_version >= '3.11' and python_full_version < '3.14'",
   "timecopilot-timesfm>=0.4.0",

tests/models/conftest.py

@@ -58,6 +58,14 @@
     from foundationforecast.models.t0 import T0
 
     models.append(T0(context_length=256, batch_size=2))
+    models.append(
+        T0(
+            repo_id="theforecastingcompany/t0-beta",
+            alias="t0-beta",
+            context_length=256,
+            batch_size=2,
+        )
+    )
 
 if (3, 11) <= sys.version_info < (3, 14):
     from foundationforecast.models.flowstate import FlowState

tests/models/test_t0.py

@@ -0,0 +1,68 @@
+import sys
+from unittest.mock import MagicMock
+
+import numpy as np
+import pytest
+
+if sys.version_info < (3, 11) or sys.version_info >= (3, 14):
+    pytest.skip(
+        "T0 requires Python >= 3.11 and < 3.14",
+        allow_module_level=True,
+    )
+
+import pandas as pd  # noqa: E402
+
+from tests.helpers import generate_series  # noqa: E402
+from foundationforecast.models.t0 import T0  # noqa: E402
+
+pytestmark = pytest.mark.models
+
+
+def test_t0_forecast_passes_quantile_levels(mocker):
+    model = T0(context_length=64, batch_size=2)
+    df = generate_series(2, freq="D", min_length=32, max_length=32)
+
+    mock_forecaster = MagicMock()
+    mock_out = MagicMock()
+    mock_out.quantiles.cpu.return_value.numpy.return_value = np.ones((2, 3, 3))
+    mock_forecaster.predict.return_value = mock_out
+
+    mock_cm = MagicMock()
+    mock_cm.__enter__.return_value = mock_forecaster
+    mock_cm.__exit__.return_value = None
+    mocker.patch.object(model, "_get_model", return_value=mock_cm)
+
+    model.forecast(df, h=3, freq="D", quantiles=[0.1, 0.9])
+
+    kwargs = mock_forecaster.predict.call_args.kwargs
+    assert kwargs["quantile_levels"] == [0.1, 0.5, 0.9]
+    assert "quantiles" not in kwargs
+
+
+def test_t0_beta_forecast():
+    df = generate_series(2, freq="D", min_length=50, max_length=50)
+    model = T0(
+        repo_id="theforecastingcompany/t0-beta",
+        alias="t0-beta",
+        context_length=256,
+        batch_size=2,
+    )
+    fcst = model.forecast(df, h=3, freq="D")
+    assert fcst.shape == (6, 3)
+    assert "t0-beta" in fcst.columns
+
+
+def test_t0_beta_quantile_forecast():
+    df = generate_series(2, freq="D", min_length=50, max_length=50)
+    model = T0(
+        repo_id="theforecastingcompany/t0-beta",
+        alias="t0-beta",
+        context_length=256,
+        batch_size=2,
+    )
+    fcst = model.forecast(df, h=3, freq="D", quantiles=[0.1, 0.9])
+    assert fcst.shape == (6, 5)
+    assert "t0-beta" in fcst.columns
+    assert "t0-beta-q-10" in fcst.columns
+    assert "t0-beta-q-90" in fcst.columns
+    assert pd.api.types.is_numeric_dtype(fcst["t0-beta-q-10"])