Upload 5 files

config.json

@@ -0,0 +1,283 @@
+{
+  "K": 52,
+  "S": 66,
+  "W": 6,
+  "architectures": [
+    "YieldForRegression"
+  ],
+  "crop_emb_dim": 8,
+  "d_model": 64,
+  "dim_ff": 128,
+  "dropout": 0.4,
+  "dtype": "float32",
+  "eval_cutoffs": [
+    20,
+    24,
+    28,
+    32,
+    36,
+    40,
+    44,
+    48,
+    52
+  ],
+  "model_type": "yield-weather-soil",
+  "nhead": 4,
+  "num_layers": 4,
+  "pool": "last",
+  "s_mean": [
+    138.40794372558594,
+    152.29786682128906,
+    154.56175231933594,
+    160.80038452148438,
+    164.22401428222656,
+    165.8285675048828,
+    232.1925506591797,
+    211.66586303710938,
+    208.61048889160156,
+    210.4038543701172,
+    196.25576782226562,
+    169.04379272460938,
+    16.51355743408203,
+    16.052227020263672,
+    16.29838752746582,
+    18.991479873657227,
+    26.101285934448242,
+    35.37116241455078,
+    293.3885192871094,
+    296.4434814453125,
+    315.0303039550781,
+    331.96435546875,
+    329.5356750488281,
+    309.2357177734375,
+    3859.735107421875,
+    2151.462158203125,
+    1574.3883056640625,
+    1057.6832275390625,
+    792.9168090820312,
+    640.056884765625,
+    401.9707946777344,
+    282.0995178222656,
+    218.99769592285156,
+    119.24910736083984,
+    71.83614349365234,
+    44.02671432495117,
+    61.37840270996094,
+    61.37840270996094,
+    61.37840270996094,
+    61.37840270996094,
+    61.37840270996094,
+    61.37840270996094,
+    64.16217041015625,
+    64.32339477539062,
+    64.31382751464844,
+    65.56702423095703,
+    68.18431854248047,
+    71.56515502929688,
+    197.78048706054688,
+    196.43638610839844,
+    191.7803955078125,
+    190.5934295654297,
+    205.2471466064453,
+    238.66366577148438,
+    508.832275390625,
+    507.11798095703125,
+    493.1898193359375,
+    477.44244384765625,
+    465.2177429199219,
+    452.10076904296875,
+    347.4835510253906,
+    207.85842895507812,
+    140.189453125,
+    88.9273681640625,
+    50.258174896240234,
+    35.850830078125
+  ],
+  "s_std": [
+    5.658433437347412,
+    4.963406562805176,
+    4.842231273651123,
+    6.4795145988464355,
+    7.407203674316406,
+    8.543693542480469,
+    32.758174896240234,
+    43.4786262512207,
+    41.363216400146484,
+    36.315101623535156,
+    37.40601348876953,
+    44.32140350341797,
+    14.699563980102539,
+    14.964685440063477,
+    15.500894546508789,
+    18.559709548950195,
+    22.95189666748047,
+    28.39887046813965,
+    55.16016387939453,
+    54.736392974853516,
+    54.40840148925781,
+    57.90495681762695,
+    59.78011703491211,
+    68.35675048828125,
+    1374.9456787109375,
+    633.04736328125,
+    509.03704833984375,
+    386.8252258300781,
+    327.8870849609375,
+    311.1480407714844,
+    57.87489318847656,
+    59.43294906616211,
+    50.55767059326172,
+    27.027706146240234,
+    13.878527641296387,
+    10.90867805480957,
+    14.719537734985352,
+    14.719537734985352,
+    14.719537734985352,
+    14.719537734985352,
+    14.719537734985352,
+    14.719537734985352,
+    3.2984049320220947,
+    3.6471333503723145,
+    4.528029441833496,
+    6.537323951721191,
+    7.542074203491211,
+    6.777342796325684,
+    131.77452087402344,
+    133.4365692138672,
+    134.30154418945312,
+    137.3269500732422,
+    143.42730712890625,
+    147.23556518554688,
+    106.86735534667969,
+    106.61359405517578,
+    99.10890197753906,
+    93.6629638671875,
+    95.26378631591797,
+    95.61508178710938,
+    88.6065444946289,
+    62.81496047973633,
+    45.79389190673828,
+    31.114017486572266,
+    18.991355895996094,
+    19.752763748168945
+  ],
+  "soil_vars": [
+    "bdod_mean_0-5cm",
+    "bdod_mean_5-15cm",
+    "bdod_mean_15-30cm",
+    "bdod_mean_30-60cm",
+    "bdod_mean_60-100cm",
+    "bdod_mean_100-200cm",
+    "cec_mean_0-5cm",
+    "cec_mean_5-15cm",
+    "cec_mean_15-30cm",
+    "cec_mean_30-60cm",
+    "cec_mean_60-100cm",
+    "cec_mean_100-200cm",
+    "cfvo_mean_0-5cm",
+    "cfvo_mean_5-15cm",
+    "cfvo_mean_15-30cm",
+    "cfvo_mean_30-60cm",
+    "cfvo_mean_60-100cm",
+    "cfvo_mean_100-200cm",
+    "clay_mean_0-5cm",
+    "clay_mean_5-15cm",
+    "clay_mean_15-30cm",
+    "clay_mean_30-60cm",
+    "clay_mean_60-100cm",
+    "clay_mean_100-200cm",
+    "nitrogen_mean_0-5cm",
+    "nitrogen_mean_5-15cm",
+    "nitrogen_mean_15-30cm",
+    "nitrogen_mean_30-60cm",
+    "nitrogen_mean_60-100cm",
+    "nitrogen_mean_100-200cm",
+    "ocd_mean_0-5cm",
+    "ocd_mean_5-15cm",
+    "ocd_mean_15-30cm",
+    "ocd_mean_30-60cm",
+    "ocd_mean_60-100cm",
+    "ocd_mean_100-200cm",
+    "ocs_mean_0-5cm",
+    "ocs_mean_5-15cm",
+    "ocs_mean_15-30cm",
+    "ocs_mean_30-60cm",
+    "ocs_mean_60-100cm",
+    "ocs_mean_100-200cm",
+    "phh2o_mean_0-5cm",
+    "phh2o_mean_5-15cm",
+    "phh2o_mean_15-30cm",
+    "phh2o_mean_30-60cm",
+    "phh2o_mean_60-100cm",
+    "phh2o_mean_100-200cm",
+    "sand_mean_0-5cm",
+    "sand_mean_5-15cm",
+    "sand_mean_15-30cm",
+    "sand_mean_30-60cm",
+    "sand_mean_60-100cm",
+    "sand_mean_100-200cm",
+    "silt_mean_0-5cm",
+    "silt_mean_5-15cm",
+    "silt_mean_15-30cm",
+    "silt_mean_30-60cm",
+    "silt_mean_60-100cm",
+    "silt_mean_100-200cm",
+    "soc_mean_0-5cm",
+    "soc_mean_5-15cm",
+    "soc_mean_15-30cm",
+    "soc_mean_30-60cm",
+    "soc_mean_60-100cm",
+    "soc_mean_100-200cm"
+  ],
+  "train_cutoffs": [
+    20,
+    24,
+    28,
+    32,
+    36,
+    40,
+    44,
+    48,
+    52
+  ],
+  "transformers_version": "5.13.0",
+  "use_crop": true,
+  "w_mean": [
+    2.6932172775268555,
+    319.30352783203125,
+    5.36029577255249,
+    15.82466983795166,
+    3.981907606124878,
+    1018.5076904296875
+  ],
+  "w_std": [
+    3.367875814437866,
+    91.14374542236328,
+    13.447235107421875,
+    11.749135971069336,
+    11.175808906555176,
+    684.6181030273438
+  ],
+  "weather_vars": [
+    "prcp",
+    "srad",
+    "swe",
+    "tmax",
+    "tmin",
+    "vp"
+  ],
+  "y_mean": 166.3075714111328,
+  "y_std": 30.291048049926758,
+  "auto_map": {
+    "AutoConfig": "configuration_yield.YieldConfig",
+    "AutoModel": "modeling_yield.YieldForRegression"
+  },
+  "custom_pipelines": {
+    "yield-estimation": {
+      "impl": "pipeline_yield.YieldEstimationPipeline",
+      "pt": [
+        "AutoModel"
+      ]
+    }
+  }
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e4b7d06037041ac9e734375b88d384eaee73ee22c87ce016606facf29e85425
+size 733108

pipeline_yield.py

@@ -0,0 +1,329 @@
+import numpy as np
+import torch
+from transformers import Pipeline
+
+
+class YieldEstimationPipeline(Pipeline):
+    """
+    Hugging Face pipeline for county-level corn yield estimation.
+
+    Inputs:
+      - 52 weekly observations for each configured weather variable
+      - static soil properties
+      - crop = "corn"
+      - optional cutoff
+
+    Weather values are already weekly in the public Corn Belt dataset.
+    No daily-to-weekly aggregation or interpolation is performed.
+    """
+
+    def _sanitize_parameters(self, cutoff=None, **kwargs):
+        preprocess_kwargs = {}
+
+        if cutoff is not None:
+            preprocess_kwargs["cutoff"] = cutoff
+
+        return preprocess_kwargs, {}, {}
+
+    def preprocess(self, inputs, cutoff=None):
+        cfg = self.model.config
+
+        if not isinstance(inputs, dict):
+            raise ValueError("Input must be a dictionary.")
+
+        if "weather" not in inputs:
+            raise ValueError(
+                "Input must contain a 'weather' dictionary."
+            )
+
+        if "soil" not in inputs:
+            raise ValueError(
+                "Input must contain a 'soil' dictionary."
+            )
+
+        # ----------------------------------------------------
+        # Crop
+        # ----------------------------------------------------
+
+        crop = str(
+            inputs.get("crop", "corn")
+        ).strip().lower()
+
+        if crop != "corn":
+            raise ValueError(
+                "This model was trained only for corn. "
+                f"Received crop='{crop}'."
+            )
+
+        crop_id = 0
+
+        # ----------------------------------------------------
+        # Weather
+        # ----------------------------------------------------
+
+        weather_dict = inputs["weather"]
+
+        weather_cols = []
+
+        for v in cfg.weather_vars:
+
+            if v not in weather_dict:
+                raise ValueError(
+                    f"Missing weather variable: '{v}'"
+                )
+
+            arr = np.asarray(
+                weather_dict[v],
+                dtype=np.float32,
+            )
+
+            if arr.ndim != 1:
+                raise ValueError(
+                    f"Weather variable '{v}' must be a "
+                    f"1D weekly sequence. Got shape {arr.shape}."
+                )
+
+            if len(arr) != int(cfg.K):
+                raise ValueError(
+                    f"Weather variable '{v}' must contain "
+                    f"{cfg.K} weekly values. Got {len(arr)}."
+                )
+
+            weather_cols.append(arr)
+
+        # [K, W] = [52, 6]
+        weather = np.stack(
+            weather_cols,
+            axis=1,
+        ).astype(np.float32)
+
+        if weather.shape != (
+            int(cfg.K),
+            int(cfg.W),
+        ):
+            raise ValueError(
+                f"Weather shape mismatch. "
+                f"Expected ({cfg.K}, {cfg.W}), "
+                f"got {weather.shape}."
+            )
+
+        # ----------------------------------------------------
+        # Cutoff
+        # ----------------------------------------------------
+
+        requested_cutoff = inputs.get(
+            "cutoff",
+            cutoff,
+        )
+
+        if requested_cutoff is None:
+            requested_cutoff = max(
+                cfg.eval_cutoffs
+                if cfg.eval_cutoffs
+                else [cfg.K]
+            )
+
+        requested_cutoff = int(
+            requested_cutoff
+        )
+
+        allowed_cutoffs = (
+            [int(x) for x in cfg.eval_cutoffs]
+            if cfg.eval_cutoffs
+            else [int(cfg.K)]
+        )
+
+        if requested_cutoff not in allowed_cutoffs:
+            raise ValueError(
+                f"Unsupported cutoff={requested_cutoff}. "
+                f"This released model was evaluated at "
+                f"{allowed_cutoffs}."
+            )
+
+        t_eff = requested_cutoff
+
+        # ----------------------------------------------------
+        # Soil
+        # ----------------------------------------------------
+
+        soil_dict = inputs["soil"]
+
+        soil = []
+
+        for v in cfg.soil_vars:
+
+            if v not in soil_dict:
+                raise ValueError(
+                    f"Missing soil variable: '{v}'"
+                )
+
+            soil.append(
+                float(soil_dict[v])
+            )
+
+        soil = np.asarray(
+            soil,
+            dtype=np.float32,
+        )
+
+        if len(soil) != int(cfg.S):
+            raise ValueError(
+                f"Soil feature mismatch. "
+                f"Expected {cfg.S}, got {len(soil)}."
+            )
+
+        # ----------------------------------------------------
+        # Training normalization
+        # ----------------------------------------------------
+
+        w_mean = np.asarray(
+            cfg.w_mean,
+            dtype=np.float32,
+        )
+
+        w_std = np.asarray(
+            cfg.w_std,
+            dtype=np.float32,
+        )
+
+        s_mean = np.asarray(
+            cfg.s_mean,
+            dtype=np.float32,
+        )
+
+        s_std = np.asarray(
+            cfg.s_std,
+            dtype=np.float32,
+        )
+
+        if len(w_mean) != int(cfg.W):
+            raise ValueError(
+                f"w_mean has {len(w_mean)} values, "
+                f"but model W={cfg.W}."
+            )
+
+        if len(w_std) != int(cfg.W):
+            raise ValueError(
+                f"w_std has {len(w_std)} values, "
+                f"but model W={cfg.W}."
+            )
+
+        if len(s_mean) != int(cfg.S):
+            raise ValueError(
+                f"s_mean has {len(s_mean)} values, "
+                f"but model S={cfg.S}."
+            )
+
+        if len(s_std) != int(cfg.S):
+            raise ValueError(
+                f"s_std has {len(s_std)} values, "
+                f"but model S={cfg.S}."
+            )
+
+        # Impute NaNs with training means.
+        weather = np.where(
+            np.isnan(weather),
+            w_mean[None, :],
+            weather,
+        )
+
+        soil = np.where(
+            np.isnan(soil),
+            s_mean,
+            soil,
+        )
+
+        # Normalize with TRAINING statistics.
+        weather = (
+            weather - w_mean[None, :]
+        ) / w_std[None, :]
+
+        soil = (
+            soil - s_mean
+        ) / s_std
+
+        # Use only data through requested calendar-week cutoff.
+        weather = weather[:t_eff]
+
+        return {
+            "weather": torch.from_numpy(
+                weather
+            ).unsqueeze(0),
+
+            "soil": torch.from_numpy(
+                soil
+            ).unsqueeze(0),
+
+            "crop_id": torch.tensor(
+                [crop_id],
+                dtype=torch.long,
+            ),
+
+            "horizon_idx": t_eff,
+
+            "cutoff": requested_cutoff,
+
+            "crop": crop,
+
+            "weather_format": "weekly",
+
+            "weather_shape": list(
+                weather.shape
+            ),
+        }
+
+    def _forward(self, model_inputs):
+        cutoff = model_inputs.pop(
+            "cutoff"
+        )
+
+        crop = model_inputs.pop(
+            "crop"
+        )
+
+        weather_format = model_inputs.pop(
+            "weather_format"
+        )
+
+        weather_shape = model_inputs.pop(
+            "weather_shape"
+        )
+
+        outputs = self.model(
+            weather=model_inputs["weather"],
+            soil=model_inputs["soil"],
+            crop_id=model_inputs["crop_id"],
+            horizon_idx=model_inputs["horizon_idx"],
+            causal=True,
+            return_sequence=False,
+        )
+
+        return {
+            "outputs": outputs,
+            "cutoff": cutoff,
+            "crop": crop,
+            "weather_format": weather_format,
+            "weather_shape": weather_shape,
+        }
+
+    def postprocess(self, model_outputs):
+        pred = float(
+            model_outputs[
+                "outputs"
+            ].predictions.detach().cpu().item()
+        )
+
+        return {
+            "predicted_yield": pred,
+            "yield_units": "bu/acre",
+            "cutoff": int(
+                model_outputs["cutoff"]
+            ),
+            "crop": model_outputs["crop"],
+            "weather_format": model_outputs[
+                "weather_format"
+            ],
+            "weather_shape": model_outputs[
+                "weather_shape"
+            ],
+        }

sample_input_weekly.json

@@ -0,0 +1,399 @@
+{
+  "crop": "corn",
+  "weather_format": "weekly",
+  "cutoff": 52,
+  "weather": {
+    "prcp": [
+      0.0,
+      0.701298713684082,
+      0.4545454680919647,
+      2.155844211578369,
+      0.9350649118423462,
+      0.11688311398029327,
+      0.0,
+      3.753246784210205,
+      0.15584415197372437,
+      2.2337663173675537,
+      0.0,
+      0.4285714328289032,
+      2.402597427368164,
+      2.3896102905273438,
+      5.753246784210205,
+      0.9220778942108154,
+      0.0,
+      1.1298701763153076,
+      1.2077921628952026,
+      1.2207791805267334,
+      2.441558361053467,
+      5.402597427368164,
+      2.909090995788574,
+      3.6883115768432617,
+      5.1688313484191895,
+      5.233766078948975,
+      7.389610290527344,
+      4.883116722106934,
+      1.5194804668426514,
+      0.8311688303947449,
+      4.051948070526123,
+      0.03896103799343109,
+      0.0,
+      5.350649356842041,
+      2.0389609336853027,
+      4.103896141052246,
+      0.012987012974917889,
+      15.49350643157959,
+      1.1168831586837769,
+      3.142857074737549,
+      6.155844211578369,
+      0.3246753215789795,
+      1.6753246784210205,
+      0.5714285969734192,
+      2.8831169605255127,
+      0.10389610379934311,
+      0.9610389471054077,
+      1.1168831586837769,
+      0.0,
+      0.0,
+      0.051948051899671555,
+      4.181818008422852
+    ],
+    "srad": [
+      187.0129852294922,
+      181.23635864257812,
+      184.8935089111328,
+      179.7818145751953,
+      255.91688537597656,
+      267.4701232910156,
+      323.158447265625,
+      282.09869384765625,
+      347.76104736328125,
+      274.28570556640625,
+      384.70648193359375,
+      332.7168884277344,
+      384.0831298828125,
+      455.7298583984375,
+      359.5636291503906,
+      525.6727294921875,
+      584.2701416015625,
+      522.7221069335938,
+      385.9532470703125,
+      420.5298767089844,
+      386.36883544921875,
+      335.12725830078125,
+      327.6467590332031,
+      329.35064697265625,
+      319.7090759277344,
+      341.6103820800781,
+      374.150634765625,
+      387.20001220703125,
+      379.0545349121094,
+      412.1766357421875,
+      353.08050537109375,
+      397.0493469238281,
+      376.4363708496094,
+      312.2701416015625,
+      312.4363708496094,
+      313.7662353515625,
+      341.07012939453125,
+      249.10130310058594,
+      292.32208251953125,
+      190.7532501220703,
+      189.13246154785156,
+      301.589599609375,
+      214.69090270996094,
+      198.7324676513672,
+      148.0311737060547,
+      213.77662658691406,
+      168.64414978027344,
+      116.23896026611328,
+      148.4051971435547,
+      169.22596740722656,
+      130.86753845214844,
+      157.7818145751953
+    ],
+    "swe": [
+      18.909090042114258,
+      18.753246307373047,
+      23.220779418945312,
+      32.93506622314453,
+      37.24675369262695,
+      41.45454406738281,
+      41.45454406738281,
+      51.22077941894531,
+      67.84415435791016,
+      62.181819915771484,
+      61.610389709472656,
+      59.844154357910156,
+      54.80519485473633,
+      71.94805145263672,
+      78.9610366821289,
+      100.98701477050781,
+      84.467529296875,
+      50.54545593261719,
+      7.636363506317139,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      1.402597427368164,
+      0.0,
+      0.0,
+      0.0,
+      9.454545021057129
+    ],
+    "tmax": [
+      -13.350648880004883,
+      -3.694805145263672,
+      -2.4740259647369385,
+      -0.1428571492433548,
+      -7.298701286315918,
+      -10.746753692626953,
+      -1.3701298236846924,
+      -2.844155788421631,
+      5.285714149475098,
+      0.7272727489471436,
+      4.675324440002441,
+      3.948051929473877,
+      4.324675559997559,
+      -0.4285714328289032,
+      3.9285714626312256,
+      8.292207717895508,
+      16.785715103149414,
+      23.915584564208984,
+      19.662338256835938,
+      25.34415626525879,
+      28.39610481262207,
+      28.39610481262207,
+      24.863636016845703,
+      27.415584564208984,
+      25.766233444213867,
+      28.10389518737793,
+      28.441558837890625,
+      30.61688232421875,
+      26.87013053894043,
+      26.402597427368164,
+      26.571428298950195,
+      29.324674606323242,
+      29.714284896850586,
+      25.02597427368164,
+      25.441558837890625,
+      23.467533111572266,
+      28.746753692626953,
+      20.162338256835938,
+      15.733766555786133,
+      12.071428298950195,
+      8.357142448425293,
+      12.409090995788574,
+      11.629870414733887,
+      9.253246307373047,
+      -0.012987012974917889,
+      -0.6623376607894897,
+      0.9155844449996948,
+      -2.0,
+      -3.344155788421631,
+      0.5064935088157654,
+      1.7402597665786743,
+      -2.0568182468414307
+    ],
+    "tmin": [
+      -23.694805145263672,
+      -14.48051929473877,
+      -12.051947593688965,
+      -7.941558361053467,
+      -19.376623153686523,
+      -21.123376846313477,
+      -14.344156265258789,
+      -14.02597427368164,
+      -7.7207794189453125,
+      -7.7272725105285645,
+      -6.344155788421631,
+      -4.116883277893066,
+      -5.246753215789795,
+      -11.681818008422852,
+      -3.909090995788574,
+      -3.649350643157959,
+      1.4220778942108154,
+      8.98051929473877,
+      8.870129585266113,
+      10.454545021057129,
+      15.129870414733887,
+      16.363636016845703,
+      14.461038589477539,
+      17.175325393676758,
+      16.701297760009766,
+      18.720779418945312,
+      17.922077178955078,
+      19.519479751586914,
+      16.305194854736328,
+      14.642857551574707,
+      15.694805145263672,
+      17.422077178955078,
+      18.220779418945312,
+      15.175324440002441,
+      15.558441162109375,
+      13.49350643157959,
+      17.201297760009766,
+      11.655843734741211,
+      5.7727274894714355,
+      4.551948070526123,
+      2.2272727489471436,
+      -0.16233766078948975,
+      1.7727272510528564,
+      1.0324674844741821,
+      -5.954545497894287,
+      -10.149351119995117,
+      -6.623376846313477,
+      -7.201298713684082,
+      -10.422078132629395,
+      -8.344156265258789,
+      -4.707792282104492,
+      -9.943181991577148
+    ],
+    "vp": [
+      100.2597427368164,
+      259.22076416015625,
+      287.2727355957031,
+      344.41558837890625,
+      137.14285278320312,
+      115.84415435791016,
+      211.42857360839844,
+      215.06494140625,
+      359.48052978515625,
+      358.4415588378906,
+      377.6623229980469,
+      454.5454406738281,
+      409.8701171875,
+      249.87013244628906,
+      451.4285583496094,
+      438.4415588378906,
+      540.7792358398438,
+      887.792236328125,
+      1041.0389404296875,
+      1108.052001953125,
+      1534.54541015625,
+      1750.6492919921875,
+      1589.6103515625,
+      1931.4285888671875,
+      1902.337646484375,
+      2178.181884765625,
+      2062.857177734375,
+      2275.32470703125,
+      1851.947998046875,
+      1670.1298828125,
+      1802.077880859375,
+      1990.1298828125,
+      2097.662353515625,
+      1738.1817626953125,
+      1810.9090576171875,
+      1587.012939453125,
+      2006.7532958984375,
+      1414.0260009765625,
+      942.337646484375,
+      852.467529296875,
+      731.9480590820312,
+      614.5454711914062,
+      715.8441772460938,
+      658.7012939453125,
+      416.1038818359375,
+      297.1428527832031,
+      392.7272644042969,
+      367.7922058105469,
+      289.35064697265625,
+      336.6233825683594,
+      435.3246765136719,
+      295.4545593261719
+    ]
+  },
+  "soil": {
+    "bdod_mean_0-5cm": 127.71450805664062,
+    "bdod_mean_5-15cm": 148.86965942382812,
+    "bdod_mean_15-30cm": 150.70364379882812,
+    "bdod_mean_30-60cm": 158.2148895263672,
+    "bdod_mean_60-100cm": 162.8789825439453,
+    "bdod_mean_100-200cm": 165.12025451660156,
+    "cec_mean_0-5cm": 252.34054565429688,
+    "cec_mean_5-15cm": 174.4131317138672,
+    "cec_mean_15-30cm": 142.52432250976562,
+    "cec_mean_30-60cm": 129.1745147705078,
+    "cec_mean_60-100cm": 119.97528839111328,
+    "cec_mean_100-200cm": 103.29729461669922,
+    "cfvo_mean_0-5cm": 21.431659698486328,
+    "cfvo_mean_5-15cm": 22.905019760131836,
+    "cfvo_mean_15-30cm": 21.820077896118164,
+    "cfvo_mean_30-60cm": 24.725868225097656,
+    "cfvo_mean_60-100cm": 34.54826354980469,
+    "cfvo_mean_100-200cm": 45.90888214111328,
+    "clay_mean_0-5cm": 150.02239990234375,
+    "clay_mean_5-15cm": 149.1436309814453,
+    "clay_mean_15-30cm": 150.81776428222656,
+    "clay_mean_30-60cm": 150.43707275390625,
+    "clay_mean_60-100cm": 134.29266357421875,
+    "clay_mean_100-200cm": 125.47721862792969,
+    "nitrogen_mean_0-5cm": 4723.78857421875,
+    "nitrogen_mean_5-15cm": 2148.763671875,
+    "nitrogen_mean_15-30cm": 1729.8802490234375,
+    "nitrogen_mean_30-60cm": 776.644775390625,
+    "nitrogen_mean_60-100cm": 547.14208984375,
+    "nitrogen_mean_100-200cm": 534.0015258789062,
+    "ocd_mean_0-5cm": 435.4494323730469,
+    "ocd_mean_5-15cm": 255.3752899169922,
+    "ocd_mean_15-30cm": 181.84786987304688,
+    "ocd_mean_30-60cm": 93.57837677001953,
+    "ocd_mean_60-100cm": 60.80463409423828,
+    "ocd_mean_100-200cm": 44.90888214111328,
+    "ocs_mean_0-5cm": 57.79845428466797,
+    "ocs_mean_5-15cm": 57.79845428466797,
+    "ocs_mean_15-30cm": 57.79845428466797,
+    "ocs_mean_30-60cm": 57.79845428466797,
+    "ocs_mean_60-100cm": 57.79845428466797,
+    "ocs_mean_100-200cm": 57.79845428466797,
+    "phh2o_mean_0-5cm": 61.812355041503906,
+    "phh2o_mean_5-15cm": 61.73050308227539,
+    "phh2o_mean_15-30cm": 61.63166046142578,
+    "phh2o_mean_30-60cm": 62.09111785888672,
+    "phh2o_mean_60-100cm": 64.0540542602539,
+    "phh2o_mean_100-200cm": 69.0687255859375,
+    "sand_mean_0-5cm": 625.2216186523438,
+    "sand_mean_5-15cm": 626.4702758789062,
+    "sand_mean_15-30cm": 631.22314453125,
+    "sand_mean_30-60cm": 649.113525390625,
+    "sand_mean_60-100cm": 697.4996337890625,
+    "sand_mean_100-200cm": 712.781494140625,
+    "silt_mean_0-5cm": 224.77297973632812,
+    "silt_mean_5-15cm": 224.3876495361328,
+    "silt_mean_15-30cm": 217.95135498046875,
+    "silt_mean_30-60cm": 200.43319702148438,
+    "silt_mean_60-100cm": 168.21775817871094,
+    "silt_mean_100-200cm": 161.7498016357422,
+    "soc_mean_0-5cm": 446.14508056640625,
+    "soc_mean_5-15cm": 223.34988403320312,
+    "soc_mean_15-30cm": 144.84561157226562,
+    "soc_mean_30-60cm": 107.50271606445312,
+    "soc_mean_60-100cm": 53.54926300048828,
+    "soc_mean_100-200cm": 47.3972053527832
+  }
+}

sample_test.py

@@ -0,0 +1,59 @@
+import json
+
+from transformers import pipeline
+
+
+MODEL_DIR = "."
+
+
+with open(
+    "sample_input_weekly.json",
+    "r",
+) as f:
+    sample = json.load(f)
+
+
+pipe = pipeline(
+    "yield-estimation",
+    model=MODEL_DIR,
+    trust_remote_code=True,
+)
+
+
+print("Testing released cutoffs...\n")
+
+for cutoff in [
+    20,
+    24,
+    28,
+    32,
+    36,
+    40,
+    44,
+    48,
+    52,
+]:
+    x = dict(sample)
+    x["cutoff"] = cutoff
+
+    output = pipe(x)
+
+    print(
+        f"Cutoff {cutoff}: "
+        f"{output['predicted_yield']:.2f} "
+        f"{output['yield_units']}"
+    )
+
+
+print("\nFull output at cutoff 52:")
+
+sample["cutoff"] = 52
+
+output = pipe(sample)
+
+print(
+    json.dumps(
+        output,
+        indent=2,
+    )
+)