shaped 2.0.1__py3-none-any.whl → 2.0.4__py3-none-any.whl

shaped/autogen/models/time_window_in_days.py

@@ -0,0 +1,154 @@
+# coding: utf-8
+
+"""
+    Shaped API
+
+    Welcome to Shaped's API reference docs. These provide a  detailed view of the endpoints and CLI commands that Shaped provides and brief  explanations of how they should be used.  The Shaped API has four main endpoints:  **Tables** - Provision and manage batch and real-time data connectors.   **Views** - Configure SQL transformations and AI enrichment on your input data. Use SQL to combine multiple data sources or use an LLM to add new categories, extract specific attributes from descriptions, etc.  **Engines** - Deploy and manage your relevance engines. The Engine API exposes configuration for indexing logic, input datasets, externam embeddings, and more.  **Query** - Execute queries against your engines, to return data based on an input query or rerank an existing list. The Query API exposes the retrieve, filter, score, and ranking steps of the 4-stage ranking architecture.  The base URL for each endpoint is: `https://api.shaped.ai/v2`
+
+    The version of the OpenAPI document: 2.0.0
+    Generated by OpenAPI Generator (https://openapi-generator.tech)
+
+    Do not edit the class manually.
+"""  # noqa: E501
+
+
+from __future__ import annotations
+
+import json
+import pprint
+import re  # noqa: F401
+from typing import TYPE_CHECKING, Any, Dict, Optional, Set, Union
+
+from pydantic import BaseModel, StrictInt, ValidationError, field_validator
+from shaped.autogen.models.tunable_int import TunableInt
+from typing_extensions import Self
+
+TIMEWINDOWINDAYS_ANY_OF_SCHEMAS = ["TunableInt", "int"]
+
+
+class TimeWindowInDays(BaseModel):
+    """
+    TimeWindowInDays
+    """
+
+    # data type: int
+    anyof_schema_1_validator: Optional[StrictInt] = None
+    # data type: TunableInt
+    anyof_schema_2_validator: Optional[TunableInt] = None
+    if TYPE_CHECKING:
+        actual_instance: Optional[Union[TunableInt, int]] = None
+    else:
+        actual_instance: Any = None
+    any_of_schemas: Set[str] = {"TunableInt", "int"}
+
+    model_config = {
+        "validate_assignment": True,
+        "protected_namespaces": (),
+    }
+
+    def __init__(self, *args, **kwargs) -> None:
+        if args:
+            if len(args) > 1:
+                raise ValueError(
+                    "If a position argument is used, only 1 is allowed to set `actual_instance`"
+                )
+            if kwargs:
+                raise ValueError(
+                    "If a position argument is used, keyword arguments cannot be used."
+                )
+            super().__init__(actual_instance=args[0])
+        else:
+            super().__init__(**kwargs)
+
+    @field_validator("actual_instance")
+    def actual_instance_must_validate_anyof(cls, v):
+        if v is None:
+            return v
+
+        instance = TimeWindowInDays.model_construct()
+        error_messages = []
+        # validate data type: int
+        try:
+            instance.anyof_schema_1_validator = v
+            return v
+        except (ValidationError, ValueError) as e:
+            error_messages.append(str(e))
+        # validate data type: TunableInt
+        if not isinstance(v, TunableInt):
+            error_messages.append(f"Error! Input type `{type(v)}` is not `TunableInt`")
+        else:
+            return v
+
+        if error_messages:
+            # no match
+            raise ValueError(
+                "No match found when setting the actual_instance in TimeWindowInDays with anyOf schemas: TunableInt, int. Details: "
+                + ", ".join(error_messages)
+            )
+        else:
+            return v
+
+    @classmethod
+    def from_dict(cls, obj: Dict[str, Any]) -> Self:
+        return cls.from_json(json.dumps(obj))
+
+    @classmethod
+    def from_json(cls, json_str: str) -> Self:
+        """Returns the object represented by the json string"""
+        instance = cls.model_construct()
+        if json_str is None:
+            return instance
+
+        error_messages = []
+        # deserialize data into int
+        try:
+            # validation
+            instance.anyof_schema_1_validator = json.loads(json_str)
+            # assign value to actual_instance
+            instance.actual_instance = instance.anyof_schema_1_validator
+            return instance
+        except (ValidationError, ValueError) as e:
+            error_messages.append(str(e))
+        # anyof_schema_2_validator: Optional[TunableInt] = None
+        try:
+            instance.actual_instance = TunableInt.from_json(json_str)
+            return instance
+        except (ValidationError, ValueError) as e:
+            error_messages.append(str(e))
+
+        if error_messages:
+            # no match
+            raise ValueError(
+                "No match found when deserializing the JSON string into TimeWindowInDays with anyOf schemas: TunableInt, int. Details: "
+                + ", ".join(error_messages)
+            )
+        else:
+            return instance
+
+    def to_json(self) -> str:
+        """Returns the JSON representation of the actual instance"""
+        if self.actual_instance is None:
+            return "null"
+
+        if hasattr(self.actual_instance, "to_json") and callable(
+            self.actual_instance.to_json
+        ):
+            return self.actual_instance.to_json()
+        else:
+            return json.dumps(self.actual_instance)
+
+    def to_dict(self) -> Optional[Union[Dict[str, Any], TunableInt, int]]:
+        """Returns the dict representation of the actual instance"""
+        if self.actual_instance is None:
+            return None
+
+        if hasattr(self.actual_instance, "to_dict") and callable(
+            self.actual_instance.to_dict
+        ):
+            return self.actual_instance.to_dict()
+        else:
+            return self.actual_instance
+
+    def to_str(self) -> str:
+        """Returns the string representation of the actual instance"""
+        return pprint.pformat(self.model_dump())

shaped/autogen/models/tokenizer.py

@@ -0,0 +1,149 @@
+# coding: utf-8
+
+"""
+    Shaped API
+
+    Welcome to Shaped's API reference docs. These provide a  detailed view of the endpoints and CLI commands that Shaped provides and brief  explanations of how they should be used.  The Shaped API has four main endpoints:  **Tables** - Provision and manage batch and real-time data connectors.   **Views** - Configure SQL transformations and AI enrichment on your input data. Use SQL to combine multiple data sources or use an LLM to add new categories, extract specific attributes from descriptions, etc.  **Engines** - Deploy and manage your relevance engines. The Engine API exposes configuration for indexing logic, input datasets, externam embeddings, and more.  **Query** - Execute queries against your engines, to return data based on an input query or rerank an existing list. The Query API exposes the retrieve, filter, score, and ranking steps of the 4-stage ranking architecture.  The base URL for each endpoint is: `https://api.shaped.ai/v2`
+
+    The version of the OpenAPI document: 2.0.4
+    Generated by OpenAPI Generator (https://openapi-generator.tech)
+
+    Do not edit the class manually.
+"""  # noqa: E501
+
+
+from __future__ import annotations
+from inspect import getfullargspec
+import json
+import pprint
+import re  # noqa: F401
+from pydantic import BaseModel, ConfigDict, Field, StrictStr, ValidationError, field_validator
+from typing import Optional
+from shaped.autogen.models.ngram_tokenizer import NgramTokenizer
+from shaped.autogen.models.stemmer_tokenizer import StemmerTokenizer
+from shaped.autogen.models.whitespace_tokenizer import WhitespaceTokenizer
+from typing import Union, Any, List, Set, TYPE_CHECKING, Optional, Dict
+from typing_extensions import Literal, Self
+from pydantic import Field
+
+TOKENIZER_ANY_OF_SCHEMAS = ["NgramTokenizer", "StemmerTokenizer", "WhitespaceTokenizer"]
+
+class Tokenizer(BaseModel):
+    """
+    The tokenizer to use for text indexing.
+    """
+
+    # data type: StemmerTokenizer
+    anyof_schema_1_validator: Optional[StemmerTokenizer] = None
+    # data type: NgramTokenizer
+    anyof_schema_2_validator: Optional[NgramTokenizer] = None
+    # data type: WhitespaceTokenizer
+    anyof_schema_3_validator: Optional[WhitespaceTokenizer] = None
+    if TYPE_CHECKING:
+        actual_instance: Optional[Union[NgramTokenizer, StemmerTokenizer, WhitespaceTokenizer]] = None
+    else:
+        actual_instance: Any = None
+    any_of_schemas: Set[str] = { "NgramTokenizer", "StemmerTokenizer", "WhitespaceTokenizer" }
+
+    model_config = {
+        "validate_assignment": True,
+        "protected_namespaces": (),
+    }
+
+    def __init__(self, *args, **kwargs) -> None:
+        if args:
+            if len(args) > 1:
+                raise ValueError("If a position argument is used, only 1 is allowed to set `actual_instance`")
+            if kwargs:
+                raise ValueError("If a position argument is used, keyword arguments cannot be used.")
+            super().__init__(actual_instance=args[0])
+        else:
+            super().__init__(**kwargs)
+
+    @field_validator('actual_instance')
+    def actual_instance_must_validate_anyof(cls, v):
+        instance = Tokenizer.model_construct()
+        error_messages = []
+        # validate data type: StemmerTokenizer
+        if not isinstance(v, StemmerTokenizer):
+            error_messages.append(f"Error! Input type `{type(v)}` is not `StemmerTokenizer`")
+        else:
+            return v
+
+        # validate data type: NgramTokenizer
+        if not isinstance(v, NgramTokenizer):
+            error_messages.append(f"Error! Input type `{type(v)}` is not `NgramTokenizer`")
+        else:
+            return v
+
+        # validate data type: WhitespaceTokenizer
+        if not isinstance(v, WhitespaceTokenizer):
+            error_messages.append(f"Error! Input type `{type(v)}` is not `WhitespaceTokenizer`")
+        else:
+            return v
+
+        if error_messages:
+            # no match
+            raise ValueError("No match found when setting the actual_instance in Tokenizer with anyOf schemas: NgramTokenizer, StemmerTokenizer, WhitespaceTokenizer. Details: " + ", ".join(error_messages))
+        else:
+            return v
+
+    @classmethod
+    def from_dict(cls, obj: Dict[str, Any]) -> Self:
+        return cls.from_json(json.dumps(obj))
+
+    @classmethod
+    def from_json(cls, json_str: str) -> Self:
+        """Returns the object represented by the json string"""
+        instance = cls.model_construct()
+        error_messages = []
+        # anyof_schema_1_validator: Optional[StemmerTokenizer] = None
+        try:
+            instance.actual_instance = StemmerTokenizer.from_json(json_str)
+            return instance
+        except (ValidationError, ValueError) as e:
+             error_messages.append(str(e))
+        # anyof_schema_2_validator: Optional[NgramTokenizer] = None
+        try:
+            instance.actual_instance = NgramTokenizer.from_json(json_str)
+            return instance
+        except (ValidationError, ValueError) as e:
+             error_messages.append(str(e))
+        # anyof_schema_3_validator: Optional[WhitespaceTokenizer] = None
+        try:
+            instance.actual_instance = WhitespaceTokenizer.from_json(json_str)
+            return instance
+        except (ValidationError, ValueError) as e:
+             error_messages.append(str(e))
+
+        if error_messages:
+            # no match
+            raise ValueError("No match found when deserializing the JSON string into Tokenizer with anyOf schemas: NgramTokenizer, StemmerTokenizer, WhitespaceTokenizer. Details: " + ", ".join(error_messages))
+        else:
+            return instance
+
+    def to_json(self) -> str:
+        """Returns the JSON representation of the actual instance"""
+        if self.actual_instance is None:
+            return "null"
+
+        if hasattr(self.actual_instance, "to_json") and callable(self.actual_instance.to_json):
+            return self.actual_instance.to_json()
+        else:
+            return json.dumps(self.actual_instance)
+
+    def to_dict(self) -> Optional[Union[Dict[str, Any], NgramTokenizer, StemmerTokenizer, WhitespaceTokenizer]]:
+        """Returns the dict representation of the actual instance"""
+        if self.actual_instance is None:
+            return None
+
+        if hasattr(self.actual_instance, "to_dict") and callable(self.actual_instance.to_dict):
+            return self.actual_instance.to_dict()
+        else:
+            return self.actual_instance
+
+    def to_str(self) -> str:
+        """Returns the string representation of the actual instance"""
+        return pprint.pformat(self.model_dump())
+
+

shaped/autogen/models/trained_model_encoder.py

@@ -0,0 +1,99 @@
+# coding: utf-8
+
+"""
+    Shaped API
+
+    Welcome to Shaped's API reference docs. These provide a  detailed view of the endpoints and CLI commands that Shaped provides and brief  explanations of how they should be used.  The Shaped API has four main endpoints:  **Tables** - Provision and manage batch and real-time data connectors.   **Views** - Configure SQL transformations and AI enrichment on your input data. Use SQL to combine multiple data sources or use an LLM to add new categories, extract specific attributes from descriptions, etc.  **Engines** - Deploy and manage your relevance engines. The Engine API exposes configuration for indexing logic, input datasets, externam embeddings, and more.  **Query** - Execute queries against your engines, to return data based on an input query or rerank an existing list. The Query API exposes the retrieve, filter, score, and ranking steps of the 4-stage ranking architecture.  The base URL for each endpoint is: `https://api.shaped.ai/v2`
+
+    The version of the OpenAPI document: 2.0.4
+    Generated by OpenAPI Generator (https://openapi-generator.tech)
+
+    Do not edit the class manually.
+"""  # noqa: E501
+
+
+from __future__ import annotations
+import pprint
+import re  # noqa: F401
+import json
+
+from pydantic import BaseModel, ConfigDict, Field, StrictStr, field_validator
+from typing import Any, ClassVar, Dict, List, Optional
+from typing import Optional, Set
+from typing_extensions import Self
+
+class TrainedModelEncoder(BaseModel):
+    """
+    TrainedModelEncoder
+    """ # noqa: E501
+    model_ref: StrictStr = Field(description="Name of the trained model to use as encoder.")
+    type: Optional[StrictStr] = Field(default='trained_model', description="Encoder type discriminator.")
+    __properties: ClassVar[List[str]] = ["model_ref", "type"]
+
+    @field_validator('type')
+    def type_validate_enum(cls, value):
+        """Validates the enum"""
+        if value is None:
+            return value
+
+        if value not in set(['trained_model']):
+            raise ValueError("must be one of enum values ('trained_model')")
+        return value
+
+    model_config = ConfigDict(
+        populate_by_name=True,
+        validate_assignment=True,
+        protected_namespaces=(),
+    )
+
+
+    def to_str(self) -> str:
+        """Returns the string representation of the model using alias"""
+        return pprint.pformat(self.model_dump(by_alias=True))
+
+    def to_json(self) -> str:
+        """Returns the JSON representation of the model using alias"""
+        # TODO: pydantic v2: use .model_dump_json(by_alias=True, exclude_unset=True) instead
+        return json.dumps(self.to_dict())
+
+    @classmethod
+    def from_json(cls, json_str: str) -> Optional[Self]:
+        """Create an instance of TrainedModelEncoder from a JSON string"""
+        return cls.from_dict(json.loads(json_str))
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Return the dictionary representation of the model using alias.
+
+        This has the following differences from calling pydantic's
+        `self.model_dump(by_alias=True)`:
+
+        * `None` is only added to the output dict for nullable fields that
+          were set at model initialization. Other fields with value `None`
+          are ignored.
+        """
+        excluded_fields: Set[str] = set([
+        ])
+
+        _dict = self.model_dump(
+            by_alias=True,
+            exclude=excluded_fields,
+            exclude_none=True,
+        )
+        return _dict
+
+    @classmethod
+    def from_dict(cls, obj: Optional[Dict[str, Any]]) -> Optional[Self]:
+        """Create an instance of TrainedModelEncoder from a dict"""
+        if obj is None:
+            return None
+
+        if not isinstance(obj, dict):
+            return cls.model_validate(obj)
+
+        _obj = cls.model_validate({
+            "model_ref": obj.get("model_ref"),
+            "type": obj.get("type") if obj.get("type") is not None else 'trained_model'
+        })
+        return _obj
+
+

shaped/autogen/models/training_compute_config.py

@@ -0,0 +1,99 @@
+# coding: utf-8
+
+"""
+    Shaped API
+
+    Welcome to Shaped's API reference docs. These provide a  detailed view of the endpoints and CLI commands that Shaped provides and brief  explanations of how they should be used.  The Shaped API has four main endpoints:  **Tables** - Provision and manage batch and real-time data connectors.   **Views** - Configure SQL transformations and AI enrichment on your input data. Use SQL to combine multiple data sources or use an LLM to add new categories, extract specific attributes from descriptions, etc.  **Engines** - Deploy and manage your relevance engines. The Engine API exposes configuration for indexing logic, input datasets, externam embeddings, and more.  **Query** - Execute queries against your engines, to return data based on an input query or rerank an existing list. The Query API exposes the retrieve, filter, score, and ranking steps of the 4-stage ranking architecture.  The base URL for each endpoint is: `https://api.shaped.ai/v2`
+
+    The version of the OpenAPI document: 2.0.4
+    Generated by OpenAPI Generator (https://openapi-generator.tech)
+
+    Do not edit the class manually.
+"""  # noqa: E501
+
+
+from __future__ import annotations
+import pprint
+import re  # noqa: F401
+import json
+
+from pydantic import BaseModel, ConfigDict, Field, StrictBool, StrictInt, StrictStr
+from typing import Any, ClassVar, Dict, List, Optional
+from typing import Optional, Set
+from typing_extensions import Self
+
+class TrainingComputeConfig(BaseModel):
+    """
+    Defines the compute resources for a single training job. These settings are used for both regular training and for each individual tuning job.
+    """ # noqa: E501
+    gpu_type: Optional[StrictStr] = Field(default='T4', description="GPU type for training (e.g., T4, L4).")
+    gpu_count: Optional[StrictInt] = Field(default=1, description="Number of GPUs per training job.")
+    cpu_memory_gb: Optional[StrictInt] = Field(default=16, description="CPU memory in GB per training job.")
+    cpu_count: Optional[StrictInt] = Field(default=4, description="Number of CPU cores per training job.")
+    force_gpu: Optional[StrictBool] = Field(default=False, description="Whether to force GPU usage even if not needed.")
+    disk_size_gb: Optional[StrictInt] = Field(default=64, description="Disk size in GB for training job storage.")
+    use_kubetorch: Optional[StrictBool] = Field(default=True, description="Whether to use kubetorch for running remote jobs.")
+    __properties: ClassVar[List[str]] = ["gpu_type", "gpu_count", "cpu_memory_gb", "cpu_count", "force_gpu", "disk_size_gb", "use_kubetorch"]
+
+    model_config = ConfigDict(
+        populate_by_name=True,
+        validate_assignment=True,
+        protected_namespaces=(),
+    )
+
+
+    def to_str(self) -> str:
+        """Returns the string representation of the model using alias"""
+        return pprint.pformat(self.model_dump(by_alias=True))
+
+    def to_json(self) -> str:
+        """Returns the JSON representation of the model using alias"""
+        # TODO: pydantic v2: use .model_dump_json(by_alias=True, exclude_unset=True) instead
+        return json.dumps(self.to_dict())
+
+    @classmethod
+    def from_json(cls, json_str: str) -> Optional[Self]:
+        """Create an instance of TrainingComputeConfig from a JSON string"""
+        return cls.from_dict(json.loads(json_str))
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Return the dictionary representation of the model using alias.
+
+        This has the following differences from calling pydantic's
+        `self.model_dump(by_alias=True)`:
+
+        * `None` is only added to the output dict for nullable fields that
+          were set at model initialization. Other fields with value `None`
+          are ignored.
+        """
+        excluded_fields: Set[str] = set([
+        ])
+
+        _dict = self.model_dump(
+            by_alias=True,
+            exclude=excluded_fields,
+            exclude_none=True,
+        )
+        return _dict
+
+    @classmethod
+    def from_dict(cls, obj: Optional[Dict[str, Any]]) -> Optional[Self]:
+        """Create an instance of TrainingComputeConfig from a dict"""
+        if obj is None:
+            return None
+
+        if not isinstance(obj, dict):
+            return cls.model_validate(obj)
+
+        _obj = cls.model_validate({
+            "gpu_type": obj.get("gpu_type") if obj.get("gpu_type") is not None else 'T4',
+            "gpu_count": obj.get("gpu_count") if obj.get("gpu_count") is not None else 1,
+            "cpu_memory_gb": obj.get("cpu_memory_gb") if obj.get("cpu_memory_gb") is not None else 16,
+            "cpu_count": obj.get("cpu_count") if obj.get("cpu_count") is not None else 4,
+            "force_gpu": obj.get("force_gpu") if obj.get("force_gpu") is not None else False,
+            "disk_size_gb": obj.get("disk_size_gb") if obj.get("disk_size_gb") is not None else 64,
+            "use_kubetorch": obj.get("use_kubetorch") if obj.get("use_kubetorch") is not None else True
+        })
+        return _obj
+
+

shaped/autogen/models/training_config.py

@@ -0,0 +1,121 @@
+# coding: utf-8
+
+"""
+    Shaped API
+
+    Welcome to Shaped's API reference docs. These provide a  detailed view of the endpoints and CLI commands that Shaped provides and brief  explanations of how they should be used.  The Shaped API has four main endpoints:  **Tables** - Provision and manage batch and real-time data connectors.   **Views** - Configure SQL transformations and AI enrichment on your input data. Use SQL to combine multiple data sources or use an LLM to add new categories, extract specific attributes from descriptions, etc.  **Engines** - Deploy and manage your relevance engines. The Engine API exposes configuration for indexing logic, input datasets, externam embeddings, and more.  **Query** - Execute queries against your engines, to return data based on an input query or rerank an existing list. The Query API exposes the retrieve, filter, score, and ranking steps of the 4-stage ranking architecture.  The base URL for each endpoint is: `https://api.shaped.ai/v2`
+
+    The version of the OpenAPI document: 2.0.4
+    Generated by OpenAPI Generator (https://openapi-generator.tech)
+
+    Do not edit the class manually.
+"""  # noqa: E501
+
+
+from __future__ import annotations
+import pprint
+import re  # noqa: F401
+import json
+
+from pydantic import BaseModel, ConfigDict, Field, StrictStr
+from typing import Any, ClassVar, Dict, List, Optional
+from shaped.autogen.models.data_split_config import DataSplitConfig
+from shaped.autogen.models.evaluation_config import EvaluationConfig
+from shaped.autogen.models.models_inner import ModelsInner
+from shaped.autogen.models.training_compute_config import TrainingComputeConfig
+from shaped.autogen.models.tuning_config import TuningConfig
+from typing import Optional, Set
+from typing_extensions import Self
+
+class TrainingConfig(BaseModel):
+    """
+    Defines the entire model training and evaluation workflow.
+    """ # noqa: E501
+    schedule: Optional[StrictStr] = Field(default='@daily', description="Model retraining frequency.")
+    compute: Optional[TrainingComputeConfig] = Field(default=None, description="Compute resources for training jobs.")
+    data_split: Optional[DataSplitConfig] = Field(default=None, description="Configure train/test split.")
+    evaluation: Optional[EvaluationConfig] = Field(default=None, description="Model evaluation strategy configuration.")
+    models: Optional[List[ModelsInner]] = Field(default=None, description="List of scoring policy models to train.")
+    tuning: Optional[TuningConfig] = Field(default=None, description="Hyperparameter tuning configuration.")
+    __properties: ClassVar[List[str]] = ["schedule", "compute", "data_split", "evaluation", "models", "tuning"]
+
+    model_config = ConfigDict(
+        populate_by_name=True,
+        validate_assignment=True,
+        protected_namespaces=(),
+    )
+
+
+    def to_str(self) -> str:
+        """Returns the string representation of the model using alias"""
+        return pprint.pformat(self.model_dump(by_alias=True))
+
+    def to_json(self) -> str:
+        """Returns the JSON representation of the model using alias"""
+        # TODO: pydantic v2: use .model_dump_json(by_alias=True, exclude_unset=True) instead
+        return json.dumps(self.to_dict())
+
+    @classmethod
+    def from_json(cls, json_str: str) -> Optional[Self]:
+        """Create an instance of TrainingConfig from a JSON string"""
+        return cls.from_dict(json.loads(json_str))
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Return the dictionary representation of the model using alias.
+
+        This has the following differences from calling pydantic's
+        `self.model_dump(by_alias=True)`:
+
+        * `None` is only added to the output dict for nullable fields that
+          were set at model initialization. Other fields with value `None`
+          are ignored.
+        """
+        excluded_fields: Set[str] = set([
+        ])
+
+        _dict = self.model_dump(
+            by_alias=True,
+            exclude=excluded_fields,
+            exclude_none=True,
+        )
+        # override the default output from pydantic by calling `to_dict()` of compute
+        if self.compute:
+            _dict['compute'] = self.compute.to_dict()
+        # override the default output from pydantic by calling `to_dict()` of data_split
+        if self.data_split:
+            _dict['data_split'] = self.data_split.to_dict()
+        # override the default output from pydantic by calling `to_dict()` of evaluation
+        if self.evaluation:
+            _dict['evaluation'] = self.evaluation.to_dict()
+        # override the default output from pydantic by calling `to_dict()` of each item in models (list)
+        _items = []
+        if self.models:
+            for _item_models in self.models:
+                if _item_models:
+                    _items.append(_item_models.to_dict())
+            _dict['models'] = _items
+        # override the default output from pydantic by calling `to_dict()` of tuning
+        if self.tuning:
+            _dict['tuning'] = self.tuning.to_dict()
+        return _dict
+
+    @classmethod
+    def from_dict(cls, obj: Optional[Dict[str, Any]]) -> Optional[Self]:
+        """Create an instance of TrainingConfig from a dict"""
+        if obj is None:
+            return None
+
+        if not isinstance(obj, dict):
+            return cls.model_validate(obj)
+
+        _obj = cls.model_validate({
+            "schedule": obj.get("schedule") if obj.get("schedule") is not None else '@daily',
+            "compute": TrainingComputeConfig.from_dict(obj["compute"]) if obj.get("compute") is not None else None,
+            "data_split": DataSplitConfig.from_dict(obj["data_split"]) if obj.get("data_split") is not None else None,
+            "evaluation": EvaluationConfig.from_dict(obj["evaluation"]) if obj.get("evaluation") is not None else None,
+            "models": [ModelsInner.from_dict(_item) for _item in obj["models"]] if obj.get("models") is not None else None,
+            "tuning": TuningConfig.from_dict(obj["tuning"]) if obj.get("tuning") is not None else None
+        })
+        return _obj
+
+