nemo-platform-plugin 0.1.0__py3-none-any.whl

nemo_platform/types/evaluation/context_recall_metric_param_param.py

@@ -0,0 +1,67 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Dict, List, Union
+from typing_extensions import Literal, Required, TypeAlias, TypedDict
+
+from .model_ref import ModelRef
+from .model_param import ModelParam
+from .inference_params_param import InferenceParamsParam
+
+__all__ = ["ContextRecallMetricParamParam", "JudgeModel"]
+
+JudgeModel: TypeAlias = Union[ModelParam, ModelRef]
+
+
+class ContextRecallMetricParamParam(TypedDict, total=False):
+    """Request type for ContextRecall metrics."""
+
+    judge_model: Required[JudgeModel]
+    """The judge model configuration."""
+
+    description: str
+    """Human-readable description of the metric."""
+
+    ignore_request_failure: bool
+    """
+    If True, request failures to the judge model are ignored and the metric result
+    is marked as NaN. Parse/output formatting failures are always converted to NaN.
+    """
+
+    inference: InferenceParamsParam
+    """Parameters for model inference.
+
+    Extra fields can be supplied for additional options applied to the inference
+    request directly. Fields not supported by the model may cause inference errors
+    during evaluation.
+    """
+
+    input_template: Dict[str, object]
+    """Optional Jinja template for rendering the input payload for RAGAS evaluation."""
+
+    labels: Dict[str, str]
+    """Labels are key-value pairs that can be used for grouping and filtering."""
+
+    supported_job_types: List[Literal["online", "offline"]]
+    """
+    A metric can evaluate model outputs for online evaluations or pre-generated
+    outputs for offline evaluations.
+    """
+
+    type: Literal["context_recall"]

nemo_platform/types/evaluation/context_recall_metric_response.py

@@ -0,0 +1,85 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Dict, List, Union, Optional
+from datetime import datetime
+from typing_extensions import Literal, TypeAlias
+
+from .model import Model
+from ..._models import BaseModel
+from .model_ref import ModelRef
+from .inference_params import InferenceParams
+
+__all__ = ["ContextRecallMetricResponse", "JudgeModel"]
+
+JudgeModel: TypeAlias = Union[Model, ModelRef]
+
+
+class ContextRecallMetricResponse(BaseModel):
+    """Response type for ContextRecall metrics."""
+
+    judge_model: JudgeModel
+    """The judge model configuration."""
+
+    id: Optional[str] = None
+    """Entity name within the workspace"""
+
+    created_at: Optional[datetime] = None
+
+    description: Optional[str] = None
+    """Human-readable description of the metric."""
+
+    ignore_request_failure: Optional[bool] = None
+    """
+    If True, request failures to the judge model are ignored and the metric result
+    is marked as NaN. Parse/output formatting failures are always converted to NaN.
+    """
+
+    inference: Optional[InferenceParams] = None
+    """Parameters for model inference.
+
+    Extra fields can be supplied for additional options applied to the inference
+    request directly. Fields not supported by the model may cause inference errors
+    during evaluation.
+    """
+
+    input_template: Optional[Dict[str, object]] = None
+    """Optional Jinja template for rendering the input payload for RAGAS evaluation."""
+
+    labels: Optional[Dict[str, str]] = None
+    """Labels are key-value pairs that can be used for grouping and filtering."""
+
+    name: Optional[str] = None
+    """Entity name within the workspace"""
+
+    parent: Optional[str] = None
+
+    project: Optional[str] = None
+    """The name of the project associated with this entity."""
+
+    supported_job_types: Optional[List[Literal["online", "offline"]]] = None
+    """
+    A metric can evaluate model outputs for online evaluations or pre-generated
+    outputs for offline evaluations.
+    """
+
+    type: Optional[Literal["context_recall"]] = None
+
+    updated_at: Optional[datetime] = None
+
+    workspace: Optional[str] = None
+    """Workspace identifier"""

nemo_platform/types/evaluation/context_relevance_metric.py

@@ -0,0 +1,89 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Dict, List, Optional
+from datetime import datetime
+from typing_extensions import Literal
+
+from .model import Model
+from ..._models import BaseModel
+from .inference_params import InferenceParams
+
+__all__ = ["ContextRelevanceMetric"]
+
+
+class ContextRelevanceMetric(BaseModel):
+    """RAGAS metric for measuring context relevance."""
+
+    id: str
+
+    created_at: datetime
+
+    created_by: Optional[str] = None
+
+    entity_id: str
+    """Alias for id for backwards compatibility."""
+
+    judge_model: Model
+    """Model definition for use without persisting to the Models API."""
+
+    parent: str
+    """Parent entity ID for nested entities."""
+
+    updated_at: datetime
+
+    updated_by: Optional[str] = None
+
+    workspace: str
+    """Workspace identifier"""
+
+    description: Optional[str] = None
+    """Human-readable description of the metric."""
+
+    ignore_request_failure: Optional[bool] = None
+    """
+    If True, request failures to the judge model are ignored and the metric result
+    is marked as NaN. Parse/output formatting failures are always converted to NaN.
+    """
+
+    inference: Optional[InferenceParams] = None
+    """Parameters for model inference.
+
+    Extra fields can be supplied for additional options applied to the inference
+    request directly. Fields not supported by the model may cause inference errors
+    during evaluation.
+    """
+
+    input_template: Optional[Dict[str, object]] = None
+    """Optional Jinja template for rendering the input payload for RAGAS evaluation."""
+
+    labels: Optional[Dict[str, str]] = None
+    """Labels are key-value pairs that can be used for grouping and filtering."""
+
+    name: Optional[str] = None
+    """Entity name within the workspace"""
+
+    project: Optional[str] = None
+    """The name of the project associated with this entity."""
+
+    supported_job_types: Optional[List[Literal["online", "offline"]]] = None
+    """
+    A metric can evaluate model outputs for online evaluations or pre-generated
+    outputs for offline evaluations.
+    """
+
+    type: Optional[Literal["context_relevance"]] = None

nemo_platform/types/evaluation/context_relevance_metric_param.py

@@ -0,0 +1,66 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Dict, List, Union, Optional
+from typing_extensions import Literal, TypeAlias
+
+from .model import Model
+from ..._models import BaseModel
+from .model_ref import ModelRef
+from .inference_params import InferenceParams
+
+__all__ = ["ContextRelevanceMetricParam", "JudgeModel"]
+
+JudgeModel: TypeAlias = Union[Model, ModelRef]
+
+
+class ContextRelevanceMetricParam(BaseModel):
+    """Request type for ContextRelevance metrics."""
+
+    judge_model: JudgeModel
+    """The judge model configuration."""
+
+    description: Optional[str] = None
+    """Human-readable description of the metric."""
+
+    ignore_request_failure: Optional[bool] = None
+    """
+    If True, request failures to the judge model are ignored and the metric result
+    is marked as NaN. Parse/output formatting failures are always converted to NaN.
+    """
+
+    inference: Optional[InferenceParams] = None
+    """Parameters for model inference.
+
+    Extra fields can be supplied for additional options applied to the inference
+    request directly. Fields not supported by the model may cause inference errors
+    during evaluation.
+    """
+
+    input_template: Optional[Dict[str, object]] = None
+    """Optional Jinja template for rendering the input payload for RAGAS evaluation."""
+
+    labels: Optional[Dict[str, str]] = None
+    """Labels are key-value pairs that can be used for grouping and filtering."""
+
+    supported_job_types: Optional[List[Literal["online", "offline"]]] = None
+    """
+    A metric can evaluate model outputs for online evaluations or pre-generated
+    outputs for offline evaluations.
+    """
+
+    type: Optional[Literal["context_relevance"]] = None

nemo_platform/types/evaluation/context_relevance_metric_param_param.py

@@ -0,0 +1,67 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Dict, List, Union
+from typing_extensions import Literal, Required, TypeAlias, TypedDict
+
+from .model_ref import ModelRef
+from .model_param import ModelParam
+from .inference_params_param import InferenceParamsParam
+
+__all__ = ["ContextRelevanceMetricParamParam", "JudgeModel"]
+
+JudgeModel: TypeAlias = Union[ModelParam, ModelRef]
+
+
+class ContextRelevanceMetricParamParam(TypedDict, total=False):
+    """Request type for ContextRelevance metrics."""
+
+    judge_model: Required[JudgeModel]
+    """The judge model configuration."""
+
+    description: str
+    """Human-readable description of the metric."""
+
+    ignore_request_failure: bool
+    """
+    If True, request failures to the judge model are ignored and the metric result
+    is marked as NaN. Parse/output formatting failures are always converted to NaN.
+    """
+
+    inference: InferenceParamsParam
+    """Parameters for model inference.
+
+    Extra fields can be supplied for additional options applied to the inference
+    request directly. Fields not supported by the model may cause inference errors
+    during evaluation.
+    """
+
+    input_template: Dict[str, object]
+    """Optional Jinja template for rendering the input payload for RAGAS evaluation."""
+
+    labels: Dict[str, str]
+    """Labels are key-value pairs that can be used for grouping and filtering."""
+
+    supported_job_types: List[Literal["online", "offline"]]
+    """
+    A metric can evaluate model outputs for online evaluations or pre-generated
+    outputs for offline evaluations.
+    """
+
+    type: Literal["context_relevance"]

nemo_platform/types/evaluation/context_relevance_metric_response.py

@@ -0,0 +1,85 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Dict, List, Union, Optional
+from datetime import datetime
+from typing_extensions import Literal, TypeAlias
+
+from .model import Model
+from ..._models import BaseModel
+from .model_ref import ModelRef
+from .inference_params import InferenceParams
+
+__all__ = ["ContextRelevanceMetricResponse", "JudgeModel"]
+
+JudgeModel: TypeAlias = Union[Model, ModelRef]
+
+
+class ContextRelevanceMetricResponse(BaseModel):
+    """Response type for ContextRelevance metrics."""
+
+    judge_model: JudgeModel
+    """The judge model configuration."""
+
+    id: Optional[str] = None
+    """Entity name within the workspace"""
+
+    created_at: Optional[datetime] = None
+
+    description: Optional[str] = None
+    """Human-readable description of the metric."""
+
+    ignore_request_failure: Optional[bool] = None
+    """
+    If True, request failures to the judge model are ignored and the metric result
+    is marked as NaN. Parse/output formatting failures are always converted to NaN.
+    """
+
+    inference: Optional[InferenceParams] = None
+    """Parameters for model inference.
+
+    Extra fields can be supplied for additional options applied to the inference
+    request directly. Fields not supported by the model may cause inference errors
+    during evaluation.
+    """
+
+    input_template: Optional[Dict[str, object]] = None
+    """Optional Jinja template for rendering the input payload for RAGAS evaluation."""
+
+    labels: Optional[Dict[str, str]] = None
+    """Labels are key-value pairs that can be used for grouping and filtering."""
+
+    name: Optional[str] = None
+    """Entity name within the workspace"""
+
+    parent: Optional[str] = None
+
+    project: Optional[str] = None
+    """The name of the project associated with this entity."""
+
+    supported_job_types: Optional[List[Literal["online", "offline"]]] = None
+    """
+    A metric can evaluate model outputs for online evaluations or pre-generated
+    outputs for offline evaluations.
+    """
+
+    type: Optional[Literal["context_relevance"]] = None
+
+    updated_at: Optional[datetime] = None
+
+    workspace: Optional[str] = None
+    """Workspace identifier"""

nemo_platform/types/evaluation/dataset_rows.py

@@ -0,0 +1,35 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Dict, List
+
+from ..._models import BaseModel
+
+__all__ = ["DatasetRows"]
+
+
+class DatasetRows(BaseModel):
+    """Inline dataset definition with embedded rows.
+
+    Use this for quick evaluations without persisting the dataset first.
+    """
+
+    rows: List[Dict[str, object]]
+    """Array of data rows.
+
+    Each row can be any valid JSON value (object, string, array, etc.).
+    """

nemo_platform/types/evaluation/dataset_rows_param.py

@@ -0,0 +1,36 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Dict, Iterable
+from typing_extensions import Required, TypedDict
+
+__all__ = ["DatasetRowsParam"]
+
+
+class DatasetRowsParam(TypedDict, total=False):
+    """Inline dataset definition with embedded rows.
+
+    Use this for quick evaluations without persisting the dataset first.
+    """
+
+    rows: Required[Iterable[Dict[str, object]]]
+    """Array of data rows.
+
+    Each row can be any valid JSON value (object, string, array, etc.).
+    """

nemo_platform/types/evaluation/evaluate_dataset_rows_param.py

@@ -0,0 +1,33 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from __future__ import annotations
+
+from typing import Dict, Iterable
+from typing_extensions import Required, TypedDict
+
+__all__ = ["EvaluateDatasetRowsParam"]
+
+
+class EvaluateDatasetRowsParam(TypedDict, total=False):
+    """Inline dataset for evaluation with a maximum of 10 rows."""
+
+    rows: Required[Iterable[Dict[str, object]]]
+    """Array of data rows.
+
+    Each row can be any valid JSON value (object, string, array, etc.).
+    """

nemo_platform/types/evaluation/exact_match_metric.py

@@ -0,0 +1,79 @@
+# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+from typing import Dict, List, Optional
+from datetime import datetime
+from typing_extensions import Literal
+
+from ..._models import BaseModel
+
+__all__ = ["ExactMatchMetric"]
+
+
+class ExactMatchMetric(BaseModel):
+    """Persisted Exact Match metric."""
+
+    id: str
+
+    created_at: datetime
+
+    created_by: Optional[str] = None
+
+    entity_id: str
+    """Alias for id for backwards compatibility."""
+
+    parent: str
+    """Parent entity ID for nested entities."""
+
+    reference: str
+    """
+    The template for the ground truth reference to calculate the exact match metric
+    with.
+    """
+
+    updated_at: datetime
+
+    updated_by: Optional[str] = None
+
+    workspace: str
+    """Workspace identifier"""
+
+    candidate: Optional[str] = None
+    """The template for the candidate to evaluate the exact match metric on.
+
+    If not provided, the output text from the model is used.
+    """
+
+    description: Optional[str] = None
+    """Human-readable description of the metric."""
+
+    labels: Optional[Dict[str, str]] = None
+    """Labels are key-value pairs that can be used for grouping and filtering."""
+
+    name: Optional[str] = None
+    """Entity name within the workspace"""
+
+    project: Optional[str] = None
+    """The name of the project associated with this entity."""
+
+    supported_job_types: Optional[List[Literal["online", "offline"]]] = None
+    """
+    A metric can evaluate model outputs for online evaluations or pre-generated
+    outputs for offline evaluations.
+    """
+
+    type: Optional[Literal["exact-match"]] = None