From e709101f5b9f2731f7083ce2cb76f8dd327c172c Mon Sep 17 00:00:00 2001
From: Xi Yan <xiyan@meta.com>
Date: Tue, 22 Oct 2024 08:24:14 -0700
Subject: [PATCH] datasets api

---
 llama_stack/apis/dataset/dataset.py           | 63 ------------------
 llama_stack/apis/datasetio/__init__.py        |  7 ++
 llama_stack/apis/datasetio/datasetio.py       |  5 ++
 .../apis/{dataset => datasets}/__init__.py    |  0
 llama_stack/apis/datasets/datasets.py         | 65 +++++++++++++++++++
 5 files changed, 77 insertions(+), 63 deletions(-)
 delete mode 100644 llama_stack/apis/dataset/dataset.py
 create mode 100644 llama_stack/apis/datasetio/__init__.py
 create mode 100644 llama_stack/apis/datasetio/datasetio.py
 rename llama_stack/apis/{dataset => datasets}/__init__.py (100%)
 create mode 100644 llama_stack/apis/datasets/datasets.py

diff --git a/llama_stack/apis/dataset/dataset.py b/llama_stack/apis/dataset/dataset.py
deleted file mode 100644
index 2fa8bb4e5..000000000
--- a/llama_stack/apis/dataset/dataset.py
+++ /dev/null
@@ -1,63 +0,0 @@
-# Copyright (c) Meta Platforms, Inc. and affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the terms described in the LICENSE file in
-# the root directory of this source tree.
-
-from enum import Enum
-from typing import Any, Dict, Optional, Protocol
-
-from llama_models.llama3.api.datatypes import URL
-
-from llama_models.schema_utils import json_schema_type, webmethod
-
-from pydantic import BaseModel
-
-
-@json_schema_type
-class TrainEvalDatasetColumnType(Enum):
-    dialog = "dialog"
-    text = "text"
-    media = "media"
-    number = "number"
-    json = "json"
-
-
-@json_schema_type
-class TrainEvalDataset(BaseModel):
-    """Dataset to be used for training or evaluating language models."""
-
-    # TODO(ashwin): figure out if we need to add an enum for a "dataset type"
-
-    columns: Dict[str, TrainEvalDatasetColumnType]
-    content_url: URL
-    metadata: Optional[Dict[str, Any]] = None
-
-
-@json_schema_type
-class CreateDatasetRequest(BaseModel):
-    """Request to create a dataset."""
-
-    uuid: str
-    dataset: TrainEvalDataset
-
-
-class Datasets(Protocol):
-    @webmethod(route="/datasets/create")
-    def create_dataset(
-        self,
-        uuid: str,
-        dataset: TrainEvalDataset,
-    ) -> None: ...
-
-    @webmethod(route="/datasets/get")
-    def get_dataset(
-        self,
-        dataset_uuid: str,
-    ) -> TrainEvalDataset: ...
-
-    @webmethod(route="/datasets/delete")
-    def delete_dataset(
-        self,
-        dataset_uuid: str,
-    ) -> None: ...
diff --git a/llama_stack/apis/datasetio/__init__.py b/llama_stack/apis/datasetio/__init__.py
new file mode 100644
index 000000000..378afbba8
--- /dev/null
+++ b/llama_stack/apis/datasetio/__init__.py
@@ -0,0 +1,7 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from .datasetio import *  # noqa: F401 F403
diff --git a/llama_stack/apis/datasetio/datasetio.py b/llama_stack/apis/datasetio/datasetio.py
new file mode 100644
index 000000000..756f351d8
--- /dev/null
+++ b/llama_stack/apis/datasetio/datasetio.py
@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
diff --git a/llama_stack/apis/dataset/__init__.py b/llama_stack/apis/datasets/__init__.py
similarity index 100%
rename from llama_stack/apis/dataset/__init__.py
rename to llama_stack/apis/datasets/__init__.py
diff --git a/llama_stack/apis/datasets/datasets.py b/llama_stack/apis/datasets/datasets.py
new file mode 100644
index 000000000..addc59952
--- /dev/null
+++ b/llama_stack/apis/datasets/datasets.py
@@ -0,0 +1,65 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import Any, Dict, Optional, Protocol
+
+from llama_models.llama3.api.datatypes import URL
+
+from llama_models.schema_utils import json_schema_type, webmethod
+
+from pydantic import BaseModel
+
+from llama_stack.apis.common.type_system import ParamType
+
+
+@json_schema_type
+class DatasetSchema(BaseModel):
+    columns: Dict[str, ParamType]
+
+
+@json_schema_type
+class DatasetDef(BaseModel):
+    identifier: str = Field(
+        description="A unique name for the dataset",
+    )
+    schema: DatasetSchema = Field(
+        description="The schema definition for this dataset",
+    )
+    url: URL
+    metadata: Dict[str, Any] = Field(
+        default_factory=dict,
+        description="Any additional metadata for this dataset",
+    )
+
+
+@json_schema_type
+class DatasetDefWithProvider(DatasetDef):
+    provider_id: str = Field(
+        description="ID of the provider which serves this dataset",
+    )
+
+
+class Datasets(Protocol):
+    @webmethod(route="/datasets/register", method="POST")
+    async def register_dataset(
+        self,
+        dataset_def: DatasetDefWithProvider,
+    ) -> None: ...
+
+    @webmethod(route="/datasets/get", method="GET")
+    async def get_dataset(
+        self,
+        dataset_identifier: str,
+    ) -> Optional[DatasetDefWithProvider]: ...
+
+    @webmethod(route="/datasets/delete")
+    async def delete_dataset(
+        self,
+        dataset_identifier: str,
+    ) -> None: ...
+
+    @webmethod(route="/datasets/list", method="GET")
+    async def list_datasets(self) -> List[DatasetDefWithProvider]: ...