seekrai 0.1.0__tar.gz → 0.2.0__tar.gz

{seekrai-0.1.0 → seekrai-0.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: seekrai
-Version: 0.1.0
+Version: 0.2.0
 Summary: Python client for SeekrAI
 Home-page: https://gitlab.cb.ntent.com/ml/seekr-py
 License: Apache-2.0
@@ -17,7 +17,7 @@ Classifier: Programming Language :: Python :: 3.12
 Requires-Dist: click (>=8.1.7,<9.0.0)
 Requires-Dist: eval-type-backport (>=0.1.3,<0.3.0)
 Requires-Dist: filelock (>=3.13.1,<4.0.0)
-Requires-Dist: httpx (>=0.27.0,<0.28.0)
+Requires-Dist: httpx[http2] (>=0.27.0,<0.28.0)
 Requires-Dist: numpy (>=1.23.5) ; python_version < "3.12"
 Requires-Dist: numpy (>=1.26.0) ; python_version >= "3.12"
 Requires-Dist: pillow (>=10.3.0,<11.0.0)
@@ -59,8 +59,6 @@ from seekrai import SeekrFlow
 client = SeekrFlow(api_key="xxxxx")
 ```
 
-This library contains both a python library and a CLI. We'll demonstrate how to use both below.
-
 # Usage – Python Client
 
 ## Chat Completions
@@ -128,7 +126,7 @@ asyncio.run(async_chat_completion(messages))
 
 ## Files
 
-The files API is used for fine-tuning and allows developers to upload data to fine-tune on. It also has several methods to list all files, retrive files, and delete files. Please refer to our fine-tuning docs [here](https://docs.seekrflow.ai/docs/fine-tuning-python).
+The files API is used for fine-tuning and allows developers to upload data to fine-tune on. It also has several methods to list all files, retrieve files, and delete files
 
 ```python
 import os
@@ -136,15 +134,14 @@ from seekrai import SeekrFlow
 
 client = SeekrFlow(api_key=os.environ.get("SEEKR_API_KEY"))
 
-client.files.upload(file="somedata.jsonl")  # uploads a file
+client.files.upload(file="somedata.parquet")  # uploads a file
 client.files.list()  # lists all uploaded files
-client.files.retrieve(id="file-d0d318cb-b7d9-493a-bd70-1cfe089d3815")  # retrieves a specific file
 client.files.delete(id="file-d0d318cb-b7d9-493a-bd70-1cfe089d3815")  # deletes a file
 ```
 
 ## Fine-tunes
 
-The finetune API is used for fine-tuning and allows developers to create finetuning jobs. It also has several methods to list all jobs, retrive statuses and get checkpoints. Please refer to our fine-tuning docs [here](https://docs.seekrflow.ai/docs/fine-tuning-python).
+The finetune API is used for fine-tuning and allows developers to create finetuning jobs. It also has several methods to list all jobs, retrieve statuses and get checkpoints.
 
 ```python
 import os
@@ -160,10 +157,8 @@ client.fine_tuning.create(
     batch_size=4,
     learning_rate=1e-5,
     suffix='my-demo-finetune',
-    wandb_api_key='1a2b3c4d5e.......',
 )
 client.fine_tuning.list()  # lists all fine-tuned jobs
 client.fine_tuning.retrieve(id="ft-c66a5c18-1d6d-43c9-94bd-32d756425b4b")  # retrieves information on finetune event
-client.fine_tuning.list_events(id="ft-c66a5c18-1d6d-43c9-94bd-32d756425b4b")  # Lists events of a fine-tune job
 ```
 

{seekrai-0.1.0 → seekrai-0.2.0}/README.md

@@ -26,8 +26,6 @@ from seekrai import SeekrFlow
 client = SeekrFlow(api_key="xxxxx")
 ```
 
-This library contains both a python library and a CLI. We'll demonstrate how to use both below.
-
 # Usage – Python Client
 
 ## Chat Completions
@@ -95,7 +93,7 @@ asyncio.run(async_chat_completion(messages))
 
 ## Files
 
-The files API is used for fine-tuning and allows developers to upload data to fine-tune on. It also has several methods to list all files, retrive files, and delete files. Please refer to our fine-tuning docs [here](https://docs.seekrflow.ai/docs/fine-tuning-python).
+The files API is used for fine-tuning and allows developers to upload data to fine-tune on. It also has several methods to list all files, retrieve files, and delete files
 
 ```python
 import os
@@ -103,15 +101,14 @@ from seekrai import SeekrFlow
 
 client = SeekrFlow(api_key=os.environ.get("SEEKR_API_KEY"))
 
-client.files.upload(file="somedata.jsonl")  # uploads a file
+client.files.upload(file="somedata.parquet")  # uploads a file
 client.files.list()  # lists all uploaded files
-client.files.retrieve(id="file-d0d318cb-b7d9-493a-bd70-1cfe089d3815")  # retrieves a specific file
 client.files.delete(id="file-d0d318cb-b7d9-493a-bd70-1cfe089d3815")  # deletes a file
 ```
 
 ## Fine-tunes
 
-The finetune API is used for fine-tuning and allows developers to create finetuning jobs. It also has several methods to list all jobs, retrive statuses and get checkpoints. Please refer to our fine-tuning docs [here](https://docs.seekrflow.ai/docs/fine-tuning-python).
+The finetune API is used for fine-tuning and allows developers to create finetuning jobs. It also has several methods to list all jobs, retrieve statuses and get checkpoints.
 
 ```python
 import os
@@ -127,9 +124,7 @@ client.fine_tuning.create(
     batch_size=4,
     learning_rate=1e-5,
     suffix='my-demo-finetune',
-    wandb_api_key='1a2b3c4d5e.......',
 )
 client.fine_tuning.list()  # lists all fine-tuned jobs
 client.fine_tuning.retrieve(id="ft-c66a5c18-1d6d-43c9-94bd-32d756425b4b")  # retrieves information on finetune event
-client.fine_tuning.list_events(id="ft-c66a5c18-1d6d-43c9-94bd-32d756425b4b")  # Lists events of a fine-tune job
 ```

{seekrai-0.1.0 → seekrai-0.2.0}/pyproject.toml

@@ -14,7 +14,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "seekrai"
-version = "0.1.0"
+version = "0.2.0"
 authors = [
     "SeekrFlow <support@seekr.com>"
 ]
@@ -46,7 +46,7 @@ numpy = [
     { version = ">=1.23.5", python = "<3.12" },
     { version = ">=1.26.0", python = ">=3.12" },
 ]
-httpx = "^0.27.0"
+httpx = {extras = ["http2"], version = "^0.27.0"}
 
 [tool.poetry.group.quality]
 optional = true

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/abstract/api_requestor.py

@@ -519,6 +519,9 @@ class APIRequestor:
             if "text/plain" in rheaders.get("Content-Type", ""):
                 data: Dict[str, Any] = {"message": rbody}
             else:
+                if rbody.strip().endswith("[DONE]"):
+                    # TODO
+                    rbody = rbody.replace("data: [DONE]", "")
                 data = json.loads(rbody)
         except (JSONDecodeError, UnicodeDecodeError) as e:
             raise error.APIError(

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/client.py

@@ -18,6 +18,7 @@ class SeekrFlow:
     images: resources.Images
     models: resources.Models
     fine_tuning: resources.FineTuning
+    alignment: resources.Alignment
 
     # client options
     client: SeekrFlowClient
@@ -77,6 +78,7 @@ class SeekrFlow:
         self.images = resources.Images(self.client)
         self.models = resources.Models(self.client)
         self.fine_tuning = resources.FineTuning(self.client)
+        self.alignment = resources.Alignment(self.client)
 
 
 class AsyncSeekrFlow:
@@ -87,6 +89,7 @@ class AsyncSeekrFlow:
     images: resources.AsyncImages
     models: resources.AsyncModels
     fine_tuning: resources.AsyncFineTuning
+    alignment: resources.AsyncAlignment
 
     # client options
     client: SeekrFlowClient
@@ -146,6 +149,7 @@ class AsyncSeekrFlow:
         self.images = resources.AsyncImages(self.client)
         self.models = resources.AsyncModels(self.client)
         self.fine_tuning = resources.AsyncFineTuning(self.client)
+        self.alignment = resources.AsyncAlignment(self.client)
 
 
 Client = SeekrFlow

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/constants.py

@@ -19,7 +19,7 @@ INITIAL_RETRY_DELAY = float(env_or_default("INITIAL_RETRY_DELAY", 0.5))
 MAX_RETRY_DELAY = float(env_or_default("MAX_RETRY_DELAY", 8.0))
 
 # API defaults
-BASE_URL = env_or_default("BASE_URL", "https://build.seekr.com/v1")
+BASE_URL = env_or_default("BASE_URL", "https://flow.seekr.com/v1")
 
 # Download defaults
 DOWNLOAD_BLOCK_SIZE = int(

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/filemanager.py

@@ -12,6 +12,7 @@ import httpx
 import requests
 from requests.structures import CaseInsensitiveDict
 from tqdm import tqdm
+from tqdm.utils import CallbackIOWrapper
 
 import seekrai.utils
 from seekrai.abstract import api_requestor
@@ -42,15 +43,15 @@ def chmod_and_replace(src: Path, dst: Path) -> None:
     # Get umask by creating a temporary file in the cache folder.
     tmp_file = dst.parent / f"tmp_{uuid.uuid4()}"
 
-    try:
-        tmp_file.touch()
+    # try:
+    tmp_file.touch()
 
-        cache_dir_mode = Path(tmp_file).stat().st_mode
+    cache_dir_mode = Path(tmp_file).stat().st_mode
 
-        os.chmod(src.as_posix(), stat.S_IMODE(cache_dir_mode))
+    os.chmod(src.as_posix(), stat.S_IMODE(cache_dir_mode))
 
-    finally:
-        tmp_file.unlink()
+    # finally:
+    #     tmp_file.unlink()
 
     shutil.move(src.as_posix(), dst.as_posix())
 
@@ -186,10 +187,7 @@ class DownloadManager:
             url, output, remote_name, fetch_metadata
         )
 
-        # Prevent parallel downloads of the same file with a lock.
-        lock_path = Path(file_path.as_posix() + ".lock")
-
-        with tempfile.NamedTemporaryFile() as temp_file:
+        with tempfile.NamedTemporaryFile(delete=False) as temp_file:
             response = requestor.request_raw(
                 options=SeekrFlowRequest(
                     method="GET",
@@ -201,7 +199,6 @@ class DownloadManager:
             try:
                 response.raise_for_status()
             except Exception as e:
-                os.remove(lock_path)
                 raise APIError(
                     "Error downloading file", http_status=response.status_code
                 ) from e
@@ -234,8 +231,6 @@ class DownloadManager:
             # Moves temp file to output file path
             chmod_and_replace(Path(temp_file.name), file_path)
 
-        os.remove(lock_path)
-
         return str(file_path.resolve()), file_size
 
 
@@ -323,13 +318,6 @@ class UploadManager:
             client=self._client,
         )
 
-        if redirect:
-            if file.suffix not in [".jsonl", ".parquet", ".pt"]:
-                raise FileTypeError(
-                    f"Unknown extension of file {file}. "
-                    "Only files with extensions .jsonl, .parquet, and .pt are supported."
-                )
-
         file_size = os.stat(file.as_posix()).st_size
 
         with tqdm(
@@ -338,13 +326,14 @@ class UploadManager:
             unit_scale=True,
             desc=f"Uploading file {file.name}",
             disable=bool(DISABLE_TQDM),
-        ):
+        ) as t:
             with file.open("rb") as f:
+                reader_wrapper = CallbackIOWrapper(t.update, f, "read")
                 response, _, _ = requestor.request(
                     options=SeekrFlowRequest(
                         method="PUT",
                         url=url,
-                        files={"files": f, "filename": file.name},
+                        files={"files": reader_wrapper, "filename": file.name},
                         params={"purpose": purpose.value},
                     ),
                 )

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/resources/__init__.py

@@ -1,3 +1,4 @@
+from seekrai.resources.alignment import Alignment, AsyncAlignment
 from seekrai.resources.chat import AsyncChat, Chat
 from seekrai.resources.completions import AsyncCompletions, Completions
 from seekrai.resources.embeddings import AsyncEmbeddings, Embeddings
@@ -8,6 +9,8 @@ from seekrai.resources.models import AsyncModels, Models
 
 
 __all__ = [
+    "AsyncAlignment",
+    "Alignment",
     "AsyncCompletions",
     "Completions",
     "AsyncChat",

seekrai-0.2.0/src/seekrai/resources/alignment.py

@@ -0,0 +1,175 @@
+from typing import List
+
+from seekrai.abstract import api_requestor
+from seekrai.seekrflow_response import SeekrFlowResponse
+from seekrai.types import (
+    AlignmentList,
+    AlignmentRequest,
+    AlignmentResponse,
+    SeekrFlowClient,
+    SeekrFlowRequest,
+)
+
+
+class Alignment:
+    def __init__(self, client: SeekrFlowClient) -> None:
+        self._client = client
+
+    def generate(
+        self,
+        instructions: str,
+        files: List[str],
+    ) -> AlignmentResponse:
+        requestor = api_requestor.APIRequestor(
+            client=self._client,
+        )
+
+        parameter_payload = AlignmentRequest(
+            instructions=instructions,
+            files=files,
+        ).model_dump()
+
+        response, _, _ = requestor.request(
+            options=SeekrFlowRequest(
+                method="POST",
+                url="flow/alignment/generate",
+                params=parameter_payload,
+            ),
+            stream=False,
+        )
+
+        assert isinstance(response, SeekrFlowResponse)
+        return AlignmentResponse(**response.data)
+
+    def list(self) -> AlignmentList:
+        """
+        Lists alignment job history
+
+        Returns:
+            AlignmentList: Object containing a list of alignment jobs
+        """
+
+        requestor = api_requestor.APIRequestor(
+            client=self._client,
+        )
+
+        response, _, _ = requestor.request(
+            options=SeekrFlowRequest(
+                method="GET",
+                url="flow/alignment",
+            ),
+            stream=False,
+        )
+
+        assert isinstance(response, SeekrFlowResponse)
+
+        return AlignmentList(**response.data)
+
+    def retrieve(self, id: str) -> AlignmentResponse:
+        """
+        Retrieves alignment job details
+
+        Args:
+            id (str): Alignment job ID to retrieve.
+
+        Returns:
+            AlignmentResponse: Object containing information about alignment job.
+        """
+
+        requestor = api_requestor.APIRequestor(
+            client=self._client,
+        )
+
+        response, _, _ = requestor.request(
+            options=SeekrFlowRequest(
+                method="GET",
+                url=f"flow/alignment/{id}",
+            ),
+            stream=False,
+        )
+
+        assert isinstance(response, SeekrFlowResponse)
+
+        return AlignmentResponse(**response.data)
+
+
+class AsyncAlignment:
+    def __init__(self, client: SeekrFlowClient) -> None:
+        self._client = client
+
+    async def generate(
+        self,
+        instructions: str,
+        files: List[str],
+    ) -> AlignmentResponse:
+        requestor = api_requestor.APIRequestor(
+            client=self._client,
+        )
+
+        parameter_payload = AlignmentRequest(
+            instructions=instructions,
+            files=files,
+        ).model_dump()
+
+        response, _, _ = await requestor.arequest(
+            options=SeekrFlowRequest(
+                method="POST",
+                url="flow/alignment/generate",
+                params=parameter_payload,
+            ),
+            stream=False,
+        )
+
+        assert isinstance(response, SeekrFlowResponse)
+        return AlignmentResponse(**response.data)
+
+    async def list(self) -> AlignmentList:
+        """
+        Lists alignment job history
+
+        Returns:
+            AlignmentList: Object containing a list of alignment jobs
+        """
+
+        requestor = api_requestor.APIRequestor(
+            client=self._client,
+        )
+
+        response, _, _ = await requestor.arequest(
+            options=SeekrFlowRequest(
+                method="GET",
+                url="flow/alignment",
+            ),
+            stream=False,
+        )
+
+        assert isinstance(response, SeekrFlowResponse)
+
+        return AlignmentList(**response.data)
+
+    async def retrieve(self, id: str) -> AlignmentResponse:
+        """
+        Retrieves alignment job details
+
+        Args:
+            id (str): Alignment job ID to retrieve.
+
+        Returns:
+            AlignmentResponse: Object containing information about alignment job.
+        """
+
+        requestor = api_requestor.APIRequestor(
+            client=self._client,
+        )
+
+        response, _, _ = await requestor.arequest(
+            options=SeekrFlowRequest(
+                method="GET",
+                url=f"flow/alignment/{id}",
+            ),
+            stream=False,
+        )
+
+        assert isinstance(response, SeekrFlowResponse)
+
+        return AlignmentResponse(**response.data)

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/resources/files.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 from pathlib import Path
+from typing import Any, Dict
 
 from seekrai.abstract import api_requestor
 from seekrai.filemanager import DownloadManager, UploadManager
@@ -14,6 +15,10 @@ from seekrai.types import (
     SeekrFlowClient,
     SeekrFlowRequest,
 )
+from seekrai.types.files import (
+    AlignFileMetadataValidationReq,
+    AlignFileMetadataValidationResp,
+)
 from seekrai.utils import normalize_key
 
 
@@ -21,11 +26,18 @@ class Files:
     def __init__(self, client: SeekrFlowClient) -> None:
         self._client = client
 
+    def _get_local_file_metadata(self, file_path: Path) -> Dict[str, Any]:
+        suffix = file_path.suffix.lstrip(".")
+        size_bytes = int(file_path.stat().st_size)
+        return {
+            "suffix": suffix,
+            "size_bytes": size_bytes,
+            "filename": file_path.name,
+        }
+
     def upload(
         self, file: Path | str, *, purpose: FilePurpose | str = FilePurpose.FineTune
     ) -> FileResponse:
-        upload_manager = UploadManager(self._client)
-
         if isinstance(file, str):
             file = Path(file)
 
@@ -34,7 +46,30 @@ class Files:
 
         assert isinstance(purpose, FilePurpose)
 
-        return upload_manager.upload("flow/files", file, purpose=purpose, redirect=True)
+        # Do the metadata validation (fail fast before uploading) for Alignment purpose
+        if purpose == FilePurpose.Alignment:
+            file_metadata = self._get_local_file_metadata(file)
+            suffix = file_metadata["suffix"]
+            size = file_metadata["size_bytes"]
+            metadata_validation = self.validate_align_file_metadata(
+                purpose,
+                suffix,
+                size,
+            )
+
+            if not metadata_validation.is_valid:
+                assert metadata_validation.errors is not None  # To appease linter
+                raise ValueError(
+                    f"Alignment file metadata validation failed: {metadata_validation.errors}"
+                )
+
+        # Upload the file to s3
+        upload_manager = UploadManager(self._client)
+        file_response = upload_manager.upload(
+            "flow/files", file, purpose=purpose, redirect=True
+        )
+
+        return file_response
 
     def list(self) -> FileList:
         requestor = api_requestor.APIRequestor(
@@ -87,7 +122,7 @@ class Files:
             output = Path(output)
 
         downloaded_filename, file_size = download_manager.download(
-            f"flow/files/{id}/content", output, normalize_key(f"{id}.jsonl")
+            f"flow/files/{id}/content", output, normalize_key(id)
         )
 
         return FileObject(
@@ -114,6 +149,31 @@ class Files:
 
         return FileDeleteResponse(**response.data)
 
+    def validate_align_file_metadata(
+        self,
+        purpose: FilePurpose,
+        suffix: str,
+        size: int,
+    ) -> AlignFileMetadataValidationResp:
+        requestor = api_requestor.APIRequestor(client=self._client)
+
+        request_body = AlignFileMetadataValidationReq(
+            purpose=purpose,
+            suffix=suffix,
+            size=size,
+        )
+
+        response, _, _ = requestor.request(
+            options=SeekrFlowRequest(
+                method="POST",
+                url="flow/files/validate_metadata",
+                params=request_body.dict(),
+            ),
+            stream=False,
+        )
+
+        return AlignFileMetadataValidationResp(**response.data)
+
 
 class AsyncFiles:
     def __init__(self, client: SeekrFlowClient) -> None:

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/resources/models.py

@@ -5,6 +5,7 @@ from pathlib import Path
 from typing import Any, List
 
 from tqdm import tqdm
+from tqdm.utils import CallbackIOWrapper
 
 from seekrai.abstract import api_requestor
 from seekrai.constants import DISABLE_TQDM
@@ -35,15 +36,16 @@ class Models:
             total=file_size,
             unit="B",
             unit_scale=True,
-            desc=f"Uploading file {file.name}",
+            desc=f"Uploading model file {file.name}",
             disable=bool(DISABLE_TQDM),
-        ):
+        ) as t:
             with file.open("rb") as f:
+                reader_wrapper = CallbackIOWrapper(t.update, f, "read")
                 response, _, _ = requestor.request(
                     options=SeekrFlowRequest(
                         method="PUT",
                         url="flow/pt-models",
-                        files={"files": f, "filename": file.name},
+                        files={"files": reader_wrapper, "filename": file.name},
                         params={"purpose": model_type},
                     ),
                 )

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/types/__init__.py

@@ -1,4 +1,10 @@
 from seekrai.types.abstract import SeekrFlowClient
+from seekrai.types.alignment import (
+    AlignmentJobStatus,
+    AlignmentList,
+    AlignmentRequest,
+    AlignmentResponse,
+)
 from seekrai.types.chat_completions import (
     ChatCompletionChunk,
     ChatCompletionRequest,
@@ -65,4 +71,8 @@ __all__ = [
     "ImageResponse",
     "ModelResponse",
     "ModelList",
+    "AlignmentRequest",
+    "AlignmentResponse",
+    "AlignmentJobStatus",
+    "AlignmentList",
 ]

seekrai-0.2.0/src/seekrai/types/alignment.py

@@ -0,0 +1,39 @@
+from datetime import datetime
+from enum import Enum
+from typing import List, Literal, Optional
+
+from pydantic import Field
+
+from seekrai.types.abstract import BaseModel
+
+
+class AlignmentRequest(BaseModel):
+    instructions: str = Field(
+        default=..., description="Task description/instructions for the alignment task"
+    )
+    files: List[str] = Field(
+        default=..., description="List of file ids to use for alignment"
+    )
+
+
+class AlignmentJobStatus(str, Enum):
+    STATUS_PENDING = "pending"
+    STATUS_QUEUED = "queued"
+    STATUS_RUNNING = "running"
+    STATUS_CANCEL_REQUESTED = "cancel_requested"
+    STATUS_CANCELLED = "cancelled"
+    STATUS_FAILED = "failed"
+    STATUS_COMPLETED = "completed"
+
+
+class AlignmentResponse(BaseModel):
+    id: Optional[str] = Field(default=..., description="Alignment job ID")
+    created_at: datetime | None = None
+    status: AlignmentJobStatus | None = None
+
+
+class AlignmentList(BaseModel):
+    # object type
+    object: Literal["list"] | None = None
+    # list of fine-tune job objects
+    data: List[AlignmentResponse] | None = None

{seekrai-0.1.0 → seekrai-0.2.0}/src/seekrai/types/files.py

@@ -2,7 +2,7 @@ from __future__ import annotations
 
 from datetime import datetime
 from enum import Enum
-from typing import List, Literal
+from typing import List, Literal, Optional, Union
 
 from seekrai.types.abstract import BaseModel
 from seekrai.types.common import (
@@ -13,12 +13,42 @@ from seekrai.types.common import (
 class FilePurpose(str, Enum):
     FineTune = "fine-tune"
     PreTrain = "pre-train"
+    Alignment = "alignment"
 
 
-class FileType(str, Enum):
+class TrainingFileType(str, Enum):
     jsonl = "jsonl"
     parquet = "parquet"
-    pytorch = "pt"
+    pytorch = "pt"  # TODO - this doesnt belong here
+
+
+class AlignmentFileType(str, Enum):
+    HTML = "html"
+    MD = "md"
+    RST = "rst"
+    RTF = "rtf"
+    TXT = "txt"
+    XML = "xml"
+    JSON = "json"
+    JSONL = "jsonl"
+    CSV = "csv"
+    DOC = "doc"
+    DOCX = "docx"
+    PDF = "pdf"
+
+
+FileType = Union[TrainingFileType, AlignmentFileType]
+
+
+class AlignFileMetadataValidationReq(BaseModel):
+    purpose: str
+    suffix: str
+    size: int
+
+
+class AlignFileMetadataValidationResp(BaseModel):
+    is_valid: bool
+    errors: Optional[str] = None
 
 
 class FileRequest(BaseModel):
@@ -61,8 +91,7 @@ class FileResponse(BaseModel):
     filename: str | None = None
     # file byte size
     bytes: int | None = None
-    # JSONL/Parquet line count
-    line_count: int | None = None
+    created_by: str | None = None  # TODO - fix this later
 
 
 class FileList(BaseModel):