gmicloud 0.1.3__py3-none-any.whl → 0.1.4__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- gmicloud/_internal/_client/_http_client.py +0 -1
- gmicloud/_internal/_client/_iam_client.py +30 -7
- gmicloud/_internal/_models.py +26 -1
- {gmicloud-0.1.3.dist-info → gmicloud-0.1.4.dist-info}/METADATA +81 -45
- {gmicloud-0.1.3.dist-info → gmicloud-0.1.4.dist-info}/RECORD +7 -7
- {gmicloud-0.1.3.dist-info → gmicloud-0.1.4.dist-info}/WHEEL +0 -0
- {gmicloud-0.1.3.dist-info → gmicloud-0.1.4.dist-info}/top_level.txt +0 -0
| @@ -34,12 +34,35 @@ class IAMClient: | |
| 34 34 | 
             
                    custom_headers = {
         | 
| 35 35 | 
             
                        CLIENT_ID_HEADER: self._client_id
         | 
| 36 36 | 
             
                    }
         | 
| 37 | 
            -
                    req =  | 
| 38 | 
            -
                     | 
| 39 | 
            -
             | 
| 40 | 
            -
             | 
| 41 | 
            -
                     | 
| 42 | 
            -
                     | 
| 37 | 
            +
                    req = AuthTokenRequest(email=self._email, password=self._password)
         | 
| 38 | 
            +
                    auth_tokens_result = self.client.post("/me/auth-tokens", custom_headers, req.model_dump())
         | 
| 39 | 
            +
                    auth_tokens_resp = AuthTokenResponse.model_validate(auth_tokens_result)
         | 
| 40 | 
            +
             | 
| 41 | 
            +
                    create_session_result = None
         | 
| 42 | 
            +
                    if auth_tokens_resp.is2FARequired:
         | 
| 43 | 
            +
                        max_attempts = 3
         | 
| 44 | 
            +
                        for attempt in range(max_attempts):
         | 
| 45 | 
            +
                            code = input(f"Attempt {attempt + 1}/{max_attempts}: Please enter the 2FA code: ")
         | 
| 46 | 
            +
             | 
| 47 | 
            +
                            create_session_req = CreateSessionRequest(
         | 
| 48 | 
            +
                                type="native", authToken=auth_tokens_resp.authToken, otpCode=code
         | 
| 49 | 
            +
                            )
         | 
| 50 | 
            +
                            try:
         | 
| 51 | 
            +
                                create_session_result = self.client.post(
         | 
| 52 | 
            +
                                    "/me/sessions", custom_headers, create_session_req.model_dump()
         | 
| 53 | 
            +
                                )
         | 
| 54 | 
            +
                                break
         | 
| 55 | 
            +
                            except Exception as e:
         | 
| 56 | 
            +
                                print("Invalid 2FA code, please try again.")
         | 
| 57 | 
            +
                                if attempt == max_attempts - 1:
         | 
| 58 | 
            +
                                    raise Exception("Failed to create session after 3 incorrect 2FA attempts.") from e
         | 
| 59 | 
            +
                    else:
         | 
| 60 | 
            +
                        create_session_req = CreateSessionRequest(type="native", authToken=auth_tokens_resp.authToken, otpCode=None)
         | 
| 61 | 
            +
                        create_session_result = self.client.post("/me/sessions", custom_headers, create_session_req.model_dump())
         | 
| 62 | 
            +
             | 
| 63 | 
            +
                    create_session_resp = CreateSessionResponse.model_validate(create_session_result)
         | 
| 64 | 
            +
                    self._access_token = create_session_resp.accessToken
         | 
| 65 | 
            +
                    self._refresh_token = create_session_resp.refreshToken
         | 
| 43 66 | 
             
                    self._user_id = self.parse_user_id()
         | 
| 44 67 |  | 
| 45 68 | 
             
                def refresh_token(self):
         | 
| @@ -51,7 +74,7 @@ class IAMClient: | |
| 51 74 | 
             
                    }
         | 
| 52 75 | 
             
                    result = self.client.patch("/me/sessions", custom_headers, {"refreshToken": self._refresh_token})
         | 
| 53 76 |  | 
| 54 | 
            -
                    resp =  | 
| 77 | 
            +
                    resp = CreateSessionResponse.model_validate(result)
         | 
| 55 78 | 
             
                    self._access_token = resp.accessToken
         | 
| 56 79 | 
             
                    self._refresh_token = resp.refreshToken
         | 
| 57 80 |  | 
    
        gmicloud/_internal/_models.py
    CHANGED
    
    | @@ -307,10 +307,35 @@ class CreateTaskResponse(BaseModel): | |
| 307 307 | 
             
                upload_link: str  # URL to upload the task data.
         | 
| 308 308 |  | 
| 309 309 |  | 
| 310 | 
            -
            class  | 
| 310 | 
            +
            class AuthTokenRequest(BaseModel):
         | 
| 311 | 
            +
                """
         | 
| 312 | 
            +
                Request object for user login.
         | 
| 313 | 
            +
                """
         | 
| 314 | 
            +
                email: str  # User email.
         | 
| 315 | 
            +
                password: str  # User password.
         | 
| 316 | 
            +
             | 
| 317 | 
            +
             | 
| 318 | 
            +
            class AuthTokenResponse(BaseModel):
         | 
| 311 319 | 
             
                """
         | 
| 312 320 | 
             
                Response object for user login.
         | 
| 313 321 | 
             
                """
         | 
| 322 | 
            +
                authToken: str  # Access token for the user session.
         | 
| 323 | 
            +
                is2FARequired: bool  # Indicates if 2FA is required for the user.
         | 
| 324 | 
            +
             | 
| 325 | 
            +
             | 
| 326 | 
            +
            class CreateSessionRequest(BaseModel):
         | 
| 327 | 
            +
                """
         | 
| 328 | 
            +
                Request object for creating a user session.
         | 
| 329 | 
            +
                """
         | 
| 330 | 
            +
                type: str  # Type of the session (e.g., native).
         | 
| 331 | 
            +
                authToken: str  # Access token for the user session.
         | 
| 332 | 
            +
                otpCode: Optional[str]  # 2FA code for the user session.
         | 
| 333 | 
            +
             | 
| 334 | 
            +
             | 
| 335 | 
            +
            class CreateSessionResponse(BaseModel):
         | 
| 336 | 
            +
                """
         | 
| 337 | 
            +
                Response object for creating a user session.
         | 
| 338 | 
            +
                """
         | 
| 314 339 | 
             
                accessToken: str  # Access token for the user session.
         | 
| 315 340 | 
             
                refreshToken: str  # Refresh token for the user session.
         | 
| 316 341 |  | 
| @@ -1,6 +1,6 @@ | |
| 1 1 | 
             
            Metadata-Version: 2.2
         | 
| 2 2 | 
             
            Name: gmicloud
         | 
| 3 | 
            -
            Version: 0.1. | 
| 3 | 
            +
            Version: 0.1.4
         | 
| 4 4 | 
             
            Summary: GMI Cloud Python SDK
         | 
| 5 5 | 
             
            Author-email: GMI <gmi@gmitec.net>
         | 
| 6 6 | 
             
            License: MIT
         | 
| @@ -14,7 +14,8 @@ Description-Content-Type: text/markdown | |
| 14 14 |  | 
| 15 15 | 
             
            ## Overview
         | 
| 16 16 |  | 
| 17 | 
            -
            Before you start: Our service and GPU resource is currenly invite-only so please contact our team ( | 
| 17 | 
            +
            Before you start: Our service and GPU resource is currenly invite-only so please contact our team (
         | 
| 18 | 
            +
            getstarted@gmicloud.ai) to get invited if you don't have one yet.
         | 
| 18 19 |  | 
| 19 20 | 
             
            The GMI Inference Engine SDK provides a Python interface for deploying and managing machine learning models in
         | 
| 20 21 | 
             
            production environments. It allows users to create model artifacts, schedule tasks for serving models, and call
         | 
| @@ -64,7 +65,18 @@ client = Client(client_id="<YOUR_CLIENT_ID>", email="<YOUR_EMAIL>", password="<Y | |
| 64 65 |  | 
| 65 66 | 
             
            ## Quick Start
         | 
| 66 67 |  | 
| 67 | 
            -
            ### 1.  | 
| 68 | 
            +
            ### 1. How to run the code in the example folder
         | 
| 69 | 
            +
            ```bash
         | 
| 70 | 
            +
            cd path/to/gmicloud-sdk
         | 
| 71 | 
            +
            # Create a virtual environment
         | 
| 72 | 
            +
            python -m venv venv
         | 
| 73 | 
            +
            source venv/bin/activate
         | 
| 74 | 
            +
             | 
| 75 | 
            +
            pip install -r requirements.txt
         | 
| 76 | 
            +
            python -m examples.<example_name>
         | 
| 77 | 
            +
            ```
         | 
| 78 | 
            +
             | 
| 79 | 
            +
            ### 2. Create a Task from an Artifact Template
         | 
| 68 80 |  | 
| 69 81 | 
             
            This is the simplest example to deploy an existing artifact template:
         | 
| 70 82 |  | 
| @@ -93,24 +105,30 @@ response = call_chat_completion(client, task.task_id) | |
| 93 105 | 
             
            print(response)
         | 
| 94 106 | 
             
            ```
         | 
| 95 107 |  | 
| 96 | 
            -
            ###  | 
| 108 | 
            +
            ### 3. Step-by-Step Example: Create Artifact, Task, and Query the Endpoint
         | 
| 97 109 |  | 
| 98 110 | 
             
            #### (a) Create an Artifact from a Template
         | 
| 99 111 |  | 
| 100 112 | 
             
            First, you’ll retrieve all templates and create an artifact based on the desired template (e.g., "Llama3.1 8B"):
         | 
| 101 113 |  | 
| 102 114 | 
             
            ```python
         | 
| 103 | 
            -
             | 
| 115 | 
            +
            from gmicloud import *
         | 
| 116 | 
            +
             | 
| 117 | 
            +
            def create_artifact_from_template(client: Client) -> str:
         | 
| 104 118 | 
             
                artifact_manager = client.artifact_manager
         | 
| 105 119 |  | 
| 106 | 
            -
                #  | 
| 120 | 
            +
                # Get all artifact templates
         | 
| 107 121 | 
             
                templates = artifact_manager.get_artifact_templates()
         | 
| 108 122 | 
             
                for template in templates:
         | 
| 109 123 | 
             
                    if template.artifact_template_id == "qwen_2.5_14b_instruct_template_001":
         | 
| 110 | 
            -
                         | 
| 111 | 
            -
             | 
| 124 | 
            +
                        # Create an artifact from a template
         | 
| 125 | 
            +
                        artifact_id = artifact_manager.create_artifact_from_template(
         | 
| 126 | 
            +
                            artifact_template_id=template.artifact_template_id,
         | 
| 112 127 | 
             
                        )
         | 
| 113 | 
            -
             | 
| 128 | 
            +
             | 
| 129 | 
            +
                        return artifact_id
         | 
| 130 | 
            +
             | 
| 131 | 
            +
                return ""
         | 
| 114 132 | 
             
            ```
         | 
| 115 133 |  | 
| 116 134 | 
             
            #### (b) Create a Task from the Artifact
         | 
| @@ -118,43 +136,55 @@ def create_artifact_from_template(client): | |
| 118 136 | 
             
            Wait until the artifact becomes "ready" and then deploy it using task scheduling:
         | 
| 119 137 |  | 
| 120 138 | 
             
            ```python
         | 
| 121 | 
            -
             | 
| 122 | 
            -
             | 
| 139 | 
            +
            from gmicloud import *
         | 
| 140 | 
            +
            import time
         | 
| 141 | 
            +
            from datetime import datetime
         | 
| 123 142 |  | 
| 124 | 
            -
             | 
| 143 | 
            +
            def create_task_and_start(client: Client, artifact_id: str) -> str:
         | 
| 144 | 
            +
                artifact_manager = client.artifact_manager
         | 
| 145 | 
            +
                # Wait for the artifact to be ready
         | 
| 125 146 | 
             
                while True:
         | 
| 126 | 
            -
                     | 
| 127 | 
            -
             | 
| 128 | 
            -
                         | 
| 129 | 
            -
             | 
| 147 | 
            +
                    try:
         | 
| 148 | 
            +
                        artifact = artifact_manager.get_artifact(artifact_id)
         | 
| 149 | 
            +
                        print(f"Artifact status: {artifact.build_status}")
         | 
| 150 | 
            +
                        # Wait until the artifact is ready
         | 
| 151 | 
            +
                        if artifact.build_status == BuildStatus.SUCCESS:
         | 
| 152 | 
            +
                            break
         | 
| 153 | 
            +
                    except Exception as e:
         | 
| 154 | 
            +
                        raise e
         | 
| 155 | 
            +
                    # Wait for 2 seconds
         | 
| 130 156 | 
             
                    time.sleep(2)
         | 
| 131 | 
            -
             | 
| 132 | 
            -
             | 
| 133 | 
            -
             | 
| 134 | 
            -
             | 
| 135 | 
            -
             | 
| 136 | 
            -
             | 
| 137 | 
            -
             | 
| 138 | 
            -
             | 
| 139 | 
            -
             | 
| 140 | 
            -
             | 
| 141 | 
            -
             | 
| 142 | 
            -
             | 
| 143 | 
            -
             | 
| 144 | 
            -
             | 
| 157 | 
            +
                try:
         | 
| 158 | 
            +
                    task_manager = client.task_manager
         | 
| 159 | 
            +
                    # Create a task
         | 
| 160 | 
            +
                    task = task_manager.create_task(Task(
         | 
| 161 | 
            +
                        config=TaskConfig(
         | 
| 162 | 
            +
                            ray_task_config=RayTaskConfig(
         | 
| 163 | 
            +
                                ray_version="2.40.0-py310-gpu",
         | 
| 164 | 
            +
                                file_path="serve",
         | 
| 165 | 
            +
                                artifact_id=artifact_id,
         | 
| 166 | 
            +
                                deployment_name="app",
         | 
| 167 | 
            +
                                replica_resource=ReplicaResource(
         | 
| 168 | 
            +
                                    cpu=10,
         | 
| 169 | 
            +
                                    ram_gb=100,
         | 
| 170 | 
            +
                                    gpu=1,
         | 
| 171 | 
            +
                                ),
         | 
| 172 | 
            +
                            ),
         | 
| 173 | 
            +
                            task_scheduling=TaskScheduling(
         | 
| 174 | 
            +
                                scheduling_oneoff=OneOffScheduling(
         | 
| 175 | 
            +
                                    trigger_timestamp=int(datetime.now().timestamp()) + 10,
         | 
| 176 | 
            +
                                    min_replicas=1,
         | 
| 177 | 
            +
                                    max_replicas=10,
         | 
| 178 | 
            +
                                )
         | 
| 145 179 | 
             
                            ),
         | 
| 146 180 | 
             
                        ),
         | 
| 147 | 
            -
             | 
| 148 | 
            -
             | 
| 149 | 
            -
             | 
| 150 | 
            -
             | 
| 151 | 
            -
             | 
| 152 | 
            -
             | 
| 153 | 
            -
                        ),
         | 
| 154 | 
            -
                    ),
         | 
| 155 | 
            -
                ))
         | 
| 181 | 
            +
                    ))
         | 
| 182 | 
            +
             | 
| 183 | 
            +
                    # Start the task
         | 
| 184 | 
            +
                    task_manager.start_task(task.task_id)
         | 
| 185 | 
            +
                except Exception as e:
         | 
| 186 | 
            +
                    raise e
         | 
| 156 187 |  | 
| 157 | 
            -
                task_manager.start_task(task.task_id)
         | 
| 158 188 | 
             
                return task.task_id
         | 
| 159 189 | 
             
            ```
         | 
| 160 190 |  | 
| @@ -163,14 +193,20 @@ def create_task_and_start(client, artifact_id): | |
| 163 193 | 
             
            Once the task is running, use the endpoint for inference:
         | 
| 164 194 |  | 
| 165 195 | 
             
            ```python
         | 
| 196 | 
            +
            from gmicloud import *
         | 
| 166 197 | 
             
            from examples.completion import call_chat_completion
         | 
| 167 198 |  | 
| 168 | 
            -
             | 
| 169 | 
            -
             | 
| 170 | 
            -
            task_id = create_task_and_start(client, artifact_id)
         | 
| 199 | 
            +
            # Initialize the Client
         | 
| 200 | 
            +
            cli = Client()
         | 
| 171 201 |  | 
| 172 | 
            -
             | 
| 173 | 
            -
             | 
| 202 | 
            +
            # Create an artifact from a template
         | 
| 203 | 
            +
            artifact_id = create_artifact_from_template(cli)
         | 
| 204 | 
            +
             | 
| 205 | 
            +
            # Create a task and start it
         | 
| 206 | 
            +
            task_id = create_task_and_start(cli, artifact_id)
         | 
| 207 | 
            +
             | 
| 208 | 
            +
            # Call chat completion
         | 
| 209 | 
            +
            print(call_chat_completion(cli, task_id))
         | 
| 174 210 | 
             
            ```
         | 
| 175 211 |  | 
| 176 212 | 
             
            ## API Reference
         | 
| @@ -5,13 +5,13 @@ gmicloud/_internal/_config.py,sha256=qIH76TSyS3MQWe62LHI46RJhDnklNFisdajY75oUAqE | |
| 5 5 | 
             
            gmicloud/_internal/_constants.py,sha256=Y085dwFlqdFkCf39iBfxz39QiiB7lX59ayNJjB86_m4,378
         | 
| 6 6 | 
             
            gmicloud/_internal/_enums.py,sha256=C4oy6Ps7OGdYd3tK-Wcpi6uXMyAKbOZW8_KykYAKsgw,510
         | 
| 7 7 | 
             
            gmicloud/_internal/_exceptions.py,sha256=hScBq7n2fOit4_umlkabZJchY8zVbWSRfWM2Y0rLCbw,306
         | 
| 8 | 
            -
            gmicloud/_internal/_models.py,sha256= | 
| 8 | 
            +
            gmicloud/_internal/_models.py,sha256=4YT1MI2V1YEBWBFEvrA6pRcmxkZCgi6rzTxYMRgfeWs,13880
         | 
| 9 9 | 
             
            gmicloud/_internal/_client/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
         | 
| 10 10 | 
             
            gmicloud/_internal/_client/_artifact_client.py,sha256=py6SQ1tVoK365hIIxnla433mpCapActZ6KE6Yakds-8,5588
         | 
| 11 11 | 
             
            gmicloud/_internal/_client/_decorator.py,sha256=sy4gxzsUB6ORXHw5pqmMf7TTlK41Nmu1fhIhK2AIsbY,670
         | 
| 12 12 | 
             
            gmicloud/_internal/_client/_file_upload_client.py,sha256=1JRs4X57S3EScPIP9w2DC1Uo6_Wbcjumcw3nVM7uIGM,4667
         | 
| 13 | 
            -
            gmicloud/_internal/_client/_http_client.py,sha256= | 
| 14 | 
            -
            gmicloud/_internal/_client/_iam_client.py,sha256= | 
| 13 | 
            +
            gmicloud/_internal/_client/_http_client.py,sha256=JZvw7HnCwyk6_5-Ja-KZuYY0JBJApx9I3jZ0G_J2kWw,5680
         | 
| 14 | 
            +
            gmicloud/_internal/_client/_iam_client.py,sha256=xEDiwwpj1dAHmAOzcu9Jz7x1F3TRJLrCAqNXPQ-wyoU,4310
         | 
| 15 15 | 
             
            gmicloud/_internal/_client/_task_client.py,sha256=JgYewfDy2iblRVrc9sdA8RbcooXyAho79ESuCK8PRWI,3842
         | 
| 16 16 | 
             
            gmicloud/_internal/_manager/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
         | 
| 17 17 | 
             
            gmicloud/_internal/_manager/_artifact_manager.py,sha256=cc367Kd-W9Zn2d_wUcG6tT06544HRItIZ5IsFfA5_RQ,12201
         | 
| @@ -20,7 +20,7 @@ gmicloud/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0 | |
| 20 20 | 
             
            gmicloud/tests/test_artifacts.py,sha256=YiC1HBMS7g491Ra4acTLI9AdwyjXZfnY9f-fNKn2azQ,17108
         | 
| 21 21 | 
             
            gmicloud/tests/test_tasks.py,sha256=AY90zTJdsXk1cxn6Jxhi4TDdwXRiGxz_r_aRk_Jkl8Y,10956
         | 
| 22 22 | 
             
            gmicloud/utils/uninstall_packages.py,sha256=zzuuaJPf39oTXWZ_7tUAGseoxocuCbbkoglJSD5yDrE,1127
         | 
| 23 | 
            -
            gmicloud-0.1. | 
| 24 | 
            -
            gmicloud-0.1. | 
| 25 | 
            -
            gmicloud-0.1. | 
| 26 | 
            -
            gmicloud-0.1. | 
| 23 | 
            +
            gmicloud-0.1.4.dist-info/METADATA,sha256=hwFbxM2blSJZQnRT-9x6y-ilfcgFL6_tdQP2f12TloQ,7689
         | 
| 24 | 
            +
            gmicloud-0.1.4.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
         | 
| 25 | 
            +
            gmicloud-0.1.4.dist-info/top_level.txt,sha256=AZimLw3y0WPpLiSiOidZ1gD0dxALh-jQNk4fxC05hYE,9
         | 
| 26 | 
            +
            gmicloud-0.1.4.dist-info/RECORD,,
         | 
| 
            File without changes
         | 
| 
            File without changes
         |