Spaces:
Sleeping
Sleeping
Download models.py from Akhilllll/data_quality_agent: direct link, hf CLI and curl.
- Browser
- Download file 2.38 kB
-
https://huggingface.co/spaces/Akhilllll/data_quality_agent/resolve/main/models.py
- Command line
-
hf download hf://spaces/Akhilllll/data_quality_agent/models.py
-
curl -L -o models.py https://huggingface.co/spaces/Akhilllll/data_quality_agent/resolve/main/models.py
2.38 kB
| """ | |
| Data models for the Data Quality Agent Environment. | |
| An ML-focused environment where an AI agent must inspect, diagnose, | |
| and remediate data quality issues in realistic datasets - the kind of | |
| work every ML engineer does before training a model. | |
| """ | |
| from typing import Dict, List, Optional | |
| from openenv.core.env_server.types import Action, Observation | |
| from pydantic import Field | |
| class DataQualityAgentAction(Action): | |
| """Action for the Data Quality Agent environment - analyst commands.""" | |
| message: str = Field( | |
| ..., | |
| description=( | |
| "Data quality command(s), semicolon-separated. " | |
| "Examples: DESCRIBE, INSPECT age, CHECK_MISSING, " | |
| "FIX_MISSING age median, VALIDATE, SUBMIT_REPORT" | |
| ), | |
| ) | |
| class DataQualityAgentObservation(Observation): | |
| """Observation from the Data Quality Agent environment.""" | |
| # Main output | |
| dashboard: str = Field(default="", description="Human-readable data quality report") | |
| step_number: int = Field(default=0, description="Current step number") | |
| total_steps: int = Field(default=40, description="Max steps in episode") | |
| # Dataset overview | |
| dataset_name: str = Field(default="", description="Name of the dataset being audited") | |
| num_rows: int = Field(default=0, description="Number of rows in dataset") | |
| num_columns: int = Field(default=0, description="Number of columns in dataset") | |
| # Issue tracking | |
| issues_detected: int = Field(default=0, description="Number of distinct issues detected so far") | |
| issues_fixed: int = Field(default=0, description="Number of issues remediated so far") | |
| total_issues: int = Field(default=0, description="Total ground-truth issues (hidden)") | |
| # Quality metrics | |
| quality_score: float = Field(default=0.0, description="Current overall data quality 0-1") | |
| completeness: float = Field(default=0.0, description="Data completeness score 0-1") | |
| consistency: float = Field(default=0.0, description="Data consistency score 0-1") | |
| accuracy: float = Field(default=0.0, description="Data accuracy score 0-1") | |
| # Phase tracking | |
| phase: str = Field(default="inspection", description="Current phase: inspection / remediation / validation") | |
| task_name: str = Field(default="", description="Current scenario name") | |
| task_description: str = Field(default="", description="Scenario description") | |