Datasets:
File size: 1,500 Bytes
7825400 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 | """JSON Schema used to constrain hosted detection responses."""
import copy
import hashlib
import json
DETECTION_ARRAY_SCHEMA: dict = {
"type": "array",
"maxItems": 50,
"items": {
"type": "object",
"properties": {
"category_id": {"type": "integer", "minimum": 1, "maximum": 97},
"bbox": {
"type": "array",
"items": {"type": "number"},
"minItems": 4,
"maxItems": 4,
},
"score": {"type": "number", "minimum": 0.0, "maximum": 1.0},
},
"required": ["category_id", "bbox", "score"],
"additionalProperties": False,
},
}
def detection_response_format(*, omit_array_cap: bool = False) -> dict:
"""OpenAI-compatible strict response format for a detection array."""
schema = copy.deepcopy(DETECTION_ARRAY_SCHEMA)
if omit_array_cap:
schema.pop("maxItems")
return {
"type": "json_schema",
"json_schema": {
"name": "municipal_detections",
"strict": True,
"schema": schema,
},
}
def detection_schema_hash(*, omit_array_cap: bool = False) -> str:
"""Stable identity of the schema, independent of dict formatting."""
encoded = json.dumps(
detection_response_format(omit_array_cap=omit_array_cap)["json_schema"]["schema"],
sort_keys=True, separators=(",", ":")
).encode("utf-8")
return hashlib.sha256(encoded).hexdigest()
|