Celes Renata
a72f336ad1
Multi-stage evidence-grounded inference architecture replacing the monolithic 9B model extraction pipeline. CPU-first specialist services handle routine extraction while the 9B vLLM model is preserved for semantic adjudication of ambiguous cases. Key components: - Capability-aware inference gateway (OpenAI-compatible + Ollama) - Endpoint registry with DB migrations and REST API - Sentence-aware document segmenter (property tests) - Deterministic financial parsing with offset integrity - Symbol resolution with ambiguity detection - Specialist service (GLiNER2, dynamic batching, K8s deployment) - Company-specific sentiment (FinBERT, calibration) - Retrieval-based novelty and duplicate detection - Confidence calibration pipeline - Deterministic routing engine (property tests) - 9B adjudication layer with VRAM gating - Stock-specific impact model (features, labels, baseline, trained) - Pipeline orchestrator (state machine, queues, leases, feature flags) - Bounded parallelism (async workers, semaphore, load shedding) - Observability (tracing, metrics, alerts) - Compatibility adapter (v3→v2 golden mapping tests) - Shadow/canary promotion framework - Active learning and fine-tuning pipeline Test results: 1,161 tests pass, ruff lint clean. All 282 spec tasks completed.
57 lines
1.5 KiB
Python
57 lines
1.5 KiB
Python
"""Request/response schemas for the specialist inference service.
|
|
|
|
Re-exports from models.py for discoverability, plus BatchConfig for
|
|
deployment configuration.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
from pydantic import BaseModel, Field
|
|
|
|
from services.specialist.models import (
|
|
MODEL_VERSION,
|
|
SCHEMA_VERSION,
|
|
BatchResponse,
|
|
ClassificationRequest,
|
|
ClassificationResult,
|
|
EntityResult,
|
|
ExtractionRequest,
|
|
RelationResult,
|
|
StructuredResult,
|
|
)
|
|
|
|
__all__ = [
|
|
"BatchConfig",
|
|
"BatchResponse",
|
|
"ClassificationRequest",
|
|
"ClassificationResult",
|
|
"EntityResult",
|
|
"ExtractionRequest",
|
|
"MODEL_VERSION",
|
|
"RelationResult",
|
|
"SCHEMA_VERSION",
|
|
"SpanResult",
|
|
"StructuredResult",
|
|
]
|
|
|
|
|
|
class SpanResult(BaseModel):
|
|
"""Generic span result used across extraction types."""
|
|
|
|
text: str
|
|
start_char: int
|
|
end_char: int
|
|
label: str
|
|
score: float = Field(..., ge=0.0, le=1.0)
|
|
model_version: str = MODEL_VERSION
|
|
schema_version: str = SCHEMA_VERSION
|
|
|
|
|
|
class BatchConfig(BaseModel):
|
|
"""Configuration for dynamic batching behavior."""
|
|
|
|
max_batch_size: int = Field(default=32, ge=1, le=512, description="Maximum items per batch")
|
|
max_wait_ms: float = Field(default=50.0, ge=1.0, le=5000.0, description="Maximum wait time before flushing a partial batch (ms)")
|
|
max_queue_size: int = Field(default=256, ge=1, le=10000, description="Maximum pending requests in queue before rejection")
|
|
warm_up_on_start: bool = Field(default=True, description="Whether to run a warm-up inference on startup")
|