Buckets:

download
raw
2.13 kB
# tests/test_cost_model.py
"""Tests for veronica.cost_model -- TableCostModel."""
from __future__ import annotations
from veronica.cost_model import TableCostModel
from veronica.types import HistoryView, StepIntent
def _intent(model: str = "gpt-4") -> StepIntent:
return StepIntent(
step_id="s1",
request_id="r1",
chain_id="c1",
kind="llm",
model=model,
tool_name=None,
timeout_ms=30_000,
metadata={},
)
def _history() -> HistoryView:
return HistoryView(
chain_id="c1",
last_n=(),
rolling_cost_usd=0.0,
failure_streak=0,
depth=0,
loop_score=0.0,
)
class TestTableCostModel:
def test_known_model(self) -> None:
cm = TableCostModel()
est = cm.estimate(_intent("gpt-4"), _history(), None)
assert est.estimated_usd > 0
assert est.basis == "pricing_table"
assert est.model_used == "gpt-4"
assert 0.0 <= est.confidence <= 1.0
def test_unknown_model_fallback(self) -> None:
cm = TableCostModel()
est = cm.estimate(_intent("unknown-model-xyz"), _history(), None)
assert est.estimated_usd > 0
assert est.basis == "fallback"
def test_tool_call_zero(self) -> None:
intent = StepIntent(
step_id="s1",
request_id="r1",
chain_id="c1",
kind="tool",
model=None,
tool_name="web_search",
timeout_ms=30_000,
metadata={},
)
cm = TableCostModel()
est = cm.estimate(intent, _history(), None)
assert est.estimated_usd == 0.0
assert est.basis == "pricing_table"
def test_no_model_fallback(self) -> None:
intent = StepIntent(
step_id="s1",
request_id="r1",
chain_id="c1",
kind="llm",
model=None,
tool_name=None,
timeout_ms=30_000,
metadata={},
)
cm = TableCostModel()
est = cm.estimate(intent, _history(), None)
assert est.basis == "fallback"

Xet Storage Details

Size:
2.13 kB
·
Xet hash:
5a68461a762a7510ed912ad46c411f5ddb68e8d4216fa952787f59776427f3bd

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.