File size: 2,960 Bytes
45917c7 66f0e23 45917c7 66f0e23 45917c7 66f0e23 45917c7 | 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 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 | """Tests for the Hugging Face inference catalogue (stdlib-only, offline-safe).
The catalogue is pure data + URL building: it loads with no token, every model stays
within the ≤32B "small minds" rule (tiny ≤4B), and a binding derives the LiteLLM
OpenAI-compatible string + the HF router URL + the token from the env.
"""
from __future__ import annotations
from src.models import hf_catalogue
# Tier upper bounds (billions of params) the catalogue must respect.
_TIER_CAP = {"tiny": 4, "fast": 8, "balanced": 13, "strong": 32}
def test_entries_load_offline_and_are_well_shaped():
entries = hf_catalogue.entries()
assert entries, "the HF catalogue should not be empty"
for e in entries:
assert {"key", "provider", "served_model_id", "profile", "params_b"} <= set(e)
# The key is the repo id, and the served id matches it (router expects the repo id).
assert e["key"] == e["served_model_id"]
assert "/" in e["served_model_id"]
def test_every_model_is_within_its_tier_param_cap():
for e in hf_catalogue.entries():
tier, params = e["profile"], e["params_b"]
assert params is not None and params <= 32 # the "small minds" rule
if tier in _TIER_CAP:
assert params <= _TIER_CAP[tier], f"{e['key']} ({params}B) exceeds {tier} cap"
def test_catalogue_has_a_tiny_default():
# The catalogue is currently scoped to the one chat-capable model live on the
# enabled providers (tagged tiny). Tiers without a dedicated model fall back to
# it at the UI layer (see lab._default_model_key), so they may return None here.
assert hf_catalogue.default_key_for_profile("tiny") == "katanemo/Arch-Router-1.5B"
def test_binding_uses_router_url_and_token():
key = hf_catalogue.default_key_for_profile("tiny")
binding = hf_catalogue.binding_for(key, env={"HF_TOKEN": "hf_xyz"})
# The model pins its provider (hf-inference) so routing needs no paid auto-select.
assert binding["model"] == f"openai/{key}:hf-inference"
assert binding["base_url"] == hf_catalogue.DEFAULT_BASE_URL
assert binding["api_key"] == "hf_xyz"
def test_binding_honours_explicit_base_url_and_legacy_token_var():
key = hf_catalogue.default_key_for_profile("tiny")
binding = hf_catalogue.binding_for(
key,
env={"HF_INFERENCE_BASE_URL": "https://my-tgi.example/v1", "HUGGINGFACEHUB_API_TOKEN": "legacy"},
)
assert binding["base_url"] == "https://my-tgi.example/v1"
assert binding["api_key"] == "legacy" # the older token var is accepted
def test_has_credentials():
assert hf_catalogue.has_credentials({"HF_TOKEN": "x"}) is True
assert hf_catalogue.has_credentials({"HF_INFERENCE_BASE_URL": "https://box/v1"}) is True
assert hf_catalogue.has_credentials({}) is False
def test_unknown_key_raises():
import pytest
with pytest.raises(KeyError):
hf_catalogue.binding_for("not/a-real-model", env={"HF_TOKEN": "x"})
|