Skip to content

Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.

9.0 KB
Raw Blame
# SPDX-License-Identifier: AGPL-3.0-only
# Copyright 2026-present the Unsloth AI Inc. team. All rights reserved. See /studio/LICENSE.AGPL-3.0

"""Only RuntimeError/ValueError text reaches the client; other exceptions stay generic."""

from __future__ import annotations

import asyncio
import sys
from pathlib import Path

import pytest

_BACKEND = Path(__file__).resolve().parents[1]
if str(_BACKEND) not in sys.path:
    sys.path.insert(0, str(_BACKEND))

pytest.importorskip("fastapi")

from fastapi import HTTPException  # noqa: E402

import routes.inference as inf  # noqa: E402
from models.inference import ValidateModelRequest  # noqa: E402


async def _inline_to_thread(function, /, *args, **kwargs):
    return function(*args, **kwargs)


@pytest.fixture(autouse = True)
def _run_route_helpers_inline(monkeypatch):
    monkeypatch.setattr(inf.asyncio, "to_thread", _inline_to_thread)


def _provoke(
    monkeypatch,
    exc: BaseException,
    *,
    native: bool = False,
) -> HTTPException:
    """Drive validate_model so from_identifier raises ``exc``; return the
    HTTPException it converts that into."""
    monkeypatch.setattr(
        inf,
        "_resolve_model_identifier_for_request",
        lambda request, operation, **_kwargs: ("org/repo", "org/repo", native),
    )

    def _raise(*_args, **_kwargs):
        raise exc

    monkeypatch.setattr(inf.ModelConfig, "from_identifier", staticmethod(_raise))

    req = ValidateModelRequest(model_path = "org/repo")
    with pytest.raises(HTTPException) as excinfo:
        asyncio.run(inf.validate_model(req, current_subject = "tester"))
    return excinfo.value


def test_runtime_error_surfaces_actionable_message(monkeypatch):
    err = RuntimeError(
        "llama-server binary not found - cannot load GGUF models. "
        "Run setup.sh to build it, or set LLAMA_SERVER_PATH."
    )
    http = _provoke(monkeypatch, err)
    assert http.status_code == 400
    assert "llama-server binary not found" in http.detail
    assert http.detail != "Invalid model"


def test_value_error_not_supported_is_wrapped(monkeypatch):
    http = _provoke(monkeypatch, ValueError("architecture FooBar is not supported"))
    assert http.status_code == 400
    assert "not supported yet" in http.detail.lower()
    assert "FooBar" in http.detail


def test_unexpected_exception_stays_generic(monkeypatch):
    # A non-user-facing exception type must NOT have its message surfaced.
    http = _provoke(monkeypatch, KeyError("secret-internal-detail"))
    assert http.status_code == 400
    assert http.detail == "Invalid model"
    assert "secret-internal-detail" not in http.detail


def test_empty_runtime_error_falls_back_to_generic(monkeypatch):
    http = _provoke(monkeypatch, RuntimeError(""))
    assert http.status_code == 400
    assert http.detail == "Invalid model"


def _drive_validate(monkeypatch, *, is_gguf: bool):
    """Run validate_model with both security helpers forced True; return the response."""
    from types import SimpleNamespace

    import utils.models.model_config as mc

    monkeypatch.setattr(
        inf,
        "_resolve_model_identifier_for_request",
        lambda request, operation, **_kwargs: ("org/mixed-repo", "org/mixed-repo", False),
    )
    config = SimpleNamespace(
        identifier = "org/mixed-repo",
        display_name = "org/mixed-repo",
        is_gguf = is_gguf,
        is_lora = False,
        is_vision = False,
        gguf_file = None,
    )
    monkeypatch.setattr(inf.ModelConfig, "from_identifier", staticmethod(lambda **_kw: config))
    monkeypatch.setattr(mc, "get_base_model_from_lora_identifier", lambda *_a, **_k: None)
    # Both gates WOULD flag this repo (mixed repo with auto_map + an unsafe pickle).
    monkeypatch.setattr(inf, "_requires_trust_remote_code_for_model", lambda *_a, **_k: True)
    monkeypatch.setattr(inf, "_requires_security_review_for_model", lambda *_a, **_k: True)

    req = ValidateModelRequest(model_path = "org/mixed-repo")
    return asyncio.run(inf.validate_model(req, current_subject = "tester"))


def test_selected_gguf_variant_skips_trc_and_security_review(monkeypatch):
    # GGUF loads via llama.cpp: auto_map and root pickles are inert, so neither gate fires.
    resp = _drive_validate(monkeypatch, is_gguf = True)
    assert resp.is_gguf is True
    assert resp.requires_trust_remote_code is False
    assert resp.requires_security_review is False


def test_non_gguf_load_still_runs_trc_and_security_review(monkeypatch):
    resp = _drive_validate(monkeypatch, is_gguf = False)
    assert resp.is_gguf is False
    assert resp.requires_trust_remote_code is True
    assert resp.requires_security_review is True


def test_resolve_loaded_trc_prefers_stored_value():
    # A value stored at load time wins, so a status refresh does not re-derive it.
    assert (
        inf._resolve_loaded_trust_remote_code("org/m", {"requires_trust_remote_code": True}, {})
        is True
    )
    assert (
        inf._resolve_loaded_trust_remote_code(
            "org/m", {"requires_trust_remote_code": False}, {"trust_remote_code": True}
        )
        is False
    )


def test_resolve_loaded_trc_uses_runtime_and_yaml():
    # No stored value: the trust_remote_code the load used, then the YAML default.
    assert (
        inf._resolve_loaded_trust_remote_code("org/m", {}, {}, trust_remote_code_used = True) is True
    )
    assert inf._resolve_loaded_trust_remote_code("org/m", {}, {"trust_remote_code": True}) is True


def test_resolve_loaded_trc_falls_back_to_raw_auto_map(monkeypatch):
    monkeypatch.setattr(inf, "_requires_trust_remote_code_for_model", lambda *_a, **_k: True)
    assert inf._resolve_loaded_trust_remote_code("org/custom", {}, {}) is True
    monkeypatch.setattr(inf, "_requires_trust_remote_code_for_model", lambda *_a, **_k: False)
    assert inf._resolve_loaded_trust_remote_code("org/plain", {}, {}) is False


@pytest.mark.parametrize(
    "model_identifier, expected_target",
    [
        ("Spark-TTS-0.5B/LLM", "unsloth/Spark-TTS-0.5B"),
        ("unsloth/Spark-TTS-0.5B", "unsloth/Spark-TTS-0.5B"),
    ],
)
def test_requires_trc_checks_bicodec_load_subdirectory(
    monkeypatch, model_identifier, expected_target
):
    import utils.inference as inference_utils
    import utils.models.model_config as model_config
    import utils.security.consent as consent

    calls = []
    monkeypatch.setattr(inference_utils, "load_inference_config", lambda *_a, **_k: {})
    monkeypatch.setattr(model_config, "detect_audio_type", lambda *_a, **_k: "bicodec")
    monkeypatch.setattr(
        model_config,
        "load_model_defaults",
        lambda *_a, **_k: {"audio_type": "bicodec"},
    )

    def config_has_auto_map(
        target,
        token,
        *,
        load_subdirs = (),
    ):
        calls.append((target, token, load_subdirs))
        return True

    monkeypatch.setattr(consent, "_config_has_auto_map", config_has_auto_map)

    assert inf._requires_trust_remote_code_for_model(model_identifier, "hf_test") is True
    assert calls == [(expected_target, "hf_test", ("LLM",))]


def _drive_validate_lora(monkeypatch, *, adapter_needs_trc, base_needs_trc):
    """Run validate_model for a LoRA adapter whose base resolves, with per-target
    trust_remote_code answers; return the response."""
    from types import SimpleNamespace

    import utils.models.model_config as mc

    adapter, base = "org/lora-adapter", "org/base-model"
    monkeypatch.setattr(
        inf,
        "_resolve_model_identifier_for_request",
        lambda request, operation, **_kwargs: (adapter, adapter, False),
    )
    config = SimpleNamespace(
        identifier = adapter,
        display_name = adapter,
        is_gguf = False,
        is_lora = True,
        is_vision = False,
        gguf_file = None,
    )
    monkeypatch.setattr(inf.ModelConfig, "from_identifier", staticmethod(lambda **_kw: config))
    monkeypatch.setattr(mc, "get_base_model_from_lora_identifier", lambda *_a, **_k: base)
    trc = {adapter: adapter_needs_trc, base: base_needs_trc}
    monkeypatch.setattr(
        inf,
        "_requires_trust_remote_code_for_model",
        lambda target, *_a, **_k: trc.get(target, False),
    )
    monkeypatch.setattr(inf, "_requires_security_review_for_model", lambda *_a, **_k: False)
    req = ValidateModelRequest(model_path = adapter)
    return asyncio.run(inf.validate_model(req, current_subject = "tester"))


def test_validate_lora_flags_trc_from_adapter_only(monkeypatch):
    # Adapter ships auto_map, base does not: the requirement follows either repo.
    resp = _drive_validate_lora(monkeypatch, adapter_needs_trc = True, base_needs_trc = False)
    assert resp.requires_trust_remote_code is True


def test_validate_lora_flags_trc_from_base_only(monkeypatch):
    resp = _drive_validate_lora(monkeypatch, adapter_needs_trc = False, base_needs_trc = True)
    assert resp.requires_trust_remote_code is True


def test_validate_lora_clean_when_neither_needs_trc(monkeypatch):
    resp = _drive_validate_lora(monkeypatch, adapter_needs_trc = False, base_needs_trc = False)
    assert resp.requires_trust_remote_code is False