Shaun ZhouandClaude Opus 5
92c88b785c
用户反馈"聚类配置里改不了 lr/weight_decay"，实测是两件事：
1. --set 的键打错字母会凭空造出一个谁也不读的键，训练照跑但改动无效。
实例：simclr.optmizer.weight_decay=0.0（optimizer 少个 i）被静默接受。
现在 apply_overrides 默认只允许覆盖已存在的键，报错时给出同级可用键
与最接近的拼写（"是不是想写 'optimizer'？"）。需要新增键时可 strict=False。
2. lr 其实生效了，但从 2229/2500 续训时 cosine 只剩满值的 3%，
新设的 1e-3 落到 2.87e-5。新增 restart_schedule 开关：按剩余轮数重开曲线，
新 lr 从满值开始；SimCLR 与检测训练都支持。续训日志补上调度进度，
并在 lr 不足 initial_lr 的 10% 时警告并给出两种解法。
顺带：严格校验立刻暴露出 configs/cluster/film.yaml 里根本没有 scatter 段
（代码 cfg.get("scatter", {}) 兜底走默认值，所以功能正常但配置项形同虚设），已补上。
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
190 lines
7.5 KiB
Python
190 lines
7.5 KiB
Python
from pathlib import Path
|
||
|
||
import pytest
|
||
|
||
pytest.importorskip("yaml")
|
||
|
||
from utils.config import Config, apply_overrides, deep_merge, load_config, parse_override # noqa: E402
|
||
|
||
|
||
def test_deep_merge_is_recursive():
|
||
merged = deep_merge({"a": {"b": 1, "c": 2}}, {"a": {"c": 3}, "d": 4})
|
||
assert merged == {"a": {"b": 1, "c": 3}, "d": 4}
|
||
|
||
|
||
def test_parse_override_literals():
|
||
assert parse_override("train.lr=1e-3") == ("train.lr", 1e-3)
|
||
assert parse_override("train.amp=False") == ("train.amp", False)
|
||
assert parse_override("model.backbone.name=tiny_conv") == ("model.backbone.name", "tiny_conv")
|
||
assert parse_override("simclr.feature_stages=[5, 6]") == ("simclr.feature_stages", [5, 6])
|
||
|
||
|
||
def test_parse_override_accepts_yaml_style_scalars():
|
||
"""配置文件里写 null/true/false，命令行上也得能这么写。
|
||
|
||
回归测试：只用 ast.literal_eval 时 `false` 会变成字符串 "false"，
|
||
而非空字符串是真值——`--set xxx.enabled=false` 反而把开关打开了。
|
||
"""
|
||
assert parse_override("a=null") == ("a", None)
|
||
assert parse_override("a=~") == ("a", None)
|
||
assert parse_override("a=true") == ("a", True)
|
||
assert parse_override("a=false") == ("a", False)
|
||
assert parse_override("a=false")[1] is False # 不是字符串
|
||
|
||
# 路径、模型名这类仍应保持字符串
|
||
assert parse_override("a=resnet18") == ("a", "resnet18")
|
||
assert parse_override("a=runs/x/best.pth") == ("a", "runs/x/best.pth")
|
||
|
||
|
||
def test_apply_overrides_creates_missing_nodes():
|
||
"""非严格模式下仍可新建嵌套节点；严格模式（load_config 默认）会拦下未知键。"""
|
||
cfg = apply_overrides({}, ["a.b.c=1"], strict=False)
|
||
assert cfg == {"a": {"b": {"c": 1}}}
|
||
|
||
with pytest.raises(KeyError, match="不存在"):
|
||
apply_overrides({}, ["a.b.c=1"])
|
||
|
||
|
||
def test_load_config_with_base(tmp_path):
|
||
(tmp_path / "base.yaml").write_text("train:\n lr: 0.1\n epochs: 10\n", encoding="utf-8")
|
||
(tmp_path / "child.yaml").write_text(
|
||
"_base_: ./base.yaml\ntrain:\n lr: 0.01\nwork_dir: runs/x\n", encoding="utf-8"
|
||
)
|
||
|
||
cfg = load_config(tmp_path / "child.yaml", ["train.epochs=2"])
|
||
assert isinstance(cfg, Config)
|
||
assert cfg.train.lr == 0.01
|
||
assert cfg.train.epochs == 2
|
||
assert cfg.get_path("work_dir") == "runs/x"
|
||
|
||
|
||
def test_load_config_detects_cycle(tmp_path):
|
||
(tmp_path / "a.yaml").write_text("_base_: ./b.yaml\n", encoding="utf-8")
|
||
(tmp_path / "b.yaml").write_text("_base_: ./a.yaml\n", encoding="utf-8")
|
||
with pytest.raises(ValueError, match="循环引用"):
|
||
load_config(tmp_path / "a.yaml")
|
||
|
||
|
||
def test_project_configs_are_loadable():
|
||
cfg = load_config("configs/train/baseline.yaml")
|
||
assert cfg["model"]["name"] == "defect_net"
|
||
assert cfg["data"]["train"]["dataset"]["name"] == "film_defect"
|
||
assert len(cfg["classes"]) > 0
|
||
|
||
|
||
def test_replace_key_swaps_a_whole_subdict(tmp_path):
|
||
"""换组件时要能整块替换，否则旧组件的专有字段会残留下来传给新组件。"""
|
||
(tmp_path / "base.yaml").write_text(
|
||
"backbone:\n name: tiny_conv\n widths: [32, 64]\n depths: [1, 2]\n block: dw\n",
|
||
encoding="utf-8",
|
||
)
|
||
(tmp_path / "child.yaml").write_text(
|
||
"_base_: ./base.yaml\nbackbone:\n _replace_: true\n name: torchvision\n arch: resnet18\n",
|
||
encoding="utf-8",
|
||
)
|
||
|
||
cfg = load_config(tmp_path / "child.yaml")
|
||
|
||
assert cfg["backbone"] == {"name": "torchvision", "arch": "resnet18"}
|
||
assert "_replace_" not in cfg["backbone"] # 标记本身不能漏进构造参数
|
||
|
||
|
||
def test_without_replace_key_subdicts_still_merge(tmp_path):
|
||
(tmp_path / "base.yaml").write_text("backbone:\n name: tiny_conv\n block: dw\n", encoding="utf-8")
|
||
(tmp_path / "child.yaml").write_text(
|
||
"_base_: ./base.yaml\nbackbone:\n block: res\n", encoding="utf-8"
|
||
)
|
||
|
||
cfg = load_config(tmp_path / "child.yaml")
|
||
assert cfg["backbone"] == {"name": "tiny_conv", "block": "res"}
|
||
|
||
|
||
def test_registry_reports_incompatible_fields_actionably():
|
||
"""--set 只改 name 时旧字段会残留，报错要说清楚是哪些字段、该怎么办。"""
|
||
import pytest
|
||
|
||
from models.registry import BACKBONES
|
||
|
||
with pytest.raises(TypeError, match=r"不接受的字段.*widths"):
|
||
BACKBONES.build({"name": "torchvision", "arch": "resnet18", "widths": [32]})
|
||
|
||
|
||
@pytest.mark.parametrize(
|
||
"config_path",
|
||
sorted(str(p) for p in Path("configs").rglob("*.yaml")),
|
||
)
|
||
def test_every_shipped_config_builds_its_components(config_path):
|
||
"""仓库里每个配置都必须能真的构建出组件。
|
||
|
||
回归测试：configs/train/resnet18.yaml 曾经无法使用——它同时继承 tiny_conv 与
|
||
resnet18 的模型配置，合并后 backbone 混了两边的字段，构建时直接报错。
|
||
"""
|
||
torch = pytest.importorskip("torch")
|
||
from datasets.yolo import load_classes
|
||
from models import build_model
|
||
from models.registry import BACKBONES
|
||
|
||
cfg = load_config(config_path)
|
||
|
||
if "model" in cfg:
|
||
if "classes" not in cfg:
|
||
# configs/model/*.yaml 是片段，只能被 _base_ 组合，单独用没有 classes
|
||
pytest.skip(f"{config_path} 是配置片段，不含 classes")
|
||
model_cfg = dict(cfg["model"])
|
||
model_cfg.setdefault("num_classes", len(load_classes(cfg["classes"])))
|
||
# 单测里不下载 ImageNet 权重
|
||
if model_cfg.get("backbone", {}).get("name") == "torchvision":
|
||
model_cfg["backbone"] = {**model_cfg["backbone"], "pretrained": False}
|
||
model = build_model(model_cfg)
|
||
with torch.no_grad():
|
||
out = model(torch.randn(1, 3, 64, 64))
|
||
assert "heatmap" in out
|
||
|
||
simclr_backbone = cfg.get_path("simclr.backbone")
|
||
if simclr_backbone:
|
||
backbone_cfg = dict(simclr_backbone)
|
||
if backbone_cfg.get("name") == "torchvision":
|
||
backbone_cfg["pretrained"] = False
|
||
backbone = BACKBONES.build(backbone_cfg)
|
||
assert backbone.out_channels > 0
|
||
|
||
|
||
def test_eval_config_matches_its_training_backbone():
|
||
"""评估配置构建出的骨架必须与训练配置一致，否则权重根本加载不上。"""
|
||
for train_cfg, eval_cfg in [
|
||
("configs/train/baseline.yaml", "configs/eval/baseline.yaml"),
|
||
("configs/train/resnet18.yaml", "configs/eval/resnet18.yaml"),
|
||
]:
|
||
assert load_config(train_cfg)["model"] == load_config(eval_cfg)["model"], eval_cfg
|
||
|
||
|
||
@pytest.mark.parametrize(
|
||
"override,expected_hint",
|
||
[
|
||
("simclr.optmizer.weight_decay=0.0", "optimizer"), # 少了个 i
|
||
("simclr.optimizer.lrr=1e-3", "lr"), # 多了个 r
|
||
("num_clusers=4", "num_clusters"), # 少了个 t
|
||
],
|
||
)
|
||
def test_set_rejects_typos_with_suggestion(override, expected_hint):
|
||
"""--set 打错字母会凭空造出没人读的键，训练照跑但改动无效——必须拦下并给出建议。"""
|
||
with pytest.raises(KeyError) as excinfo:
|
||
load_config("configs/cluster/film_simclr_resnet50.yaml", [override])
|
||
assert expected_hint in str(excinfo.value)
|
||
|
||
|
||
def test_set_accepts_existing_keys():
|
||
cfg = load_config(
|
||
"configs/cluster/film_simclr_resnet50.yaml",
|
||
["simclr.optimizer.lr=0.001", "simclr.optimizer.weight_decay=0.0", "num_clusters=7"],
|
||
)
|
||
assert cfg["simclr"]["optimizer"]["lr"] == 0.001
|
||
assert cfg["simclr"]["optimizer"]["weight_decay"] == 0.0
|
||
assert cfg["num_clusters"] == 7
|
||
|
||
|
||
def test_apply_overrides_can_add_keys_when_not_strict():
|
||
from utils.config import apply_overrides
|
||
|
||
assert apply_overrides({"a": {"b": 1}}, ["a.c=2"], strict=False) == {"a": {"b": 1, "c": 2}}
|