fix(ci): clear ruff backlog + accept quantization shorthand

packages/python/servingcard/apply.py

@@ -65,7 +65,10 @@ def generate_vllm_command(card: ServingCard) -> str:
 def generate_tgi_command(card: ServingCard) -> str:
     """Generate a TGI launch command from a servingcard."""
     if not card.serving or not card.serving.engine_args:
-        return f"# No engine_args in servingcard -- cannot generate TGI command\ntext-generation-launcher --model-id {card.model}"
+        return (
+            "# No engine_args in servingcard -- cannot generate TGI command\n"
+            f"text-generation-launcher --model-id {card.model}"
+        )
 
     args = card.serving.engine_args.copy()
     model_id = args.pop("model", card.model)

packages/python/servingcard/backends.py

@@ -5,7 +5,6 @@
 import json
 import shutil
 import subprocess
-import sys
 from abc import ABC, abstractmethod
 
 

packages/python/servingcard/cli.py

@@ -2,7 +2,6 @@
 
 from __future__ import annotations
 
-import sys
 from datetime import date
 from pathlib import Path
 from typing import Optional
@@ -11,7 +10,6 @@
 import yaml
 
 from servingcard.apply import (
-    REGISTRY_BASE_URL,
     generate_launch_command,
     resolve_source,
 )
@@ -201,7 +199,7 @@ def apply(
 def _fetch_remote_card(url: str) -> ServingCard:
     """Fetch a servingcard from a URL."""
     try:
-        from urllib.request import urlopen, Request
+        from urllib.request import Request, urlopen
 
         req = Request(url, headers={"User-Agent": "servingcard-cli/0.1"})
         with urlopen(req, timeout=15) as resp:

packages/python/servingcard/schema.py

@@ -6,7 +6,7 @@
 from typing import Any
 
 import yaml
-from pydantic import BaseModel, ConfigDict, Field
+from pydantic import BaseModel, ConfigDict, Field, field_validator
 
 
 class BenchmarkObservation(BaseModel):
@@ -78,11 +78,33 @@ class HardwareDetails(BaseModel):
     architecture: str | None = None
 
 
+_QUANT_BITS: dict[str, int] = {
+    "fp4": 4,
+    "nvfp4": 4,
+    "int4": 4,
+    "fp8": 8,
+    "int8": 8,
+    "fp16": 16,
+    "bf16": 16,
+    "fp32": 32,
+}
+
+
 class QuantizationSection(BaseModel):
-    """Quantization configuration."""
+    """Quantization configuration.
+
+    Accepts either a structured form (`{method: fp8, bits: 8}`) or a bare
+    string shorthand (`fp8`) coerced via `from_shorthand`. Bit width is
+    inferred from the well-known method name; unknown shorthands default
+    to 0 bits with a warning rather than failing the load.
+    """
 
     method: str
-    bits: int
+    bits: int = 0
+
+    @classmethod
+    def from_shorthand(cls, value: str) -> "QuantizationSection":
+        return cls(method=value, bits=_QUANT_BITS.get(value.lower(), 0))
 
 
 class SpeculativeDecodingSection(BaseModel):
@@ -169,6 +191,13 @@ class ServingCard(BaseModel):
 
     hardware_details: HardwareDetails | None = None
     quantization: QuantizationSection | None = None
+
+    @field_validator("quantization", mode="before")
+    @classmethod
+    def _coerce_quantization(cls, v: object) -> object:
+        if isinstance(v, str):
+            return QuantizationSection.from_shorthand(v)
+        return v
     speculative_decoding: SpeculativeDecodingSection | None = None
     benchmark: BenchmarkSection | None = None
     benchmarks: list[BenchmarkObservation] | None = None

packages/python/servingcard/validate.py

@@ -32,7 +32,7 @@ def validate_card(path: Path) -> list[str]:
     if not path.exists():
         return [f"File not found: {path}"]
 
-    if not path.suffix in (".yaml", ".yml"):
+    if path.suffix not in (".yaml", ".yml"):
         errors.append(f"Expected .yaml or .yml extension, got: {path.suffix}")
 
     try:

packages/python/tests/conftest.py

@@ -7,7 +7,6 @@
 import pytest
 import yaml
 
-
 REAL_EAGLE3_CONFIG = (
     Path(__file__).resolve().parents[3] / "registry" / "qwen3-coder" / "gb10-fp8-eagle3-spec3.yaml"
 )

packages/python/tests/test_apply.py

@@ -2,8 +2,6 @@
 
 from __future__ import annotations
 
-import pytest
-
 from servingcard.apply import (
     REGISTRY_BASE_URL,
     generate_launch_command,
@@ -12,13 +10,9 @@
     resolve_source,
 )
 from servingcard.schema import (
-    CapacitySection,
     ServingCard,
-    ServingSection,
-    SpeculativeDecodingSection,
 )
 
-
 # ---------------------------------------------------------------------------
 # Helpers
 # ---------------------------------------------------------------------------

packages/python/tests/test_backends.py

@@ -13,7 +13,6 @@
     get_backend,
 )
 
-
 # ---------------------------------------------------------------------------
 # 1. ManualBackend.run returns expected keys (mock input)
 # ---------------------------------------------------------------------------

packages/python/tests/test_schema.py

@@ -10,15 +10,13 @@
 
 from servingcard.schema import (
     BenchmarkEntry,
-    BenchmarkSection,
     CapacitySection,
     PawBenchResults,
     ServingCard,
     ServingSection,
     SpeculativeDecodingSection,
 )
 
-
 # ---------------------------------------------------------------------------
 # 1. from_yaml loads real config
 # ---------------------------------------------------------------------------

packages/python/tests/test_validate.py

@@ -4,12 +4,10 @@
 
 from pathlib import Path
 
-import pytest
 import yaml
 
 from servingcard.validate import validate_card
 
-
 # ---------------------------------------------------------------------------
 # 1. Valid config returns empty errors list
 # ---------------------------------------------------------------------------