[Refactor] Remove dead code quantization 2 (#52821)

Signed-off-by: yewentao256 <zhyanwentao@126.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
This commit is contained in:
Wentao Ye
2026-08-25 20:28:44 -07:00
committed by GitHub
co-authored by mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
parent 7156c63bef
commit a5e0045433
14 changed files with 0 additions and 526 deletions
@@ -9,7 +9,6 @@ if TYPE_CHECKING:
from vllm.model_executor.layers.fused_moe import FusedMoEMethodBase
from vllm.model_executor.layers.linear import LinearMethodBase
from vllm.model_executor.layers.quantization import QuantizationMethods
from ..config_parser import INCLayerConfig
from ..inc import INCConfig
@@ -22,7 +21,6 @@ class INCScheme(ABC):
- can_handle(): when does this scheme apply?
- get_linear_method(): required — how to quantize Linear layers
- get_moe_method(): optional — how to quantize MoE layers
- get_kvcache_method(): optional — how to quantize KV cache
Schemes that don't support MoE/KVCache inherit the default raise.
"""
@@ -56,20 +54,6 @@ class INCScheme(ABC):
f"Layer config: {layer_config}"
)
def get_kvcache_method(
self,
config: "INCConfig",
layer: "torch.nn.Module",
prefix: str,
layer_config: "INCLayerConfig",
) -> "QuantizationMethods":
"""Optional. Override if this scheme supports KV cache quantization.
Default raises NotImplementedError."""
raise NotImplementedError(
f"{type(self).__name__} does not support KV cache quantization. "
f"Layer config: {layer_config}"
)
class INCLinearScheme(ABC):
@classmethod