peft.SimulateQuantizedEinsum

class gemma.peft.SimulateQuantizedEinsum(

*,

wrapped: flax.linen.linear.Einsum,

method: gemma.peft._quantization_utils.QuantizationMethod = QuantizationMethod.NONE,

parent: flax.linen.module.Module | flax.core.scope.Scope | flax.linen.module._Sentinel | None = <flax.linen.module._Sentinel object>,

name: str | None = None,

)

Bases: flax.linen.module.Module

Wrapper around nn.Einsum which adds a Quantized adapter.

wrapped: flax.linen.linear.Einsum

method: gemma.peft._quantization_utils.QuantizationMethod = 'none'

process_einsum_str(

einsum_str: str,

) → str

处理 einsum 字符串。

name: str | None = None

parent: flax.linen.module.Module | flax.core.scope.Scope | flax.linen.module._Sentinel | None = None

scope: Scope | None = None