Skip to content

vllm.model_executor.layers.quantization.kernels.scaled_mm.triton

TritonScaledMMLinearKernel

Bases: ScaledMMLinearKernel

Source code in vllm/model_executor/layers/quantization/kernels/scaled_mm/triton.py 
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
class TritonScaledMMLinearKernel(ScaledMMLinearKernel):
    @classmethod
    def is_supported(
        cls, compute_capability: int | None = None
    ) -> tuple[bool, str | None]:
        if current_platform.is_cuda_alike():
            return True, None
        return False, "Requires ROCm or CUDA."

    @classmethod
    def can_implement(cls, c: ScaledMMLinearLayerConfig) -> tuple[bool, str | None]:
        if not c.input_symmetric:
            return False, "Only symmetric input is supported."
        return True, None

    def process_weights_after_loading(self, layer: torch.nn.Module) -> None:
        weight = getattr(layer, self.w_q_name)
        replace_parameter(
            layer,
            self.w_q_name,
            torch.nn.Parameter(weight.t().data, requires_grad=False),
        )

        # INPUT SCALE
        if self.config.is_static_input_scheme:
            input_scale = getattr(layer, self.i_s_name)
            replace_parameter(
                layer,
                self.i_s_name,
                torch.nn.Parameter(input_scale.max(), requires_grad=False),
            )
            setattr(layer, self.i_zp_name, None)
        else:
            setattr(layer, self.i_s_name, None)
            setattr(layer, self.i_zp_name, None)

        setattr(layer, self.azp_adj_name, None)

    def apply_weights(
        self,
        layer: torch.nn.Module,
        x: torch.Tensor,
        bias: torch.Tensor | None = None,
    ) -> torch.Tensor:
        w_q, w_s, i_s, i_zp, azp_adj = self._get_weight_params(layer)

        x_q, x_s, x_zp = ops.scaled_int8_quant(
            x.contiguous(), i_s, i_zp, symmetric=True
        )

        assert x_zp is None, "Triton kernel only supports symmetric quantization"

        return triton_scaled_mm(
            x_q, w_q, scale_a=x_s, scale_b=w_s, out_dtype=x.dtype, bias=bias
        )

apply_weights 

apply_weights(
    layer: Module, x: Tensor, bias: Tensor | None = None
) -> Tensor
Source code in vllm/model_executor/layers/quantization/kernels/scaled_mm/triton.py 
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
def apply_weights(
    self,
    layer: torch.nn.Module,
    x: torch.Tensor,
    bias: torch.Tensor | None = None,
) -> torch.Tensor:
    w_q, w_s, i_s, i_zp, azp_adj = self._get_weight_params(layer)

    x_q, x_s, x_zp = ops.scaled_int8_quant(
        x.contiguous(), i_s, i_zp, symmetric=True
    )

    assert x_zp is None, "Triton kernel only supports symmetric quantization"

    return triton_scaled_mm(
        x_q, w_q, scale_a=x_s, scale_b=w_s, out_dtype=x.dtype, bias=bias
    )

can_implement classmethod 

can_implement(
    c: ScaledMMLinearLayerConfig,
) -> tuple[bool, str | None]
Source code in vllm/model_executor/layers/quantization/kernels/scaled_mm/triton.py 
26
27
28
29
30
@classmethod
def can_implement(cls, c: ScaledMMLinearLayerConfig) -> tuple[bool, str | None]:
    if not c.input_symmetric:
        return False, "Only symmetric input is supported."
    return True, None

is_supported classmethod 

is_supported(
    compute_capability: int | None = None,
) -> tuple[bool, str | None]
Source code in vllm/model_executor/layers/quantization/kernels/scaled_mm/triton.py 
18
19
20
21
22
23
24
@classmethod
def is_supported(
    cls, compute_capability: int | None = None
) -> tuple[bool, str | None]:
    if current_platform.is_cuda_alike():
        return True, None
    return False, "Requires ROCm or CUDA."

process_weights_after_loading 

process_weights_after_loading(layer: Module) -> None
Source code in vllm/model_executor/layers/quantization/kernels/scaled_mm/triton.py 
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
def process_weights_after_loading(self, layer: torch.nn.Module) -> None:
    weight = getattr(layer, self.w_q_name)
    replace_parameter(
        layer,
        self.w_q_name,
        torch.nn.Parameter(weight.t().data, requires_grad=False),
    )

    # INPUT SCALE
    if self.config.is_static_input_scheme:
        input_scale = getattr(layer, self.i_s_name)
        replace_parameter(
            layer,
            self.i_s_name,
            torch.nn.Parameter(input_scale.max(), requires_grad=False),
        )
        setattr(layer, self.i_zp_name, None)
    else:
        setattr(layer, self.i_s_name, None)
        setattr(layer, self.i_zp_name, None)

    setattr(layer, self.azp_adj_name, None)