Skip to content

[Kernel] Use CUTLASS kernels for the FP8 layers with Bias #6270

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
robertgshaw2-redhat merged 1 commit into from
Jul 15, 2024
Merged

[Kernel] Use CUTLASS kernels for the FP8 layers with Bias #6270

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions vllm/model_executor/layers/quantization/utils/w8a8_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -112,15 +112,16 @@ def apply_fp8_linear(
# If dynamic, layer.input_scale is None and x_scale computed from x.
# If static, layer.input_scale is scalar and x_scale is input_scale.

if bias is None and cutlass_fp8_supported:
if cutlass_fp8_supported:
qinput, x_scale = ops.scaled_fp8_quant(input, input_scale)

# Fused GEMM_DQ
output = ops.cutlass_scaled_mm(qinput,
weight,
out_dtype=input.dtype,
scale_a=x_scale,
scale_b=weight_scale)
scale_b=weight_scale,
bias=bias)

else:
qinput, x_scale = ops.scaled_fp8_quant(input,
Expand Down