fix: correct quantize function name in fused router test
This commit is contained in:
@@ -31,8 +31,8 @@ def test_fused_router():
|
||||
|
||||
# Create random BF16 gate weight and quantize to NVFP4
|
||||
W_gate_bf16 = torch.randn(K, N, dtype=torch.bfloat16, device=device)
|
||||
from dsv4.ops.quantize import quantize_weight_nvfp4
|
||||
w_fp4, w_sf, ws2_val, _ = quantize_weight_nvfp4(W_gate_bf16)
|
||||
from dsv4.ops.quantize import quantize_weight_to_nvfp4
|
||||
w_fp4, w_sf, ws2_val, _ = quantize_weight_to_nvfp4(W_gate_bf16)
|
||||
|
||||
# Build Nvfp4Linear for reference path
|
||||
gate_lin = Nvfp4Linear(in_features=K, out_features=N, device=device)
|
||||
|
||||
Reference in New Issue
Block a user