[Kernel] Pass a device pointer into the quantize kernel for the scales (#5159)
This commit is contained in:
+1
-1
@@ -265,7 +265,7 @@ def scaled_fp8_quant(
|
||||
|
||||
# int8
|
||||
def static_scaled_int8_quant(input: torch.Tensor,
|
||||
scale: float) -> torch.Tensor:
|
||||
scale: torch.Tensor) -> torch.Tensor:
|
||||
"""
|
||||
Quantize the input tensor to int8 and return the quantized tensor.
|
||||
|
||||
|
||||
Reference in New Issue
Block a user