Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #24530
Annotations
2 errors
|
|
|
Build and Test
The operation was canceled.
|
The logs for this run have expired and are no longer available.
Loading