Skip to content

[float8] Allow specifying arbitrary dtype for each tensor #5332

[float8] Allow specifying arbitrary dtype for each tensor

[float8] Allow specifying arbitrary dtype for each tensor #5332

test (CUDA Nightly, linux.g5.12xlarge.nvidia.gpu, --pre torch==2.6.0.dev20241101 --index-url http...  /  linux-job

succeeded Nov 22, 2024 in 1h 9m 33s