Skip to content

Enable FP8 Per-Tensor Scales and Integrate Marlin/MoE Kernels Repo for ROCm #2062

Enable FP8 Per-Tensor Scales and Integrate Marlin/MoE Kernels Repo for ROCm

Enable FP8 Per-Tensor Scales and Integrate Marlin/MoE Kernels Repo for ROCm #2062

Annotations

2 errors

build (cuda-trtllm)  /  build-and-push

cancelled Dec 18, 2024 in 7m 48s