ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 ROCm devices:
  Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32
| model                          |       size |     params | backend    | ngl | n_ubatch | fa | mmap |            test |                  t/s |
| ------------------------------ | ---------: | ---------: | ---------- | --: | -------: | -: | ---: | --------------: | -------------------: |
| llama 70B Q8_0                 |  75.65 GiB |    70.55 B | ROCm       |  99 |     2048 |  1 |    0 | pp2048 @ d32768 |         15.05 ± 0.00 |
HW Exception by GPU node-1 (Agent handle: 0x13295700) reason :GPU Hang
✖ ! [rocm7_rc-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1__longctx32768 failed (exit 134)