ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 2 ROCm devices: Device 0: AMD Radeon Graphics, gfx1201 (0x1201), VMM: no, Wave Size: 32 Device 1: AMD Radeon Graphics, gfx1201 (0x1201), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | n_ubatch | fa | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -------: | -: | --------------: | -------------------: | /opt/llama.cpp/ggml/src/ggml-cuda/ggml-cuda.cu:92: ROCm error /usr/local/lib64/libggml-base.so.0(+0x3565) [0x7f0224efc565] /usr/local/lib64/libggml-base.so.0(ggml_print_backtrace+0x1eb) [0x7f0224efc92b] /usr/local/lib64/libggml-base.so.0(ggml_abort+0x11f) [0x7f0224efcaaf] /usr/local/lib64/libggml-hip.so.0(+0x2fd1ef2) [0x7f0227f8aef2] /usr/local/lib64/libggml-hip.so.0(+0x2fe6757) [0x7f0227f9f757] /usr/local/lib64/libggml-hip.so.0(+0x2fe4d2c) [0x7f0227f9dd2c] /usr/local/lib64/libggml-hip.so.0(+0x2fe3ef7) [0x7f0227f9cef7] /usr/local/lib64/libggml-hip.so.0(+0x2fde9db) [0x7f0227f979db] /usr/local/lib64/libggml-hip.so.0(+0x2fdacab) [0x7f0227f93cab] /usr/local/lib64/libggml-hip.so.0(+0x2fd716f) [0x7f0227f9016f] /usr/local/lib64/libggml-base.so.0(ggml_backend_sched_graph_compute_async+0x7f3) [0x7f0224f16f63] /usr/local/lib64/libllama.so.0(_ZN13llama_context13graph_computeEP11ggml_cgraphb+0xa0) [0x7f0228658a90] /usr/local/lib64/libllama.so.0(_ZN13llama_context14process_ubatchERK12llama_ubatch14llm_graph_typeP22llama_memory_context_iR11ggml_status+0xe2) [0x7f022865a722] /usr/local/lib64/libllama.so.0(_ZN13llama_context6decodeERK11llama_batch+0x3bf) [0x7f022865f5df] /usr/local/lib64/libllama.so.0(llama_decode+0xe) [0x7f02286603fe] /usr/local/bin/llama-bench() [0x40acdb] /usr/local/bin/llama-bench() [0x4087ec] /lib64/libc.so.6(+0x35b5) [0x7f02248925b5] /lib64/libc.so.6(__libc_start_main+0x88) [0x7f0224892668] /usr/local/bin/llama-bench() [0x409b65] ✖ ! [rocm-7.9] gemma-3-27b-it-BF16-00001-of-00002__hblt0__fa1 __longctx16384 failed (exit 0)