ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 2 CUDA devices: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes Device 1: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes | model | size | params | backend | ngl | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | --------------: | -------------------: | | qwen3 4B Q5_K - Medium | 2.69 GiB | 4.02 B | CUDA | 35 | pp8 | 385.17 ± 7.14 | | qwen3 4B Q5_K - Medium | 2.69 GiB | 4.02 B | CUDA | 35 | tg128 | 75.17 ± 0.92 | build: 92bb442ad (7040)