Open felixgeen opened 1 year ago
(RVM) C:\Users\7950X\RobustVideoMatting>python inference.py --variant mobilenetv3 --checkpoint checkpoint/rvm_mobilenetv3.pth --device cuda --input-source input/16by9.mp4 --output-type video --output-composition output/com.mp4 --output-alpha output/alp.mp4 --output-foreground output/for.mp4 --output-video-mbps 15 --seq-chunk 1 0%|▏ | 1/503 [00:02<17:01, 2.03s/it]Traceback (most recent call last): File "inference.py", line 204, in <module> progress=not args.disable_progress File "inference.py", line 169, in convert convert_video(self.model, device=self.device, dtype=torch.float32, *args, **kwargs) File "inference.py", line 127, in convert_video fgr, pha, *rec = model(src, *rec, downsample_ratio) File "C:\Users\7950X\anaconda3\envs\RVM\lib\site-packages\torch\nn\modules\module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) RuntimeError: nvrtc: error: invalid value for --gpu-architecture (-arch) nvrtc compilation failed: #define NAN __int_as_float(0x7fffffff) #define POS_INFINITY __int_as_float(0x7f800000) #define NEG_INFINITY __int_as_float(0xff800000) template<typename T> __device__ T maximum(T a, T b) { return isnan(a) ? a : (a > b ? a : b); } template<typename T> __device__ T minimum(T a, T b) { return isnan(a) ? a : (a < b ? a : b); } extern "C" __global__ void fused_sigmoid_mul(float* t0, float* t1, float* aten_mul) { { float v = __ldg(t0 + (576 * (((512 * blockIdx.x + threadIdx.x) / 576) % 128) + 32 * (((512 * blockIdx.x + threadIdx.x) / 32) % 18)) + (512 * blockIdx.x + threadIdx.x) % 32); float v_1 = __ldg(t1 + ((512 * blockIdx.x + threadIdx.x) / 576) % 128); aten_mul[((((512 * blockIdx.x + threadIdx.x) / 73728) * 73728 + 576 * (((512 * blockIdx.x + threadIdx.x) / 576) % 128)) + 32 * (((512 * blockIdx.x + threadIdx.x) / 32) % 18)) + (512 * blockIdx.x + threadIdx.x) % 32] = v * (1.f / (1.f + (expf(0.f - v_1)))); } } 0%| | 1/503 [00:02<21:32, 2.57s/it] (RVM) C:\Users\7950X\RobustVideoMatting>
Any progress?
没gpu吧