(base) root@c53a81a8cfce:~/data1/JNeRF# python tools/run_net.py --config-file ./projects/ngp/configs/ngp_kitti.py
[i 1006 05:47:24.488160 08 compiler.py:955] Jittor(1.3.5.18) src: /opt/conda/lib/python3.7/site-packages/jittor
[i 1006 05:47:24.494011 08 compiler.py:956] g++ at /usr/bin/g++(7.5.0)
[i 1006 05:47:24.494112 08 compiler.py:957] cache_path: /root/.cache/jittor/jt1.3.5/g++7.5.0/py3.7.11/Linux-3.10.0-1xb1/IntelRXeonRSilx06/default
[i 1006 05:47:24.501204 08 __init__.py:411] Found /usr/local/cuda/bin/nvcc(11.3.109) at /usr/local/cuda/bin/nvcc.
[i 1006 05:47:24.507581 08 __init__.py:411] Found addr2line(2.30) at /usr/bin/addr2line.
[i 1006 05:47:24.732417 08 compiler.py:1010] cuda key:cu11.3.109_sm_86
[i 1006 05:47:25.065434 08 __init__.py:227] Total mem: 251.58GB, using 16 procs for compiling.
[i 1006 05:47:25.270457 08 jit_compiler.cc:28] Load cc_path: /usr/bin/g++
[i 1006 05:47:25.378972 08 py_var_tracer.cc:22] Load trace_py_var: 3
[i 1006 05:47:25.379371 08 init.cc:62] Found cuda archs: [86,]
[i 1006 05:47:26.155324 08 compile_extern.py:517] mpicc not found, distribution disabled.
[i 1006 05:47:26.222870 08 compile_extern.py:30] found /usr/local/cuda/include/cublas.h
[i 1006 05:47:26.239217 08 compile_extern.py:30] found /usr/local/cuda/lib64/libcublas.so
[i 1006 05:47:26.239343 08 compile_extern.py:30] found /usr/local/cuda/lib64/libcublasLt.so.11
[i 1006 05:47:27.208520 08 compile_extern.py:30] found /usr/include/cudnn.h
[i 1006 05:47:27.260239 08 compile_extern.py:30] found /usr/lib/x86_64-linux-gnu/libcudnn.so.8
[i 1006 05:47:27.260388 08 compile_extern.py:30] found /usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8
[i 1006 05:47:27.263994 08 compile_extern.py:30] found /usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8
[i 1006 05:47:27.264625 08 compile_extern.py:30] found /usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8
[i 1006 05:47:27.304770 08 compile_extern.py:30] found /usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8
[i 1006 05:47:28.222796 08 compile_extern.py:30] found /usr/local/cuda/include/curand.h
[i 1006 05:47:28.293317 08 compile_extern.py:30] found /usr/local/cuda/lib64/libcurand.so
[i 1006 05:47:28.314973 08 compile_extern.py:30] found /usr/local/cuda/include/cufft.h
[i 1006 05:47:28.395065 08 compile_extern.py:30] found /usr/local/cuda/lib64/libcufft.so
[i 1006 05:47:29.840524 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.836172 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.837645 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.837810 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.837954 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.838093 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.838240 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.838404 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.838558 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.838738 08 cuda_flags.cc:32] CUDA enabled.
[i 1006 05:47:33.840907 08 cuda_flags.cc:32] CUDA enabled.
Loading config from: ./projects/ngp/configs/ngp_kitti.py
Loading /root/data1/PanopticNeRF/datasets/KITTI-360/data_3d_bboxes/train/2013_05_28_drive_0000_sync.xml…
Loaded 25359 instances
load meta_00 done
0%| | 0/40000 [00:00<?, ?it/s][4096,3,]
[e 1006 05:47:50.810397 08 mem_info.cc:96] {1:169, } 0
0%| | 0/40000 [00:00<?, ?it/s]
Traceback (most recent call last):
File “tools/run_net.py”, line 55, in
main()
File “tools/run_net.py”, line 48, in main
runner.train()
File “/root/data1/JNeRF/python/jnerf/runner/p_runner.py”, line 69, in train
print(img_ids[0])
File “/opt/conda/lib/python3.7/site-packages/jittor/init.py”, line 1819, in vtos
data_str = f"jt.Var({v.data}, dtype={v.dtype})"
RuntimeError: [f 1006 05:47:50.810465 08 executor.cc:666]
Execute fused operator(189/195) failed.
[JIT Source]: /root/.cache/jittor/jt1.3.5/g++7.5.0/py3.7.11/Linux-3.10.0-1xb1/IntelRXeonRSilx06/default/cu11.3.109_sm_86/jit/getitem__Ti_int32__IDIM_1__ODIM_1__FOV_0__VD_1__IV0_0__IO0__1__VS0_1__VST0_int32__LO0_41_____hash_9fe891ca90685233_op.cc
[OP TYPE]: getitem
[Input]: int32[64,], int32[4096,],
[Output]: int32[4096,],
[Async Backtrace]: —
tools/run_net.py:55 <>
tools/run_net.py:48
/root/data1/JNeRF/python/jnerf/runner/p_runner.py:65
/root/data1/JNeRF/python/jnerf/dataset/p_dataset.py:327 <next>
/opt/conda/lib/python3.7/site-packages/jittor/contrib.py:192
[Reason]: [f 1006 05:47:50.810157 08 helper_cuda.h:128] CUDA error at /opt/conda/lib/python3.7/site-packages/jittor/src/executor.cc:619 code=700( cudaErrorIllegalAddress ) cudaDeviceSynchronize()
[e 1006 05:47:51.952985 08 helper_cuda.h:117] Peek CUDA error at /opt/conda/lib/python3.7/site-packages/jittor/src/mem/allocator/cuda_dual_allocator.h:101 code=700( cudaErrorIllegalAddress ) _cudaLaunchHostFunc(0, &to_free_allocation, 0)
terminate called after throwing an instance of ‘std::runtime_error’
what(): [f 1006 05:47:52.099821 08 helper_cuda.h:128] CUDA error at /opt/conda/lib/python3.7/site-packages/jittor/extern/cuda/cudnn/src/cudnn_wrapper.cc:34 code=4( CUDNN_STATUS_INTERNAL_ERROR ) cudnnDestroy(cudnn_handle)
Aborted (core dumped)
你好，我用如下代码读取图像
rays_rgb_np.append(rays_rgb)
self.rays_rgb_np = jt.array(np.concatenate(rays_rgb_np))
img_index=self.shuffle_index[self.idx_now:self.idx_now+self.batch_size]
rays_rgb = self.rays_rgb_np[img_index]
print(rays_rgb.shape)可以运行，结果正确 (B,3)
但是print(rays_rgb[0])就运行不了，报如上错误
