해결하고자 하는 문제
YOLOV4를 통해이미지 파일을 object detect 시도하였는데 weights 모델 관련 OS Error가 발생합니다.
코드 혹은 오류
(tf2) jaker@jaker:~/다운로드/tensorflow-yolov4-tflite-master$ python detect.py --weights ./data/yolov4.weights --framework tf --size 608 --image ./data/kite.jpg
2020-08-18 16:05:39.603963: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.257904: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcuda.so.1
2020-08-18 16:05:40.288791: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.289049: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce RTX 2070 SUPER computeCapability: 7.5
coreClock: 1.77GHz coreCount: 40 deviceMemorySize: 7.79GiB deviceMemoryBandwidth: 417.29GiB/s
2020-08-18 16:05:40.289067: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.290120: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-18 16:05:40.291104: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-18 16:05:40.291275: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-18 16:05:40.292310: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-18 16:05:40.292886: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-18 16:05:40.295438: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-18 16:05:40.295573: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.295956: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.296211: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-18 16:05:40.309082: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN)to use the following CPU instructions in performance-critical operations: AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2020-08-18 16:05:40.313220: I tensorflow/core/platform/profile_utils/cpu_utils.cc:104] CPU Frequency: 3699850000 Hz
2020-08-18 16:05:40.313384: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55cdeab04e90 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-08-18 16:05:40.313396: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
2020-08-18 16:05:40.390526: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.390833: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55cdeab90d80 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2020-08-18 16:05:40.390847: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): GeForce RTX 2070 SUPER, Compute Capability 7.5
2020-08-18 16:05:40.390988: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.391201: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce RTX 2070 SUPER computeCapability: 7.5
coreClock: 1.77GHz coreCount: 40 deviceMemorySize: 7.79GiB deviceMemoryBandwidth: 417.29GiB/s
2020-08-18 16:05:40.391220: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.391233: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-18 16:05:40.391242: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-18 16:05:40.391250: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-18 16:05:40.391258: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-18 16:05:40.391266: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-18 16:05:40.391274: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-18 16:05:40.391303: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.391519: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.391714: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-18 16:05:40.391730: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.706537: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1257] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-08-18 16:05:40.706565: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1263] 0
2020-08-18 16:05:40.706570: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1276] 0: N
2020-08-18 16:05:40.706700: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.706963: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.707180: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1402] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 6125 MB memory) -> physical GPU (device: 0, name: GeForce RTX 2070 SUPER, pci bus id: 0000:01:00.0, compute capability: 7.5)
2020-08-18 16:05:40.736659: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.736910: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce RTX 2070 SUPER computeCapability: 7.5
coreClock: 1.77GHz coreCount: 40 deviceMemorySize: 7.79GiB deviceMemoryBandwidth: 417.29GiB/s
2020-08-18 16:05:40.736955: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.736971: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-18 16:05:40.736979: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-18 16:05:40.736986: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-18 16:05:40.736993: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-18 16:05:40.737000: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-18 16:05:40.737008: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-18 16:05:40.737044: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737296: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737488: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-18 16:05:40.737507: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1257] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-08-18 16:05:40.737512: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1263] 0
2020-08-18 16:05:40.737515: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1276] 0: N
2020-08-18 16:05:40.737557: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737773: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737971: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1402] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 6125 MB memory) -> physical GPU (device: 0, name: GeForce RTX 2070 SUPER, pci bus id: 0000:01:00.0, compute capability: 7.5)
Traceback (most recent call last):
File "detect.py", line 90, in
app.run(main)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/absl/app.py", line 299, in run
_run_main(main, args)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/absl/app.py", line 250, in _run_main
sys.exit(main(argv))
File "detect.py", line 63, in main
saved_model_loaded = tf.saved_model.load(FLAGS.weights, tags=[tag_constants.SERVING])
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/load.py", line 603, in load
return load_internal(export_dir, tags, options)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/load.py", line 614, in load_internal
loader_impl.parse_saved_model_with_debug_info(export_dir))
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/loader_impl.py", line 56, in parse_saved_model_with_debug_info
saved_model = _parse_saved_model(export_dir)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/loader_impl.py", line 113, in parse_saved_model
constants.SAVED_MODEL_FILENAME_PB))
OSError: SavedModel file does not exist at: ./data/yolov4.weights/{saved_model.pbtxt|saved_model.pb}
환경
OS : Ubuntu 20.04
언어 : Python 3.7.7
Library : Yolo V4, tensorflow 2.3.0(GPU), OpenCV 4.2.0, CUDA 10.1, CuDNN 7.6.5
GPU : RTX2070 super
시도해본 방법
살려주세요 ㅠㅠㅠㅠ
해결하고자 하는 문제
YOLOV4를 통해이미지 파일을 object detect 시도하였는데 weights 모델 관련 OS Error가 발생합니다.
코드 혹은 오류
2020-08-18 16:05:39.603963: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.257904: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcuda.so.1
2020-08-18 16:05:40.288791: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.289049: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce RTX 2070 SUPER computeCapability: 7.5
coreClock: 1.77GHz coreCount: 40 deviceMemorySize: 7.79GiB deviceMemoryBandwidth: 417.29GiB/s
2020-08-18 16:05:40.289067: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.290120: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-18 16:05:40.291104: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-18 16:05:40.291275: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-18 16:05:40.292310: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-18 16:05:40.292886: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-18 16:05:40.295438: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-18 16:05:40.295573: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.295956: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.296211: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-18 16:05:40.309082: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN)to use the following CPU instructions in performance-critical operations: AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2020-08-18 16:05:40.313220: I tensorflow/core/platform/profile_utils/cpu_utils.cc:104] CPU Frequency: 3699850000 Hz
2020-08-18 16:05:40.313384: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55cdeab04e90 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-08-18 16:05:40.313396: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
2020-08-18 16:05:40.390526: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.390833: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55cdeab90d80 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2020-08-18 16:05:40.390847: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): GeForce RTX 2070 SUPER, Compute Capability 7.5
2020-08-18 16:05:40.390988: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.391201: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce RTX 2070 SUPER computeCapability: 7.5
coreClock: 1.77GHz coreCount: 40 deviceMemorySize: 7.79GiB deviceMemoryBandwidth: 417.29GiB/s
2020-08-18 16:05:40.391220: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.391233: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-18 16:05:40.391242: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-18 16:05:40.391250: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-18 16:05:40.391258: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-18 16:05:40.391266: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-18 16:05:40.391274: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-18 16:05:40.391303: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.391519: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.391714: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-18 16:05:40.391730: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.706537: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1257] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-08-18 16:05:40.706565: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1263] 0
2020-08-18 16:05:40.706570: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1276] 0: N
2020-08-18 16:05:40.706700: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.706963: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.707180: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1402] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 6125 MB memory) -> physical GPU (device: 0, name: GeForce RTX 2070 SUPER, pci bus id: 0000:01:00.0, compute capability: 7.5)
2020-08-18 16:05:40.736659: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.736910: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce RTX 2070 SUPER computeCapability: 7.5
coreClock: 1.77GHz coreCount: 40 deviceMemorySize: 7.79GiB deviceMemoryBandwidth: 417.29GiB/s
2020-08-18 16:05:40.736955: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-18 16:05:40.736971: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-18 16:05:40.736979: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-18 16:05:40.736986: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-18 16:05:40.736993: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-18 16:05:40.737000: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-18 16:05:40.737008: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-18 16:05:40.737044: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737296: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737488: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-18 16:05:40.737507: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1257] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-08-18 16:05:40.737512: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1263] 0
2020-08-18 16:05:40.737515: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1276] 0: N
2020-08-18 16:05:40.737557: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737773: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-18 16:05:40.737971: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1402] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 6125 MB memory) -> physical GPU (device: 0, name: GeForce RTX 2070 SUPER, pci bus id: 0000:01:00.0, compute capability: 7.5)
Traceback (most recent call last):
File "detect.py", line 90, in
app.run(main)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/absl/app.py", line 299, in run
_run_main(main, args)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/absl/app.py", line 250, in _run_main
sys.exit(main(argv))
File "detect.py", line 63, in main
saved_model_loaded = tf.saved_model.load(FLAGS.weights, tags=[tag_constants.SERVING])
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/load.py", line 603, in load
return load_internal(export_dir, tags, options)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/load.py", line 614, in load_internal
loader_impl.parse_saved_model_with_debug_info(export_dir))
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/loader_impl.py", line 56, in parse_saved_model_with_debug_info
saved_model = _parse_saved_model(export_dir)
File "/home/jaker/.pyenv/versions/tf2/lib/python3.7/site-packages/tensorflow/python/saved_model/loader_impl.py", line 113, in parse_saved_model
constants.SAVED_MODEL_FILENAME_PB))
OSError: SavedModel file does not exist at: ./data/yolov4.weights/{saved_model.pbtxt|saved_model.pb}
환경
OS : Ubuntu 20.04
언어 : Python 3.7.7
Library : Yolo V4, tensorflow 2.3.0(GPU), OpenCV 4.2.0, CUDA 10.1, CuDNN 7.6.5
GPU : RTX2070 super
시도해본 방법
살려주세요 ㅠㅠㅠㅠ