does this omniglot thing mds look right to you?

[brando90]

(metalearning_gpu) brando9~ $ python -m meta_dataset.dataset_conversion.convert_datasets_to_records \
>   --dataset=omniglot \
>   --omniglot_data_root=$MDS_DATA_PATH/omniglot \
>   --splits_root=$SPLITS \
>   --records_root=$RECORDS
2023-01-04 14:05:08.829399: I tensorflow/core/platform/cpu_feature_guard.cc:193] 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.
2023-01-04 14:05:13.321368: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:
2023-01-04 14:05:13.321582: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:
2023-01-04 14:05:13.321596: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.

2023-01-04 14:05:30.511938: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudnn.so.8'; dlerror: libcudnn.so.8: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:
2023-01-04 14:05:30.511989: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1934] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
I0104 14:05:30.512454 140076921520768 convert_datasets_to_records.py:151] Creating Omniglot specification and records in directory /lfs/ampere4/0/brando9/data/mds/records/omniglot...

?

[brando90]

not sure what it's complaining about lib64 seems there.

(metalearning_gpu) brando9/usr/local/cuda-11.7 $ python -m meta_dataset.dataset_conversion.convert_datasets_to_records   --dataset=omniglot   --omniglot_data_root=$MDS_DATA_PATH/omniglot   --splits_root=$SPLITS   --records_root=$RECORDS
2023-01-04 14:15:01.939017: I tensorflow/core/platform/cpu_feature_guard.cc:193] 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.
2023-01-04 14:15:03.542706: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:
2023-01-04 14:15:03.542819: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:
2023-01-04 14:15:03.542834: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
2023-01-04 14:15:06.476005: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudnn.so.8'; dlerror: libcudnn.so.8: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:/usr/local/cuda-11.7/lib64:
2023-01-04 14:15:06.476069: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1934] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
I0104 14:15:06.476591 139645946860160 convert_datasets_to_records.py:151] Creating Omniglot specification and records in directory /lfs/ampere4/0/brando9/data/mds/records/omniglot...
(metalearning_gpu) brando9/usr/local/cuda-11.7 $ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Tue_May__3_18:49:52_PDT_2022
Cuda compilation tools, release 11.7, V11.7.64
Build cuda_11.7.r11.7/compiler.31294372_0
(metalearning_gpu) brando9/usr/local/cuda-11.7 $ ls /usr/local/cuda-11.7/
bin  compute-sanitizer  DOCS  EULA.txt  extras  include  lib64  libnvvp  nsight-compute-2022.2.0  nsightee_plugins  nsight-systems-2022.1.3  nvml  nvvm  README  share  src  targets  tools  version.json
(metalearning_gpu) brando9/usr/local/cuda-11.7 $

will ignore it for now.

Until you report whats up.

[patricks-lab]

I recall having the same issue when I did mines (even with compiling w/ GPU), seems like it's ignoring GPU when creating tfrecords. I think it's fine for now? They look like warnings to me.