cudnn 7.6.4 for cuda 10.1 doesn't work for cuda 10.2

2021-06-11 02:13:58.205671: E tensorflow/stream_executor/cuda/cuda_dnn.cc:319] Loaded runtime CuDNN library: 7.4.2 but source was compiled with: 7.6.4.  CuDNN library major and minor version needs to match or have higher minor version in case of CuDNN 7.0 or later version. If using a binary install, upgrade your CuDNN library.  If building from sources, make sure the library loaded at runtime is compatible with the version specified during compile configuration.
Traceback (most recent call last):
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/client/session.py", line 1365, in _do_call
    return fn(*args)
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/client/session.py", line 1349, in _run_fn
    return self._call_tf_sessionrun(options, feed_dict, fetch_list,
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/client/session.py", line 1441, in _call_tf_sessionrun
    return tf_session.TF_SessionRun_wrapper(self._session, options, feed_dict,
tensorflow.python.framework.errors_impl.UnknownError: 2 root error(s) found.
  (0) Unknown: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
	 [[{{node transform_net1/tconv1/Conv2D}}]]
	 [[conv7/Relu/_225]]
  (1) Unknown: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
	 [[{{node transform_net1/tconv1/Conv2D}}]]
0 successful operations.
0 derived errors ignored.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "Feature_extraction.py", line 98, in <module>
    evaluate()
  File "Feature_extraction.py", line 76, in evaluate
    eval_one_epoch(sess, ops)
  File "Feature_extraction.py", line 92, in eval_one_epoch
    feat = sess.run([ops['feat']], feed_dict=feed_dict)
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/client/session.py", line 957, in run
    result = self._run(None, fetches, feed_dict, options_ptr,
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/client/session.py", line 1180, in _run
    results = self._do_run(handle, final_targets, final_fetches,
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/client/session.py", line 1358, in _do_run
    return self._do_call(_run_fn, feeds, fetches, targets, options,
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/client/session.py", line 1384, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.UnknownError: 2 root error(s) found.
  (0) Unknown: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
	 [[node transform_net1/tconv1/Conv2D (defined at /scratch3/research/code/DJ-RN-dawnlight/pointnet/utils/tf_util.py:158) ]]
	 [[conv7/Relu/_225]]
  (1) Unknown: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
	 [[node transform_net1/tconv1/Conv2D (defined at /scratch3/research/code/DJ-RN-dawnlight/pointnet/utils/tf_util.py:158) ]]
0 successful operations.
0 derived errors ignored.

Errors may have originated from an input operation.
Input Source operations connected to node transform_net1/tconv1/Conv2D:
 transform_net1/ExpandDims (defined at /scratch3/research/code/DJ-RN-dawnlight/pointnet/models/transform_nets.py:18)
 transform_net1/tconv1/weights/read (defined at /scratch3/research/code/DJ-RN-dawnlight/pointnet/utils/tf_util.py:22)

Input Source operations connected to node transform_net1/tconv1/Conv2D:
 transform_net1/ExpandDims (defined at /scratch3/research/code/DJ-RN-dawnlight/pointnet/models/transform_nets.py:18)
 transform_net1/tconv1/weights/read (defined at /scratch3/research/code/DJ-RN-dawnlight/pointnet/utils/tf_util.py:22)

Original stack trace for 'transform_net1/tconv1/Conv2D':
  File "Feature_extraction.py", line 98, in <module>
    evaluate()
  File "Feature_extraction.py", line 57, in evaluate
    feat = MODEL.get_model(pointclouds_pl, is_training_pl)
  File "/scratch3/research/code/DJ-RN-dawnlight/pointnet/models/pointnet_hico.py", line 27, in get_model
    transform = input_transform_net(point_cloud, is_training, bn_decay, K=3)
  File "/scratch3/research/code/DJ-RN-dawnlight/pointnet/models/transform_nets.py", line 19, in input_transform_net
    net = tf_util.conv2d(input_image, 64, [1,3],
  File "/scratch3/research/code/DJ-RN-dawnlight/pointnet/utils/tf_util.py", line 158, in conv2d
    outputs = tf.nn.conv2d(inputs, kernel,
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/ops/nn_ops.py", line 2006, in conv2d
    return gen_nn_ops.conv2d(input,  # pylint: disable=redefined-builtin
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/ops/gen_nn_ops.py", line 965, in conv2d
    _, _, _op, _outputs = _op_def_library._apply_op_helper(
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/framework/op_def_library.py", line 742, in _apply_op_helper
    op = g._create_op_internal(op_type_name, inputs, dtypes=None,
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/framework/ops.py", line 3319, in _create_op_internal
    ret = Operation(
  File "/scratch3/venv/djrn/lib/python3.8/site-packages/tensorflow/python/framework/ops.py", line 1791, in __init__
    self._traceback = tf_stack.extract_stack()