我定义了tensorflow 2 tensorflow_hub的yamnet模型：

yamnet_model = hub.load('https://tfhub.dev/google/yamnet/1')
wav_data, sample_rate = librosa.load(wav_file_name, sr=16000)

scores, spectogram = yamnet_model(wav_data)

我使用Cuda 11.4
调试上面最后一行时：

NotFoundError
Graph execution error:
2023-02-28 09:56:18.617965: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1532] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 1603 MB memory:  -> device: 0, name: NVIDIA GeForce MX350, pci bus id: 0000:2b:00.0, compute capability: 6.1
2023-02-28 09:56:34.527922: I tensorflow/stream_executor/cuda/cuda_dnn.cc:384] Loaded cuDNN version 8204
2023-02-28 09:56:34.761531: W tensorflow/core/framework/op_kernel.cc:1745] OP_REQUIRES failed at conv_ops.cc:1130 : NOT_FOUND: No algorithm worked!  Error messages:
  <br>Profiling failure on CUDNN engine 1: UNKNOWN: CUDNN_STATUS_EXECUTION_FAILED<br>
in tensorflow/stream_executor/cuda/cuda_dnn.cc(4023): 'cudnnConvolutionForward( cudnn.handle(), alpha, input_nd_.handle(), input_data.opaque(), filter_.handle(), filter_data.opaque(), conv_.handle(), ToConvForwardAlgo(algo), scratch_memory.opaque(), scratch_memory.size(), beta, output_nd_.handle(), output_data.opaque())'
  <br>Profiling failure on CUDNN engine 0: UNKNOWN: CUDNN_STATUS_EXECUTION_FAILED<br>
in tensorflow/stream_executor/cuda/cuda_dnn.cc(4023): 'cudnnConvolutionForward( cudnn.handle(), alpha, input_nd_.handle(), input_data.opaque(), filter_.handle(), filter_data.opaque(), conv_.handle(), ToConvForwardAlgo(algo), scratch_memory.opaque(), scratch_memory.size(), beta, output_nd_.handle(), output_data.opaque())'
  <br>Profiling failure on CUDNN engine 2: UNKNOWN: CUDNN_STATUS_EXECUTION_FAILED<br>
in tensorflow/stream_executor/cuda/cuda_dnn.cc(4023): 'cudnnConvolutionForward( cudnn.handle(), alpha, input_nd_.handle(), input_data.opaque(), filter_.handle(), filter_data.opaque(), conv_.handle(), ToConvForwardAlgo(algo), scratch_memory.opaque(), scratch_memory.size(), beta, output_nd_.handle(), output_data.opaque())'
  <br>Profiling failure on CUDNN engine 5: UNKNOWN: CUDNN_STATUS_EXECUTION_FAILED<br>
in tensorflow/stream_executor/cuda/cuda_dnn.cc(4023): 'cudnnConvolutionForward( cudnn.handle(), alpha, input_nd_.handle(), input_data.opaque(), filter_.handle(), filter_data.opaque(), conv_.handle(), ToConvForwardAlgo(algo), scratch_memory.opaque(), scratch_memory.size(), beta, output_nd_.handle(), output_data.opaque())'