Could not load library cudnn_cnn_infer64_8.dll. Error code 126 Please make sure cudnn_cnn_infer64_8.dll is in your library path!

Sigo recibiendo este error cuando trato de usar TensorFlow con GPU, instalé CUDA, cuDNN y todos los controladores varias veces según las instrucciones. Pero nada parece funcionar. Si uso una computadora portátil, entonces TensorFlow usa la CPU, con la extensión de computadora portátil con código VS puedo usar la gpu, pero detiene la sesión en la primera época, cuando traté de ejecutarla como un archivo de Python normal. se produjo el error anterior.

Salida terminal completa:

 Found 14630 validated image filenames belonging to 3 classes. Found 1500 validated image filenames belonging to 3 classes. 2021-11-08 11:03:58.000354: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: AVX AVX2 To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags. 2021-11-08 11:03:58.603592: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 2775 MB memory: -> device: 0, name: NVIDIA GeForce GTX 1050 Ti, pci bus id: 0000:01:00.0, compute capability: 6.1 Epoch 1/10 2021-11-08 11:04:07.306011: I tensorflow/stream_executor/cuda/cuda_dnn.cc:366] Loaded cuDNN version 8300 Could not load library cudnn_cnn_infer64_8.dll. Error code 126 Please make sure cudnn_cnn_infer64_8.dll is in your library path! E:\MyWorkSpace\animal_detect>

El fragmento de código:

 import tensorflow as tf from tensorflow.keras.preprocessing.image import ImageDataGenerator from tensorflow.keras import layers from tensorflow.keras import Model from tensorflow.keras.preprocessing.image import ImageDataGenerator from tensorflow.keras.applications.vgg16 import VGG16 import pandas as pd import numpy as np train_df = pd.read_csv('train.csv') test_df = pd.read_csv('test.csv') train_gen = ImageDataGenerator(rescale = 1./255.,rotation_range = 40, width_shift_range = 0.2, height_shift_range = 0.2, shear_range = 0.2, zoom_range = 0.2, horizontal_flip = True) test_gen = ImageDataGenerator( rescale = 1.0/255. ) train_set = train_gen.flow_from_dataframe(train_df,x_col='loc',y_col='label',batch_size=20,target_size=(224,224)) test_set = train_gen.flow_from_dataframe(test_df,x_col='loc',y_col='label',batch_size=20,target_size=(224,224)) base_model = VGG16(input_shape = (224, 224, 3), include_top = False, weights = 'imagenet') for layer in base_model.layers: layer.trainable = False x = layers.Flatten()(base_model.output) x = layers.Dense(512, activation='relu')(x) x = layers.Dropout(0.5)(x) x = layers.Dense(3, activation='sigmoid')(x) model = tf.keras.models.Model(base_model.input, x) model.compile(optimizer = tf.keras.optimizers.RMSprop(learning_rate=0.0001), loss = 'categorical_crossentropy',metrics = ['acc']) vgghist = model.fit(train_set, validation_data = test_set, steps_per_epoch = 100, epochs = 10)

el mismo código se ha utilizado para Jupyter-notebook, la extensión del cuaderno de código VS y como un archivo de python normal

Especificaciones del dispositivo:

Procesador: Intel i5 GPU: Nvidia Geforce 1050ti

Cuda versión: 11.5 cuDNN versión: 8.3