由于 CUDA 错误，无法从 Keras 模型进行预测

Question

我是 Python 新手。我按照this website 作为指导来做一些未来的预测。在我做完所有事情后，图表没有显示出来，我得到了这些错误：

2020-10-09 08:27:09.619051: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found
2020-10-09 08:27:09.620905: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2020-10-09 08:27:16.105403: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'nvcuda.dll'; dlerror: nvcuda.dll not found
2020-10-09 08:27:16.107108: W tensorflow/stream_executor/cuda/cuda_driver.cc:312] failed call to cuInit: UNKNOWN ERROR (303)
2020-10-09 08:27:16.110329: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:169] retrieving CUDA diagnostic information for host: kwk-tech-05
2020-10-09 08:27:16.110968: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:176] hostname: kwk-tech-05
2020-10-09 08:27:16.111746: 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
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2020-10-09 08:27:16.119507: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x2624680e450 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-10-09 08:27:16.120408: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version

这是我写的代码：

import pandas as pd
import numpy as np
import tensorflow as tf
import keras
from keras.preprocessing.sequence import TimeseriesGenerator
from keras.models import Sequential
from keras.layers import LSTM, Dense
import plotly.graph_objects as go

df = pd.read_excel('T:/Python/NRN-Netze_Python/Stromverbrauch2Jahren.xlsx')
print(df.info())

df['Datum'] = pd.to_datetime(df['Datum'])
df.set_axis(df['Datum'], inplace = True)
df.drop(columns = ['Datum_u_Uhrzeit', 'Stunden', 'Minuten', 'Uhrzeit'], inplace = True)

value = df['Werte'].values
value = value.reshape((-1, 1))

split_percent = 0.80
split = int(split_percent * len(value))

value_train = value[:split]
value_test = value[split:]
date_train = df['Datum'][:split]
date_test = df['Datum'][split:]

print('')
print(len(value_train))
print(len(value_test))

look_back = 15

train_generator = TimeseriesGenerator(value_train, value_train, length = look_back, batch_size = 20)
test_generator = TimeseriesGenerator(value_test, value_test, length = look_back, batch_size = 1)

model = Sequential()
model.add(LSTM(10, activation = 'relu', input_shape = (look_back, 1)))
model.add(Dense(1))
model.compile(optimizer = 'adam', loss = 'mse')

num_epochs = 25
model.fit(train_generator, epochs = num_epochs, verbose = 1)

prediction = model.predict(test_generator)

value_train = value_train.reshape((-1))
value_test = value_test.reshape((-1))
prediction = prediction.reshape((-1))

trace1 = go.Scatter(
    x = date_train,
    y = value_train,
    mode = 'lines',
    name = 'Original'
)

trace2 = go.Scatter(
    x = date_test,
    y = value_test,
    mode = 'lines',
    name = 'Prediction'
)

layout = go.Layout(
    title = 'Strom Verbrauch In Der Wasserversorgung',
    xaxis = {'title' : 'Datum'},
    yaxis = {'title' : 'Werte'}
)

value = value.reshape((-1))

def predict(num_prediction, model):
    prediction_list = value[-look_back:]
    
    for _ in range(num_prediction):
        x = prediction_list[-look_back:]
        x = x.reshape((1, look_back, 1))
        out = model.predict(x)[0][0]
        prediction_list = np.append(prediction_list, out)
    prediction_list = prediction_list[look_back - 1:]
    
    return prediction_list

def predict_dates(num_prediction):
    last_date = df['Datum'].values[-1]
    prediction_dates = pd.date_range(last_date, periods = num_prediction + 1).tolist()
    return prediction_dates

num_prediction = 365
forecast = predict(num_prediction, model)
forecast_dates = predict_dates(num_prediction)
fig = go.Figure(data = [trace1, trace2], layout = layout)
fig.show()

Answer 1

从错误消息来看，您似乎没有为您的显卡安装 CUDA 驱动程序。