How can I improve the speed of data generated process? It's too slow and it's hard to work with a big dataset.

[Cabbagehust2507]

I tried to use your code to generate TFrecord from 2.4 million images (1024x64), but it takes 3s for 1 image generation, hence I need 1 month at least for generating that dataset. So I want to know how to improve this issue?

[docongminh]

You could use GPUs for up speech conversion process
same code here

 config = tf.ConfigProto()

 np_data = np.array(img)

 image = tf.image.convert_image_dtype(np_data, dtype=tf.uint8)

 image = tf.image.encode_jpeg(image)

 with tf.Session(config=config) as sess:

         image_data = sess.run(image)

         sess.close()