how can save model by tensorflowlite

43 Views Asked by At

i am try to save model by tesorflow lite but it give me this error how can solve this error or what it mean

Summary on the non-converted ops:

  • Accepted dialects: tfl, builtin, func
  • Non-Converted Ops: 63, Total Ops 191, % non-converted = 32.98 %
  • 63 ARITH ops
  • arith.constant: 63 occurrences (f32: 60, i32: 3)

    (f32: 10) (f32: 35) (f32: 17) (f32: 4) (f32: 1) (f32: 4) (uq_8: 52) (f32: 1) (f32: 1)

i am try to save as tf and covert it to tflite put give me same error this my saved model as tf

mobilenet_model.save("saved_model", save_format="tf")
path = '/kaggle/working/saved_model'
saved_model = tf.saved_model.load(path)

converter = tf.lite.TFLiteConverter.from_saved_model(path)
converter.optimizations = [tf.lite.Optimize.DEFAULT]
tflite_model = converter.convert()
# Save the TensorFlow Lite model to a file
with open("EGY_TOUR_model.tflite", "wb") as f:
        f.write(tflite_model)

saved_model

variables

assets

keras_metadata.pb

fingerprint.pb

saved_model.pb

0

There are 0 best solutions below