clean up sandbox code

examples/model_server.py

@@ -45,6 +45,8 @@ def predict():
         })
 
     except Exception as e:
+        logging.error(f"An error occurred: {str(e)}")
+        print(f"An error occurred: {str(e)}")
         return jsonify({'error': str(e)}), 500
 
 # Main entry point

examples/quantized_model.py

@@ -1,48 +1,18 @@
-import tempfile
-import os
-import csv
-
 import tensorflow as tf
 import numpy as np
-import tflite as tfl
 
-# import tensorflow.keras.utils
-from tensorflow import keras
-# import tensorflow.compat.v1 as tf
 import tensorflow as tf
-import tensorflow_hub as hub
-import itertools
-import soundfile as sf
-import librosa
-import math
-import time
-# from prune import prune
-# import tensorflow_model_optimization as tfmot
 
 
 interpreter=tf.lite.Interpreter(model_path='data/model/quantized_model1.tflite')
+
 interpreter.allocate_tensors()
 input_details = interpreter.get_input_details()
 output_details = interpreter.get_output_details()
 
-i=0
-whales_identified=0
-whale_samples=0
-other_samples_identified=0
-other_samples=0
-directory_mn='./clips/mn/'
-directory_not_mn='./clips/not_mn/'
-
-# filename = os.fsdecode(file)
-
-# # try opening the file
-# waveform, sample_rate = tf.audio.decode_wav(tf.io.read_file(directory_mn+filename))
-
-# waveform = tf.slice(tf.squeeze(tf.expand_dims(waveform, 0),[2]),[1,0],[1,15600])  # makes a batch of size 1
 waveform = tf.convert_to_tensor(np.random.random((1, 15600)).astype(np.float32))
 print(waveform.shape)
 
-
 # Create input tensor out of raw features
 interpreter.set_tensor(input_details[0]['index'], waveform)