Merging

llvm/projects/keras/src/mobilenet_imagenet.py

 import os
 import glob
+import random
 
 import scipy
 import scipy.io
 import cv2
 import numpy as np
 
+import tensorflow as tf
 import keras
 from keras.models import Sequential, Model
 from keras.layers import *
@@ -35,21 +37,21 @@ VAL_SIZE = 100
 
 
 
-# def relu6(x):
-#     return K.relu(x, max_value=6)
+def relu6(x):
+    return K.relu(x, max_value=6)
 
 
 def _conv_block(inputs, filters, alpha, kernel=(3, 3), strides=(1, 1)):
     channel_axis = 1
     filters = int(filters * alpha)
         
-    x = ZeroPadding2D(padding=((0, 1), (0, 1)))(inputs)
+    x = ZeroPadding2D(padding=((1, 1), (1, 1)))(inputs)
     x = Conv2D(filters, kernel,
                       padding='valid',
                       use_bias=False,
                       strides=strides)(x)
     x = BatchNormalization(axis=channel_axis)(x)
-    return Activation('relu')(x)
+    return Activation(relu6)(x)
     
     
 def _depthwise_conv_block(inputs, pointwise_conv_filters, alpha,
@@ -58,7 +60,7 @@ def _depthwise_conv_block(inputs, pointwise_conv_filters, alpha,
     pointwise_conv_filters = int(pointwise_conv_filters * alpha)
 
     if strides != (1, 1):
-        x = ZeroPadding2D(padding=(1, 1))(inputs)
+        x = ZeroPadding2D(padding=((1, 1), (1, 1)))(inputs)
     else:
         x = inputs
     
@@ -68,14 +70,14 @@ def _depthwise_conv_block(inputs, pointwise_conv_filters, alpha,
                                strides=strides,
                                use_bias=False)(x)
     x = BatchNormalization(axis=channel_axis)(x)
-    x = Activation('relu')(x)
+    x = Activation(relu6)(x)
 
     x = Conv2D(pointwise_conv_filters, (1, 1),
                       padding='same',
                       use_bias=False,
                       strides=(1, 1))(x)
     x = BatchNormalization(axis=channel_axis)(x)
-    return Activation('relu')(x)
+    return Activation(relu6)(x)
     
     
     
@@ -112,11 +114,13 @@ def get_mobilenet_nchw_keras():
     x = _depthwise_conv_block(x, 1024, alpha, depth_multiplier, block_id=13)
 
 
-    x = GlobalAveragePooling2D()(x)
-    x = Dropout(dropout)(x)
-    x = Dense(1000)(x)
+    x = AveragePooling2D((7, 7))(x)    
+    x = Conv2D(1000, (1, 1),
+                      padding='same')(x)
+    x = Flatten()(x)
     x = Activation('softmax')(x)
 
+
     model = Model(img_input, x)
     
     
@@ -129,6 +133,7 @@ def get_mobilenet_nchw_keras():
         try:
             model.layers[j].set_weights(original_model.layers[i].get_weights())
             print (j, 'loaded')
+#             model.layers[j].trainable = False
             j += 1
         except:
             print (j, 'skipped', model.layers[j])
@@ -205,6 +210,53 @@ y_true = np.array(y_true)
 
 
 
+def train_helper(x):
+    
+    try:
+        x = x.decode('utf-8')
+    except:
+        pass
+    
+    image = load_image(x)
+    
+    y = np.zeros(1000, dtype=np.uint8)
+        
+    y[synset_to_keras_idx[x.split('/')[-2]]]= 1
+        
+    return image, y
+
+
+
+
+train_images = glob.glob(IMAGENET_DIR + 'train/*/*')
+random.shuffle(train_images)
+
+dataset = tf.data.Dataset().from_tensor_slices(train_images)
+dataset = dataset.map(
+    lambda x : tf.py_func(train_helper, [x], [tf.float32, tf.uint8]), 
+    num_parallel_calls=16
+)
+
+dataset = dataset.shuffle(buffer_size=1000)
+dataset = dataset.batch(64)
+dataset = dataset.repeat()
+
+next_element = dataset.make_one_shot_iterator().get_next()
+
+sess = tf.Session()
+
+def generate():
+    while True:
+        yield sess.run(next_element)
+    
+
+
+model.compile(optimizer=keras.optimizers.Adam(lr=0.00001), loss='categorical_crossentropy', metrics=['acc'])
+
+
+model.fit_generator(generate(), steps_per_epoch=1000, validation_data=(X_test, to_categorical(y_true, num_classes=1000)), epochs=5)
+
+
 translate_to_approxhpvm(model, OUTPUT_DIR, X_test[:VAL_SIZE], y_true[:VAL_SIZE], 1000)
 
 dumpCalibrationData(OUTPUT_DIR + 'test_input.bin', X_test, OUTPUT_DIR + 'test_labels.bin', y_true)