diff --git a/apps/object-detection/download_model.sh b/apps/object-detection/download_model.sh
index 9cea0e8e91e..76d2f4782e2 100755
--- a/apps/object-detection/download_model.sh
+++ b/apps/object-detection/download_model.sh
@@ -39,7 +39,7 @@ then
 fi
 
 echo "Downloading pretrained model"
-wget https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model
+wget https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model
 
 
 echo "Finished"
diff --git a/apps/object-detection/object-detection.ipynb b/apps/object-detection/object-detection.ipynb
index 7ff80412e62..889f84c6918 100644
--- a/apps/object-detection/object-detection.ipynb
+++ b/apps/object-detection/object-detection.ipynb
@@ -100,7 +100,7 @@
    "source": [
     "* Here we use a SSD-MobileNet pretrained by PASCAL VOC dataset. \n",
     "\n",
-    "Download [model](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model)"
+    "Download [model](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model)"
    ]
   },
   {
diff --git a/apps/run-app-tests-pip.sh b/apps/run-app-tests-pip.sh
index fd0d95599d6..68ce432a759 100755
--- a/apps/run-app-tests-pip.sh
+++ b/apps/run-app-tests-pip.sh
@@ -88,7 +88,7 @@ if [ -f "$FILENAME" ]
 then
     echo "$FILENAME already exists"
 else
-    wget https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
+    wget https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
 fi
 FILENAME="${ANALYTICS_ZOO_HOME}/apps/object-detection/train_dog.mp4"
 if [ -f "$FILENAME" ]
@@ -101,7 +101,7 @@ if [ -f "$FILENAME" ]
 then
     echo "$FILENAME already exists"
 else
-    wget https://s3.amazonaws.com/analytics-zoo-data/train_dog.mp4 -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
+    wget $FTP_URI/analytics-zoo-data/apps/object-detection/train_dog.mp4  -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
 fi
 FILENAME="~/.imageio/ffmpeg/ffmpeg-linux64-v3.3.1"
 if [ -f "$FILENAME" ]
diff --git a/apps/run-app-tests.sh b/apps/run-app-tests.sh
index a7c1cb157b4..a947ec2c1f2 100644
--- a/apps/run-app-tests.sh
+++ b/apps/run-app-tests.sh
@@ -79,7 +79,7 @@ if [ -f "$FILENAME" ]
 then
     echo "$FILENAME already exists"
 else
-    wget https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
+    wget https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
 fi
 FILENAME="${ANALYTICS_ZOO_HOME}/apps/object-detection/train_dog.mp4"
 if [ -f "$FILENAME" ]
@@ -92,7 +92,7 @@ if [ -f "$FILENAME" ]
 then
     echo "$FILENAME already exists"
 else
-    wget https://s3.amazonaws.com/analytics-zoo-data/train_dog.mp4 -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
+    wget $FTP_URI/analytics-zoo-data/apps/object-detection/train_dog.mp4  -P ${ANALYTICS_ZOO_HOME}/apps/object-detection/
 fi
 
 FILENAME="~/.imageio/ffmpeg/ffmpeg-linux64-v3.3.1"
diff --git a/docker/hyperzoo/Dockerfile b/docker/hyperzoo/Dockerfile
index 146ba503553..b64333cd810 100644
--- a/docker/hyperzoo/Dockerfile
+++ b/docker/hyperzoo/Dockerfile
@@ -123,8 +123,8 @@ RUN mkdir -p /opt/cluster-serving && \
 ADD https://raw.githubusercontent.com/intel-analytics/analytics-zoo/master/pyzoo/test/zoo/resources/serving_quick_start/cat1.jpeg /opt/cluster-serving/test_image 
 ADD https://raw.githubusercontent.com/intel-analytics/analytics-zoo/master/pyzoo/test/zoo/resources/serving_quick_start/dog1.jpeg /opt/cluster-serving/test_image
 ADD https://raw.githubusercontent.com/intel-analytics/analytics-zoo/master/pyzoo/test/zoo/resources/serving_quick_start/fish1.jpeg /opt/cluster-serving/test_image 
-ADD https://analytics-zoo-models.s3-ap-southeast-1.amazonaws.com/tf-res50/frozen_inference_graph.pb /opt/cluster-serving/model/tf_res50/frozen_inference_graph.pb
-ADD https://analytics-zoo-models.s3-ap-southeast-1.amazonaws.com/tf-res50/graph_meta.json /opt/cluster-serving/model/tf_res50/graph_meta.json
+ADD https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/tf-res50/frozen_inference_graph.pb /opt/cluster-serving/model/tf_res50/frozen_inference_graph.pb
+ADD https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/tf-res50/graph_meta.json /opt/cluster-serving/model/tf_res50/graph_meta.json
 RUN chmod a+x ${ANALYTICS_ZOO_HOME}/bin/cluster-serving/*
 RUN cd /opt/cluster-serving && cluster-serving-init && cd ..
 
diff --git a/docs/docs/ProgrammingGuide/image-classification.md b/docs/docs/ProgrammingGuide/image-classification.md
index 76789cb24bb..34dbc544e70 100644
--- a/docs/docs/ProgrammingGuide/image-classification.md
+++ b/docs/docs/ProgrammingGuide/image-classification.md
@@ -58,23 +58,23 @@ For preprocessors for Image Classification models, please check [Image Classific
 
 ## Download link
 
-* [Alexnet](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_alexnet_imagenet_0.1.0.model)
-* [Alexnet Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_alexnet-quantize_imagenet_0.1.0.model)
-* [Inception-V1](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_inception-v1_imagenet_0.1.0.model)
-* [Inception-V1 Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model)
-* [Inception-V3](https://analytics-zoo-models.s3-ap-southeast-1.amazonaws.com/analytics-zoo_inception-v3_imagenet_0.1.0.model)
-* [Inception-V3 Quantize](https://analytics-zoo-models.s3-ap-southeast-1.amazonaws.com/analytics-zoo_inception-v3-quantize_imagenet_0.1.0.model)
-* [VGG-16](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_vgg-16_imagenet_0.1.0.model)
-* [VGG-16 Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_vgg-16-quantize_imagenet_0.1.0.model)
-* [VGG-19](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_vgg-19_imagenet_0.1.0.model)
-* [VGG-19 Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_vgg-19-quantize_imagenet_0.1.0.model)
-* [Resnet-50](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_resnet-50_imagenet_0.1.0.model)
-* [Resnet-50 Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_resnet-50-quantize_imagenet_0.1.0.model)
-* [Resnet-50 Int8](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_resnet-50-int8_imagenet_0.5.0.model)
-* [Densenet-161](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_densenet-161_imagenet_0.1.0.model)
-* [Densenet-161 Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_densenet-161-quantize_imagenet_0.1.0.model)
-* [Mobilenet](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_mobilenet_imagenet_0.1.0.model)
-* [Mobilenet-V2](https://analytics-zoo-models.s3-ap-southeast-1.amazonaws.com/analytics-zoo_mobilenet-v2_imagenet_0.1.0.model)
-* [Mobilenet-V2 Quantize](https://analytics-zoo-models.s3-ap-southeast-1.amazonaws.com/analytics-zoo_mobilenet-v2-quantize_imagenet_0.1.0.model)
-* [Squeezenet](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_squeezenet_imagenet_0.1.0.model)
-* [Squeezenet Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_squeezenet-quantize_imagenet_0.1.0.model)
+* [Alexnet](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_alexnet_imagenet_0.1.0.model)
+* [Alexnet Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_alexnet-quantize_imagenet_0.1.0.model)
+* [Inception-V1](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_inception-v1_imagenet_0.1.0.model)
+* [Inception-V1 Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model)
+* [Inception-V3](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/analytics-zoo_inception-v3_imagenet_0.1.0.model)
+* [Inception-V3 Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/analytics-zoo_inception-v3-quantize_imagenet_0.1.0.model)
+* [VGG-16](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_vgg-16_imagenet_0.1.0.model)
+* [VGG-16 Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_vgg-16-quantize_imagenet_0.1.0.model)
+* [VGG-19](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_vgg-19_imagenet_0.1.0.model)
+* [VGG-19 Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_vgg-19-quantize_imagenet_0.1.0.model)
+* [Resnet-50](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_resnet-50_imagenet_0.1.0.model)
+* [Resnet-50 Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_resnet-50-quantize_imagenet_0.1.0.model)
+* [Resnet-50 Int8](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_resnet-50-int8_imagenet_0.5.0.model)
+* [Densenet-161](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_densenet-161_imagenet_0.1.0.model)
+* [Densenet-161 Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_densenet-161-quantize_imagenet_0.1.0.model)
+* [Mobilenet](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_mobilenet_imagenet_0.1.0.model)
+* [Mobilenet-V2](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/analytics-zoo_mobilenet-v2_imagenet_0.1.0.model)
+* [Mobilenet-V2 Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/analytics-zoo_mobilenet-v2-quantize_imagenet_0.1.0.model)
+* [Squeezenet](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_squeezenet_imagenet_0.1.0.model)
+* [Squeezenet Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_squeezenet-quantize_imagenet_0.1.0.model)
diff --git a/docs/docs/ProgrammingGuide/object-detection.md b/docs/docs/ProgrammingGuide/object-detection.md
index fee5be0710f..43851c14bb1 100644
--- a/docs/docs/ProgrammingGuide/object-detection.md
+++ b/docs/docs/ProgrammingGuide/object-detection.md
@@ -63,22 +63,22 @@ For preprocessors for Object Detection models, please check [Object Detection Co
 
 **PASCAL VOC models**
 
-* [SSD 300x300 MobileNet](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model)
-* [SSD 300x300 VGG](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_PASCAL_0.1.0.model)
-* [SSD 300x300 VGG Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300-quantize_PASCAL_0.1.0.model)
-* [SSD 512x512 VGG](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512_PASCAL_0.1.0.model)
-* [SSD 512x512 VGG Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512-quantize_PASCAL_0.1.0.model)
-* [Faster-RCNN VGG](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_frcnn-vgg16_PASCAL_0.1.0.model)
-* [Faster-RCNN VGG Compress](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_frcnn-vgg16-compress_PASCAL_0.1.0.model)
-* [Faster-RCNN VGG Compress Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_frcnn-vgg16-compress-quantize_PASCAL_0.1.0.model)
-* [Faster-RCNN PvaNet](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_frcnn-pvanet_PASCAL_0.1.0.model)
-* [Faster-RCNN PvaNet Compress](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_frcnn-pvanet-compress_PASCAL_0.1.0.model)
-* [Faster-RCNN PvaNet Compress Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_frcnn-pvanet-compress-quantize_PASCAL_0.1.0.model)
+* [SSD 300x300 MobileNet](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet-300x300_PASCAL_0.1.0.model)
+* [SSD 300x300 VGG](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_PASCAL_0.1.0.model)
+* [SSD 300x300 VGG Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300-quantize_PASCAL_0.1.0.model)
+* [SSD 512x512 VGG](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512_PASCAL_0.1.0.model)
+* [SSD 512x512 VGG Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512-quantize_PASCAL_0.1.0.model)
+* [Faster-RCNN VGG](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_frcnn-vgg16_PASCAL_0.1.0.model)
+* [Faster-RCNN VGG Compress](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_frcnn-vgg16-compress_PASCAL_0.1.0.model)
+* [Faster-RCNN VGG Compress Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_frcnn-vgg16-compress-quantize_PASCAL_0.1.0.model)
+* [Faster-RCNN PvaNet](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_frcnn-pvanet_PASCAL_0.1.0.model)
+* [Faster-RCNN PvaNet Compress](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_frcnn-pvanet-compress_PASCAL_0.1.0.model)
+* [Faster-RCNN PvaNet Compress Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_frcnn-pvanet-compress-quantize_PASCAL_0.1.0.model)
 
 
 **COCO models**
 
-* [SSD 300x300 VGG](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
-* [SSD 300x300 VGG Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300-quantize_COCO_0.1.0.model)
-* [SSD 512x512 VGG](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512_COCO_0.1.0.model)
-* [SSD 512x512 VGG Quantize](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512-quantize_COCO_0.1.0.model)
+* [SSD 300x300 VGG](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
+* [SSD 300x300 VGG Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300-quantize_COCO_0.1.0.model)
+* [SSD 512x512 VGG](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512_COCO_0.1.0.model)
+* [SSD 512x512 VGG Quantize](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-512x512-quantize_COCO_0.1.0.model)
diff --git a/pyzoo/zoo/examples/streaming/README.md b/pyzoo/zoo/examples/streaming/README.md
index b447b7f1039..4485dceec77 100644
--- a/pyzoo/zoo/examples/streaming/README.md
+++ b/pyzoo/zoo/examples/streaming/README.md
@@ -15,7 +15,7 @@ Quick example about integrating analytics-zoo inference/predict service into str
 ## Datasets and pre-trained models
 **1. Streaming Object Detection**
 * Datasets: [COCO](http://cocodataset.org/#home)
-* Pre-trained model: [SSD 300x300](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
+* Pre-trained model: [SSD 300x300](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
 
 **2. Streaming Text Classification**
 * Pre-trained model: Save trained text classification model and word2index in [Text Classification](https://github.com/intel-analytics/analytics-zoo/blob/master/docs/docs/ProgrammingGuide/text-classification.md).
diff --git a/pyzoo/zoo/examples/streaming/objectdetection/README.md b/pyzoo/zoo/examples/streaming/objectdetection/README.md
index 8c8139da0d2..ddcf36fa3de 100644
--- a/pyzoo/zoo/examples/streaming/objectdetection/README.md
+++ b/pyzoo/zoo/examples/streaming/objectdetection/README.md
@@ -10,7 +10,7 @@ So, there are two applications in this example: image_path_writer and streaming_
 
 ## Datasets and pre-trained models
 * Datasets: [COCO](http://cocodataset.org/#home)
-* Pre-trained model: [SSD 300x300](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
+* Pre-trained model: [SSD 300x300](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
 
 ## Run this example
 Make sure all nodes can access image files, model and text files. Local file system/HDFS/Amazon S3 are supported.
diff --git a/pyzoo/zoo/examples/vnni/bigdl/README.md b/pyzoo/zoo/examples/vnni/bigdl/README.md
index d132b30a5a3..7055a811ad2 100644
--- a/pyzoo/zoo/examples/vnni/bigdl/README.md
+++ b/pyzoo/zoo/examples/vnni/bigdl/README.md
@@ -5,7 +5,7 @@ Int8 quantized models can generally give you better performance on Intel Xeon sc
 
 ## Download Analytics Zoo and the pre-trained model
 - You can download Analytics Zoo prebuilt release and nightly build package from [here](https://analytics-zoo.github.io/master/#release-download/) and extract it.
-- Download pre-trained int8 quantized ResNet50 model from [here](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_resnet-50-int8_imagenet_0.5.0.model).
+- Download pre-trained int8 quantized ResNet50 model from [here](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_resnet-50-int8_imagenet_0.5.0.model).
 
 ## Examples
 This folder contains three examples for BigDL VNNI support:
diff --git a/scripts/data/NAB/nyc_taxi/get_nyc_taxi.sh b/scripts/data/NAB/nyc_taxi/get_nyc_taxi.sh
index 0fd53908d88..39b75766ead 100755
--- a/scripts/data/NAB/nyc_taxi/get_nyc_taxi.sh
+++ b/scripts/data/NAB/nyc_taxi/get_nyc_taxi.sh
@@ -41,5 +41,4 @@ fi
 echo "Downloading nyc_taxi.csv"
 wget http://10.239.45.10:8081/repository/raw/analytics-zoo-data/data/NAB/nyc_taxi/nyc_taxi.csv
 
-
 echo "Finished"
diff --git a/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/README.md b/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/README.md
index bf87f631f4a..cc270fd6f79 100644
--- a/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/README.md
+++ b/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/README.md
@@ -14,7 +14,7 @@ Quick example about integrating analytics-zoo inference/predict service into str
 ## Datasets and pre-trained models
 **1. Streaming Object Detection**
 * Datasets: [COCO](http://cocodataset.org/#home)
-* Pre-trained model: [SSD 300x300](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
+* Pre-trained model: [SSD 300x300](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
 
 **2. Streaming Text Classification**
 * Pre-trained model: Save trained text classification model and word2index in [Text Classification](https://github.com/intel-analytics/analytics-zoo/blob/master/docs/docs/ProgrammingGuide/text-classification.md).
diff --git a/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/objectdetection/README.md b/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/objectdetection/README.md
index 6d1c85b3141..6af6d87e1b5 100644
--- a/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/objectdetection/README.md
+++ b/zoo/src/main/scala/com/intel/analytics/zoo/examples/streaming/objectdetection/README.md
@@ -9,7 +9,7 @@ So, there are two applications in this example: ImagePathWriter and StreamingObj
 
 ## Datasets and pre-trained models
 * Datasets: [COCO](http://cocodataset.org/#home)
-* Pre-trained model: [SSD 300x300](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
+* Pre-trained model: [SSD 300x300](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/object-detection/analytics-zoo_ssd-vgg16-300x300_COCO_0.1.0.model)
 
 ## Run this example
 Make sure all nodes can access image files, model and text files. Local file system/HDFS/Amazon S3 are supported.
diff --git a/zoo/src/main/scala/com/intel/analytics/zoo/examples/vnni/bigdl/README.md b/zoo/src/main/scala/com/intel/analytics/zoo/examples/vnni/bigdl/README.md
index 8769fb33c25..ee88aab7b9b 100644
--- a/zoo/src/main/scala/com/intel/analytics/zoo/examples/vnni/bigdl/README.md
+++ b/zoo/src/main/scala/com/intel/analytics/zoo/examples/vnni/bigdl/README.md
@@ -5,7 +5,7 @@ Int8 quantized models can generally give you better performance on Intel Xeon sc
 
 ## Download Analytics Zoo and the pre-trained model
 - You can download Analytics Zoo prebuilt release and nightly build package from [here](https://analytics-zoo.github.io/master/#release-download/) and extract it.
-- Download pre-trained int8 quantized ResNet50 model from [here](https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/imagenet/analytics-zoo_resnet-50-int8_imagenet_0.5.0.model).
+- Download pre-trained int8 quantized ResNet50 model from [here](https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/image-classification/analytics-zoo_resnet-50-int8_imagenet_0.5.0.model).
 
 ## Examples
 This folder contains three examples for BigDL VNNI support:
diff --git a/zoo/src/test/scala/com/intel/analytics/zoo/models/image/imageclassification/ImageClassificationSpec.scala b/zoo/src/test/scala/com/intel/analytics/zoo/models/image/imageclassification/ImageClassificationSpec.scala
index 7e997b2a0d5..acac1e104c6 100644
--- a/zoo/src/test/scala/com/intel/analytics/zoo/models/image/imageclassification/ImageClassificationSpec.scala
+++ b/zoo/src/test/scala/com/intel/analytics/zoo/models/image/imageclassification/ImageClassificationSpec.scala
@@ -132,15 +132,15 @@ class ImageClassificationSpec extends ZooSpecHelper {
   }
 
   "ImageClassifier" should "predict inception-v1-quantize locally" in {
-    predictLocal("https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/" +
-      "imageclassification/imagenet/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model",
+    predictLocal("https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/" +
+      "image-classification/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model",
       "analytics-zoo", resource.getFile)
   }
 
   "ImageClassifier" should "predict inception-v1-quantize" in {
-    predict("https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/imageclassification/" +
-      "imagenet/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model", "analytics-zoo",
-      resource.getFile)
+    predict("https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/" +
+      "image-classification/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model",
+      "analytics-zoo", resource.getFile)
   }
 
   "ImageClassifier" should "predict bigdl inception-v1-quantize locally" in {
@@ -156,8 +156,8 @@ class ImageClassificationSpec extends ZooSpecHelper {
   "ImageClassifier" should "throw exception if input is empty locally" in {
     val tempDir = createTmpDir()
     val thrown = intercept[Exception]{
-      predictLocal("https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/" +
-        "imageclassification/imagenet/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model",
+      predictLocal("https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/" +
+        "image-classification/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model",
         "analytics-zoo", tempDir.getAbsolutePath)
     }
     assert("requirement failed: ImageModel.predictImageSet: input is empty, please check your " +
@@ -168,8 +168,8 @@ class ImageClassificationSpec extends ZooSpecHelper {
   "ImageClassifier" should "throw exception if input is empty" in {
     val tempDir = createTmpDir()
     val thrown = intercept[Exception]{
-      predict("https://s3-ap-southeast-1.amazonaws.com/analytics-zoo-models/" +
-        "imageclassification/imagenet/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model",
+      predict("https://sourceforge.net/projects/analytics-zoo/files/analytics-zoo-models/" +
+        "image-classification/analytics-zoo_inception-v1-quantize_imagenet_0.1.0.model",
         "analytics-zoo", tempDir.getAbsolutePath)
     }
     assert("requirement failed: ImageModel.predictImageSet: input is empty, please check your " +
diff --git a/zoo/src/test/scala/com/intel/analytics/zoo/models/image/objectdetection/ObjectDetectionSpec.scala b/zoo/src/test/scala/com/intel/analytics/zoo/models/image/objectdetection/ObjectDetectionSpec.scala
index 7f611e4cf47..db7ca78aff6 100644
--- a/zoo/src/test/scala/com/intel/analytics/zoo/models/image/objectdetection/ObjectDetectionSpec.scala
+++ b/zoo/src/test/scala/com/intel/analytics/zoo/models/image/objectdetection/ObjectDetectionSpec.scala
@@ -35,7 +35,7 @@ class ObjectDetectionSpec extends FlatSpec with Matchers with BeforeAndAfter {
   before {
     val conf = new SparkConf().setAppName("Test ObjectDetector").setMaster("local[1]")
     sc = NNContext.initNNContext(conf)
-    val s = s"wget -q -O ./ssd.model https://s3-ap-southeast-1.amazonaws.com/" +
+    val s = s"wget -q -O ./ssd.model https://sourceforge.net/projects/analytics-zoo/files/" +
       s"analytics-zoo-models/object-detection/analytics-zoo_ssd-mobilenet" +
       s"-300x300_PASCAL_0.1.0.model" !!
     val s2 = s"wget -q -O ./bigdl_ssd.model https://s3-ap-southeast-1.amazonaws.com/" +