Update README files with int8 docker image (#193)

* Update README files with int8 docker image

* Update READMEs to make note of the image and how it was built.

Author: dmsuehir

benchmarks/README.md

@@ -26,7 +26,7 @@ dependencies to be installed:
 | Image Segmentation     | TensorFlow    | Mask R-CNN          | Inference | [FP32](image_segmentation/tensorflow/maskrcnn/README.md#fp32-inference-instructions) |
 | Image Segmentation     | TensorFlow    | UNet                | Inference | [FP32](image_segmentation/tensorflow/unet/README.md#fp32-inference-instructions) |
 | Language Translation   | TensorFlow    | Transformer Language| Inference | [FP32](language_translation/tensorflow/transformer_language/README.md#fp32-inference-instructions) |
-| Object Detection       | TensorFlow    | Fast R-CNN          | Inference | [FP32](object_detection/tensorflow/fastrcnn/README.md#fp32-inference-instructions) |
+| Object Detection       | TensorFlow    | Fast R-CNN          | Inference | [Int8](object_detection/tensorflow/fastrcnn/README.md#int8-inference-instructions) [FP32](object_detection/tensorflow/fastrcnn/README.md#fp32-inference-instructions) |
 | Object Detection       | TensorFlow    | R-FCN               | Inference | [FP32](object_detection/tensorflow/rfcn/README.md#fp32-inference-instructions) |
 | Object Detection       | TensorFlow    | SSD-MobileNet       | Inference | [FP32](object_detection/tensorflow/ssd-mobilenet/README.md#fp32-inference-instructions) |
 | Recommendation         | TensorFlow    | NCF                 | Inference | [FP32](recommendation/tensorflow/ncf/README.md#fp32-inference-instructions) |

benchmarks/image_recognition/tensorflow/inceptionv3/README.md

@@ -35,12 +35,7 @@ requires.
 $ wget https://storage.googleapis.com/intel-optimized-tensorflow/models/inceptionv3_int8_pretrained_model.pb
 ```
 
-4. Build a docker image using master of the official
-[TensorFlow](https://github.com/tensorflow/tensorflow) repository with
-`--config=mkl`. More instructions on
-[how to build from source](https://software.intel.com/en-us/articles/intel-optimization-for-tensorflow-installation-guide#inpage-nav-5).
-
-5. If you would like to run Inception V3 inference and test for
+4. If you would like to run Inception V3 inference and test for
 accuracy, you will need the ImageNet dataset. Benchmarking for latency
 and throughput do not require the ImageNet dataset.
 
@@ -72,19 +67,19 @@ $ ll /home/myuser/datasets/ImageNet_TFRecords
 -rw-r--r--. 1 user  55292089 Jun 20 15:09 validation-00127-of-00128
 ```
 
-6. Next, navigate to the `benchmarks` directory in your local clone of
+5. Next, navigate to the `benchmarks` directory in your local clone of
 the [intelai/models](https://github.com/IntelAI/models) repo from step 1.
 The `launch_benchmark.py` script in the `benchmarks` directory is
 used for starting a benchmarking run in a optimized TensorFlow docker
 container. It has arguments to specify which model, framework, mode,
 precision, and docker image to use, along with your path to the ImageNet
-TF Records that you generated in step 5.
+TF Records that you generated in step 4.
 
-Substitute in your own `--data-location` (from step 5, for accuracy
-only), `--in-graph` pretrained model file path (from step 3),
+Substitute in your own `--data-location` (from step 4, for accuracy
+only), `--in-graph` pretrained model file path (from step 3) and
 `--model-source-dir` for the location where you cloned the
 [tensorflow/models](https://github.com/tensorflow/models) repo
-(from step 2), and the name/tag for your docker image (from step 4).
+(from step 2).
 
 Inception V3 can be run for accuracy, latency benchmarking, or throughput
 benchmarking. Use one of the following examples below, depending on
@@ -101,7 +96,7 @@ python launch_benchmark.py \
     --framework tensorflow \
     --accuracy-only \
     --batch-size 100 \
-    --docker-image tf_int8_docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
     --in-graph /home/myuser/inceptionv3_int8_pretrained_model.pb \
     --data-location /home/myuser/datasets/ImageNet_TFRecords \
     -- input_height=299 input_width=299
@@ -123,7 +118,7 @@ python launch_benchmark.py \
     --benchmark-only \
     --batch-size 1 \
     --socket-id 0 \
-    --docker-image tf_int8_docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
     --in-graph /home/myuser/inceptionv3_int8_pretrained_model.pb \
     --data-location /home/myuser/datasets/ImageNet_TFRecords \
     -- input_height=299 input_width=299 warmup_steps=50 steps=500
@@ -140,12 +135,18 @@ python launch_benchmark.py \
     --benchmark-only \
     --batch-size 128 \
     --socket-id 0 \
-    --docker-image tf_int8_docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
     --in-graph /home/myuser/inceptionv3_int8_pretrained_model.pb \
     --data-location /home/myuser/datasets/ImageNet_TFRecords \
     -- input_height=299 input_width=299 warmup_steps=50 steps=500
 ```
 
+The docker image (`intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl`)
+used in the commands above were built using
+[TensorFlow]([email protected]:tensorflow/tensorflow.git) master
+([e889ea1](https://github.com/tensorflow/tensorflow/commit/e889ea1dd965c31c391106aa3518fc23d2689954)) and
+[PR #25765](https://github.com/tensorflow/tensorflow/pull/25765).
+
 Note that the `--verbose` or `--output-dir` flag can be added to any of the above commands
 to get additional debug output or change the default output location..
 

benchmarks/image_recognition/tensorflow/resnet101/README.md

@@ -22,12 +22,7 @@ an optimized version of the ResNet101 model code.
 $ wget https://storage.googleapis.com/intel-optimized-tensorflow/models/resnet101_int8_pretrained_model.pb
 ```
 
-3. Build a docker image using master of the official
-[TensorFlow](https://github.com/tensorflow/tensorflow) repository with
-`--config=mkl`. More instructions on
-[how to build from source](https://software.intel.com/en-us/articles/intel-optimization-for-tensorflow-installation-guide#inpage-nav-5).
-
-4. If you would like to run ResNet101 inference and test for
+3. If you would like to run ResNet101 inference and test for
 accurancy, you will need the full ImageNet dataset.
 
 Register and download the
@@ -58,17 +53,16 @@ $ ll /home/myuser/datasets/ImageNet_TFRecords
 -rw-r--r--. 1 user  55292089 Jun 20 15:09 validation-00127-of-00128
 ```
 
-5. Next, navigate to the `benchmarks` directory in your local clone of
+4. Next, navigate to the `benchmarks` directory in your local clone of
 the [intelai/models](https://github.com/IntelAI/models) repo from step 1.
 The `launch_benchmark.py` script in the `benchmarks` directory is
 used for starting a benchmarking run in a optimized TensorFlow docker
 container. It has arguments to specify which model, framework, mode,
 precision, and docker image to use, along with your path to the ImageNet
-TF Records that you generated in step 4.
+TF Records that you generated in step 3.
 
-Substitute in your own `--data-location` (from step 4, for accuracy
-only), `--in-graph` pre-trained model file path (from step 2),
-and the name/tag for your docker image (from step 3).
+Substitute in your own `--data-location` (from step 3, for accuracy
+only) and `--in-graph` pre-trained model file path (from step 2).
 
 ResNet101 can be run for accuracy or performance benchmarking. Use one of
 the following examples below, depending on your use case.
@@ -86,7 +80,7 @@ $ python launch_benchmark.py \
     --framework tensorflow \
     --accuracy-only \
     --batch-size 100 \
-    --docker-image tf_int8_docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
     --data-location /home/myuser/dataset/FullImageNetData_directory \
     --in-graph=/home/myuser/resnet101_int8_pretrained_model.pb
 ```
@@ -107,7 +101,7 @@ python launch_benchmark.py \
     --benchmark-only \
     --batch-size 1 \
     --socket-id 0 \
-    --docker-image tf_int8_docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
     --in-graph=/home/myuser/resnet101_int8_pretrained_model.pb \
     -- warmup_steps=50 steps=500
 ```
@@ -123,11 +117,17 @@ python launch_benchmark.py \
     --benchmark-only \
     --batch-size 128 \
     --socket-id 0 \
-    --docker-image tf_int8_docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
     --in-graph=/home/myuser/resnet101_int8_pretrained_model.pb \
     -- warmup_steps=50 steps=500
 ```
 
+The docker image (`intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl`)
+used in the commands above were built using
+[TensorFlow]([email protected]:tensorflow/tensorflow.git) master
+([e889ea1](https://github.com/tensorflow/tensorflow/commit/e889ea1dd965c31c391106aa3518fc23d2689954)) and
+[PR #25765](https://github.com/tensorflow/tensorflow/pull/25765).
+
 Note that the `--verbose` or `--output-dir` flag can be added to any of the above commands
 to get additional debug output or change the default output location..
 

benchmarks/image_recognition/tensorflow/resnet50/README.md

@@ -58,7 +58,7 @@ $ python launch_benchmark.py \
     --mode inference \
     --batch-size=100 \
     --accuracy-only \
-    --docker-image docker_image
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl
 ```
 The log file is saved to the value of `--output-dir`.
 
@@ -95,7 +95,7 @@ $ python launch_benchmark.py \
     --mode inference \
     --batch-size=128 \
     --benchmark-only \
-    --docker-image docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl
     -- warmup_steps=50 steps=500
 ```
 The tail of the log output when the benchmarking completes should look
@@ -112,6 +112,12 @@ Ran inference with batch size 128
 Log location outside container: {--output-dir value}/benchmark_resnet50_inference_int8_20190223_180546.log
 ```
 
+The docker image (`intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl`)
+used in the commands above were built using
+[TensorFlow]([email protected]:tensorflow/tensorflow.git) master
+([e889ea1](https://github.com/tensorflow/tensorflow/commit/e889ea1dd965c31c391106aa3518fc23d2689954)) and
+[PR #25765](https://github.com/tensorflow/tensorflow/pull/25765).
+
 Note that the `--verbose` or `--output-dir` flag can be added to any of the above commands
 to get additional debug output or change the default output location..
 

benchmarks/object_detection/tensorflow/fastrcnn/README.md

@@ -255,8 +255,8 @@ $ python launch_benchmark.py \
     --mode inference \
     --socket-id 0 \
     --in-graph /home/myuser/fastrcnn_int8_pretrained_model.pb \
-    --docker-image tf_int8_docker_image
-    --benchmark-only
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
+    --benchmark-only \
     -- number_of_steps=5000
 ```
 
@@ -270,13 +270,19 @@ python launch_benchmark.py \
     --precision int8 \
     --framework tensorflow \
     --socket-id 0 \
-    --docker-image tf_int8_docker_image \
+    --docker-image intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl \
     --model-source-dir /home/myuser/tensorflow/models \
     --data-location /home/myuser/coco_dataset/coco_val.record \
     --in-graph /home/myuser/fastrcnn_int8_pretrained_model.pb  \
     --accuracy-only
 ```
 
+The docker image (`intelaipg/intel-optimized-tensorflow:PR25765-devel-mkl`)
+used in the commands above were built using
+[TensorFlow]([email protected]:tensorflow/tensorflow.git) master
+([e889ea1](https://github.com/tensorflow/tensorflow/commit/e889ea1dd965c31c391106aa3518fc23d2689954)) and
+[PR #25765](https://github.com/tensorflow/tensorflow/pull/25765).
+
 5. The log file is saved to the value of `--output-dir`.
 
 Below is a sample log file tail when running benchmarking for throughput