updated the numbers on K80

Author: vapaunic

README.md

@@ -40,7 +40,7 @@ The notebooks are executed on an Azure [Deep Learning Virtual Machine](https://a
 | [PyTorch](notebooks/PyTorch_CNN.ipynb)                |        169         |         51          |
 | [Julia - Knet](notebooks/Knet_CNN.ipynb)              |        159         |         ??          |
 | [R - MXNet](notebooks/.ipynb)                         |        ???         |         ??          |
-| [R - Keras(TF)](notebooks/KerasR_TF_CNN.ipynb)        |        211         |         72          |
+| [R - Keras(TF)](notebooks/KerasR_TF_CNN.ipynb)        |        205         |         72          |
 
 
 *Note: It is recommended to use higher level APIs where possible; see these notebooks for examples with [Tensorflow](notebooks/Tensorflow_CNN_highAPI.ipynb), [MXNet](notebooks/MXNet_CNN_highAPI.ipynb) and [CNTK](notebooks/CNTK_CNN_highAPI.ipynb). They are not linked in the table to keep the common-structure-for-all approach*
@@ -94,7 +94,7 @@ Input for this model is 112,120 PNGs of chest X-rays resized to (264, 264). **No
 | [PyTorch](notebooks/PyTorch_Inference.ipynb)        | 7.7                | 1.9                 |
 | [Julia - Knet](notebooks/Knet_Inference.ipynb)      | 6.3                | ???                 |
 | [R - MXNet](notebooks/.ipynb)                       | ???                | ???                 |
-| [R - Keras(TF)](notebooks/KerasR_TF_Inference.ipynb)| 16                 | 7.4                 |
+| [R - Keras(TF)](notebooks/KerasR_TF_Inference.ipynb)| 17                 | 7.4                 |
 
 
 A pre-trained ResNet50 model is loaded and chopped just after the avg_pooling at the end (7, 7), which outputs a 2048D dimensional vector. This can be plugged into a softmax layer or another classifier such as a boosted tree to perform transfer learning. Allowing for a warm start; this forward-only pass to the avg_pool layer is timed. *Note: batch-size remains constant, however filling the RAM on a GPU would produce further performance boosts (greater for GPUs with more RAM).*
@@ -111,7 +111,7 @@ A pre-trained ResNet50 model is loaded and chopped just after the avg_pooling at
 | [Tensorflow](notebooks/Tensorflow_RNN.ipynb)       | 30                 | 22                  | Yes          |
 | [Julia - Knet](notebooks/Knet_RNN.ipynb)           | 29                 | ??                  | Yes          |
 | [R - MXNet](notebooks/.ipynb)                      | ??                 | ??                  | ???          |
-| [R - Keras(TF)](notebooks/KerasR_TF_RNN.ipynb)     | ??                 | 25                  | Yes          |
+| [R - Keras(TF)](notebooks/KerasR_TF_RNN.ipynb)     | 35                 | 25                  | Yes          |
 
 
 Input for this model is the standard [IMDB movie review dataset](http://ai.stanford.edu/~amaas/data/sentiment/) containing 25k training reviews and 25k test reviews, uniformly split across 2 classes (positive/negative). Processing follows [Keras](https://github.com/fchollet/keras/blob/master/keras/datasets/imdb.py) approach where start-character is set as 1, out-of-vocab (vocab size of 30k is used) represented as 2 and thus word-index starts from 3. Zero-padded / truncated to fixed axis of 150 words per review.