note on resnet

examples/ResNet/README.md

 
-## ResNet
+ImageNet training code coming soon.
+
+## cifar10-resnet.py
 
 Reproduce the results in paper "Deep Residual Learning for Image Recognition", [http://arxiv.org/abs/1512.03385](http://arxiv.org/abs/1512.03385)
-with the variants proposed in "Identity Mappings in Deep Residual Networks", [https://arxiv.org/abs/1603.05027](https://arxiv.org/abs/1603.05027).
+with the variants proposed in "Identity Mappings in Deep Residual Networks", [https://arxiv.org/abs/1603.05027](https://arxiv.org/abs/1603.05027) on CIFAR10.
 
 The train error shown here is a moving average of the error rate of each batch in training.
 The validation error here is computed on test set.
@@ -13,3 +15,14 @@ Download model:
 [Cifar10 ResNet-110 (n=18)](https://drive.google.com/open?id=0B9IPQTvr2BBkTXBlZmh1cmlnQ0k)
 
 Also see an implementation of [DenseNet](https://github.com/YixuanLi/densenet-tensorflow) from [Densely Connected Convolutional Networks](https://arxiv.org/abs/1608.06993).
+
+## load-resnet.py
+
+A script to convert and run ResNet{50,101,152} ImageNet models released by Kaiming.
+
+Example usage:
+```bash
+python -m tensorpack.utils.loadcaffe PATH/TO/{ResNet-101-deploy.prototxt,ResNet-101-model.caffemodel} ResNet101.npy
+./load-resnet.py --load ResNet-101.npy --input cat.png --depth 101
+```
+

examples/ResNet/load-resnet.py

@@ -22,7 +22,7 @@ from tensorpack.dataflow.dataset import ILSVRCMeta
 
 """
 Usage:
-    python -m tensorpack.utils.loadcaffe PATH/TO/CAFFE/{ResNet-101-deploy.prototxt,ResNet-101-model.caffemodel} ResNet101.npy
+    python -m tensorpack.utils.loadcaffe PATH/TO/{ResNet-101-deploy.prototxt,ResNet-101-model.caffemodel} ResNet101.npy
     ./load-resnet.py --load ResNet-101.npy --input cat.png --depth 101
 """
 MODEL_DEPTH = None

tensorpack/train/config.py

@@ -23,8 +23,7 @@ class TrainConfig(object):
         :param optimizer: a `tf.train.Optimizer` instance defining the optimizer for trainig.
         :param callbacks: a `callback.Callbacks` instance. Define
             the callbacks to perform during training.
-        :param session_config: a `tf.ConfigProto` instance to instantiate the
-            session. default to a session running 1 GPU.
+        :param session_config: a `tf.ConfigProto` instance to instantiate the session.
         :param session_init: a `sessinit.SessionInit` instance to
             initialize variables of a session. default to a new session.
         :param model: a `ModelDesc` instance.