update docs

.github/ISSUE_TEMPLATE.md

@@ -19,6 +19,7 @@ For any unexpected problems, __PLEASE ALWAYS INCLUDE__:
 	+ TF version: `python -c 'import tensorflow as tf; print(tf.GIT_VERSION, tf.VERSION)'`.
 	+ Tensorpack version: `python -c 'import tensorpack; print(tensorpack.__version__)'`.
       You can install Tensorpack master by `pip install -U git+https://github.com/ppwwyyxx/tensorpack.git`.:
+  + Hardware information, if relevant.
 5. About efficiency, PLEASE first read http://tensorpack.readthedocs.io/en/latest/tutorial/performance-tuning.html
 
 Feature Requests:

examples/FasterRCNN/coco.py

@@ -18,8 +18,6 @@ from config import config as cfg
 
 __all__ = ['COCODetection', 'COCOMeta']
 
-COCO_NUM_CATEGORY = 80
-
 
 class _COCOMeta(object):
     INSTANCE_TO_BASEDIR = {
@@ -39,7 +37,7 @@ class _COCOMeta(object):
         cat_names: list of names
         """
         assert not self.valid()
-        assert len(cat_ids) == COCO_NUM_CATEGORY and len(cat_names) == COCO_NUM_CATEGORY
+        assert len(cat_ids) == cfg.DATA.NUM_CATEGORY and len(cat_names) == cfg.DATA.NUM_CATEGORY
         self.cat_names = cat_names
         self.class_names = ['BG'] + self.cat_names
 

examples/FasterRCNN/config.py

@@ -63,7 +63,7 @@ _C.MODE_FPN = False
 _C.DATA.BASEDIR = '/path/to/your/COCO/DIR'
 _C.DATA.TRAIN = ['train2014', 'valminusminival2014']   # i.e., trainval35k
 _C.DATA.VAL = 'minival2014'   # For now, only support evaluation on single dataset
-_C.DATA.NUM_CATEGORY = 80    # 80 categories
+_C.DATA.NUM_CATEGORY = 80    # 80 categories.
 _C.DATA.CLASS_NAMES = []  # NUM_CLASS strings. Needs to be populated later by data loader
 
 # basemodel ----------------------

examples/FasterRCNN/train.py

@@ -399,7 +399,6 @@ class ResNetFPNModel(DetectionModel):
 def visualize(model, model_path, nr_visualize=100, output_dir='output'):
     """
     Visualize some intermediate results (proposals, raw predictions) inside the pipeline.
-    Does not support FPN.
     """
     df = get_train_dataflow()   # we don't visualize mask stuff
     df.reset_state()

examples/ImageNetModels/imagenet_utils.py

@@ -57,7 +57,7 @@ def fbresnet_augmentor(isTrain):
     if isTrain:
         augmentors = [
             GoogleNetResize(),
-            # It's OK to remove these augs if your CPU is not fast enough.
+            # It's OK to remove the following augs if your CPU is not fast enough.
             # Removing brightness/contrast/saturation does not have a significant effect on accuracy.
             # Removing lighting leads to a tiny drop in accuracy.
             imgaug.RandomOrderAug(

tensorpack/input_source/input_source.py

@@ -48,7 +48,14 @@ class PlaceholderInput(InputSource):
 
 
 class FeedInput(InputSource):
-    """ Input by iterating over a DataFlow and feed datapoints. """
+    """
+    Input by iterating over a DataFlow and feed datapoints.
+
+    Note:
+        If `get_input_tensors()` is called more than one time, it will return the same placeholders (i.e. feed points)
+        as the first time.
+        Therefore you can't use it for data-parallel training.
+    """
 
     class _FeedCallback(Callback):
         def __init__(self, ds, placeholders):

tensorpack/train/trainers.py

@@ -14,7 +14,7 @@ from ..utils.develop import HIDE_DOC
 from ..tfutils import get_global_step_var
 from ..tfutils.distributed import get_distributed_session_creator
 from ..tfutils.tower import TrainTowerContext
-from ..input_source import QueueInput
+from ..input_source import QueueInput, FeedfreeInput
 
 from ..graph_builder.training import (
     SyncMultiGPUParameterServerBuilder,
@@ -59,7 +59,7 @@ class SimpleTrainer(SingleCostTrainer):
 # Only exists for type check & back-compatibility
 class QueueInputTrainer(SimpleTrainer):
     def _setup_graph(self, input, get_cost_fn, get_opt_fn):
-        assert isinstance(input, QueueInput)
+        assert isinstance(input, QueueInput), input
         return super(QueueInputTrainer, self)._setup_graph(input, get_cost_fn, get_opt_fn)
 
 
@@ -87,6 +87,8 @@ class SyncMultiGPUTrainerParameterServer(SingleCostTrainer):
         super(SyncMultiGPUTrainerParameterServer, self).__init__()
 
     def _setup_graph(self, input, get_cost_fn, get_opt_fn):
+        if len(self.devices) > 1:
+            assert isinstance(input, FeedfreeInput), input
         self.train_op = self._builder.build(
             self._make_get_grad_fn(input, get_cost_fn, get_opt_fn), get_opt_fn)
         return []
@@ -124,6 +126,8 @@ class AsyncMultiGPUTrainer(SingleCostTrainer):
         super(AsyncMultiGPUTrainer, self).__init__()
 
     def _setup_graph(self, input, get_cost_fn, get_opt_fn):
+        if len(self.devices) > 1:
+            assert isinstance(input, FeedfreeInput), input
         self.train_op = self._builder.build(
             self._make_get_grad_fn(input, get_cost_fn, get_opt_fn), get_opt_fn)
         return []
@@ -162,6 +166,8 @@ class SyncMultiGPUTrainerReplicated(SingleCostTrainer):
         super(SyncMultiGPUTrainerReplicated, self).__init__()
 
     def _setup_graph(self, input, get_cost_fn, get_opt_fn):
+        if len(self.devices) > 1:
+            assert isinstance(input, FeedfreeInput), input
         self.train_op, post_init_op = self._builder.build(
             self._make_get_grad_fn(input, get_cost_fn, get_opt_fn), get_opt_fn)
 
@@ -220,6 +226,7 @@ class DistributedTrainerParameterServer(DistributedTrainerBase):
         self.is_chief = self._builder.is_chief
 
     def _setup_graph(self, input, get_cost_fn, get_opt_fn):
+        assert isinstance(input, FeedfreeInput), input
         self.train_op = self._builder.build(
             self._make_get_grad_fn(input, get_cost_fn, get_opt_fn), get_opt_fn)
         return []
@@ -255,6 +262,7 @@ class DistributedTrainerReplicated(DistributedTrainerBase):
             return input.setup(inputs_desc)
 
     def _setup_graph(self, input, get_cost_fn, get_opt_fn):
+        assert isinstance(input, FeedfreeInput), input
         self.train_op, initial_sync_op, model_sync_op = self._builder.build(
             self._make_get_grad_fn(input, get_cost_fn, get_opt_fn), get_opt_fn)