brew is Caffe2’s new API for building models. The CNNModelHelper filled this role in the past, but since Caffe2 has expanded well beyond excelling at CNNs it made sense to provide a ModelHelper object that is more generic. You may notice that the new ModelHelper has much the same functionality as CNNModelHelper. brew wraps the new ModelHelper making building models even easier than before.

CNNModelHelper is being deprecated. Refer to brew and python/model_helper.py for its replacements.

CNNModelHelper

Code: cnn.py | API Docs: cnn.CNNModelHelper

CNNModelHelper is a helper class so you can write CNN models more easily, without having to manually define parameter initializations and operators separately. You will find many built-in helper functions as well as automatic support for a collection of operators that are listed below.

Example Usage

# Create the input data
data = np.random.rand(16, 100).astype(np.float32)

# Create labels for the data as integers [0, 9].
label = (np.random.rand(16) * 10).astype(np.int32)

workspace.FeedBlob("data", data)
workspace.FeedBlob("label", label)

# Create model using a model helper
m = cnn.CNNModelHelper(name="my first net")
fc_1 = m.FC("data", "fc1", dim_in=100, dim_out=10)
pred = m.Sigmoid(fc_1, "pred")
[softmax, loss] = m.SoftmaxWithLoss([pred, "label"], ["softmax", "loss"])

Arguments
order=”NCHW”,	ws_nbytes_limit=None, init_params=True,
name=None,	skip_sparse_optim=False,
use_cudnn=True,	param_model=None
cudnn_exhaustive_search=False,

Functions
Accuracy	GetWeights
AddWeightDecay	ImageInput
AveragePool	InstanceNorm
Concat	Iter
Conv	LRN
ConvTranspose	LSTM
DepthConcat	MaxPool
Dropout	PackedFC
FC	PadImage
FC_Decomp	PRelu
FC_Prune	Relu
FC_Sparse	SpatialBN
GetBiases	Sum
GroupConv	Transpose

Operators
Accuracy	NCCLAllreduce
Adam	NHWC2NCHW
Add	PackSegments
Adagrad	Print
SparseAdagrad	PRelu
AveragedLoss	Scale
Cast	ScatterWeightedSum
Checkpoint	Sigmoid
ConstantFill	SortedSegmentSum
Copy	Snapshot # Note: snapshot is deprecated use Checkpoint
CopyGPUToCPU	Softmax
CopyCPUToGPU	SoftmaxWithLoss
DequeueBlobs	SquaredL2Distance
EnsureCPUOutput	Squeeze
Flatten	StopGradient
FlattenToVec	Summarize
LabelCrossEntropy	Tanh
LearningRate	UnpackSegments
MakeTwoClass	WeightedSum
MatMul

Edit on GitHub