shuokay / mxnet-ir Goto Github PK

the code as follows:
"from symbol_resnet import resnet"
but where is the symbol_resnet

I read the definition of class Auc(), but i do not understand why it can evaluate the model.

程序在运行时报了这个错误：

/Users/nali/PycharmProjects/mxnet-ir/test.py:96: DeprecationWarning: mxnet.model.FeedForward has been deprecated. Please use mxnet.mod.Module instead.
optimizer=optimizer)
/Library/Python/2.7/site-packages/mxnet/initializer.py:267: DeprecationWarning: Calling initializer with init(str, NDArray) has been deprecated.please use init(mx.init.InitDesc(...), NDArray) instead.
self.default_init(name, arr)
Traceback (most recent call last):
File "/Users/nali/PycharmProjects/mxnet-ir/test.py", line 105, in
test_train()
File "/Users/nali/PycharmProjects/mxnet-ir/test.py", line 101, in test_train
epoch_end_callback=mx.callback.do_checkpoint("models/ir-blur"))
File "/Library/Python/2.7/site-packages/mxnet/model.py", line 782, in fit
self._init_params(data.provide_data+data.provide_label)
File "/Library/Python/2.7/site-packages/mxnet/model.py", line 526, in _init_params
self.initializer(k, v)
File "/Library/Python/2.7/site-packages/mxnet/initializer.py", line 267, in call
self.default_init(name, arr)
File "/Library/Python/2.7/site-packages/mxnet/initializer.py", line 72, in call
self._legacy_init(desc, arr)
File "/Library/Python/2.7/site-packages/mxnet/initializer.py", line 138, in _legacy_init
self._init_default(name, arr)
File "/Library/Python/2.7/site-packages/mxnet/initializer.py", line 180, in _init_default
'Please use mx.sym.Variable(init=mx.init.) to set initialization pattern' % name)
ValueError: Unknown initialization pattern for softmax_label. Default initialization is now limited to "weight", "bias", "gamma" (1.0), and "beta" (0.0).Please use mx.sym.Variable(init=mx.init.) to set initialization pattern

train_resnet.py:238: DeprecationWarning: mxnet.model.FeedForward has been deprecated. Please use mxnet.mod.Module instead.
optimizer=optimizer,
2017-10-31 13:42:05,376 Start training with [gpu(0), gpu(1), gpu(2), gpu(3)]
[13:42:05] /media/zhaijunzhi/mxnet/dmlc-core/include/dmlc/logging.h:308: [13:42:05] src/operator/tensor/./matrix_op-inl.h:867: Check failed: (*end <= axis_size) && (*end >= 0) Invalid begin, end, get begin=0, end=64

Stack trace returned 10 entries:
[bt] (0) /usr/local/lib/python2.7/dist-packages/mxnet-0.11.1-py2.7.egg/mxnet/libmxnet.so(_ZN4dmlc15LogMessageFatalD1Ev+0x3c) [0x7fea14bf39ac]
[bt] (1) /usr/local/lib/python2.7/dist-packages/mxnet-0.11.1-py2.7.egg/mxnet/libmxnet.so(ZN5mxnet2op18GetSliceAxisParamsERKNS0_14SliceAxisParamERKN4nnvm6TShapeEPiS8_S8+0x2f1) [0x7fea15d71591]
[bt] (2) /usr/local/lib/python2.7/dist-packages/mxnet-0.11.1-py2.7.egg/mxnet/libmxnet.so(ZN5mxnet2op14SliceAxisShapeERKN4nnvm9NodeAttrsEPSt6vectorINS1_6TShapeESaIS6_EES9+0x1ad) [0x7fea15d7b70d]
[bt] (3) /usr/local/lib/python2.7/dist-packages/mxnet-0.11.1-py2.7.egg/mxnet/libmxnet.so(+0x1b7911e) [0x7fea15f7911e]
[bt] (4) /usr/local/lib/python2.7/dist-packages/mxnet-0.11.1-py2.7.egg/mxnet/libmxnet.so(_ZN5mxnet4exec10InferShapeEON4nnvm5GraphEOSt6vectorINS1_6TShapeESaIS5_EERKSs+0x1670) [0x7fea15f7bc50]
[bt] (5) /usr/local/lib/python2.7/dist-packages/mxnet-0.11.1-py2.7.egg/mxnet/libmxnet.so(MXSymbolInferShape+0x1539) [0x7fea15f07b79]
[bt] (6) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call_unix64+0x4c) [0x7fea30a2fadc]
[bt] (7) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call+0x1fc) [0x7fea30a2f40c]
[bt] (8) /usr/lib/python2.7/lib-dynload/_ctypes.x86_64-linux-gnu.so(_ctypes_callproc+0x48e) [0x7fea30c465fe]
[bt] (9) /usr/lib/python2.7/lib-dynload/_ctypes.x86_64-linux-gnu.so(+0x15f9e) [0x7fea30c47f9e]

I get an infer_shape error, and I am new to the use of triplet loss an I would like to know if I need to separate my images into differents folder depending on their batch and label?

How do I specify which image is same of different for each anchor
(which are also images and could also be same of different later on)

我在运行时，发现ir-resnet.py中get_network函数调用了resnet函数。

我在网上找到了这个版本:https://github.com/tornadomeet/ResNet/blob/master/symbol_resnet.py

但是这里的resnet参数跟你的调用不符。你的调用中多了data参数
请问这个resnet的代码哪里可以找到？

非常感谢你的博客与源码，受益颇多，谢谢！