Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Can't find the params #43

Open
zmHaiNan opened this issue Jun 8, 2018 · 5 comments
Open

Can't find the params #43

zmHaiNan opened this issue Jun 8, 2018 · 5 comments

Comments

@zmHaiNan
Copy link

zmHaiNan commented Jun 8, 2018

Hello,I can't find these:
http://homes.cs.washington.edu/~jxie/download/deep3d-0050.params &&
eep3d-0050.params
I don't know what's the matter.
can you send me your params? Tank you !
mail to : i@hnxh.ai

@n1ckfg
Copy link

n1ckfg commented Jun 8, 2018

@zmHaiNan
Copy link
Author

zmHaiNan commented Jun 8, 2018

I've tried.This url is still wrong...

@n1ckfg
Copy link

n1ckfg commented Jun 8, 2018

No, I mean the file you're looking for is preserved in my fork

@zmHaiNan
Copy link
Author

zmHaiNan commented Jun 8, 2018

Thank you,I have just downloaded it.

@zmHaiNan
Copy link
Author

hello,I open your deep3d.ipython when i load the params:
"""
model = mx.model.FeedForward.load('./deep3d',50,ctx=ctx)

MXNetError Traceback (most recent call last)
in ()
5 print(ctx)
6 #model = mx.model.FeedForward.load('deep3d',50,ctx=ctx)
----> 7 model = mx.model.FeedForward.load('./deep3d',50,ctx=ctx)

/home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/model.pyc in load(prefix, epoch, ctx, **kwargs)
921 - prefix-epoch.params will be saved for parameters.
922 """
--> 923 symbol, arg_params, aux_params = load_checkpoint(prefix, epoch)
924 return FeedForward(symbol, ctx=ctx,
925 arg_params=arg_params, aux_params=aux_params,

/home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/model.pyc in load_checkpoint(prefix, epoch)
418 - Parameters will be loaded from prefix-epoch.params.
419 """
--> 420 symbol = sym.load('%s-symbol.json' % prefix)
421 save_dict = nd.load('%s-%04d.params' % (prefix, epoch))
422 arg_params = {}

/home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/symbol/symbol.pyc in load(fname)
2516 raise TypeError('fname need to be string')
2517 handle = SymbolHandle()
-> 2518 check_call(_LIB.MXSymbolCreateFromFile(c_str(fname), ctypes.byref(handle)))
2519 return Symbol(handle)
2520

/home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/base.pyc in check_call(ret)
144 """
145 if ret != 0:
--> 146 raise MXNetError(py_str(_LIB.MXGetLastError()))
147
148

MXNetError: Failed loading Op depth of type DepthDot: [09:30:47] src/core/op.cc:55: Check failed: op != nullptr Operator DepthDot is not registered

Stack trace returned 10 entries:
[bt] (0) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x2ab998) [0x7f194f624998]
[bt] (1) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x6234ee5) [0x7f19555adee5]
[bt] (2) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x6266c3b) [0x7f19555dfc3b]
[bt] (3) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x626bf60) [0x7f19555e4f60]
[bt] (4) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x626580f) [0x7f19555de80f]
[bt] (5) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x262d28f) [0x7f19519a628f]
[bt] (6) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x62381dd) [0x7f19555b11dd]
[bt] (7) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x262e421) [0x7f19519a7421]
[bt] (8) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x262d28f) [0x7f19519a628f]
[bt] (9) /home/quantumliu/anaconda3/envs/mx27/lib/python2.7/site-packages/mxnet/libmxnet.so(+0x62381dd) [0x7f19555b11dd]

"""
It occuers an error,I don't know whether is the param too old to follow the new mxnet api .
Can you help me ? Thank you.

# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants