search

ijkguo / mx-rcnn

Parallel Faster R-CNN implementation with MXNet.
Other
671 stars 290 forks source link

Error in downloading pre-trained model #47

Closed dakshvar22 closed 7 years ago

dakshvar22 commented 7 years ago

Hi,

I am trying to download the basic pre-trained model given in the README from this link - https://www.dropbox.com/s/jrr83q0ai2ckltq/final-0000.params.tar.gz?dl=0 I tried multiple times but the download fails after downloading around 320 MB of the file. Tried switching to different networks also, but no luck. Can someone confirm if the file is not corrupted on the link?

For the different link posted - http://pan.baidu.com/share/init?shareid=2372813683&uk=3979418934 , the page seems to be in chinese I think and it asks for some password I guess. Can I get access to that?

Thanks

ijkguo commented 7 years ago

The password is in the brackets Baidu Yun (ixiw). I will check dropbox link soon.

dakshvar22 commented 7 years ago

@precedenceguo Similar problem seems to happen with the Baidu link too. Can you check that too please?

dakshvar22 commented 7 years ago

@precedenceguo I managed to download from the dropbox link. Thank you so much :)

dakshvar22 commented 7 years ago

Running the demo program gives me this error -

Traceback (most recent call last): File "demo.py", line 139, in main() File "demo.py", line 119, in main demo_net(predictor, args.image) File "demo.py", line 70, in demo_net scores, boxes, data_dict = im_detect(predictor, data_batch, data_names, im_scale) File "/home/tarun/daksh/mx-rcnn-master/rcnn/core/tester.py", line 106, in im_detect output = predictor.predict(data_batch) File "/home/tarun/daksh/mx-rcnn-master/rcnn/core/tester.py", line 25, in predict self._mod.forward(data_batch) File "/home/tarun/daksh/mx-rcnn-master/rcnn/core/module.py", line 187, in forward shared_module=self._curr_module) File "/usr/local/lib/python2.7/dist-packages/mxnet-0.9.1-py2.7-linux-x86_64.egg/mxnet/module/module.py", line 344, in bind grad_req=grad_req, input_types=input_types) File "/usr/local/lib/python2.7/dist-packages/mxnet-0.9.1-py2.7-linux-x86_64.egg/mxnet/module/executor_group.py", line 188, in init self.label_layouts = self.decide_slices(label_shapes) File "/usr/local/lib/python2.7/dist-packages/mxnet-0.9.1-py2.7-linux-x86_64.egg/mxnet/module/executor_group.py", line 203, in decide_slices assert len(data_shapes) > 0 AssertionError

Can you help out?

ijkguo commented 7 years ago

Yeah this is new error from recent update to nnvm. I will fix it soon.

dakshvar22 commented 7 years ago

@precedenceguo okay. In the mean time, I really want to test demo.py and see the results. I assume it should work on your forked version of mxnet. I tried installing that version following the steps of this link - https://no2147483647.wordpress.com/2015/12/07/deep-learning-for-hackers-with-mxnet-1/ (replaced the repo of git clone with that of yours). But the installation failed. Can you give the instructions on how to install your forked version of mxnet?

ijkguo commented 7 years ago

It should be the same as in the guide but make sure you are on the legacy branch.

ijkguo commented 7 years ago

OK I have updated. Please try again. I recommend http://mxnet.io/get_started/ubuntu_setup.html as the installation guide.

ghost commented 7 years ago

A json file "resnet-101-symbol.json" may be better for defining the resnet model.

ijkguo commented 7 years ago

We do not use json file for historic reasons.