How to get the theano models?

[meisa233]

The link is unavailable.
Can anyone help me?

[Junaid112]

after training via "vsum_train.py' model should be saved and you can use for inference. Did you train successfully or you are looking for already trained model?

[meisa233]

after training via "vsum_train.py' model should be saved and you can use for inference. Did you train successfully or you are looking for already trained model?

Thanks for your reply. I am looking for already trained model in order to compare with my method.

[Junaid112]

I think there is no link to already trained model but I have just trained using pytorch implementation of this by same developer "https://github.com/KaiyangZhou/pytorch-vsumm-reinforce" so either you can train using the instruction or I can share trained model files.

[meisa233]

I think there is no link to already trained model but I have just trained using pytorch implementation of this by same developer "https://github.com/KaiyangZhou/pytorch-vsumm-reinforce" so either you can train using the instruction or I can share trained model files.

Can you share your model? Thank you very much.

[meisa233]

I think there is no link to already trained model but I have just trained using pytorch implementation of this by same developer "https://github.com/KaiyangZhou/pytorch-vsumm-reinforce" so either you can train using the instruction or I can share trained model files.

I have trained my model using Theano with same results in this paper. Thanks for your reply.^ ^

[jayagupta678]

I am running "vsum_train.py' to regenerate model files, but getting following error message. please help me to recreate results used in the paper.

(bennett) C:\Users\Administrator\Downloads\vsumm-reinforce-master\vsumm-reinforce-master>python vsum_train.py
WARNING (theano.configdefaults): g++ not available, if using conda: conda install m2w64-toolchain
WARNING (theano.configdefaults): g++ not detected ! Theano will be unable to execute optimized C-implementations (for both CPU and GPU) and will default to Python implementations. Performance will be severely degraded. To remove this warning, set Theano flags cxx to an empty string.
[13/09/2021 06:20:40] model options: {'n_episodes': 5, 'input_dim': 1024, 'hidden_dim': 256, 'W_init': 'normal', 'U_init': 'normal', 'weight_decay': 1e-05, 'regularizer': 'L2', 'optimizer': 'adam', 'base_lr': 1e-05, 'decay_rate': 0.1, 'max_epochs': 60, 'decay_stepsize': -1, 'ignore_distant_sim': False, 'distant_sim_thre': 20, 'alpha': 0.01, 'model_file': None, 'disp_freq': 1, 'train_dataset_path': 'datasets/eccv16_dataset_summe_google_pool5.h5'}
[13/09/2021 06:20:40] initializing net model
Traceback (most recent call last):
File "vsum_train.py", line 145, in
train(n_episodes=args.n_epi,
File "vsum_train.py", line 59, in train
net = reinforceRNN(model_options)
File "C:\Users\Administrator\Downloads\vsumm-reinforce-master\vsumm-reinforce-master\model_reinforceRNN.py", line 37, in init
self.fwd_rnn = LSTM(
File "C:\Users\Administrator\Downloads\vsumm-reinforce-master\vsumm-reinforce-master\theano_nets.py", line 343, in init
self.output = self.step(self.state_below)
File "C:\Users\Administrator\Downloads\vsumm-reinforce-master\vsumm-reinforce-master\theano_nets.py", line 393, in step
rval,_ = theano.scan(
File "C:\Users\Administrator\anaconda3\envs\bennett\lib\site-packages\theano\scan_module\scan.py", line 1071, in scan
scan_outs = local_op(*scan_inputs)
File "C:\Users\Administrator\anaconda3\envs\bennett\lib\site-packages\theano\gof\op.py", line 615, in call
node = self.make_node(*inputs, **kwargs)
File "C:\Users\Administrator\anaconda3\envs\bennett\lib\site-packages\theano\scan_module\scan_op.py", line 568, in make_node
raise ValueError(err_msg2 %
ValueError: When compiling the inner function of scan the following error has been encountered: The initial state (outputs_info in scan nomenclature) of variable IncSubtensor{Set;:int64:}.0 (argument number 1) has dtype float32, while the result of the inner function (fn) has dtype float64. This can happen if the inner function of scan results in an upcast or downcast.