Lesson 4 IMDB Test Part Fails


#1

When running this code in IMDB lesson notebook:

s = "Cars have already started the march toward autonomy"
m = learner.model

# def proc_str(s): return TEXT.preprocess(TEXT.tokenize(s))
def proc_str(s): return [spacy_tok(s)]
def num_str(s): return TEXT.numericalize([proc_str(s)])

def sample_model(m, s, l=50):
    t = num_str(s)
    m[0].bs=1
    m.eval()
    m.reset()
    res,*_ = m(t)
    print('...', end='')

    for i in range(l):
        n=res[-1].topk(2)[1]
        n = n[1] if n.data[0]==0 else n[0]
        word = TEXT.vocab.itos[n.data[0]]
        print(word, end=' ')
        if word=='<eos>': break
        res,*_ = m(n[0].unsqueeze(0))

    m[0].bs=bs

sample_model(m,s)

I got the error:

ValueError                    Traceback (most recent call last)
<ipython-input-109-6b9c36a1a857> in <module>()
     24     m[0].bs=bs
     25 
---> 26 sample_model(m,s)

<ipython-input-109-6b9c36a1a857> in sample_model(m, s, l)
     20         print(word, end=' ')
     21         if word=='<eos>': break
---> 22         res,*_ = m(n[0].unsqueeze(0))
     23 
     24     m[0].bs=bs

~/anaconda3/envs/fastai/lib/python3.6/site-packages/torch/nn/modules/module.py in __call__(self, *input, **kwargs)
    489             result = self._slow_forward(*input, **kwargs)
    490         else:
--> 491             result = self.forward(*input, **kwargs)
    492         for hook in self._forward_hooks.values():
    493             hook_result = hook(self, input, result)

~/anaconda3/envs/fastai/lib/python3.6/site-packages/torch/nn/modules/container.py in forward(self, input)
     89     def forward(self, input):
     90         for module in self._modules.values():
---> 91             input = module(input)
     92         return input
     93 

~/anaconda3/envs/fastai/lib/python3.6/site-packages/torch/nn/modules/module.py in __call__(self, *input, **kwargs)
    489             result = self._slow_forward(*input, **kwargs)
    490         else:
--> 491             result = self.forward(*input, **kwargs)
    492         for hook in self._forward_hooks.values():
    493             hook_result = hook(self, input, result)

~/fastai/courses/dl1/fastai/lm_rnn.py in forward(self, input)
     91             dropouth, list of tensors evaluated from each RNN layer using dropouth,
     92         """
---> 93         sl,bs = input.size()
     94         if bs!=self.bs:
     95             self.bs=bs

ValueError: not enough values to unpack (expected 2, got 1)

I figured that the result of n[0].unsqueeze(0) is a 1 dimensional (?) tensor when the code expects it to be 2 values. Any ideas on how to fix that?


Wiki: Lesson 4
Unofficial pytorch 0.4 support
(Leon Dummer) #2

I’m stuck at the same problem…


(Borys Zibrov) #3

I hit the same problem and I think it’s due to another pytorch version used by me, after brief investigation it seems like n[0].unsqueeze(0) is just single number (like [34]) whereas model expects two dimensions (like [ [34] ]). So unsqueeze actually did nothing in this case.


(Borys Zibrov) #4

Temporarily solved it like this

        n=res[-1].topk(2)[1]
        n = n[1] if n.data[0]==0 else n[0]
        word = TEXT.vocab.itos[n.item()]
        print(word, end=' ')
        if word=='<eos>': break
        res,*_ = m(torch.tensor([n]).unsqueeze(0).cuda())

I think new version of fast.ai will be using new pytorch version.

Also, (not yet released) torchtext 0.3 is required for pytorch 0.4