wanchaol

graph(%alist.1 : Tensor):
  %26 : int = prim::Constant[value=1]()
  %18 : int = prim::Constant[value=0]()
  %16 : bool = prim::Constant[value=1]() # /home/wanchaol/test.py:194:4
  %4 : None = prim::Constant()
  %1 : int = prim::Constant[value=3]() # /home/wanchaol/test.py:193:22
  %2 : int = prim::Constant[value=4]() # /home/wanchaol/test.py:193:25
  %9 : int = prim::Constant[value=9223372036854775807]() # /home/wanchaol/test.py:194:23
  %3 : int[] = prim::ListConstruct(%1, %2)
  %res.1 : Tensor = aten::randn(%3, %4, %4, %4, %4) # /home/wanchaol/test.py:193:10

wanchaol

graph(%self : ClassType<ModelTrainer>,
      %data.1 : Float(*, *, *, *, *),
      %target.1 : Long(*, *)):
  %829 : int[] = prim::Constant[value=[0, 0]]()
  %828 : int[] = prim::Constant[value=[1, 1]]()
  %3 : str = prim::Constant[value="params"]() # test.py:140:38
  %4 : str = prim::Constant[value="step"]() # test.py:152:18
  %5 : str = prim::Constant[value="weight_decay"]() # test.py:155:29
  %6 : str = prim::Constant[value="lr"]() # test.py:166:32
  %7 : str = prim::Constant[value="lr_decay"]() # test.py:167:56

wanchaol

    class eagerNet(nn.Module):
            def __init__(self):
                super(eagerNet, self).__init__()
                self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
                self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
                self.fc1 = nn.Linear(320, 50)
                self.fc2 = nn.Linear(50, 10)

            def forward(self, x):

wanchaol

```python
    class eagerNet(nn.Module):
            def __init__(self):
                super(eagerNet, self).__init__()
                self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
                self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
                self.fc1 = nn.Linear(320, 50)
                self.fc2 = nn.Linear(50, 10)

            def forward(self, x):

wanchaol

class GreedySearchDecoder(nn.Module):
    def __init__(self, encoder, decoder, decoder_n_layers):
        super(GreedySearchDecoder, self).__init__()
        self.encoder = encoder
        self.decoder = decoder
        self._device = device
        self._SOS_token = SOS_token
        self._decoder_n_layers = decoder_n_layers

    __constants__ = ['_device', '_SOS_token', '_decoder_n_layers']

wanchaol

# Luong attention layer
class Attn(torch.nn.Module):
    def __init__(self, method, hidden_size):
        super(Attn, self).__init__()
        self.method = method
        if self.method not in ['dot', 'general', 'concat']:
            raise ValueError(self.method, "is not an appropriate attention method.")
        self.hidden_size = hidden_size
        if self.method == 'general':
            self.attn = torch.nn.Linear(self.hidden_size, hidden_size)

wanchaol

## first use case
inp = torch.randn(3, 3)

a = torch.jit.script(F.dropout)
a(inp)

## error:
# RuntimeError: 

# undefined value _VF:

wanchaol

pip install git+https://github.com/arraiyopensource/kornia

====
import torch
import torch.nn as nn
from torch.testing import assert_allclose

import kornia

@torch.jit.script

wanchaol

class VAE(nn.Module):
    def __init__(self):
        super(VAE, self).__init__()

        self.fc1 = nn.Linear(784, 400)
        self.fc21 = nn.Linear(400, 20)

    def forward(self, x):
        x = x.view(-1, 784)
        h1 = self.fc1(x)

wanchaol

import torch
import torch.nn as nn
import torch.nn.functional as F

class MnistNet(nn.Module):
    def __init__(self):
        super(MnistNet, self).__init__()
        # self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        # self.conv2 = nn.Conv2d(1, 20, kernel_size=5)
        # self.conv2_drop = nn.Dropout2d()