How can i get pooled output from deberta model? - pytorch

Can anyone tell how can i get pooled output form deberta model ?
Can anyone tell how can i get pooled output form deberta model ? i would like to use it from DebertaModel for my classification model without using *DebertaForSequenceClassification *

i solved it like this :
deberta_model = DebertaForSequenceClassification.from_pretrained("microsoft/deberta-base")
deberta_model.config.num_labels = 1
class DebrtaRegressor(nn.Module):
def __init__(self):
super(DebrtaRegressor, self).__init__()
self.deberta = deberta_model
self.sigmoid1 = nn.Sigmoid()
def forward(self, input_ids, attention_masks):
outputs = self.deberta(input_ids, attention_masks)
outputs = outputs.logits[:, : 1]
outputs = self.sigmoid1(outputs)
return outputs

Related

Pytorch netwrok with variable number of hidden layers

I want to create a class that creates a simple network with X fully connected layers, where X is an input given by the user. I tried this using the setattr/getattr but for some reason is not working.
class MLP(nn.Module):
def __init__(self,in_size, out_size,n_layers, hidden_size):
super(MLP,self).__init__()
self.n_layers=n_layers
for i in range(n_layers):
if i==0:
layer_in_size = in_size
else:
layer_in_size = hidden_size
if i==(n_layers-1):
layer_out_size = out_size
else:
layer_out_size = hidden_size
setattr(self,'dense_{}'.format(i), nn.Linear(layer_in_size,layer_out_size))
def forward(self,x):
out = x
for i in range(self.n_layers):
if i==(self.n_layers-1):
out = getattr(self,'dense_{}'.format(i),out)
else:
out = F.relu(getattr(self,'dense_{}'.format(i),out))
return out
This is the error im getting when trying a forward pass with the net:
enter image description here
Some insights of what's the issue will be helpful.
This seems like a problem with forward implementation with the mod2 function. Try the pytorch functions (torch.fmod and torch.remainder) or if you don't need the backprop capabilities try to do .detach() before the mod2 function.

Does Keras official sample code about Transformer applied in time-series contain Position Embedding part?

The sample code for referring from url:https://keras.io/examples/timeseries/timeseries_transformer_classification/
I could not find out any description about "Position Embedding" content in full page of above url. When I looked through Transformer applied in NLP, I can clearly see the class named "TokenAndPositionEmbedding".
If it does not contain "Position Embedding", how can I apply Position Embedding in time series in sample code?
From what I can tell it does not contain the positional embedding. Something like this should work.
class PositionEmbeddingFixedWeights(Layer):
def __init__(self, sequence_length, vocab_size, output_dim, **kwargs):
super(PositionEmbeddingFixedWeights, self).__init__(**kwargs)
word_embedding_matrix = self.get_position_encoding(vocab_size, output_dim)
position_embedding_matrix = self.get_position_encoding(sequence_length, output_dim)
self.word_embedding_layer = Embedding(
input_dim=vocab_size, output_dim=output_dim,
weights=[word_embedding_matrix],
trainable=False
)
self.position_embedding_layer = Embedding(
input_dim=sequence_length, output_dim=output_dim,
weights=[position_embedding_matrix],
trainable=False
)
def get_position_encoding(self, seq_len, d, n=10000):
P = np.zeros((seq_len, d))
for k in range(seq_len):
for i in np.arange(int(d/2)):
denominator = np.power(n, 2*i/d)
P[k, 2*i] = np.sin(k/denominator)
P[k, 2*i+1] = np.cos(k/denominator)
return P
def call(self, inputs):
position_indices = tf.range(tf.shape(inputs)[-1])
embedded_words = self.word_embedding_layer(inputs)
embedded_indices = self.position_embedding_layer(position_indices)
return embedded_words + embedded_indices
This class originated from https://machinelearningmastery.com/the-transformer-positional-encoding-layer-in-keras-part-2/

Custom layer from keras to pytorch

Coming from TensorFlow background, I am trying to convert a snippet of code of the custom layer from Keras to PyTorch.
The custom layer in Keras looks like this:
class Attention_module(tf.keras.layers.Layer):
def __init__(self, class_num):
super(Attention_module,self).__init__(class_num)
self.class_num = class_num
self.Ws = None
def build(self, input_shape):
embedding_length = int(input_shape[2])
self.Ws = self.add_weight(shape=(self.class_num, embedding_length),
initializer=tf.keras.initializers.get('glorot_uniform'), trainable=True)
super(Attention_module, self).build(input_shape)
def call(self, inputs):
sentence_trans = tf.transpose(inputs, [0, 2, 1])
at = tf.matmul(self.Ws, sentence_trans)
at = tf.math.tanh(at)
at = K.exp(at - K.max(at, axis=-1, keepdims=True))
at = at / K.sum(at, axis=-1, keepdims=True)
v = K.batch_dot(at, inputs)
return v
I want to implement the same in the torch; I have already done the forward pass block but am confused about how to do the embedding and weight initialization the same as the above layer in PyTorch?
class Attention_module(torch.nn.Module):
def __init__(self, class_num):
# how to initialize weight with same as above keras layer?
def forward(self, inputs):
sentence_trans = inputs.permute(0, 2, 1)
at = torch.mm(self.Ws, sentence_trans)
at = torch.nn.Tanh(at)
at = torch.exp(at - torch.max(torch.Tensor(at), dim=-1, keepdims=True).values)
at = at / torch.sum(at, dim = -1, keepdims=True)
v = torch.einsum('ijk,ikl->ijl', at, inputs)
return v
Thank you!
class Attention_module(torch.nn.Module):
def __init__(self, class_num, input_shape):
super().__init__()
self.class_num = class_num
embedding_length = int(input_shape[2])
self.Ws = torch.nn.Embedding(num_embeddings=class_num,
embedding_dim=embedding_length) # Embedding layer
torch.nn.init.xavier_uniform_(self.Ws.weight) # Glorot initialization
Here's the reference for layer initialization methods. Xavier init is another name for Glorot init.
The _ at the end of torch.nn.init.xavier_uniform_ is a pytorch convention that signifies an inplace operation.
You can also use torch.nn.init at runtime. It doesn't have to be within __init__(). Like:
att = Attention_module(class_num, input_shape)
torch.nn.init.xavier_uniform_(att.Ws.weight)
or :
for param in att.parameters():
torch.nn.init.xavier_uniform_(param)

Pytorch dynamic amount of Layers?

I am trying to specify a dynamic amount of layers, which I seem to be doing wrong.
My issue is that when I define the 100 layers here, I will get an error in the forward step.
But when I define the layer properly it works?
Below simplified example
class PredictFromEmbeddParaSmall(LightningModule):
def __init__(self, hyperparams={'lr': 0.0001}):
super(PredictFromEmbeddParaSmall, self).__init__()
#Input is something like tensor.size=[768*100]
self.TO_ILLUSTRATE = nn.Linear(768, 5)
self.enc_ref=[]
for i in range(100):
self.enc_red.append(nn.Linear(768, 5))
# gather the layers output sth
self.dense_simple1 = nn.Linear(5*100, 2)
self.output = nn.Sigmoid()
def forward(self, x):
# first input to enc_red
x_vecs = []
for i in range(self.para_count):
layer = self.enc_red[i]
# The first dim is the batch size here, output is correct
processed_slice = x[:, i * 768:(i + 1) * 768]
# This works and give the out of size 5
rand = self.TO_ILLUSTRATE(processed_slice)
#This will fail? Error below
ret = layer(processed_slice)
#more things happening we can ignore right now since we fail earlier
I get this error when executing "ret = layer.forward(processed_slice)"
RuntimeError: Expected object of device type cuda but got device type
cpu for argument #1 'self' in call to _th_addmm
Is there a smarter way to program this? OR solve the error?
You should use a ModuleList from pytorch instead of a list: https://pytorch.org/docs/master/generated/torch.nn.ModuleList.html . That is because Pytorch has to keep a graph with all modules of your model, if you just add them in a list they are not properly indexed in the graph, resulting in the error you faced.
Your coude should be something alike:
class PredictFromEmbeddParaSmall(LightningModule):
def __init__(self, hyperparams={'lr': 0.0001}):
super(PredictFromEmbeddParaSmall, self).__init__()
#Input is something like tensor.size=[768*100]
self.TO_ILLUSTRATE = nn.Linear(768, 5)
self.enc_ref=nn.ModuleList() # << MODIFIED LINE <<
for i in range(100):
self.enc_red.append(nn.Linear(768, 5))
# gather the layers output sth
self.dense_simple1 = nn.Linear(5*100, 2)
self.output = nn.Sigmoid()
def forward(self, x):
# first input to enc_red
x_vecs = []
for i in range(self.para_count):
layer = self.enc_red[i]
# The first dim is the batch size here, output is correct
processed_slice = x[:, i * 768:(i + 1) * 768]
# This works and give the out of size 5
rand = self.TO_ILLUSTRATE(processed_slice)
#This will fail? Error below
ret = layer(processed_slice)
#more things happening we can ignore right now since we fail earlier
Then it should work all right!
Edit: alternative way.
Instead of using ModuleList you can also just use nn.Sequential, this allows you to avoid using the for loop in the forward pass. That also means that you will not have access to intermediary activations, so that is not the solution for you if you need them.
class PredictFromEmbeddParaSmall(LightningModule):
def __init__(self, hyperparams={'lr': 0.0001}):
super(PredictFromEmbeddParaSmall, self).__init__()
#Input is something like tensor.size=[768*100]
self.TO_ILLUSTRATE = nn.Linear(768, 5)
self.enc_ref=[]
for i in range(100):
self.enc_red.append(nn.Linear(768, 5))
self.enc_red = nn.Seqential(*self.enc_ref) # << MODIFIED LINE <<
# gather the layers output sth
self.dense_simple1 = nn.Linear(5*100, 2)
self.output = nn.Sigmoid()
def forward(self, x):
# first input to enc_red
x_vecs = []
out = self.enc_red(x) # << MODIFIED LINE <<
A little bit more adjustable solution which comes down to matter of taste or complexity of your exact situation was posted here.
For reference I post an adjusted version of the code here:
import torch
from torch import nn, optim
from torch.nn.modules import Module
from implem.settings import settings
class Model(nn.Module):
def __init__(self, input_size, layers_data: list, learning_rate=0.01, optimizer=optim.Adam):
super().__init__()
self.layers = nn.ModuleList()
self.input_size = input_size # Can be useful later ...
for size, activation in layers_data:
self.layers.append(nn.Linear(input_size, size))
input_size = size # For the next layer
if activation is not None:
assert isinstance(activation, Module), \
"Each tuples should contain a size (int) and a torch.nn.modules.Module."
self.layers.append(activation)
self.device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
self.learning_rate = learning_rate
self.optimizer = optimizer(params=self.parameters(), lr=learning_rate)
def forward(self, input_data):
for layer in self.layers:
input_data = layer(input_data)
return input_data
# test that the net is working properly
if __name__ == "__main__":
data_size = 5
layer1, layer2 = 10, 10
output_size = 2
data = torch.randn(data_size)
mlp = Model(data_size, [(layer1, nn.ReLU()), (layer2, nn.ReLU()), (output_size, nn.Sigmoid())])
output = mlp(data)
print("done")

ValueError: optimizer got an empty parameter list

I create the following simple linear class:
class Decoder(nn.Module):
def __init__(self, K, h=()):
super().__init__()
h = (K,)+h+(K,)
self.layers = [nn.Linear(h1,h2) for h1,h2 in zip(h, h[1:])]
def forward(self, x):
for layer in self.layers[:-1]:
x = F.relu(layer(x))
return self.layers[-1](x)
However, when I try to put the parameters in a optimizer class I get the error ValueError: optimizer got an empty parameter list.
decoder = Decoder(4)
LR = 1e-3
opt = optim.Adam(decoder.parameters(), lr=LR)
Is there something I'm doing obviously wrong with the class definition?
Since you store your layers in a regular pythonic list inside your Decoder, Pytorch has no way of telling these members of the self.list are actually sub modules. Convert this list into pytorch's nn.ModuleList and your problem will be solved
class Decoder(nn.Module):
def __init__(self, K, h=()):
super().__init__()
h = (K,)+h+(K,)
self.layers = nn.ModuleList(nn.Linear(h1,h2) for h1,h2 in zip(h, h[1:]))

Resources