Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

type hints for pytorch models #17064

Merged
merged 4 commits into from
May 5, 2022
Merged

type hints for pytorch models #17064

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
9e72fcb
type hints for pytorch models
robotjellyzone May 3, 2022
e50750a
fixed import error
robotjellyzone May 3, 2022
19c5bc0
Merge branch 'main' into type_hint_pytorch
robotjellyzone May 4, 2022
9e32033
fixed some errors
robotjellyzone May 4, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
193 changes: 101 additions & 92 deletions src/transformers/models/canine/modeling_canine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,7 @@
import math
import os
from dataclasses import dataclass
from typing import Optional, Tuple
from typing import Optional, Tuple, Union

import torch
import torch.utils.checkpoint
Expand Down Expand Up @@ -253,11 +253,11 @@ def _embed_hash_buckets(self, input_ids, embedding_size: int, num_hashes: int, n

def forward(
self,
input_ids=None,
token_type_ids=None,
position_ids=None,
inputs_embeds=None,
):
input_ids: Optional[torch.LongTensor] = None,
token_type_ids: Optional[torch.LongTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
) -> torch.FloatTensor:
if input_ids is not None:
input_shape = input_ids.size()
else:
Expand Down Expand Up @@ -356,7 +356,11 @@ def __init__(self, config):
self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
self.dropout = nn.Dropout(config.hidden_dropout_prob)

def forward(self, inputs, final_seq_char_positions=None):
def forward(
self,
inputs: torch.Tensor,
final_seq_char_positions: Optional[torch.Tensor] = None,
) -> torch.Tensor:
# inputs has shape [batch, mol_seq, molecule_hidden_size+char_hidden_final]
# we transpose it to be [batch, molecule_hidden_size+char_hidden_final, mol_seq]
inputs = torch.transpose(inputs, 1, 2)
Expand Down Expand Up @@ -419,12 +423,12 @@ def transpose_for_scores(self, x):

def forward(
self,
from_tensor,
to_tensor,
attention_mask=None,
head_mask=None,
output_attentions=False,
):
from_tensor: torch.Tensor,
to_tensor: torch.Tensor,
attention_mask: Optional[torch.FloatTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
output_attentions: Optional[bool] = False,
) -> Tuple[torch.Tensor, Optional[torch.Tensor]]:
mixed_query_layer = self.query(from_tensor)

# If this is instantiated as a cross-attention module, the keys
Expand Down Expand Up @@ -496,7 +500,9 @@ def __init__(self, config):
self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
self.dropout = nn.Dropout(config.hidden_dropout_prob)

def forward(self, hidden_states, input_tensor):
def forward(
self, hidden_states: Tuple[torch.FloatTensor], input_tensor: torch.FloatTensor
) -> Tuple[torch.FloatTensor, torch.FloatTensor]:
hidden_states = self.dense(hidden_states)
hidden_states = self.dropout(hidden_states)
hidden_states = self.LayerNorm(hidden_states + input_tensor)
Expand Down Expand Up @@ -574,11 +580,11 @@ def prune_heads(self, heads):

def forward(
self,
hidden_states,
attention_mask=None,
head_mask=None,
output_attentions=False,
):
hidden_states: Tuple[torch.FloatTensor],
attention_mask: Optional[torch.FloatTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
output_attentions: Optional[bool] = False,
) -> Tuple[torch.FloatTensor, Optional[torch.FloatTensor]]:
if not self.local:
self_outputs = self.self(hidden_states, hidden_states, attention_mask, head_mask, output_attentions)
attention_output = self_outputs[0]
Expand Down Expand Up @@ -656,7 +662,7 @@ def __init__(self, config):
else:
self.intermediate_act_fn = config.hidden_act

def forward(self, hidden_states):
def forward(self, hidden_states: torch.FloatTensor) -> torch.FloatTensor:
hidden_states = self.dense(hidden_states)
hidden_states = self.intermediate_act_fn(hidden_states)
return hidden_states
Expand All @@ -669,7 +675,7 @@ def __init__(self, config):
self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
self.dropout = nn.Dropout(config.hidden_dropout_prob)

def forward(self, hidden_states, input_tensor):
def forward(self, hidden_states: Tuple[torch.FloatTensor], input_tensor: torch.FloatTensor) -> torch.FloatTensor:
hidden_states = self.dense(hidden_states)
hidden_states = self.dropout(hidden_states)
hidden_states = self.LayerNorm(hidden_states + input_tensor)
Expand Down Expand Up @@ -706,11 +712,11 @@ def __init__(

def forward(
self,
hidden_states,
attention_mask=None,
head_mask=None,
output_attentions=False,
):
hidden_states: Tuple[torch.FloatTensor],
attention_mask: Optional[torch.FloatTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
output_attentions: Optional[bool] = False,
) -> Tuple[torch.FloatTensor, Optional[torch.FloatTensor]]:
self_attention_outputs = self.attention(
hidden_states,
attention_mask,
Expand Down Expand Up @@ -767,13 +773,13 @@ def __init__(

def forward(
self,
hidden_states,
attention_mask=None,
head_mask=None,
output_attentions=False,
output_hidden_states=False,
return_dict=True,
):
hidden_states: Tuple[torch.FloatTensor],
attention_mask: Optional[torch.FloatTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
output_attentions: Optional[bool] = False,
output_hidden_states: Optional[bool] = False,
return_dict: Optional[bool] = True,
) -> Union[Tuple, BaseModelOutput]:
all_hidden_states = () if output_hidden_states else None
all_self_attentions = () if output_attentions else None

Expand Down Expand Up @@ -822,7 +828,7 @@ def __init__(self, config):
self.dense = nn.Linear(config.hidden_size, config.hidden_size)
self.activation = nn.Tanh()

def forward(self, hidden_states):
def forward(self, hidden_states: Tuple[torch.FloatTensor]) -> torch.FloatTensor:
# We "pool" the model by simply taking the hidden state corresponding
# to the first token.
first_token_tensor = hidden_states[:, 0]
Expand All @@ -841,7 +847,7 @@ def __init__(self, config):
self.transform_act_fn = config.hidden_act
self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)

def forward(self, hidden_states):
def forward(self, hidden_states: Tuple[torch.FloatTensor]) -> torch.FloatTensor:
hidden_states = self.dense(hidden_states)
hidden_states = self.transform_act_fn(hidden_states)
hidden_states = self.LayerNorm(hidden_states)
Expand All @@ -862,7 +868,7 @@ def __init__(self, config):
# Need a link between the two variables so that the bias is correctly resized with `resize_token_embeddings`
self.decoder.bias = self.bias

def forward(self, hidden_states):
def forward(self, hidden_states: Tuple[torch.FloatTensor]) -> torch.FloatTensor:
hidden_states = self.transform(hidden_states)
hidden_states = self.decoder(hidden_states)
return hidden_states
Expand All @@ -873,7 +879,10 @@ def __init__(self, config):
super().__init__()
self.predictions = CanineLMPredictionHead(config)

def forward(self, sequence_output):
def forward(
self,
sequence_output: Tuple[torch.Tensor],
) -> Tuple[torch.Tensor]:
prediction_scores = self.predictions(sequence_output)
return prediction_scores

Expand Down Expand Up @@ -1093,16 +1102,16 @@ def _repeat_molecules(self, molecules: torch.Tensor, char_seq_length: torch.Tens
)
def forward(
self,
input_ids=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
input_ids: Optional[torch.LongTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids: Optional[torch.LongTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, CanineModelOutputWithPooling]:
output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
output_hidden_states = (
output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
Expand Down Expand Up @@ -1275,17 +1284,17 @@ def __init__(self, config):
)
def forward(
self,
input_ids=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
labels=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
input_ids: Optional[torch.LongTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids: Optional[torch.LongTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
labels: Optional[torch.LongTensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, SequenceClassifierOutput]:
r"""
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
Expand Down Expand Up @@ -1372,17 +1381,17 @@ def __init__(self, config):
)
def forward(
self,
input_ids=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
labels=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
input_ids: Optional[torch.LongTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids: Optional[torch.LongTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
labels: Optional[torch.LongTensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, MultipleChoiceModelOutput]:
r"""
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
Expand Down Expand Up @@ -1465,17 +1474,17 @@ def __init__(self, config):
)
def forward(
self,
input_ids=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
labels=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
input_ids: Optional[torch.LongTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids: Optional[torch.LongTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
labels: Optional[torch.LongTensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, TokenClassifierOutput]:
r"""
labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`.
Expand Down Expand Up @@ -1543,18 +1552,18 @@ def __init__(self, config):
)
def forward(
self,
input_ids=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
start_positions=None,
end_positions=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
input_ids: Optional[torch.LongTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids: Optional[torch.LongTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
start_positions: Optional[torch.LongTensor] = None,
end_positions: Optional[torch.LongTensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, QuestionAnsweringModelOutput]:
r"""
start_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
Labels for position (index) of the start of the labelled span for computing the token classification loss.
Expand Down
Loading