Spaces:
Build error
Build error
wissamantoun
commited on
Commit
•
ce026c5
1
Parent(s):
5be9630
updated gpt2 to transformer 4.10
Browse filesI hope it works ( i didnt test the parralelize method)
- backend/modeling_gpt2.py +502 -99
backend/modeling_gpt2.py
CHANGED
@@ -23,42 +23,35 @@ and https://github.com/ghosthamlet/gpt2-ml-torch/blob/master/gpt2_ml_torch/model
|
|
23 |
|
24 |
import logging
|
25 |
import os
|
26 |
-
|
27 |
from dataclasses import dataclass
|
28 |
from typing import List, Optional, Tuple
|
29 |
|
30 |
import torch
|
31 |
import torch.nn as nn
|
32 |
from torch.nn import CrossEntropyLoss, MSELoss
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
from transformers.activations import ACT2FN
|
37 |
-
from transformers import
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
prune_conv1d_layer,
|
44 |
-
find_pruneable_heads_and_indices
|
45 |
)
|
46 |
-
|
47 |
-
from transformers import CONFIG_NAME, WEIGHTS_NAME, GPT2Config, GPT2Model
|
48 |
-
|
49 |
from transformers.modeling_outputs import (
|
50 |
BaseModelOutputWithPastAndCrossAttentions,
|
51 |
CausalLMOutputWithCrossAttentions,
|
52 |
-
SequenceClassifierOutputWithPast
|
|
|
53 |
)
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
replace_return_docstrings
|
61 |
)
|
|
|
62 |
|
63 |
# THe Difference from Transformers is code under _USE_GROVER
|
64 |
_USE_GROVER = True
|
@@ -83,30 +76,30 @@ console.setLevel(logging.INFO)
|
|
83 |
logger.addHandler(console)
|
84 |
|
85 |
_GPT2_ML_TF_TO_TORCH = {
|
86 |
-
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
'beta': 'bias',
|
105 |
-
'bias': 'bias',
|
106 |
}
|
107 |
|
108 |
|
109 |
-
def convert_gpt2_checkpoint_to_pytorch(
|
|
|
|
|
110 |
# Construct model
|
111 |
if gpt2_config_file == "":
|
112 |
config = GPT2Config()
|
@@ -130,10 +123,10 @@ def convert_gpt2_checkpoint_to_pytorch(gpt2_checkpoint_path, gpt2_config_file, p
|
|
130 |
# XXX: MUST do like: convert_gpt2_checkpoint_to_pytorch('./model.ckpt-100000', './mega.json', './')
|
131 |
# https://github.com/tensorflow/models/issues/2675#issuecomment-516595597
|
132 |
def load_tf_weights_in_gpt2(model, config, gpt2_checkpoint_path):
|
133 |
-
"""
|
134 |
-
"""
|
135 |
try:
|
136 |
import re
|
|
|
137 |
import tensorflow as tf
|
138 |
except ImportError:
|
139 |
logger.error(
|
@@ -154,6 +147,7 @@ def load_tf_weights_in_gpt2(model, config, gpt2_checkpoint_path):
|
|
154 |
arrays.append(array.squeeze())
|
155 |
|
156 |
import copy
|
|
|
157 |
orig_model = copy.deepcopy(model)
|
158 |
|
159 |
for name, array in zip(names, arrays):
|
@@ -161,7 +155,7 @@ def load_tf_weights_in_gpt2(model, config, gpt2_checkpoint_path):
|
|
161 |
name = name.split("/")
|
162 |
pointer = model
|
163 |
|
164 |
-
attn_layer =
|
165 |
for m_name in name:
|
166 |
if re.fullmatch(r"[A-Za-z]+\d+", m_name):
|
167 |
scope_names = re.split(r"(\d+)", m_name)
|
@@ -169,23 +163,23 @@ def load_tf_weights_in_gpt2(model, config, gpt2_checkpoint_path):
|
|
169 |
scope_names = [m_name]
|
170 |
sname = scope_names[0]
|
171 |
|
172 |
-
if sname ==
|
173 |
continue
|
174 |
elif sname not in _GPT2_ML_TF_TO_TORCH:
|
175 |
-
print(
|
176 |
-
logger.info(
|
177 |
pointer = None
|
178 |
break
|
179 |
else:
|
180 |
tname = _GPT2_ML_TF_TO_TORCH[sname]
|
181 |
-
if
|
182 |
-
parent, child = tname.split(
|
183 |
pointer = getattr(pointer, parent)
|
184 |
pointer = getattr(pointer, child)
|
185 |
else:
|
186 |
pointer = getattr(pointer, tname)
|
187 |
|
188 |
-
if tname ==
|
189 |
attn_layer = sname
|
190 |
|
191 |
if len(scope_names) >= 2:
|
@@ -194,39 +188,47 @@ def load_tf_weights_in_gpt2(model, config, gpt2_checkpoint_path):
|
|
194 |
|
195 |
if pointer is None:
|
196 |
continue
|
197 |
-
if attn_layer ==
|
198 |
try:
|
199 |
assert pointer.shape == array.shape
|
200 |
except AssertionError as e:
|
201 |
e.args += (pointer.shape, array.shape)
|
202 |
raise
|
203 |
-
logger.info(
|
204 |
-
|
|
|
|
|
|
|
|
|
205 |
pointer.data = torch.from_numpy(array)
|
206 |
else:
|
207 |
shape = pointer.shape
|
208 |
d = torch.from_numpy(array)
|
209 |
is_bias = len(shape) == 1
|
210 |
-
end = int(shape[0 if is_bias else 1]/3)
|
211 |
m = dict(
|
212 |
-
|
213 |
-
|
214 |
-
|
215 |
-
|
216 |
start = m[attn_layer]
|
217 |
end = start + end
|
218 |
if is_bias:
|
219 |
pointer.data[start:end] = d
|
220 |
else:
|
221 |
pointer.data[:, start:end] = d
|
222 |
-
logger.info(
|
|
|
|
|
|
|
|
|
223 |
|
224 |
for name, params in orig_model.named_parameters():
|
225 |
for n, p in model.named_parameters():
|
226 |
if name == n:
|
227 |
if params.equal(p):
|
228 |
-
print(
|
229 |
-
print(
|
230 |
return model
|
231 |
|
232 |
|
@@ -238,7 +240,10 @@ class Attention(nn.Module):
|
|
238 |
# [switch nx => n_state from Block to Attention to keep identical to TF implem]
|
239 |
assert n_state % config.n_head == 0
|
240 |
self.register_buffer(
|
241 |
-
"bias",
|
|
|
|
|
|
|
242 |
)
|
243 |
self.register_buffer("masked_bias", torch.tensor(-1e4))
|
244 |
self.n_head = config.n_head
|
@@ -261,7 +266,9 @@ class Attention(nn.Module):
|
|
261 |
heads, index = find_pruneable_heads_and_indices(
|
262 |
heads, self.n_head, self.split_size // self.n_head, self.pruned_heads
|
263 |
)
|
264 |
-
index_attn = torch.cat(
|
|
|
|
|
265 |
|
266 |
# Prune conv1d layers
|
267 |
self.c_attn = prune_conv1d_layer(self.c_attn, index_attn, dim=1)
|
@@ -272,7 +279,9 @@ class Attention(nn.Module):
|
|
272 |
self.n_head = self.n_head - len(heads)
|
273 |
self.pruned_heads = self.pruned_heads.union(heads)
|
274 |
|
275 |
-
def _attn(
|
|
|
|
|
276 |
w = torch.matmul(q, k)
|
277 |
if self.scale:
|
278 |
w = w / (float(v.size(-1)) ** 0.5)
|
@@ -328,7 +337,9 @@ class Attention(nn.Module):
|
|
328 |
self, "q_attn"
|
329 |
), "If class is used as cross attention, the weights `q_attn` have to be defined. Please make sure to instantiate class with `Attention(..., is_cross_attention=True)`."
|
330 |
query = self.q_attn(hidden_states)
|
331 |
-
key, value = self.c_attn(encoder_hidden_states).split(
|
|
|
|
|
332 |
attention_mask = encoder_attention_mask
|
333 |
else:
|
334 |
query, key, value = self.c_attn(hidden_states).split(self.split_size, dim=2)
|
@@ -337,16 +348,23 @@ class Attention(nn.Module):
|
|
337 |
key = self.split_heads(key, k=True)
|
338 |
value = self.split_heads(value)
|
339 |
if layer_past is not None:
|
340 |
-
past_key, past_value =
|
|
|
|
|
|
|
341 |
key = torch.cat((past_key, key), dim=-1)
|
342 |
value = torch.cat((past_value, value), dim=-2)
|
343 |
|
344 |
if use_cache is True:
|
345 |
-
present = torch.stack(
|
|
|
|
|
346 |
else:
|
347 |
present = (None,)
|
348 |
|
349 |
-
attn_outputs = self._attn(
|
|
|
|
|
350 |
a = attn_outputs[0]
|
351 |
|
352 |
a = self.merge_heads(a)
|
@@ -381,8 +399,12 @@ class Block(nn.Module):
|
|
381 |
self.attn = Attention(hidden_size, n_ctx, config, scale)
|
382 |
self.ln_2 = nn.LayerNorm(hidden_size, eps=config.layer_norm_epsilon)
|
383 |
if config.add_cross_attention:
|
384 |
-
self.crossattention = Attention(
|
385 |
-
|
|
|
|
|
|
|
|
|
386 |
self.mlp = MLP(inner_dim, config)
|
387 |
|
388 |
def forward(
|
@@ -425,7 +447,9 @@ class Block(nn.Module):
|
|
425 |
attn_output = cross_attn_outputs[0]
|
426 |
# residual connection
|
427 |
hidden_states = hidden_states + attn_output
|
428 |
-
outputs =
|
|
|
|
|
429 |
|
430 |
feed_forward_hidden_states = self.mlp(self.ln_1(hidden_states))
|
431 |
# residual connection
|
@@ -446,6 +470,7 @@ class GPT2PreTrainedModel(PreTrainedModel):
|
|
446 |
config_class = GPT2Config
|
447 |
load_tf_weights = load_tf_weights_in_gpt2
|
448 |
base_model_prefix = "transformer"
|
|
|
449 |
|
450 |
def __init__(self, *inputs, **kwargs):
|
451 |
super().__init__(*inputs, **kwargs)
|
@@ -588,6 +613,51 @@ GPT2_INPUTS_DOCSTRING = r"""
|
|
588 |
Whether or not to return a :class:`~transformers.file_utils.ModelOutput` instead of a plain tuple.
|
589 |
"""
|
590 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
591 |
|
592 |
@add_start_docstrings(
|
593 |
"The bare GPT2 Model transformer outputting raw hidden-states without any specific head on top.",
|
@@ -603,12 +673,57 @@ class GPT2Model(GPT2PreTrainedModel):
|
|
603 |
self.emb_norm = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
|
604 |
|
605 |
self.drop = nn.Dropout(config.embd_pdrop)
|
606 |
-
self.h = nn.ModuleList(
|
|
|
|
|
607 |
if not _USE_GROVER:
|
608 |
self.ln_f = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
|
609 |
|
610 |
self.init_weights()
|
611 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
612 |
def get_input_embeddings(self):
|
613 |
return self.wte
|
614 |
|
@@ -645,15 +760,25 @@ class GPT2Model(GPT2PreTrainedModel):
|
|
645 |
output_hidden_states=None,
|
646 |
return_dict=None,
|
647 |
):
|
648 |
-
output_attentions =
|
|
|
|
|
|
|
|
|
649 |
output_hidden_states = (
|
650 |
-
output_hidden_states
|
|
|
|
|
651 |
)
|
652 |
use_cache = use_cache if use_cache is not None else self.config.use_cache
|
653 |
-
return_dict =
|
|
|
|
|
654 |
|
655 |
if input_ids is not None and inputs_embeds is not None:
|
656 |
-
raise ValueError(
|
|
|
|
|
657 |
elif input_ids is not None:
|
658 |
input_shape = input_ids.size()
|
659 |
input_ids = input_ids.view(-1, input_shape[-1])
|
@@ -676,12 +801,18 @@ class GPT2Model(GPT2PreTrainedModel):
|
|
676 |
past_length = past_key_values[0][0].size(-2)
|
677 |
if position_ids is None:
|
678 |
device = input_ids.device if input_ids is not None else inputs_embeds.device
|
679 |
-
position_ids = torch.arange(
|
|
|
|
|
|
|
|
|
|
|
680 |
position_ids = position_ids.unsqueeze(0).view(-1, input_shape[-1])
|
681 |
|
682 |
# Attention mask.
|
683 |
if attention_mask is not None:
|
684 |
-
|
|
|
685 |
attention_mask = attention_mask.view(batch_size, -1)
|
686 |
# We create a 3D attention mask from a 2D tensor mask.
|
687 |
# Sizes are [batch_size, 1, 1, to_seq_length]
|
@@ -701,7 +832,11 @@ class GPT2Model(GPT2PreTrainedModel):
|
|
701 |
# If a 2D ou 3D attention mask is provided for the cross-attention
|
702 |
# we need to make broadcastable to [batch_size, num_heads, seq_length, seq_length]
|
703 |
if self.config.add_cross_attention and encoder_hidden_states is not None:
|
704 |
-
|
|
|
|
|
|
|
|
|
705 |
encoder_hidden_shape = (encoder_batch_size, encoder_sequence_length)
|
706 |
if encoder_attention_mask is None:
|
707 |
encoder_attention_mask = torch.ones(encoder_hidden_shape, device=device)
|
@@ -731,18 +866,40 @@ class GPT2Model(GPT2PreTrainedModel):
|
|
731 |
|
732 |
presents = () if use_cache else None
|
733 |
all_self_attentions = () if output_attentions else None
|
734 |
-
all_cross_attentions = (
|
|
|
|
|
735 |
all_hidden_states = () if output_hidden_states else None
|
736 |
for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
737 |
if output_hidden_states:
|
738 |
-
all_hidden_states = all_hidden_states + (
|
|
|
|
|
739 |
|
740 |
if getattr(self.config, "gradient_checkpointing", False):
|
741 |
|
742 |
def create_custom_forward(module):
|
743 |
def custom_forward(*inputs):
|
744 |
# checkpointing only works with tuple returns, not with lists
|
745 |
-
return tuple(
|
|
|
|
|
|
|
746 |
|
747 |
return custom_forward
|
748 |
|
@@ -772,9 +929,19 @@ class GPT2Model(GPT2PreTrainedModel):
|
|
772 |
presents = presents + (present,)
|
773 |
|
774 |
if output_attentions:
|
775 |
-
all_self_attentions = all_self_attentions + (
|
|
|
|
|
776 |
if self.config.add_cross_attention:
|
777 |
-
all_cross_attentions = all_cross_attentions + (
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
778 |
|
779 |
if not _USE_GROVER:
|
780 |
hidden_states = self.ln_f(hidden_states)
|
@@ -785,7 +952,17 @@ class GPT2Model(GPT2PreTrainedModel):
|
|
785 |
all_hidden_states = all_hidden_states + (hidden_states,)
|
786 |
|
787 |
if not return_dict:
|
788 |
-
return tuple(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
789 |
|
790 |
return BaseModelOutputWithPastAndCrossAttentions(
|
791 |
last_hidden_state=hidden_states,
|
@@ -813,6 +990,30 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
|
|
813 |
|
814 |
self.init_weights()
|
815 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
816 |
def get_output_embeddings(self):
|
817 |
return self.lm_head
|
818 |
|
@@ -848,7 +1049,7 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
|
|
848 |
@add_code_sample_docstrings(
|
849 |
tokenizer_class=_TOKENIZER_FOR_DOC,
|
850 |
checkpoint="gpt2",
|
851 |
-
output_type=
|
852 |
config_class=_CONFIG_FOR_DOC,
|
853 |
)
|
854 |
def forward(
|
@@ -874,7 +1075,9 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
|
|
874 |
``labels = input_ids`` Indices are selected in ``[-100, 0, ..., config.vocab_size]`` All labels set to
|
875 |
``-100`` are ignored (masked), the loss is only computed for labels in ``[0, ..., config.vocab_size]``
|
876 |
"""
|
877 |
-
return_dict =
|
|
|
|
|
878 |
|
879 |
transformer_outputs = self.transformer(
|
880 |
input_ids,
|
@@ -893,6 +1096,11 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
|
|
893 |
)
|
894 |
hidden_states = transformer_outputs[0]
|
895 |
|
|
|
|
|
|
|
|
|
|
|
896 |
lm_logits = self.lm_head(hidden_states)
|
897 |
|
898 |
loss = None
|
@@ -902,13 +1110,15 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
|
|
902 |
shift_labels = labels[..., 1:].contiguous()
|
903 |
# Flatten the tokens
|
904 |
loss_fct = CrossEntropyLoss()
|
905 |
-
loss = loss_fct(
|
|
|
|
|
906 |
|
907 |
if not return_dict:
|
908 |
output = (lm_logits,) + transformer_outputs[1:]
|
909 |
return ((loss,) + output) if loss is not None else output
|
910 |
|
911 |
-
return
|
912 |
loss=loss,
|
913 |
logits=lm_logits,
|
914 |
past_key_values=transformer_outputs.past_key_values,
|
@@ -917,6 +1127,23 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
|
|
917 |
cross_attentions=transformer_outputs.cross_attentions,
|
918 |
)
|
919 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
920 |
|
921 |
@add_start_docstrings(
|
922 |
"""
|
@@ -937,6 +1164,34 @@ class GPT2DoubleHeadsModel(GPT2PreTrainedModel):
|
|
937 |
|
938 |
self.init_weights()
|
939 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
940 |
def get_output_embeddings(self):
|
941 |
return self.lm_head
|
942 |
|
@@ -970,7 +1225,9 @@ class GPT2DoubleHeadsModel(GPT2PreTrainedModel):
|
|
970 |
}
|
971 |
|
972 |
@add_start_docstrings_to_model_forward(GPT2_INPUTS_DOCSTRING)
|
973 |
-
@replace_return_docstrings(
|
|
|
|
|
974 |
def forward(
|
975 |
self,
|
976 |
input_ids=None,
|
@@ -1029,7 +1286,9 @@ class GPT2DoubleHeadsModel(GPT2PreTrainedModel):
|
|
1029 |
>>> mc_logits = outputs.mc_logits
|
1030 |
|
1031 |
"""
|
1032 |
-
return_dict =
|
|
|
|
|
1033 |
|
1034 |
transformer_outputs = self.transformer(
|
1035 |
input_ids,
|
@@ -1047,19 +1306,28 @@ class GPT2DoubleHeadsModel(GPT2PreTrainedModel):
|
|
1047 |
|
1048 |
hidden_states = transformer_outputs[0]
|
1049 |
|
|
|
|
|
|
|
|
|
|
|
1050 |
lm_logits = self.lm_head(hidden_states)
|
1051 |
mc_logits = self.multiple_choice_head(hidden_states, mc_token_ids).squeeze(-1)
|
1052 |
|
1053 |
mc_loss = None
|
1054 |
if mc_labels is not None:
|
1055 |
loss_fct = CrossEntropyLoss()
|
1056 |
-
mc_loss = loss_fct(
|
|
|
|
|
1057 |
lm_loss = None
|
1058 |
if labels is not None:
|
1059 |
shift_logits = lm_logits[..., :-1, :].contiguous()
|
1060 |
shift_labels = labels[..., 1:].contiguous()
|
1061 |
loss_fct = CrossEntropyLoss()
|
1062 |
-
lm_loss = loss_fct(
|
|
|
|
|
1063 |
|
1064 |
if not return_dict:
|
1065 |
output = (lm_logits, mc_logits) + transformer_outputs[1:]
|
@@ -1077,6 +1345,23 @@ class GPT2DoubleHeadsModel(GPT2PreTrainedModel):
|
|
1077 |
attentions=transformer_outputs.attentions,
|
1078 |
)
|
1079 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1080 |
|
1081 |
@add_start_docstrings(
|
1082 |
"""
|
@@ -1104,6 +1389,10 @@ class GPT2ForSequenceClassification(GPT2PreTrainedModel):
|
|
1104 |
|
1105 |
self.init_weights()
|
1106 |
|
|
|
|
|
|
|
|
|
1107 |
@add_start_docstrings_to_model_forward(GPT2_INPUTS_DOCSTRING)
|
1108 |
@add_code_sample_docstrings(
|
1109 |
tokenizer_class=_TOKENIZER_FOR_DOC,
|
@@ -1132,7 +1421,9 @@ class GPT2ForSequenceClassification(GPT2PreTrainedModel):
|
|
1132 |
config.num_labels - 1]`. If :obj:`config.num_labels == 1` a regression loss is computed (Mean-Square loss),
|
1133 |
If :obj:`config.num_labels > 1` a classification loss is computed (Cross-Entropy).
|
1134 |
"""
|
1135 |
-
return_dict =
|
|
|
|
|
1136 |
|
1137 |
transformer_outputs = self.transformer(
|
1138 |
input_ids,
|
@@ -1162,7 +1453,9 @@ class GPT2ForSequenceClassification(GPT2PreTrainedModel):
|
|
1162 |
sequence_lengths = -1
|
1163 |
else:
|
1164 |
if input_ids is not None:
|
1165 |
-
sequence_lengths =
|
|
|
|
|
1166 |
else:
|
1167 |
sequence_lengths = -1
|
1168 |
logger.warning(
|
@@ -1180,7 +1473,9 @@ class GPT2ForSequenceClassification(GPT2PreTrainedModel):
|
|
1180 |
loss = loss_fct(pooled_logits.view(-1), labels.to(self.dtype).view(-1))
|
1181 |
else:
|
1182 |
loss_fct = CrossEntropyLoss()
|
1183 |
-
loss = loss_fct(
|
|
|
|
|
1184 |
|
1185 |
if not return_dict:
|
1186 |
output = (pooled_logits,) + transformer_outputs[1:]
|
@@ -1194,3 +1489,111 @@ class GPT2ForSequenceClassification(GPT2PreTrainedModel):
|
|
1194 |
attentions=transformer_outputs.attentions,
|
1195 |
)
|
1196 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
|
24 |
import logging
|
25 |
import os
|
|
|
26 |
from dataclasses import dataclass
|
27 |
from typing import List, Optional, Tuple
|
28 |
|
29 |
import torch
|
30 |
import torch.nn as nn
|
31 |
from torch.nn import CrossEntropyLoss, MSELoss
|
32 |
+
from transformers import CONFIG_NAME, WEIGHTS_NAME, GPT2Config, GPT2Model
|
|
|
|
|
33 |
from transformers.activations import ACT2FN
|
34 |
+
from transformers.file_utils import (
|
35 |
+
ModelOutput,
|
36 |
+
add_code_sample_docstrings,
|
37 |
+
add_start_docstrings,
|
38 |
+
add_start_docstrings_to_model_forward,
|
39 |
+
replace_return_docstrings,
|
|
|
|
|
40 |
)
|
|
|
|
|
|
|
41 |
from transformers.modeling_outputs import (
|
42 |
BaseModelOutputWithPastAndCrossAttentions,
|
43 |
CausalLMOutputWithCrossAttentions,
|
44 |
+
SequenceClassifierOutputWithPast,
|
45 |
+
TokenClassifierOutput,
|
46 |
)
|
47 |
+
from transformers.modeling_utils import (
|
48 |
+
Conv1D,
|
49 |
+
PreTrainedModel,
|
50 |
+
SequenceSummary,
|
51 |
+
find_pruneable_heads_and_indices,
|
52 |
+
prune_conv1d_layer,
|
|
|
53 |
)
|
54 |
+
from transformers.utils.model_parallel_utils import assert_device_map, get_device_map
|
55 |
|
56 |
# THe Difference from Transformers is code under _USE_GROVER
|
57 |
_USE_GROVER = True
|
|
|
76 |
logger.addHandler(console)
|
77 |
|
78 |
_GPT2_ML_TF_TO_TORCH = {
|
79 |
+
"LayerNorm_embed_norm": "emb_norm",
|
80 |
+
"pos_embed": "wpe.weight",
|
81 |
+
"word_embed": "wte.weight",
|
82 |
+
"layer": "h",
|
83 |
+
# Most importently This two layer norm must be put on the same position as gpt2-ml
|
84 |
+
# or generated data is bad, just repeat the last token
|
85 |
+
"LayerNorm_mlp_ln0": "ln_1",
|
86 |
+
"LayerNorm_mlp_ln1": "ln_2",
|
87 |
+
"intermediate": "mlp.c_fc",
|
88 |
+
"output": "mlp.c_proj",
|
89 |
+
"query_layer": "attn.c_attn",
|
90 |
+
"key_layer": "attn.c_attn",
|
91 |
+
"value_layer": "attn.c_attn",
|
92 |
+
"context_projection_layer": "attn.c_proj",
|
93 |
+
"gamma": "weight",
|
94 |
+
"kernel": "weight",
|
95 |
+
"beta": "bias",
|
96 |
+
"bias": "bias",
|
|
|
|
|
97 |
}
|
98 |
|
99 |
|
100 |
+
def convert_gpt2_checkpoint_to_pytorch(
|
101 |
+
gpt2_checkpoint_path, gpt2_config_file, pytorch_dump_folder_path
|
102 |
+
):
|
103 |
# Construct model
|
104 |
if gpt2_config_file == "":
|
105 |
config = GPT2Config()
|
|
|
123 |
# XXX: MUST do like: convert_gpt2_checkpoint_to_pytorch('./model.ckpt-100000', './mega.json', './')
|
124 |
# https://github.com/tensorflow/models/issues/2675#issuecomment-516595597
|
125 |
def load_tf_weights_in_gpt2(model, config, gpt2_checkpoint_path):
|
126 |
+
"""Load tf checkpoints in a pytorch model"""
|
|
|
127 |
try:
|
128 |
import re
|
129 |
+
|
130 |
import tensorflow as tf
|
131 |
except ImportError:
|
132 |
logger.error(
|
|
|
147 |
arrays.append(array.squeeze())
|
148 |
|
149 |
import copy
|
150 |
+
|
151 |
orig_model = copy.deepcopy(model)
|
152 |
|
153 |
for name, array in zip(names, arrays):
|
|
|
155 |
name = name.split("/")
|
156 |
pointer = model
|
157 |
|
158 |
+
attn_layer = ""
|
159 |
for m_name in name:
|
160 |
if re.fullmatch(r"[A-Za-z]+\d+", m_name):
|
161 |
scope_names = re.split(r"(\d+)", m_name)
|
|
|
163 |
scope_names = [m_name]
|
164 |
sname = scope_names[0]
|
165 |
|
166 |
+
if sname == "" or sname == "embeddings":
|
167 |
continue
|
168 |
elif sname not in _GPT2_ML_TF_TO_TORCH:
|
169 |
+
print("=========================================================")
|
170 |
+
logger.info("Skip var name {}".format(scope_names))
|
171 |
pointer = None
|
172 |
break
|
173 |
else:
|
174 |
tname = _GPT2_ML_TF_TO_TORCH[sname]
|
175 |
+
if "." in tname:
|
176 |
+
parent, child = tname.split(".")
|
177 |
pointer = getattr(pointer, parent)
|
178 |
pointer = getattr(pointer, child)
|
179 |
else:
|
180 |
pointer = getattr(pointer, tname)
|
181 |
|
182 |
+
if tname == "attn.c_attn":
|
183 |
attn_layer = sname
|
184 |
|
185 |
if len(scope_names) >= 2:
|
|
|
188 |
|
189 |
if pointer is None:
|
190 |
continue
|
191 |
+
if attn_layer == "":
|
192 |
try:
|
193 |
assert pointer.shape == array.shape
|
194 |
except AssertionError as e:
|
195 |
e.args += (pointer.shape, array.shape)
|
196 |
raise
|
197 |
+
logger.info(
|
198 |
+
"Initialize PyTorch weight {}, {}, {}".format(
|
199 |
+
name, array.mean(), pointer.mean()
|
200 |
+
)
|
201 |
+
)
|
202 |
+
if attn_layer == "":
|
203 |
pointer.data = torch.from_numpy(array)
|
204 |
else:
|
205 |
shape = pointer.shape
|
206 |
d = torch.from_numpy(array)
|
207 |
is_bias = len(shape) == 1
|
208 |
+
end = int(shape[0 if is_bias else 1] / 3)
|
209 |
m = dict(
|
210 |
+
query_layer=0,
|
211 |
+
key_layer=end,
|
212 |
+
value_layer=end * 2,
|
213 |
+
)
|
214 |
start = m[attn_layer]
|
215 |
end = start + end
|
216 |
if is_bias:
|
217 |
pointer.data[start:end] = d
|
218 |
else:
|
219 |
pointer.data[:, start:end] = d
|
220 |
+
logger.info(
|
221 |
+
"Initialize PyTorch weight {}, {}, {}".format(
|
222 |
+
name, array.mean(), pointer.mean()
|
223 |
+
)
|
224 |
+
)
|
225 |
|
226 |
for name, params in orig_model.named_parameters():
|
227 |
for n, p in model.named_parameters():
|
228 |
if name == n:
|
229 |
if params.equal(p):
|
230 |
+
print("--------------------------")
|
231 |
+
print(" %s not changed!" % n)
|
232 |
return model
|
233 |
|
234 |
|
|
|
240 |
# [switch nx => n_state from Block to Attention to keep identical to TF implem]
|
241 |
assert n_state % config.n_head == 0
|
242 |
self.register_buffer(
|
243 |
+
"bias",
|
244 |
+
torch.tril(torch.ones((n_ctx, n_ctx), dtype=torch.uint8)).view(
|
245 |
+
1, 1, n_ctx, n_ctx
|
246 |
+
),
|
247 |
)
|
248 |
self.register_buffer("masked_bias", torch.tensor(-1e4))
|
249 |
self.n_head = config.n_head
|
|
|
266 |
heads, index = find_pruneable_heads_and_indices(
|
267 |
heads, self.n_head, self.split_size // self.n_head, self.pruned_heads
|
268 |
)
|
269 |
+
index_attn = torch.cat(
|
270 |
+
[index, index + self.split_size, index + (2 * self.split_size)]
|
271 |
+
)
|
272 |
|
273 |
# Prune conv1d layers
|
274 |
self.c_attn = prune_conv1d_layer(self.c_attn, index_attn, dim=1)
|
|
|
279 |
self.n_head = self.n_head - len(heads)
|
280 |
self.pruned_heads = self.pruned_heads.union(heads)
|
281 |
|
282 |
+
def _attn(
|
283 |
+
self, q, k, v, attention_mask=None, head_mask=None, output_attentions=False
|
284 |
+
):
|
285 |
w = torch.matmul(q, k)
|
286 |
if self.scale:
|
287 |
w = w / (float(v.size(-1)) ** 0.5)
|
|
|
337 |
self, "q_attn"
|
338 |
), "If class is used as cross attention, the weights `q_attn` have to be defined. Please make sure to instantiate class with `Attention(..., is_cross_attention=True)`."
|
339 |
query = self.q_attn(hidden_states)
|
340 |
+
key, value = self.c_attn(encoder_hidden_states).split(
|
341 |
+
self.split_size, dim=2
|
342 |
+
)
|
343 |
attention_mask = encoder_attention_mask
|
344 |
else:
|
345 |
query, key, value = self.c_attn(hidden_states).split(self.split_size, dim=2)
|
|
|
348 |
key = self.split_heads(key, k=True)
|
349 |
value = self.split_heads(value)
|
350 |
if layer_past is not None:
|
351 |
+
past_key, past_value = (
|
352 |
+
layer_past[0].transpose(-2, -1),
|
353 |
+
layer_past[1],
|
354 |
+
) # transpose back cf below
|
355 |
key = torch.cat((past_key, key), dim=-1)
|
356 |
value = torch.cat((past_value, value), dim=-2)
|
357 |
|
358 |
if use_cache is True:
|
359 |
+
present = torch.stack(
|
360 |
+
(key.transpose(-2, -1), value)
|
361 |
+
) # transpose to have same shapes for stacking
|
362 |
else:
|
363 |
present = (None,)
|
364 |
|
365 |
+
attn_outputs = self._attn(
|
366 |
+
query, key, value, attention_mask, head_mask, output_attentions
|
367 |
+
)
|
368 |
a = attn_outputs[0]
|
369 |
|
370 |
a = self.merge_heads(a)
|
|
|
399 |
self.attn = Attention(hidden_size, n_ctx, config, scale)
|
400 |
self.ln_2 = nn.LayerNorm(hidden_size, eps=config.layer_norm_epsilon)
|
401 |
if config.add_cross_attention:
|
402 |
+
self.crossattention = Attention(
|
403 |
+
hidden_size, n_ctx, config, scale, is_cross_attention=True
|
404 |
+
)
|
405 |
+
self.ln_cross_attn = nn.LayerNorm(
|
406 |
+
hidden_size, eps=config.layer_norm_epsilon
|
407 |
+
)
|
408 |
self.mlp = MLP(inner_dim, config)
|
409 |
|
410 |
def forward(
|
|
|
447 |
attn_output = cross_attn_outputs[0]
|
448 |
# residual connection
|
449 |
hidden_states = hidden_states + attn_output
|
450 |
+
outputs = (
|
451 |
+
outputs + cross_attn_outputs[2:]
|
452 |
+
) # add cross attentions if we output attention weights
|
453 |
|
454 |
feed_forward_hidden_states = self.mlp(self.ln_1(hidden_states))
|
455 |
# residual connection
|
|
|
470 |
config_class = GPT2Config
|
471 |
load_tf_weights = load_tf_weights_in_gpt2
|
472 |
base_model_prefix = "transformer"
|
473 |
+
is_parallelizable = True
|
474 |
|
475 |
def __init__(self, *inputs, **kwargs):
|
476 |
super().__init__(*inputs, **kwargs)
|
|
|
613 |
Whether or not to return a :class:`~transformers.file_utils.ModelOutput` instead of a plain tuple.
|
614 |
"""
|
615 |
|
616 |
+
PARALLELIZE_DOCSTRING = r"""
|
617 |
+
This is an experimental feature and is a subject to change at a moment's notice.
|
618 |
+
|
619 |
+
Uses a device map to distribute attention modules of the model across several devices. If no device map is given,
|
620 |
+
it will evenly distribute blocks across all devices.
|
621 |
+
|
622 |
+
Args:
|
623 |
+
device_map (:obj:`Dict[int, list]`, optional, defaults to None):
|
624 |
+
A dictionary that maps attention modules to devices. Note that the embedding module and LMHead are always
|
625 |
+
automatically mapped to the first device (for esoteric reasons). That means that the first device should
|
626 |
+
have fewer attention modules mapped to it than other devices. For reference, the gpt2 models have the
|
627 |
+
following number of attention modules:
|
628 |
+
|
629 |
+
- gpt2: 12
|
630 |
+
- gpt2-medium: 24
|
631 |
+
- gpt2-large: 36
|
632 |
+
- gpt2-xl: 48
|
633 |
+
|
634 |
+
Example::
|
635 |
+
|
636 |
+
# Here is an example of a device map on a machine with 4 GPUs using gpt2-xl, which has a total of 48 attention modules:
|
637 |
+
model = GPT2LMHeadModel.from_pretrained('gpt2-xl')
|
638 |
+
device_map = {0: [0, 1, 2, 3, 4, 5, 6, 7, 8],
|
639 |
+
|
640 |
+
1: [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21],
|
641 |
+
2: [22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34],
|
642 |
+
3: [35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47]}
|
643 |
+
model.parallelize(device_map)
|
644 |
+
"""
|
645 |
+
DEPARALLELIZE_DOCSTRING = r"""
|
646 |
+
Moves the model to cpu from a model parallel state.
|
647 |
+
|
648 |
+
Example::
|
649 |
+
|
650 |
+
# On a 4 GPU machine with gpt2-large:
|
651 |
+
model = GPT2LMHeadModel.from_pretrained('gpt2-large')
|
652 |
+
device_map = {0: [0, 1, 2, 3, 4, 5, 6, 7],
|
653 |
+
|
654 |
+
1: [8, 9, 10, 11, 12, 13, 14, 15],
|
655 |
+
2: [16, 17, 18, 19, 20, 21, 22, 23],
|
656 |
+
3: [24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35]}
|
657 |
+
model.parallelize(device_map) # Splits the model across several devices
|
658 |
+
model.deparallelize() # Put the model back on cpu and cleans memory by calling torch.cuda.empty_cache()
|
659 |
+
"""
|
660 |
+
|
661 |
|
662 |
@add_start_docstrings(
|
663 |
"The bare GPT2 Model transformer outputting raw hidden-states without any specific head on top.",
|
|
|
673 |
self.emb_norm = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
|
674 |
|
675 |
self.drop = nn.Dropout(config.embd_pdrop)
|
676 |
+
self.h = nn.ModuleList(
|
677 |
+
[Block(config.n_ctx, config, scale=True) for _ in range(config.n_layer)]
|
678 |
+
)
|
679 |
if not _USE_GROVER:
|
680 |
self.ln_f = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
|
681 |
|
682 |
self.init_weights()
|
683 |
|
684 |
+
# Model parallel
|
685 |
+
self.model_parallel = False
|
686 |
+
self.device_map = None
|
687 |
+
|
688 |
+
@add_start_docstrings(PARALLELIZE_DOCSTRING)
|
689 |
+
def parallelize(self, device_map=None):
|
690 |
+
# Check validity of device_map
|
691 |
+
self.device_map = (
|
692 |
+
get_device_map(len(self.h), range(torch.cuda.device_count()))
|
693 |
+
if device_map is None
|
694 |
+
else device_map
|
695 |
+
)
|
696 |
+
assert_device_map(self.device_map, len(self.h))
|
697 |
+
self.model_parallel = True
|
698 |
+
self.first_device = (
|
699 |
+
"cpu"
|
700 |
+
if "cpu" in self.device_map.keys()
|
701 |
+
else "cuda:" + str(min(self.device_map.keys()))
|
702 |
+
)
|
703 |
+
self.last_device = "cuda:" + str(max(self.device_map.keys()))
|
704 |
+
self.wte = self.wte.to(self.first_device)
|
705 |
+
self.wpe = self.wpe.to(self.first_device)
|
706 |
+
# Load onto devices
|
707 |
+
for k, v in self.device_map.items():
|
708 |
+
for block in v:
|
709 |
+
cuda_device = "cuda:" + str(k)
|
710 |
+
self.h[block] = self.h[block].to(cuda_device)
|
711 |
+
# ln_f to last
|
712 |
+
self.ln_f = self.ln_f.to(self.last_device)
|
713 |
+
|
714 |
+
@add_start_docstrings(DEPARALLELIZE_DOCSTRING)
|
715 |
+
def deparallelize(self):
|
716 |
+
self.model_parallel = False
|
717 |
+
self.device_map = None
|
718 |
+
self.first_device = "cpu"
|
719 |
+
self.last_device = "cpu"
|
720 |
+
self.wte = self.wte.to("cpu")
|
721 |
+
self.wpe = self.wpe.to("cpu")
|
722 |
+
for index in range(len(self.h)):
|
723 |
+
self.h[index] = self.h[index].to("cpu")
|
724 |
+
self.ln_f = self.ln_f.to("cpu")
|
725 |
+
torch.cuda.empty_cache()
|
726 |
+
|
727 |
def get_input_embeddings(self):
|
728 |
return self.wte
|
729 |
|
|
|
760 |
output_hidden_states=None,
|
761 |
return_dict=None,
|
762 |
):
|
763 |
+
output_attentions = (
|
764 |
+
output_attentions
|
765 |
+
if output_attentions is not None
|
766 |
+
else self.config.output_attentions
|
767 |
+
)
|
768 |
output_hidden_states = (
|
769 |
+
output_hidden_states
|
770 |
+
if output_hidden_states is not None
|
771 |
+
else self.config.output_hidden_states
|
772 |
)
|
773 |
use_cache = use_cache if use_cache is not None else self.config.use_cache
|
774 |
+
return_dict = (
|
775 |
+
return_dict if return_dict is not None else self.config.use_return_dict
|
776 |
+
)
|
777 |
|
778 |
if input_ids is not None and inputs_embeds is not None:
|
779 |
+
raise ValueError(
|
780 |
+
"You cannot specify both input_ids and inputs_embeds at the same time"
|
781 |
+
)
|
782 |
elif input_ids is not None:
|
783 |
input_shape = input_ids.size()
|
784 |
input_ids = input_ids.view(-1, input_shape[-1])
|
|
|
801 |
past_length = past_key_values[0][0].size(-2)
|
802 |
if position_ids is None:
|
803 |
device = input_ids.device if input_ids is not None else inputs_embeds.device
|
804 |
+
position_ids = torch.arange(
|
805 |
+
past_length,
|
806 |
+
input_shape[-1] + past_length,
|
807 |
+
dtype=torch.long,
|
808 |
+
device=device,
|
809 |
+
)
|
810 |
position_ids = position_ids.unsqueeze(0).view(-1, input_shape[-1])
|
811 |
|
812 |
# Attention mask.
|
813 |
if attention_mask is not None:
|
814 |
+
if batch_size <= 0:
|
815 |
+
raise ValueError("batch_size has to be defined and > 0")
|
816 |
attention_mask = attention_mask.view(batch_size, -1)
|
817 |
# We create a 3D attention mask from a 2D tensor mask.
|
818 |
# Sizes are [batch_size, 1, 1, to_seq_length]
|
|
|
832 |
# If a 2D ou 3D attention mask is provided for the cross-attention
|
833 |
# we need to make broadcastable to [batch_size, num_heads, seq_length, seq_length]
|
834 |
if self.config.add_cross_attention and encoder_hidden_states is not None:
|
835 |
+
(
|
836 |
+
encoder_batch_size,
|
837 |
+
encoder_sequence_length,
|
838 |
+
_,
|
839 |
+
) = encoder_hidden_states.size()
|
840 |
encoder_hidden_shape = (encoder_batch_size, encoder_sequence_length)
|
841 |
if encoder_attention_mask is None:
|
842 |
encoder_attention_mask = torch.ones(encoder_hidden_shape, device=device)
|
|
|
866 |
|
867 |
presents = () if use_cache else None
|
868 |
all_self_attentions = () if output_attentions else None
|
869 |
+
all_cross_attentions = (
|
870 |
+
() if output_attentions and self.config.add_cross_attention else None
|
871 |
+
)
|
872 |
all_hidden_states = () if output_hidden_states else None
|
873 |
for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)):
|
874 |
+
|
875 |
+
# Model parallel
|
876 |
+
if self.model_parallel:
|
877 |
+
torch.cuda.set_device(hidden_states.device)
|
878 |
+
# Ensure layer_past is on same device as hidden_states (might not be correct)
|
879 |
+
if layer_past is not None:
|
880 |
+
layer_past = tuple(
|
881 |
+
past_state.to(hidden_states.device) for past_state in layer_past
|
882 |
+
)
|
883 |
+
# Ensure that attention_mask is always on the same device as hidden_states
|
884 |
+
if attention_mask is not None:
|
885 |
+
attention_mask = attention_mask.to(hidden_states.device)
|
886 |
+
if isinstance(head_mask, torch.Tensor):
|
887 |
+
head_mask = head_mask.to(hidden_states.device)
|
888 |
+
|
889 |
if output_hidden_states:
|
890 |
+
all_hidden_states = all_hidden_states + (
|
891 |
+
hidden_states.view(*output_shape),
|
892 |
+
)
|
893 |
|
894 |
if getattr(self.config, "gradient_checkpointing", False):
|
895 |
|
896 |
def create_custom_forward(module):
|
897 |
def custom_forward(*inputs):
|
898 |
# checkpointing only works with tuple returns, not with lists
|
899 |
+
return tuple(
|
900 |
+
output
|
901 |
+
for output in module(*inputs, use_cache, output_attentions)
|
902 |
+
)
|
903 |
|
904 |
return custom_forward
|
905 |
|
|
|
929 |
presents = presents + (present,)
|
930 |
|
931 |
if output_attentions:
|
932 |
+
all_self_attentions = all_self_attentions + (
|
933 |
+
outputs[2 if use_cache else 1],
|
934 |
+
)
|
935 |
if self.config.add_cross_attention:
|
936 |
+
all_cross_attentions = all_cross_attentions + (
|
937 |
+
outputs[3 if use_cache else 2],
|
938 |
+
)
|
939 |
+
|
940 |
+
# Model Parallel: If it's the last layer for that device, put things on the next device
|
941 |
+
if self.model_parallel:
|
942 |
+
for k, v in self.device_map.items():
|
943 |
+
if i == v[-1] and "cuda:" + str(k) != self.last_device:
|
944 |
+
hidden_states = hidden_states.to("cuda:" + str(k + 1))
|
945 |
|
946 |
if not _USE_GROVER:
|
947 |
hidden_states = self.ln_f(hidden_states)
|
|
|
952 |
all_hidden_states = all_hidden_states + (hidden_states,)
|
953 |
|
954 |
if not return_dict:
|
955 |
+
return tuple(
|
956 |
+
v
|
957 |
+
for v in [
|
958 |
+
hidden_states,
|
959 |
+
presents,
|
960 |
+
all_hidden_states,
|
961 |
+
all_self_attentions,
|
962 |
+
all_cross_attentions,
|
963 |
+
]
|
964 |
+
if v is not None
|
965 |
+
)
|
966 |
|
967 |
return BaseModelOutputWithPastAndCrossAttentions(
|
968 |
last_hidden_state=hidden_states,
|
|
|
990 |
|
991 |
self.init_weights()
|
992 |
|
993 |
+
# Model parallel
|
994 |
+
self.model_parallel = False
|
995 |
+
self.device_map = None
|
996 |
+
|
997 |
+
@add_start_docstrings(PARALLELIZE_DOCSTRING)
|
998 |
+
def parallelize(self, device_map=None):
|
999 |
+
self.device_map = (
|
1000 |
+
get_device_map(len(self.transformer.h), range(torch.cuda.device_count()))
|
1001 |
+
if device_map is None
|
1002 |
+
else device_map
|
1003 |
+
)
|
1004 |
+
assert_device_map(self.device_map, len(self.transformer.h))
|
1005 |
+
self.transformer.parallelize(self.device_map)
|
1006 |
+
self.lm_head = self.lm_head.to(self.transformer.first_device)
|
1007 |
+
self.model_parallel = True
|
1008 |
+
|
1009 |
+
@add_start_docstrings(DEPARALLELIZE_DOCSTRING)
|
1010 |
+
def deparallelize(self):
|
1011 |
+
self.transformer.deparallelize()
|
1012 |
+
self.transformer = self.transformer.to("cpu")
|
1013 |
+
self.lm_head = self.lm_head.to("cpu")
|
1014 |
+
self.model_parallel = False
|
1015 |
+
torch.cuda.empty_cache()
|
1016 |
+
|
1017 |
def get_output_embeddings(self):
|
1018 |
return self.lm_head
|
1019 |
|
|
|
1049 |
@add_code_sample_docstrings(
|
1050 |
tokenizer_class=_TOKENIZER_FOR_DOC,
|
1051 |
checkpoint="gpt2",
|
1052 |
+
output_type=CausalLMOutputWithCrossAttentions,
|
1053 |
config_class=_CONFIG_FOR_DOC,
|
1054 |
)
|
1055 |
def forward(
|
|
|
1075 |
``labels = input_ids`` Indices are selected in ``[-100, 0, ..., config.vocab_size]`` All labels set to
|
1076 |
``-100`` are ignored (masked), the loss is only computed for labels in ``[0, ..., config.vocab_size]``
|
1077 |
"""
|
1078 |
+
return_dict = (
|
1079 |
+
return_dict if return_dict is not None else self.config.use_return_dict
|
1080 |
+
)
|
1081 |
|
1082 |
transformer_outputs = self.transformer(
|
1083 |
input_ids,
|
|
|
1096 |
)
|
1097 |
hidden_states = transformer_outputs[0]
|
1098 |
|
1099 |
+
# Set device for model parallelism
|
1100 |
+
if self.model_parallel:
|
1101 |
+
torch.cuda.set_device(self.transformer.first_device)
|
1102 |
+
hidden_states = hidden_states.to(self.lm_head.weight.device)
|
1103 |
+
|
1104 |
lm_logits = self.lm_head(hidden_states)
|
1105 |
|
1106 |
loss = None
|
|
|
1110 |
shift_labels = labels[..., 1:].contiguous()
|
1111 |
# Flatten the tokens
|
1112 |
loss_fct = CrossEntropyLoss()
|
1113 |
+
loss = loss_fct(
|
1114 |
+
shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1)
|
1115 |
+
)
|
1116 |
|
1117 |
if not return_dict:
|
1118 |
output = (lm_logits,) + transformer_outputs[1:]
|
1119 |
return ((loss,) + output) if loss is not None else output
|
1120 |
|
1121 |
+
return CausalLMOutputWithCrossAttentions(
|
1122 |
loss=loss,
|
1123 |
logits=lm_logits,
|
1124 |
past_key_values=transformer_outputs.past_key_values,
|
|
|
1127 |
cross_attentions=transformer_outputs.cross_attentions,
|
1128 |
)
|
1129 |
|
1130 |
+
@staticmethod
|
1131 |
+
def _reorder_cache(
|
1132 |
+
past: Tuple[Tuple[torch.Tensor]], beam_idx: torch.Tensor
|
1133 |
+
) -> Tuple[Tuple[torch.Tensor]]:
|
1134 |
+
"""
|
1135 |
+
This function is used to re-order the :obj:`past_key_values` cache if
|
1136 |
+
:meth:`~transformers.PreTrainedModel.beam_search` or :meth:`~transformers.PreTrainedModel.beam_sample` is
|
1137 |
+
called. This is required to match :obj:`past_key_values` with the correct beam_idx at every generation step.
|
1138 |
+
"""
|
1139 |
+
return tuple(
|
1140 |
+
tuple(
|
1141 |
+
past_state.index_select(0, beam_idx.to(past_state.device))
|
1142 |
+
for past_state in layer_past
|
1143 |
+
)
|
1144 |
+
for layer_past in past
|
1145 |
+
)
|
1146 |
+
|
1147 |
|
1148 |
@add_start_docstrings(
|
1149 |
"""
|
|
|
1164 |
|
1165 |
self.init_weights()
|
1166 |
|
1167 |
+
# Model parallel
|
1168 |
+
self.model_parallel = False
|
1169 |
+
self.device_map = None
|
1170 |
+
|
1171 |
+
@add_start_docstrings(PARALLELIZE_DOCSTRING)
|
1172 |
+
def parallelize(self, device_map=None):
|
1173 |
+
self.device_map = (
|
1174 |
+
get_device_map(len(self.transformer.h), range(torch.cuda.device_count()))
|
1175 |
+
if device_map is None
|
1176 |
+
else device_map
|
1177 |
+
)
|
1178 |
+
assert_device_map(self.device_map, len(self.transformer.h))
|
1179 |
+
self.transformer.parallelize(self.device_map)
|
1180 |
+
self.lm_head = self.lm_head.to(self.transformer.first_device)
|
1181 |
+
self.multiple_choice_head = self.multiple_choice_head.to(
|
1182 |
+
self.transformer.first_device
|
1183 |
+
)
|
1184 |
+
self.model_parallel = True
|
1185 |
+
|
1186 |
+
@add_start_docstrings(DEPARALLELIZE_DOCSTRING)
|
1187 |
+
def deparallelize(self):
|
1188 |
+
self.transformer.deparallelize()
|
1189 |
+
self.transformer = self.transformer.to("cpu")
|
1190 |
+
self.lm_head = self.lm_head.to("cpu")
|
1191 |
+
self.multiple_choice_head = self.multiple_choice_head.to("cpu")
|
1192 |
+
self.model_parallel = False
|
1193 |
+
torch.cuda.empty_cache()
|
1194 |
+
|
1195 |
def get_output_embeddings(self):
|
1196 |
return self.lm_head
|
1197 |
|
|
|
1225 |
}
|
1226 |
|
1227 |
@add_start_docstrings_to_model_forward(GPT2_INPUTS_DOCSTRING)
|
1228 |
+
@replace_return_docstrings(
|
1229 |
+
output_type=GPT2DoubleHeadsModelOutput, config_class=_CONFIG_FOR_DOC
|
1230 |
+
)
|
1231 |
def forward(
|
1232 |
self,
|
1233 |
input_ids=None,
|
|
|
1286 |
>>> mc_logits = outputs.mc_logits
|
1287 |
|
1288 |
"""
|
1289 |
+
return_dict = (
|
1290 |
+
return_dict if return_dict is not None else self.config.use_return_dict
|
1291 |
+
)
|
1292 |
|
1293 |
transformer_outputs = self.transformer(
|
1294 |
input_ids,
|
|
|
1306 |
|
1307 |
hidden_states = transformer_outputs[0]
|
1308 |
|
1309 |
+
# Set device for model parallelism
|
1310 |
+
if self.model_parallel:
|
1311 |
+
torch.cuda.set_device(self.transformer.first_device)
|
1312 |
+
hidden_states = hidden_states.to(self.lm_head.weight.device)
|
1313 |
+
|
1314 |
lm_logits = self.lm_head(hidden_states)
|
1315 |
mc_logits = self.multiple_choice_head(hidden_states, mc_token_ids).squeeze(-1)
|
1316 |
|
1317 |
mc_loss = None
|
1318 |
if mc_labels is not None:
|
1319 |
loss_fct = CrossEntropyLoss()
|
1320 |
+
mc_loss = loss_fct(
|
1321 |
+
mc_logits.view(-1, mc_logits.size(-1)), mc_labels.view(-1)
|
1322 |
+
)
|
1323 |
lm_loss = None
|
1324 |
if labels is not None:
|
1325 |
shift_logits = lm_logits[..., :-1, :].contiguous()
|
1326 |
shift_labels = labels[..., 1:].contiguous()
|
1327 |
loss_fct = CrossEntropyLoss()
|
1328 |
+
lm_loss = loss_fct(
|
1329 |
+
shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1)
|
1330 |
+
)
|
1331 |
|
1332 |
if not return_dict:
|
1333 |
output = (lm_logits, mc_logits) + transformer_outputs[1:]
|
|
|
1345 |
attentions=transformer_outputs.attentions,
|
1346 |
)
|
1347 |
|
1348 |
+
@staticmethod
|
1349 |
+
def _reorder_cache(
|
1350 |
+
past: Tuple[Tuple[torch.Tensor]], beam_idx: torch.Tensor
|
1351 |
+
) -> Tuple[Tuple[torch.Tensor]]:
|
1352 |
+
"""
|
1353 |
+
This function is used to re-order the :obj:`past_key_values` cache if
|
1354 |
+
:meth:`~transformers.PreTrainedModel.beam_search` or :meth:`~transformers.PreTrainedModel.beam_sample` is
|
1355 |
+
called. This is required to match :obj:`past_key_values` with the correct beam_idx at every generation step.
|
1356 |
+
"""
|
1357 |
+
return tuple(
|
1358 |
+
tuple(
|
1359 |
+
past_state.index_select(0, beam_idx.to(past_state.device))
|
1360 |
+
for past_state in layer_past
|
1361 |
+
)
|
1362 |
+
for layer_past in past
|
1363 |
+
)
|
1364 |
+
|
1365 |
|
1366 |
@add_start_docstrings(
|
1367 |
"""
|
|
|
1389 |
|
1390 |
self.init_weights()
|
1391 |
|
1392 |
+
# Model parallel
|
1393 |
+
self.model_parallel = False
|
1394 |
+
self.device_map = None
|
1395 |
+
|
1396 |
@add_start_docstrings_to_model_forward(GPT2_INPUTS_DOCSTRING)
|
1397 |
@add_code_sample_docstrings(
|
1398 |
tokenizer_class=_TOKENIZER_FOR_DOC,
|
|
|
1421 |
config.num_labels - 1]`. If :obj:`config.num_labels == 1` a regression loss is computed (Mean-Square loss),
|
1422 |
If :obj:`config.num_labels > 1` a classification loss is computed (Cross-Entropy).
|
1423 |
"""
|
1424 |
+
return_dict = (
|
1425 |
+
return_dict if return_dict is not None else self.config.use_return_dict
|
1426 |
+
)
|
1427 |
|
1428 |
transformer_outputs = self.transformer(
|
1429 |
input_ids,
|
|
|
1453 |
sequence_lengths = -1
|
1454 |
else:
|
1455 |
if input_ids is not None:
|
1456 |
+
sequence_lengths = (
|
1457 |
+
torch.ne(input_ids, self.config.pad_token_id).sum(-1) - 1
|
1458 |
+
)
|
1459 |
else:
|
1460 |
sequence_lengths = -1
|
1461 |
logger.warning(
|
|
|
1473 |
loss = loss_fct(pooled_logits.view(-1), labels.to(self.dtype).view(-1))
|
1474 |
else:
|
1475 |
loss_fct = CrossEntropyLoss()
|
1476 |
+
loss = loss_fct(
|
1477 |
+
pooled_logits.view(-1, self.num_labels), labels.view(-1)
|
1478 |
+
)
|
1479 |
|
1480 |
if not return_dict:
|
1481 |
output = (pooled_logits,) + transformer_outputs[1:]
|
|
|
1489 |
attentions=transformer_outputs.attentions,
|
1490 |
)
|
1491 |
|
1492 |
+
|
1493 |
+
@add_start_docstrings(
|
1494 |
+
"""
|
1495 |
+
GPT2 Model with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for
|
1496 |
+
Named-Entity-Recognition (NER) tasks.
|
1497 |
+
""",
|
1498 |
+
GPT2_START_DOCSTRING,
|
1499 |
+
)
|
1500 |
+
class GPT2ForTokenClassification(GPT2PreTrainedModel):
|
1501 |
+
def __init__(self, config):
|
1502 |
+
super().__init__(config)
|
1503 |
+
self.num_labels = config.num_labels
|
1504 |
+
|
1505 |
+
self.transformer = GPT2Model(config)
|
1506 |
+
if (
|
1507 |
+
hasattr(config, "classifier_dropout")
|
1508 |
+
and config.classifier_dropout is not None
|
1509 |
+
):
|
1510 |
+
classifier_dropout = config.classifier_dropout
|
1511 |
+
elif hasattr(config, "hidden_dropout") and config.hidden_dropout is not None:
|
1512 |
+
classifier_dropout = config.hidden_dropout
|
1513 |
+
else:
|
1514 |
+
classifier_dropout = 0.1
|
1515 |
+
self.dropout = nn.Dropout(classifier_dropout)
|
1516 |
+
self.classifier = nn.Linear(config.hidden_size, config.num_labels)
|
1517 |
+
|
1518 |
+
self.init_weights()
|
1519 |
+
|
1520 |
+
# Model parallel
|
1521 |
+
self.model_parallel = False
|
1522 |
+
self.device_map = None
|
1523 |
+
|
1524 |
+
@add_start_docstrings_to_model_forward(GPT2_INPUTS_DOCSTRING)
|
1525 |
+
@add_code_sample_docstrings(
|
1526 |
+
tokenizer_class=_TOKENIZER_FOR_DOC,
|
1527 |
+
checkpoint="microsoft/DialogRPT-updown",
|
1528 |
+
output_type=TokenClassifierOutput,
|
1529 |
+
config_class=_CONFIG_FOR_DOC,
|
1530 |
+
)
|
1531 |
+
def forward(
|
1532 |
+
self,
|
1533 |
+
input_ids=None,
|
1534 |
+
past_key_values=None,
|
1535 |
+
attention_mask=None,
|
1536 |
+
token_type_ids=None,
|
1537 |
+
position_ids=None,
|
1538 |
+
head_mask=None,
|
1539 |
+
inputs_embeds=None,
|
1540 |
+
labels=None,
|
1541 |
+
use_cache=None,
|
1542 |
+
output_attentions=None,
|
1543 |
+
output_hidden_states=None,
|
1544 |
+
return_dict=None,
|
1545 |
+
):
|
1546 |
+
r"""
|
1547 |
+
labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size,)`, `optional`):
|
1548 |
+
Labels for computing the sequence classification/regression loss. Indices should be in :obj:`[0, ...,
|
1549 |
+
config.num_labels - 1]`. If :obj:`config.num_labels == 1` a regression loss is computed (Mean-Square loss),
|
1550 |
+
If :obj:`config.num_labels > 1` a classification loss is computed (Cross-Entropy).
|
1551 |
+
"""
|
1552 |
+
return_dict = (
|
1553 |
+
return_dict if return_dict is not None else self.config.use_return_dict
|
1554 |
+
)
|
1555 |
+
|
1556 |
+
transformer_outputs = self.transformer(
|
1557 |
+
input_ids,
|
1558 |
+
past_key_values=past_key_values,
|
1559 |
+
attention_mask=attention_mask,
|
1560 |
+
token_type_ids=token_type_ids,
|
1561 |
+
position_ids=position_ids,
|
1562 |
+
head_mask=head_mask,
|
1563 |
+
inputs_embeds=inputs_embeds,
|
1564 |
+
use_cache=use_cache,
|
1565 |
+
output_attentions=output_attentions,
|
1566 |
+
output_hidden_states=output_hidden_states,
|
1567 |
+
return_dict=return_dict,
|
1568 |
+
)
|
1569 |
+
|
1570 |
+
hidden_states = transformer_outputs[0]
|
1571 |
+
hidden_states = self.dropout(hidden_states)
|
1572 |
+
logits = self.classifier(hidden_states)
|
1573 |
+
|
1574 |
+
loss = None
|
1575 |
+
if labels is not None:
|
1576 |
+
loss_fct = CrossEntropyLoss()
|
1577 |
+
# Only keep active parts of the loss
|
1578 |
+
if attention_mask is not None:
|
1579 |
+
active_loss = attention_mask.view(-1) == 1
|
1580 |
+
active_logits = logits.view(-1, self.num_labels)
|
1581 |
+
active_labels = torch.where(
|
1582 |
+
active_loss,
|
1583 |
+
labels.view(-1),
|
1584 |
+
torch.tensor(loss_fct.ignore_index).type_as(labels),
|
1585 |
+
)
|
1586 |
+
loss = loss_fct(active_logits, active_labels)
|
1587 |
+
else:
|
1588 |
+
loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
|
1589 |
+
|
1590 |
+
if not return_dict:
|
1591 |
+
output = (logits,) + transformer_outputs[2:]
|
1592 |
+
return ((loss,) + output) if loss is not None else output
|
1593 |
+
|
1594 |
+
return TokenClassifierOutput(
|
1595 |
+
loss=loss,
|
1596 |
+
logits=logits,
|
1597 |
+
hidden_states=transformer_outputs.hidden_states,
|
1598 |
+
attentions=transformer_outputs.attentions,
|
1599 |
+
)
|