riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 02:02:21 +06:00
Code Issues Actions 33 Packages Projects Releases Wiki Activity

fix another one

Browse Source
This commit is contained in:
Arthur 2025-07-01 15:37:33 +02:00
parent 9fa5f266a1
commit aaae861fc8
2 changed files with 8 additions and 72 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

41
src/transformers/models/starcoder2/modeling_starcoder2.py
View File

@ -29,6 +29,8 @@ from typing import Callable, Optional, Union
import torch
from torch import nn
from transformers.utils.generic import check_model_inputs
from ...activations import ACT2FN
from ...cache_utils import Cache, DynamicCache
from ...generation import GenerationMixin
@ -343,8 +345,7 @@ class Starcoder2Model(Starcoder2PreTrainedModel):
def set_input_embeddings(self, value):
self.embed_tokens = value
@can_return_tuple
@auto_docstring
@check_model_inputs
def forward(
self,
input_ids: Optional[torch.LongTensor] = None,
@ -353,26 +354,12 @@ class Starcoder2Model(Starcoder2PreTrainedModel):
past_key_values: Optional[Union[Cache, list[torch.FloatTensor]]] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
use_cache: Optional[bool] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
cache_position: Optional[torch.LongTensor] = None,
**kwargs: Unpack[TransformersKwargs],
) -> BaseModelOutputWithPast:
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
)
use_cache = use_cache if use_cache is not None else self.config.use_cache
if (input_ids is None) ^ (inputs_embeds is not None):
raise ValueError("You must specify exactly one of input_ids or inputs_embeds")
if self.gradient_checkpointing and self.training and use_cache:
logger.warning_once(
"`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`."
)
use_cache = False
if inputs_embeds is None:
inputs_embeds = self.embed_tokens(input_ids)
@ -404,43 +391,23 @@ class Starcoder2Model(Starcoder2PreTrainedModel):
# create position embeddings to be shared across the decoder layers
position_embeddings = self.rotary_emb(hidden_states, position_ids)
# decoder layers
all_hidden_states = () if output_hidden_states else None
all_self_attns = () if output_attentions else None
for decoder_layer in self.layers[: self.config.num_hidden_layers]:
if output_hidden_states:
all_hidden_states += (hidden_states,)
layer_outputs = decoder_layer(
hidden_states = decoder_layer(
hidden_states,
attention_mask=causal_mask,
position_ids=position_ids,
past_key_value=past_key_values,
output_attentions=output_attentions,
use_cache=use_cache,
cache_position=cache_position,
position_embeddings=position_embeddings,
**kwargs,
)
hidden_states = layer_outputs[0]
if output_attentions:
all_self_attns += (layer_outputs[1],)
hidden_states = self.norm(hidden_states)
# add hidden states from the last decoder layer
if output_hidden_states:
all_hidden_states += (hidden_states,)
return BaseModelOutputWithPast(
last_hidden_state=hidden_states,
past_key_values=past_key_values if use_cache else None,
hidden_states=all_hidden_states,
attentions=all_self_attns,
)

39
src/transformers/models/starcoder2/modular_starcoder2.py
View File

@ -25,6 +25,8 @@ import torch
import torch.utils.checkpoint
from torch import nn
from transformers.utils.generic import check_model_inputs
from ...activations import ACT2FN
from ...cache_utils import Cache, DynamicCache
from ...masking_utils import create_causal_mask, create_sliding_window_causal_mask
@ -164,6 +166,7 @@ class Starcoder2Model(MistralModel):
self.norm = nn.LayerNorm(config.hidden_size, eps=config.norm_epsilon)
self.embedding_dropout = config.embedding_dropout
@check_model_inputs
def forward(
self,
input_ids: Optional[torch.LongTensor] = None,
@ -172,26 +175,12 @@ class Starcoder2Model(MistralModel):
past_key_values: Optional[Union[Cache, list[torch.FloatTensor]]] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
use_cache: Optional[bool] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
cache_position: Optional[torch.LongTensor] = None,
**kwargs: Unpack[TransformersKwargs],
) -> BaseModelOutputWithPast:
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
)
use_cache = use_cache if use_cache is not None else self.config.use_cache
if (input_ids is None) ^ (inputs_embeds is not None):
raise ValueError("You must specify exactly one of input_ids or inputs_embeds")
if self.gradient_checkpointing and self.training and use_cache:
logger.warning_once(
"`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`."
)
use_cache = False
if inputs_embeds is None:
inputs_embeds = self.embed_tokens(input_ids)
@ -223,43 +212,23 @@ class Starcoder2Model(MistralModel):
# create position embeddings to be shared across the decoder layers
position_embeddings = self.rotary_emb(hidden_states, position_ids)
# decoder layers
all_hidden_states = () if output_hidden_states else None
all_self_attns = () if output_attentions else None
for decoder_layer in self.layers[: self.config.num_hidden_layers]:
if output_hidden_states:
all_hidden_states += (hidden_states,)
layer_outputs = decoder_layer(
hidden_states = decoder_layer(
hidden_states,
attention_mask=causal_mask,
position_ids=position_ids,
past_key_value=past_key_values,
output_attentions=output_attentions,
use_cache=use_cache,
cache_position=cache_position,
position_embeddings=position_embeddings,
**kwargs,
)
hidden_states = layer_outputs[0]
if output_attentions:
all_self_attns += (layer_outputs[1],)
hidden_states = self.norm(hidden_states)
# add hidden states from the last decoder layer
if output_hidden_states:
all_hidden_states += (hidden_states,)
return BaseModelOutputWithPast(
last_hidden_state=hidden_states,
past_key_values=past_key_values if use_cache else None,
hidden_states=all_hidden_states,
attentions=all_self_attns,
)