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

update transformers imports for deepspeed and is_torch_xla_available #2012

Merged
merged 3 commits into from
Sep 7, 2024
Merged

update transformers imports for deepspeed and is_torch_xla_available #2012

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f85baca
change deepspeed to integrations.deepspeed
Rohan138 Sep 3, 2024
dba908c
add version check and change tpu to xla
Rohan138 Sep 5, 2024
4c81992
add version check
Rohan138 Sep 6, 2024
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
24 changes: 20 additions & 4 deletions optimum/onnxruntime/trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,7 +55,6 @@
from torch.utils.data import Dataset, RandomSampler
from transformers.data.data_collator import DataCollator
from transformers.debug_utils import DebugOption, DebugUnderflowOverflow
from transformers.deepspeed import deepspeed_init, deepspeed_load_checkpoint, is_deepspeed_zero3_enabled
from transformers.modeling_utils import PreTrainedModel, unwrap_model
from transformers.tokenization_utils_base import PreTrainedTokenizerBase
from transformers.trainer import Trainer
Expand All @@ -81,10 +80,10 @@
is_apex_available,
is_sagemaker_dp_enabled,
is_sagemaker_mp_enabled,
is_torch_tpu_available,
)

from ..utils import logging
from ..utils.import_utils import check_if_transformers_greater
from .training_args import ORTOptimizerNames, ORTTrainingArguments
from .utils import (
is_onnxruntime_training_available,
Expand All @@ -94,8 +93,25 @@
if is_apex_available():
from apex import amp

if is_torch_tpu_available(check_device=False):
import torch_xla.core.xla_model as xm
if check_if_transformers_greater("4.33"):
from transformers.integrations.deepspeed import (
deepspeed_init,
deepspeed_load_checkpoint,
is_deepspeed_zero3_enabled,
)
else:
from transformers.deepspeed import deepspeed_init, deepspeed_load_checkpoint, is_deepspeed_zero3_enabled

if check_if_transformers_greater("4.39"):
from transformers.utils import is_torch_xla_available

if is_torch_xla_available():
import torch_xla.core.xla_model as xm
else:
from transformers.utils import is_torch_tpu_available

if is_torch_tpu_available(check_device=False):
import torch_xla.core.xla_model as xm

if TYPE_CHECKING:
import optuna
Expand Down
7 changes: 6 additions & 1 deletion optimum/onnxruntime/trainer_seq2seq.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,10 +19,10 @@
import torch
from torch import nn
from torch.utils.data import Dataset
from transformers.deepspeed import is_deepspeed_zero3_enabled
from transformers.trainer_utils import PredictionOutput
from transformers.utils import is_accelerate_available, logging

from ..utils.import_utils import check_if_transformers_greater
from .trainer import ORTTrainer


Expand All @@ -33,6 +33,11 @@
"The package `accelerate` is required to use the ORTTrainer. Please install it following https://huggingface.co/docs/accelerate/basic_tutorials/install."
)

if check_if_transformers_greater("4.33"):
from transformers.integrations.deepspeed import is_deepspeed_zero3_enabled
else:
from transformers.deepspeed import is_deepspeed_zero3_enabled

logger = logging.get_logger(__name__)


Expand Down
Loading