SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Make torch xla available on GPU (#29334)

Browse Source 
* add USE_TORCH_XLA env

* rename torch_tpu to torch_xla

* better is_torch_xla_available; fix some fsdp and performance issues

* fix format

* fix bug when pjrt_device is cpu

* fix bug

* fix the deprecation handling

---------

Co-authored-by: anw90 <ang868@gmail.com>
Co-authored-by: wangang.wa <wangang.wa@alibaba-inc.com>
This commit is contained in:
Yitong Huang
2024-03-11 22:07:16 +08:00
committed by GitHub
parent 9a3f4d4daf
commit 873d9bb3cc
25 changed files with 120 additions and 77 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
examples/pytorch/language-modeling/run_clm.py
View File

@@ -46,7 +46,7 @@ from transformers import (
Trainer,
TrainingArguments,
default_data_collator,
is_torch_tpu_available,
is_torch_xla_available,
set_seed,
)
from transformers.testing_utils import CaptureLogger
@@ -602,9 +602,9 @@ def main():
tokenizer=tokenizer,
# Data collator will default to DataCollatorWithPadding, so we change it.
data_collator=default_data_collator,
compute_metrics=compute_metrics if training_args.do_eval and not is_torch_tpu_available() else None,
compute_metrics=compute_metrics if training_args.do_eval and not is_torch_xla_available() else None,
preprocess_logits_for_metrics=preprocess_logits_for_metrics
if training_args.do_eval and not is_torch_tpu_available()
if training_args.do_eval and not is_torch_xla_available()
else None,
)