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

[XPU] add xpu support for llama sft #9152

Merged
merged 1 commit into from
Sep 25, 2024
Merged

[XPU] add xpu support for llama sft #9152

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
10 changes: 10 additions & 0 deletions llm/run_finetune.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,6 +64,7 @@
init_chat_template,
)
from paddlenlp.utils.log import logger
from paddlenlp.utils.tools import get_env_device

# Fine-tune Environment Variables to support sharding stage1 overlap optimization.
os.environ["USE_CASUAL_MASK"] = "False"
Expand Down Expand Up @@ -105,6 +106,15 @@ def main():
"the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
)

if get_env_device() == "xpu" and training_args.gradient_accumulation_steps > 1:
try:
from paddle_xpu.layers.nn.linear import LinearConfig # noqa: F401
LinearConfig.enable_accumulate_steps_opt()
LinearConfig.set_accumulate_steps(training_args.gradient_accumulation_steps)
except ImportError:
# It's OK, not use accumulate_steps optimization
pass

# Load model
if training_args.fp16_opt_level == "O2":
if training_args.fp16:
Expand Down
Loading