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

[Distributed] enable tensor_parallel_output for finetuning #8370

Merged
merged 1 commit into from
May 10, 2024
Merged

[Distributed] enable tensor_parallel_output for finetuning #8370

Show file tree
Hide file tree
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
6 changes: 3 additions & 3 deletions llm/finetune_generation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -140,7 +140,7 @@ def main():
if not training_args.autotuner_benchmark:
model = AutoModelForCausalLMPipe.from_pretrained(
model_args.model_name_or_path,
tensor_parallel_output=False,
tensor_parallel_output=training_args.tensor_parallel_output,
tensor_parallel_degree=training_args.tensor_parallel_degree,
tensor_parallel_rank=training_args.tensor_parallel_rank,
use_flash_attention=model_args.use_flash_attention,
Expand All @@ -152,7 +152,7 @@ def main():
# NOTE(gongenlei): new add autotuner_benchmark
model_config = AutoConfig.from_pretrained(
model_args.model_name_or_path,
tensor_parallel_output=False,
tensor_parallel_output=training_args.tensor_parallel_output,
tensor_parallel_degree=training_args.tensor_parallel_degree,
tensor_parallel_rank=training_args.tensor_parallel_rank,
dtype=dtype,
Expand All @@ -163,7 +163,7 @@ def main():
else:
model_config = AutoConfig.from_pretrained(
model_args.model_name_or_path,
tensor_parallel_output=False,
tensor_parallel_output=training_args.tensor_parallel_output,
tensor_parallel_degree=training_args.tensor_parallel_degree,
tensor_parallel_rank=training_args.tensor_parallel_rank,
dtype=dtype,
Expand Down
7 changes: 7 additions & 0 deletions llm/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -211,6 +211,13 @@ def prediction_step(
# keepdim in order to maintain the same shape as logits
if isinstance(logits, (list, tuple)):
logits = logits[0]
# all gather logits when enabling tensor_parallel_output
if self.args.tensor_parallel_degree > 1 and self.args.tensor_parallel_output:
hcg = fleet.get_hybrid_communicate_group()
model_parallel_group = hcg.get_model_parallel_group()
gathered_logits = []
dist.all_gather(gathered_logits, logits, group=model_parallel_group)
logits = paddle.concat(gathered_logits, axis=-1)
return (loss, logits.argmax(axis=-1, keepdim=True), labels)

loss = None
Expand Down
4 changes: 4 additions & 0 deletions paddlenlp/trainer/training_args.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -793,6 +793,10 @@ class TrainingArguments:
default=False,
metadata={"help": "whether to run distributed training in auto parallel mode"},
)
tensor_parallel_output: Optional[bool] = field(
default=False,
metadata={"help": "whether to output logits in distributed status"},
)

def __post_init__(self):
env_local_rank = int(os.environ.get("PADDLE_RANK_IN_NODE", -1))
Expand Down
Loading