Skip to content

Fix rng_state in llm models #8396

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

Merged
wawltor merged 1 commit into from
May 27, 2024
Merged

Fix rng_state in llm models #8396

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
36 changes: 27 additions & 9 deletions llm/ernie-3.5-se/modeling.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1181,17 +1181,35 @@ def __init__(self, config):
else:
vocab_size = config.vocab_size

self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size] if config.tie_word_embeddings else [config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
if config.weight_share_add_bias and config.use_bias:
self.bias = self.create_parameter(
shape=[vocab_size],
if vocab_size != config.vocab_size:
with get_rng_state_tracker().rng_state():
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size]
if config.tie_word_embeddings
else [config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
if config.weight_share_add_bias and config.use_bias:
self.bias = self.create_parameter(
shape=[vocab_size],
dtype=paddle.get_default_dtype(),
)
else:
self.bias = None
else:
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size]
if config.tie_word_embeddings
else [config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
else:
self.bias = None
if config.weight_share_add_bias and config.use_bias:
self.bias = self.create_parameter(
shape=[vocab_size],
dtype=paddle.get_default_dtype(),
)
else:
self.bias = None

# Must set distributed attr for Tensor Parallel !
self.weight.is_distributed = True if (vocab_size != config.vocab_size) else False
Expand Down
20 changes: 15 additions & 5 deletions paddlenlp/transformers/gemma/modeling.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1366,11 +1366,21 @@ def __init__(self, config: GemmaConfig):
else:
vocab_size = config.vocab_size

self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size] if config.tie_word_embeddings else [config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)

if vocab_size != config.vocab_size:
with get_rng_state_tracker().rng_state():
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size]
if config.tie_word_embeddings
else [config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
else:
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size]
if config.tie_word_embeddings
else [config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
# Must set distributed attr for Tensor Parallel !
self.weight.is_distributed = True if (vocab_size != config.vocab_size) else False
if self.weight.is_distributed:
Expand Down
15 changes: 11 additions & 4 deletions paddlenlp/transformers/gpt/modeling.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1432,10 +1432,17 @@ def __init__(self, config: GPTConfig, embedding_weights=None):
else:
vocab_size = config.vocab_size

self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size],
dtype=paddle.get_default_dtype(),
)
if vocab_size != config.vocab_size:
with get_rng_state_tracker().rng_state():
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size],
dtype=paddle.get_default_dtype(),
)
else:
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size],
dtype=paddle.get_default_dtype(),
)
# Must set distributed attr for Tensor Parallel !
self.weight.is_distributed = True if (vocab_size != config.vocab_size) else False
if self.weight.is_distributed:
Expand Down
15 changes: 11 additions & 4 deletions paddlenlp/transformers/gpt/modeling_auto.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1138,10 +1138,17 @@ def __init__(self, config: GPTConfig, embedding_weights=None, ipp=None):
else:
vocab_size = config.vocab_size

self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size],
dtype=paddle.get_default_dtype(),
)
if vocab_size != config.vocab_size:
with get_rng_state_tracker().rng_state():
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size],
dtype=paddle.get_default_dtype(),
)
else:
self.weight = self.create_parameter(
shape=[vocab_size, config.hidden_size],
dtype=paddle.get_default_dtype(),
)
# Must set distributed attr for Tensor Parallel !
self.weight.is_distributed = True if (vocab_size != config.vocab_size) else False
if self.weight.is_distributed:
Expand Down
15 changes: 11 additions & 4 deletions paddlenlp/transformers/llama/modeling.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1765,10 +1765,17 @@ def __init__(self, config: LlamaConfig):
else:
vocab_size = config.vocab_size

self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
if vocab_size != config.vocab_size:
with get_rng_state_tracker().rng_state():
self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
else:
self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
# Must set distributed attr for Tensor Parallel !
self.weight.is_distributed = True if (vocab_size != config.vocab_size) else False
if self.weight.is_distributed:
Expand Down
15 changes: 11 additions & 4 deletions paddlenlp/transformers/mixtral/modeling.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1338,10 +1338,17 @@ def __init__(self, config: MixtralConfig):
else:
vocab_size = config.vocab_size

self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
if vocab_size != config.vocab_size:
with get_rng_state_tracker().rng_state():
self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
else:
self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
# Must set distributed attr for Tensor Parallel !
self.weight.is_distributed = True if (vocab_size != config.vocab_size) else False
if self.weight.is_distributed:
Expand Down
16 changes: 12 additions & 4 deletions paddlenlp/transformers/qwen/modeling.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,7 @@
import paddle.nn.functional as F
from paddle import Tensor, nn
from paddle.distributed import fleet
from paddle.distributed.fleet.layers.mpu.random import get_rng_state_tracker
from paddle.distributed.fleet.utils import recompute
from paddle.utils import try_import

Expand Down Expand Up @@ -784,10 +785,17 @@ def __init__(self, config: QWenConfig):
else:
vocab_size = config.vocab_size

self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
if vocab_size != config.vocab_size:
with get_rng_state_tracker().rng_state():
self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
else:
self.weight = self.create_parameter(
shape=[config.hidden_size, vocab_size],
dtype=paddle.get_default_dtype(),
)
# Must set distributed attr for Tensor Parallel !
self.weight.is_distributed = True if (vocab_size != config.vocab_size) else False
if self.weight.is_distributed:
Expand Down