Skip to content

[dev] feat(moe): Cherry-pick #1989 back to dev #3011

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
FDecaYed merged 1 commit into from
Jan 21, 2026
Merged

[dev] feat(moe): Cherry-pick #1989 back to dev #3011

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
27 changes: 22 additions & 5 deletions gpt_builders.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,8 @@
get_gpt_decoder_layer_specs,
)
from megatron.core.models.gpt.experimental_attention_variant_module_specs import (
is_linear_attention_variant,
get_transformer_block_with_experimental_attention_variant_spec,
get_transformer_layer_with_experimental_attention_variant_spec,
)
from megatron.core.models.gpt.heterogeneous.heterogeneous_layer_specs import (
get_gpt_heterogeneous_layer_spec,
Expand Down Expand Up @@ -46,7 +47,13 @@ def gpt_builder(args, pre_process, post_process, vp_stage=None, config=None, pg_
else:
use_te = args.transformer_impl == "transformer_engine"

if args.num_experts or is_linear_attention_variant(args.experimental_attention_variant):
if args.experimental_attention_variant is not None:
transformer_layer_spec = (
get_transformer_block_with_experimental_attention_variant_spec(
config=config, vp_stage=vp_stage
)
)
elif args.num_experts:
assert not (config.transformer_impl == "inference_optimized")
# Define the decoder block spec
transformer_layer_spec = get_gpt_decoder_block_spec(
Expand All @@ -70,9 +77,19 @@ def gpt_builder(args, pre_process, post_process, vp_stage=None, config=None, pg_
mtp_transformer_layer_spec = import_module(args.spec)
else:
# Define the decoder block spec
decoder_layer_specs = get_gpt_decoder_layer_specs(
config, use_transformer_engine=use_te, normalization=args.normalization, qk_l2_norm=args.qk_l2_norm, vp_stage=vp_stage
)
if args.experimental_attention_variant is not None:
decoder_layer_specs = (
get_transformer_layer_with_experimental_attention_variant_spec(
config=config
)
)
else:
decoder_layer_specs = get_gpt_decoder_layer_specs(
config,
use_transformer_engine=use_te,
normalization=args.normalization,
qk_l2_norm=args.qk_l2_norm,
)
mtp_transformer_layer_spec = decoder_layer_specs[-1]
# Use spec of the last layer in decoder block as spec of the transformer layer in MTP
mtp_block_spec = get_gpt_mtp_block_spec(
Expand Down
Loading
Loading