Skip to content

Using DTensor to handle local num_heads change while TP is applied #1373

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
merged 1 commit into from
Jul 16, 2025
Merged

Using DTensor to handle local num_heads change while TP is applied #1373

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
15 changes: 5 additions & 10 deletions distributed/tensor_parallelism/fsdp_tp_example.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -121,12 +121,12 @@
layer_tp_plan = {
"attention_norm": SequenceParallel(),
"attention": PrepareModuleInput(
input_layouts=(Shard(1), None),
desired_input_layouts=(Replicate(), None),
input_layouts=(Shard(1), Replicate()),
desired_input_layouts=(Replicate(), Replicate()),
),
"attention.wq": ColwiseParallel(),
"attention.wk": ColwiseParallel(),
"attention.wv": ColwiseParallel(),
"attention.wq": ColwiseParallel(use_local_output=False),
"attention.wk": ColwiseParallel(use_local_output=False),
"attention.wv": ColwiseParallel(use_local_output=False),
"attention.wo": RowwiseParallel(output_layouts=Shard(1)),
"ffn_norm": SequenceParallel(),
"feed_forward": PrepareModuleInput(
Expand All @@ -138,11 +138,6 @@
"feed_forward.w3": ColwiseParallel(),
}

# Adjust attention module to use the local number of heads
attn_layer = transformer_block.attention
attn_layer.n_heads = attn_layer.n_heads // tp_mesh.size()
attn_layer.n_kv_heads = attn_layer.n_kv_heads // tp_mesh.size()

# Custom parallelization plan for the model
parallelize_module(
module=transformer_block,
Expand Down