Skip to content

[draft] print blockmask sprsity #1901

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

Draft
wants to merge 1 commit into
base: gh/XilunWu/32/base
Choose a base branch
from
Draft

[draft] print blockmask sprsity #1901

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
7 changes: 6 additions & 1 deletion torchtitan/distributed/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -467,15 +467,18 @@ def cp_shard(
from torch.nn.attention.flex_attention import BlockMask

load_balancer = None
"""

seq_length = inputs.shape[1]
load_balancer = _HeadTailLoadBalancer(
seq_length, cp_mesh.size(0), cp_mesh.device_type
)

"""
assert isinstance(attention_masks, BlockMask)
load_balancer = _PTRRLoadBalancer(attention_masks, cp_mesh.size(0))
"""
assert isinstance(attention_masks, BlockMask)
logger.info(f"global block_mask sprsity = {attention_masks.sparsity()}")

inputs, labels = _context_parallel_shard(
mesh=cp_mesh,
Expand Down Expand Up @@ -510,5 +513,7 @@ def cp_shard(
if isinstance(attention_masks, BlockMask)
else {k: v for k, v in zip(attention_masks.keys(), masks)}
)
assert isinstance(attention_masks, BlockMask)
logger.info(f"cp sharded block_mask sprsity = {attention_masks.sparsity()}")

return inputs, labels, attention_masks, order_sensitive_buffers
Loading