Skip to content

fix masking error when using mlp_bias=True causing NaN during gradien… #3699

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

Open
snehalv2002 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix masking error when using mlp_bias=True causing NaN during gradien… #3699

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
4 changes: 4 additions & 0 deletions src/maxtext/layers/moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1237,6 +1237,7 @@ def wrapper(x, logits, pre_bias_logits, w0, w1, wo, w0_bias, w1_bias, wo_bias, r
global_sorted_experts=selected_experts,
use_custom_sort_vjp=self.config.use_custom_sort_vjp,
)
mask = jnp.arange(x.shape[0]) < jnp.sum(group_sizes)

if self.config.mlp_bias:
w0_bias, w1_bias, wo_bias = self.transform_bias(selected_experts, w0_bias, w1_bias, wo_bias)
Expand Down Expand Up @@ -1300,6 +1301,7 @@ def get_active_sharding_axes(pspec_dim_axes, tensor_dim_index):
layer_w0 = jax.lax.psum(layer_w0, "tensor_transpose")
if self.config.mlp_bias:
layer_w0 = layer_w0 + w0_bias
layer_w0 = jnp.where(mask[:, None], layer_w0, 0)
Copy link
Copy Markdown
Collaborator

@RissyRan RissyRan Apr 20, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is default mask defined without EP sharding?

layer_w0 = adc.checkpoint_name(layer_w0, "moe_mlpwi_0")

layer_w1 = gmm_fn(
Expand All @@ -1312,6 +1314,7 @@ def get_active_sharding_axes(pspec_dim_axes, tensor_dim_index):
layer_w1 = jax.lax.psum(layer_w1, "tensor_transpose")
if self.config.mlp_bias:
layer_w1 = layer_w1 + w1_bias
layer_w1 = jnp.where(mask[:, None], layer_w1, 0)
layer_w1 = adc.checkpoint_name(layer_w1, "moe_mlpwi_1")
intermediate_layer = self.apply_ffn_activation(layer_w0, layer_w1)

Expand All @@ -1327,6 +1330,7 @@ def get_active_sharding_axes(pspec_dim_axes, tensor_dim_index):
)
if self.config.mlp_bias:
intermediate_output = intermediate_output + wo_bias
intermediate_output = jnp.where(mask[:, None], intermediate_output, 0)
intermediate_output = adc.checkpoint_name(intermediate_output, "moe_mlpwo")

if self.config.use_ring_of_experts:
Expand Down
Loading