[Perf] RMSNorm uncontiguous support, 1.2~3.1x kernel performance improvement (#49750)
Signed-off-by: yewentao256 <zhyanwentao@126.com> Signed-off-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
This commit is contained in:
@@ -211,8 +211,8 @@ class ApertusAttention(nn.Module):
|
||||
) -> torch.Tensor:
|
||||
qkv, _ = self.qkv_proj(hidden_states)
|
||||
q, k, v = qkv.split([self.q_size, self.kv_size, self.kv_size], dim=-1)
|
||||
q = self.q_norm(q.contiguous().view(-1, self.head_dim)).view_as(q)
|
||||
k = self.k_norm(k.contiguous().view(-1, self.head_dim)).view_as(k)
|
||||
q = self.q_norm(q.view(-1, self.num_heads, self.head_dim)).view_as(q)
|
||||
k = self.k_norm(k.view(-1, self.num_kv_heads, self.head_dim)).view_as(k)
|
||||
q, k = self.rotary_emb(positions, q, k)
|
||||
attn_output = self.attn(q, k, v)
|
||||
output, _ = self.o_proj(attn_output)
|
||||
|
||||
Reference in New Issue
Block a user