zhangyiss/mlx-examples
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx-examples.git synced 2025-08-30 02:53:41 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

fix sed

Browse Source
This commit is contained in:
Alex Barron 2024-10-31 12:06:34 -07:00
parent 2e0690374e
commit 1d53354b51
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
llms/mlx_lm/models/llama.py
View File

@ -191,7 +191,7 @@ class Attention(nn.Module):
keys = self.rope(keys) keys = self.rope(keys)
output = scaled_dot_product_attention( output = scaled_dot_product_attention(
queries, keys, values, cache=cache, cache=cache, scale=self.scale, mask=mask queries, keys, values, cache=cache, scale=self.scale, mask=mask
) )
output = output.transpose(0, 2, 1, 3).reshape(B, L, -1) output = output.transpose(0, 2, 1, 3).reshape(B, L, -1)

2
llms/mlx_lm/models/qwen2.py
View File

@ -90,7 +90,7 @@ class Attention(nn.Module):
keys = self.rope(keys) keys = self.rope(keys)
output = scaled_dot_product_attention( output = scaled_dot_product_attention(
queries, keys, values, cache=cache, cache=cache, scale=self.scale, mask=mask queries, keys, values, cache=cache, scale=self.scale, mask=mask
) )
output = output.transpose(0, 2, 1, 3).reshape(B, L, -1) output = output.transpose(0, 2, 1, 3).reshape(B, L, -1)
return self.o_proj(output) return self.o_proj(output)