1
0
mirror of https://github.com/osmarks/nanogpt-experiments.git synced 2024-12-22 08:00:28 +00:00

Merge pull request from python273/patch-1

Fix GPT.crop_block_size when flash attention is available
This commit is contained in:
Andrej 2023-04-12 22:13:01 -07:00 committed by GitHub
commit ea24604b29
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -207,6 +207,7 @@ class GPT(nn.Module):
self.config.block_size = block_size self.config.block_size = block_size
self.transformer.wpe.weight = nn.Parameter(self.transformer.wpe.weight[:block_size]) self.transformer.wpe.weight = nn.Parameter(self.transformer.wpe.weight[:block_size])
for block in self.transformer.h: for block in self.transformer.h:
if hasattr(block.attn, 'bias'):
block.attn.bias = block.attn.bias[:,:,:block_size,:block_size] block.attn.bias = block.attn.bias[:,:,:block_size,:block_size]
@classmethod @classmethod