tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

Respect sliding_window=None (#1214)

Browse Source
This commit is contained in:
DreamGenX
2024-01-26 13:43:37 +01:00
committed by GitHub
parent 5407ddd233
commit 62ca4a2b71
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/axolotl/monkeypatch/mistral_attn_hijack_flash.py
View File

@@ -94,7 +94,7 @@ def _prepare_decoder_attention_mask(
sliding_window, sliding_window,
): # pylint: disable=unused-argument ): # pylint: disable=unused-argument
# [bsz, seq_len] # [bsz, seq_len]
if attention_mask is None: if attention_mask is None or sliding_window is None:
return attention_mask return attention_mask
# NOTE: attention mask and sliding masks are only broadcastable in certain scenarios. # NOTE: attention mask and sliding masks are only broadcastable in certain scenarios.
@@ -151,7 +151,7 @@ def flashattn_forward(
) )
use_sliding_windows = ( use_sliding_windows = (
hasattr(self.config, "sliding_window") is not None getattr(self.config, "sliding_window") is not None
and kv_seq_len > self.config.sliding_window and kv_seq_len > self.config.sliding_window
) )