biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Update rope_scaling to rope_parameters in preparation for Transformers v5 (#28542)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-11-19 18:06:36 +01:00
committed by GitHub
parent d44e9df7d4
commit a8b70304d6
104 changed files with 542 additions and 910 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
vllm/model_executor/models/olmo.py
View File

@@ -87,7 +87,6 @@ class OlmoAttention(nn.Module):
self.num_heads = self.total_num_heads // tensor_model_parallel_world_size
self.head_dim = self.hidden_size // self.total_num_heads
self.max_position_embeddings = config.max_position_embeddings
self.rope_theta = config.rope_theta
self.clip_qkv = config.clip_qkv
# Attention input projection. Projects x -> (q, k, v)
@@ -105,7 +104,7 @@ class OlmoAttention(nn.Module):
self.head_dim,
rotary_dim=self.head_dim,
max_position=self.max_position_embeddings,
base=self.rope_theta,
rope_parameters=config.rope_parameters,
)
self.scaling = self.head_dim**-0.5
self.attn = Attention(