A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 03:15 UTC
Remove print statement in DeepseekScalingRotaryEmbedding (#11604)
Michael Goin committed
Dec 29, 2024 at 15:16 UTC
0aa38d16f56327622c1689d7510171662757deee
1 file changed
-1
vllm/model_executor/layers/rotary_embedding.py
-1
@@ -668,7 +668,6 @@ class DeepseekScalingRotaryEmbedding(RotaryEmbedding):
668
cos = (freqs.cos() * self.mscale)
669
sin = (freqs.sin() * self.mscale)
670
cache = torch.cat((cos, sin), dim=-1)
671
- print("Cache shape", cache.shape)
671
return cache
672
673
def forward(