@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / d189170b6c
Fork 0

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 17, 06:45 UTC

remove useless statements (#1408)

Wang Ran (汪然) committed Oct 20, 2023 at 23:52 UTC d189170b6c5a143e493c3f5cb7e8c976e8da62c7
1 file changed -1
vllm/model_executor/input_metadata.py
-1
@@ -66,7 +66,6 @@ class InputMetadata:
66 else:
67 self.max_num_blocks_per_seq = 0
68 assert block_tables.shape[0] == self.num_generation_tokens
69 - assert context_lens.shape[0] == self.num_generation_tokens
69
70 # Set during the execution of the first attention op.
71 self.attn_bias: Optional[AttentionBias] = None