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 15, 17:15 UTC
[V1][Minor] fix comments (#15392)
Signed-off-by: chenjincong <chenjincong@baidu.com> Signed-off-by: Chen-0210 <chenjincong11@gmail.com> Co-authored-by: chenjincong <chenjincong@baidu.com>
Chen1022 committed
Mar 24, 2025 at 23:45 UTC
0893567db9350f4760434ae9e140881c58c5f1d7
1 file changed
+1
-1
vllm/v1/sample/sampler.py
+1
-1
@@ -137,7 +137,7 @@ class Sampler(nn.Module):
137
Gather logprobs for topk and sampled/prompt token.
138
139
Args:
140
- logits: (num tokens) x (vocab) tensor
140
+ logprobs: (num tokens) x (vocab) tensor
141
num_logprobs: minimum number of logprobs to
142
retain per token
143
token_ids: prompt tokens (if prompt logprobs)