wassname/vllm
1
0
Fork 0
mirror of https://github.com/wassname/vllm.git synced 2026-08-12 12:30:29 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix doc (#622)

Browse Source
This commit is contained in:
Chaofan Lin
2023-07-31 13:11:57 -07:00
committed by GitHub
parent 953f28cf9a
commit aa39e42c5a
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
vllm/config.py
+1 -1
View File
@@ -224,7 +224,7 @@ class SchedulerConfig:
a single iteration.
max_num_seqs: Maximum number of sequences to be processed in a single
iteration.
max_seq_len: Maximum length of a sequence (including prompt
max_model_len: Maximum length of a sequence (including prompt
and generated text).
"""
vllm/engine/llm_engine.py
+1 -1
View File
@@ -353,7 +353,7 @@ class LLMEngine:
if stopped:
continue
# Check if the sequence has reached max_seq_len.
# Check if the sequence has reached max_model_len.
if seq.get_len() > self.scheduler_config.max_model_len:
self.scheduler.free_seq(
seq, SequenceStatus.FINISHED_LENGTH_CAPPED)