Higher priority for user input of max_prefill_tokens & format (#540)

2024-06-12 21:48:40 -07:00
parent 1374334d38
commit fb9296f0ed
50 changed files with 817 additions and 569 deletions
--- a/python/sglang/srt/conversation.py
+++ b/python/sglang/srt/conversation.py
@@ -1,4 +1,5 @@
 """Conversation templates."""
+
 # Adapted from
 # https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py
 import dataclasses