feat: use model state cache to achieve 5x - 50x faster preparation time for generation

2023-05-28 23:52:38 +08:00
parent 822f2d729c
commit 3e11128c9d
7 changed files with 160 additions and 5 deletions
--- a/backend-python/dep_check.py
+++ b/backend-python/dep_check.py
@@ -1,3 +1,4 @@
+import cyac
 import GPUtil
 import torch
 import rwkv