Repository navigation
Replies: 4 comments
|
我也遇到这个问题了,这有好的解决办法么 |
0 replies
|
对话中一个模型用到底就很少有这种问题,但要切别的模型后就有机率出现这种问题,如果能设计出换模型后也能对前面的上下文以新的模型上下文所支持容量压缩,那么这种问题应该就难出现了。这点zcode做得好一些。 |
0 replies
|
右键复制session id,在新会话粘贴继续 |
0 replies
|
补救的话,好像没法读session id,可以在右上角下载session日志,然后新建一个会话让他重新读一遍。 提前预防,可以手动配置一下模型的上下文上限 contextWindow ,然后配置一下自动压缩插件,让他到达一定阈值就自动压缩会话。 - id: compaction-basic
name: "@deepseek-ai/dsh-compaction-basic"
config:
thresholdRatio: 0.6 # 上下文用到 60% 时触发压缩
retainTokens: XX # 保留最近 XX token 原文
auto: true |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
本轮运行失败This model's maximum context length is 1048576 tokens. However, you requested 1048638 tokens (792638 in the messages, 256000 in the completion). Please reduce the length of the messages or completion.
CONTEXT_WINDOW_EXCEEDED
压缩上下文
8月28日 19:32
本轮运行失败This model's maximum context length is 1048576 tokens. However, you requested 1048641 tokens (792641 in the messages, 256000 in the completion). Please reduce the length of the messages or completion.
CONTEXT_WINDOW_EXCEEDED
/compact
8月28日 19:35
本轮运行失败This model's maximum context length is 1048576 tokens. However, you requested 1048644 tokens (792644 in the messages, 256000 in the completion). Please reduce the length of the messages or completion.
CONTEXT_WINDOW_EXCEEDED
/compact
8月28日 19:35
本轮运行失败This model's maximum context length is 1048576 tokens. However, you requested 1048647 tokens (792647 in the messages, 256000 in the completion). Please reduce the length of the messages or completion.
All reactions