The token_size bug where it is not propagating the context_size correctly from #306 still exists during compaction
Error: compaction.failed APIContextOverflowError: 400 This endpoint's maximum context length is 262144 tokens. However, you requested about 264896 tokens (52139 of text input, 12757 of tool input, 200000 in the output). Please reduce the length of either one.
my config looks like this:
[models. "openrouter/kimi-k2.6"]
provider = "openrouter"
model = "moonshotai/kimi-k2.6:baidu"
max_context_size = 200000
max_output_size = 16384
max_tokens = 16384
It shouldn't be sending the max_context_size for the max_output_size when doing /compaction, it works normally for regular input, but crashes as soon as it tries to issue a /compaction command
The token_size bug where it is not propagating the context_size correctly from #306 still exists during compaction
Error: compaction.failed APIContextOverflowError: 400 This endpoint's maximum context length is 262144 tokens. However, you requested about 264896 tokens (52139 of text input, 12757 of tool input, 200000 in the output). Please reduce the length of either one.my config looks like this:
It shouldn't be sending the max_context_size for the max_output_size when doing /compaction, it works normally for regular input, but crashes as soon as it tries to issue a
/compactioncommand