yep 在 oenapi接入github的gpt-4o会限制8000tokens长度?使用模型别名调用则没问题 中发帖
将github models中的gpt-4o和gpt-4o-mini接入oneapi中,使用gpt-4o和gpt-4o-mini名称调用时,上下文较长时会提示如下:
{"error":{"message":"Request body too large for gpt-4o model. Max size: 8000 tokens. (request id: 202411071816306112813384245475)","type":"","param":"","code":"tokens_limit_reached"}}
如果在oneapi中将模型名称重定向后:
{
"github-gpt-4o": "gpt-4o",
"github-gpt-4o-mini": "gpt-4o-mini"
}
使用"github-gpt-4o"和 "github-gpt-4o-min...