coai
coai copied to clipboard
coaidev
能不能搞个渠道自动获取最新model？
- [x] 我已确认目前没有类似 issue
- [x] 我已确认我已升级到最新版本
- [x] 我已完整浏览项目 README 和项目文档并未找到解决方案
- [x] 我理解并愿意跟进此 issue，协助测试和提供反馈
- [x] 我将以礼貌和尊重的态度提问，不得使用不文明用语 (包括在此发布评论的所有人同样适用, 不遵守的人将被 block)
- [x] 我理解并认可上述内容，并理解项目维护者精力有限，不遵循规则的 issue 可能会被无视或直接关闭
功能描述
如题，希望能有一个能设置定时更新获取最新模型的功能，还有能不能将api接入口和获取model的输入口分开？有的模型站接入上游的时候是这个口子，获取模型的就是那个口子了
相关截图 (如果有)
意思就是接入点，然后再加个模型更新接口和定时自动获取最新模型功能，没办法，现在ai模型迭代太快了，隔几天chatgpt这边就出新模型，gemini那边又有新模型，所以希望有个功能能自动更新，免得自己手动每次都要获取上游
Bot detected the issue body's language is not English, translate it automatically.
Title: Is it possible to create a channel to automatically get the latest model?
- [x] I have confirmed that there is currently no similar issue
- [x] I have confirmed that I have upgraded to the latest version
- [x] I have fully browsed the project README and project documentation and did not find a solution
- [x] I understand and am willing to follow up on this issue, assist in testing and provide feedback
- [x] I will ask questions in a polite and respectful manner and shall not use uncivilized language (the same applies to everyone who posts comments here, those who do not comply will be blocked)
- [x] I understand and agree with the above content, and understand that project maintainers have limited energy. Issues that do not follow the rules may be ignored or closed directly
Function description
Like the title, I hope there is a function that can set up regular updates to obtain the latest model. Is it possible to separate the API access port and the input port for obtaining the model? When some model stations are connected to the upstream, this is the opening, and the model is obtained through that opening.
Relevant screenshots (if any)
It means an access point, and then add a model update interface and a function to automatically obtain the latest model at a scheduled time. There is no way, the ai model iteration is too fast now, new models will be released on chatgpt every few days, and new models will be released on gemini. , so I hope there is a function that can automatically update, so that I don’t have to manually get the upstream every time
个人感觉具有生产危险性。 企业版的解决方案是设置通用模型组，渠道绑定模型组id。可通过更新模型组模型以更新所有绑定此模型组的渠道。 自动任务请求/v1/models以监测是否有新模型变动，并提供建议merge建议将新模型加入模型组。 开源版后续会考虑做出简化调整并同步。
Bot detected the issue body's language is not English, translate it automatically.
I personally feel it is dangerous to produce. The solution for the enterprise version is to set up a general model group and bind the channel to the model group ID. You can update all channels bound to this model group by updating the model group model. Automate task request models and provide merge suggestions. The open source version will consider making simplified adjustments and synchronization in the future.
个人感觉具有生产危险性。 企业版的解决方案是设置通用模型组，渠道绑定模型组id。可通过更新模型组模型以更新所有绑定此模型组的渠道。 自动任务请求/v1/models以监测是否有新模型变动，并提供建议merge建议将新模型加入模型组。 开源版后续会考虑做出简化调整并同步。
主要是某些api接口获取最新模型不是默认的/v1/models/ 能不能整个自定义获取模型接口的设置？
Bot detected the issue body's language is not English, translate it automatically.
Personally feel that it is dangerous to produce. The solution for the enterprise version is to set up a general model group and bind the channel to the model group ID. You can update all channels bound to this model group by updating the model group model. The automatic task requests /v1/models to monitor whether there are new model changes and provide merge suggestions to add the new model to the model group. The open source version will consider making simplified adjustments and synchronization in the future. The main reason is that some API interfaces to obtain the latest models are not the default /v1/models/. Can the entire settings of the model interface be customized?
Bot detected the issue body's language is not English, translate it automatically.
This is not the core of the problem. The open source version of the adapter can easily solve this problem.
Bot detected the issue body's language is not English, translate it automatically.
Hopefully, like other (similar to lobechat), we can automatically obtain the channel model without inputting one by one.