does this lib support api mode

[Nyx2022]

after i found this lib , i'm very excited.it offer me a very convenient way to integrate ai model into my app, however, the offline mode is still not stronger than full model online,and more and more company may wil choose deploy full model in intranet,so chat with the intranet or internet model api is a very important feature or need,so does this lib support api mode? if not, can you add this featur,thank you in advance !

[sauravpanda]

We dont support API yet but we might soon. I haven't found the right application on why people would use it and how we should design the system. Would love to know your use case! We can for sure add something on the lines of API!

[Nyx2022]

We dont support API yet but we might soon. I haven't found the right application on why people would use it and how we should design the system. Would love to know your use case! We can for sure add something on the lines of API!

i'm willing to help you to make up your minds to add api feature,
in my opnion,use offline model is for privacy,this is also the goal of this lib, however,the method for protect privacy may not be only on user'smachine,it also can be in a range of intranet which can prevent the message that user provided frome internet in ,othersides, for user's machine ,it actually can not run big model,so it will not get a full experience of full model,so if at intranet,we have only one env to run a full model,which can called by some users in that intrane ,the model and value will be maximized,and the cost for all users in intranet will be minimize. do you think so?

and now.typesctipt or javascript has been used in many ares and scenes of computers, so this lib will be more popular if you have more choice for users and maintain MIT license

[sauravpanda]

This does make sense, will put it on the roadmap and start working in a couple of weeks (Once we are done with current priorities)