maaft
maaft
@rose-a Are you sure that this is defined in the GQL spec? Because I can use (and have used) following schema without problems in every tool from the GQL world...
Thank you for the amazing work. How is the progress on publishing the code for using the trained models with an external VO SLAM system to generate dense point clouds?
You are right. What about ```python worker_init_fn=lambda work_id:np.random.seed(work_id+args.workers*epoch)) ``` where `epoch` is incremented on every epoch. This would, of course, require to rebuild the train_loader on each epoch unless there...
I'd also be very interested in this.
@AlecAivazis Hi! Is there any progress on this topic? Is there an official roadmap tracker or similar?
Yes, please share them again. Baidu is not very easy to use for non-chinese people.
Hi, I'd also be very interested in this feature!
@Yamashou I got it working by using a slightly modified websocket client from hasura: https://gist.github.com/maaft/ac197c72fe5d18eec5916f3f5a75943f How to use with generated structs (generated.SubscribeProjectsQuery, generated.SubscribeProjects): ```golang func main() { subscriptionClient := subscription.NewSubscriptionClient("ws://localhost:8080/graphql")...
Hi! I also just copy-pasted the client I found and it works fine so far. The only issue is 100% CPU after some time, but that should be fixable. I...
Just to document it for later: The >100% CPU issue is related to this function here: ```go func (wh *websocketHandler) ReadJSON(v interface{}) error { ctx, cancel := context.WithTimeout(wh.ctx, wh.timeout) defer...