Sem
Sem
Although its been ask quite a while but thank you for your response. The dev team decided to use web as UI instead of JavaFX. This way, we have a...
JavaFX has many unique things to offer and we are not totally backing out because of this. We are still waiting for this to fix.
Thanks @frankfliu for the response. However, running the **CocoTest** gives me this error: ``` java.lang.IllegalStateException: Expected BEGIN_OBJECT but was BEGIN_ARRAY at line 1 column 1369436 path $.annotations[0].bbox com.google.gson.JsonSyntaxException: java.lang.IllegalStateException: Expected...
> Here is the code in CocoDetection dataset which is similar to what you need: https://github.com/awslabs/djl/blob/master/basicdataset/src/main/java/ai/djl/basicdataset/CocoDetection.java#L137-L156 I have checked the actual **CocoDetection** class and modified my code accordingly which is...
Thanks for the response @frankfliu. I confirmed that the **CocoDetection** has been fixed. However, have you taken a look at the error I am getting? ``` MXNet engine call failed:...
Hi @stu1130, this is the class that I made which is kind of a combination of **PikachuDetection** class and the **CocoDetection** class: https://gist.github.com/androuino/00095ba5be3d10cab765bd2447d236cf Then this is my TrainCustomModel class: https://gist.github.com/androuino/6e8e014e3e70b107f4ebb9cf6c9387ac...
@stu1130 So far this is the only stack trace that I could get: > ``` > MXNet engine call failed: MXNetError: Check failed: in_type->at(i) == mshadow: :default_type_flag || in_type->at(i) ==...
@stu1130 I tried your suggestion and this is the error am getting: This is the function looks like after I cast the label to float: ``` private float[] convertRecToFloatList(String[] anno)...
@stu1130 I've some update about the training attempt when I changed the argument's value like this: ``` args = new String[] {"-e", "8", "-m", "1", "-b", "1"}; ``` It runs...
Looks like it's working now with this arguments settings: ```args = new String[] {"-e", "8", "-b", "1"};```.