You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
hi,
i habe trained object detection model, i have trained my custom dataset with this http://download.tensorflow.org/models/object_detection/tf2/20200711/ssd_mobilenet_v2_320x320_coco17_tpu-8.tar.gz as pre trained model in tensorflow 2.3,
and now trying to convert model to tflite, how can i manage my dataset, in your test dataset there are three types and in a single image there is only one object so this is may be object classifier, am i right?
just converting my model to tflite gives more latency than the normal pb file.
i want to try Full integer quantizations but i tried your code and its not creating tflite in the dataset generator it says "found 0 images belonging to 0 classes" in my test folder i have directly put all images with voc xml.
please help.
The text was updated successfully, but these errors were encountered:
hi,
i habe trained object detection model, i have trained my custom dataset with this http://download.tensorflow.org/models/object_detection/tf2/20200711/ssd_mobilenet_v2_320x320_coco17_tpu-8.tar.gz as pre trained model in tensorflow 2.3,
and now trying to convert model to tflite, how can i manage my dataset, in your test dataset there are three types and in a single image there is only one object so this is may be object classifier, am i right?
just converting my model to tflite gives more latency than the normal pb file.
i want to try Full integer quantizations but i tried your code and its not creating tflite in the dataset generator it says "found 0 images belonging to 0 classes" in my test folder i have directly put all images with voc xml.
please help.
The text was updated successfully, but these errors were encountered: