TensorFlow Object Detection API trains only one class instead of multiple classes
I have been trying to fine-tune the SSD Inception v2 model with my own dataset of different kinds of drinks. Unfortunately, when I am training, I found that only one category of class is getting trained but not the rest. I am unsure what I was doing wrong. I have recreated my dataset thrice and tried to train with it, but still I end up with the same results.
Is there any particular procedure to be followed, when training Object Detection API with several classes?
Will the tensorflow train one class after another?
In my case, the problem is caused due to the config file. I have assigned the num_classes with one instead of three. After i initialized it with three, everything works well now. Sometimes, this kind of issue might also arise due t the issue with the dataset.
- Training a Keras model to identify leap years
- Can batch normalization be considered a linear transformation?
- Reduce inference time of object detection model by retraining with subset of original dataset
- How to save a Dataset in multiple shards using `tf.data.Dataset.save`
- why explain logit as 'unscaled log probabililty' in sotfmax_cross_entropy_with_logits?
- How to improve the performance of CNN Model for a specific Dataset? Getting Low Accuracy on both training and Testing Dataset
- InvalidArgumentError: No DNN in stream executor while training a TensorFlow RetinaNet model on Google Colab
- how to improve the accuracy of autoencoder?
- TypeError: Only integers, slices, ellipsis, tf.newaxis and scalar tf.int32/tf.int64 tensors are valid indices
- tensorflow.keras only runs correctly once
- Install Tensorflow in MacOs M1
- Could not find a version that satisfies the requirement tensorflow
- How do I use distributed DNN training in TensorFlow?
- Loading tf.keras model, ValueError: The two structures don't have the same nested structure
- Tensorflow is unable to train to predict simple multiplication
- Why does tensorflow loss go to infinity with larger training set?
- Tensorflow Probability MixtureNormal layer example not working as in example
- how to get string value out of tf.tensor which dtype is string
- How to predict list elements outside the bounds of a py dataframe?
- Load model from model.weights.h5 file stored in Azure Blob
- the number and the name of the event files in tensorflow?
- Meaning of sparse in "sparse cross entropy loss"?
- Error --accelerator unrecognized argument when launching gcloud beta ai-platform versions create API
- Change the threshold value of the keras RELU activation function
- How to implement tf.gather_nd in Pytorch with the argument batch_dims?
- Pipenv fails locking when installing TensorFlow 2.4.1
- Tensorflow dataset splitted sizing parameter problem: Local rendezvous is aborting with status: OUT_OF_RANGE: End of sequence
- Difference between "compute capability" "cuda architecture" clarification for using Tensorflow v2.3.0
- problem with importing @tensorflow/tfjs-node while working with face-api.js package (node.js)
- How to optimize multiple loss functions separately in Keras?