I am trying to quantize MobileFacenet (code from sirius-ai) according to the suggestion
and I think I met the same issue as this one
When I add tf.contrib.quantize.create_training_graph() into training graph
(train_nets.py ln.187: before train_op = train(...) or in train() utils/common.py ln.38 before gradients)
It did not add quantize-aware ops into the graph to collect dynamic range max\min.
I assume that I should see some additional nodes in tensorboard, but I did not, thus I think I did not successfully add quantize-aware ops in training graph.
And I try to trace tensorflow, found that I got nothing with _FindLayersToQuantize().
However when I add tf.contrib.quantize.create_eval_graph() to refine the training graph. I can see some quantize-aware ops as act_quant...
Since I did not add ops in training graph successfully, I have no weights to load in eval graph.
Thus I got some error message as
Key MobileFaceNet/Logits/LinearConv1x1/act_quant/max not found in checkpoint
or
tensorflow.python.framework.errors_impl.FailedPreconditionError: Attempting to use uninitialized value MobileFaceNet/Logits/LinearConv1x1/act_quant/max
Does anyone know how to fix this error? or how to get quantized MobileFacenet with good accuracy?
Thanks!
H,
Unfortunately, the contrib/quantize tool is now deprecated. It won't be able to support newer models, and we are not working on it anymore.
If you are interested in QAT, I would recommend trying the new TF/Keras QAT API. We are actively developing that and providing support for it.
Related
I am having a keras model trained on my own dataset. However after loading weights the summary shows None as the first dimension(the batch size).
I want to know the process to fix the shape to batch size of 1, as it is compulsory for me to fix it so i can convert the model to tflite with GPU support.
What worked for me was to specify batch size to the Input layer, like this:
input = layers.Input(shape=input_shape, batch_size=1, dtype='float32', name='images')
This then carried through the rest of the layers.
The bad news is that despite this "fix" the tfl runtime still complains about dynamic tensors. I get these non-fatal errors in logcat when it runs:
E/tflite: third_party/tensorflow/lite/core/subgraph.cc:801 tensor.data.raw != nullptr was not true.
E/tflite: Attempting to use a delegate that only supports static-sized tensors with a graph that has dynamic-sized tensors (tensor#26 is a dynamic-sized tensor).
E/tflite: Ignoring failed application of the default TensorFlow Lite delegate indexed at 0.
The good news is that despite these errors it seems to be using the GPU anyway, based on performance testing.
I'm using:
tensorflow-lite-support:0.2.0'
tensorflow-lite-metadata:0.2.1'
tensorflow-lite:2.6.0'
tensorflow:tensorflow-lite-gpu:2.3.0'
Hopefully, they'll fix the runtime so it doesn't matter whether the batch size is 'None'. It shouldn't matter for doing inference.
I used the transfer learning approach to develop a detection model using the faster_rcnn algorithm.
To evaluate my model, I used the following commands-
!python model_main_tf2.py --model_dir=models/faster_rcnn_inception_resnet_v2 --pipeline_config_path=models/faster_rcnn_inception_resnet_v2/pipeline.config --checkpoint_dir=models/faster_rcnn_inception_resnet_v2
However, I have been getting the following error/info message: -
INFO:tensorflow:Waiting for new checkpoint at models/faster_rcnn_inception_resnet_v2
I0331 23:23:11.699681 140426971481984 checkpoint_utils.py:139] Waiting for new checkpoint at models/faster_rcnn_inception_resnet_v2
I checked the path to the checkpoint_dir is correct. What could be the problem and how can I resolve it?
Thanks in advance.
You need to run another script for training to generate new checkpoint. model_main_tf2.py does not do both at once, i.e., it won't train model and evaluate the model at the end of each epoch.
One way to get what you want modifying checkpoint_max_to_keep in https://github.com/tensorflow/models/blob/13ec3c1460b928301d208115aed0c94fb47538b7/research/object_detection/model_lib_v2.py#L445
to keep all checkpoints, then evaluate separately. This does not work exactly same as you want, but it generates the curves.
A similar situation happened to me. I don't know if this is the solution or just a workaround but it did work for me. I simply exported my model and provided the path to that checkpoint folder.
fintune_checkpoint_model_directory
|
\---checkpoint(folder)
|
\---checkpoint(file with no extension)
\---ckpt-1.data0000of0001
\---ckpt-1.index
and then simply run the model_main_tf.py file for evaluation.
if you trained your model with a few number of steps it can be a problem maybe a few checkpoints can affect that TensorFlow can't generate evaluation so try to increase the number of steps
I am pretty new to tensorflow and I am struggling to get tensorboard to display some of my custom metrics. The model I am working with is a tf.estimator.Estimator, with an associated EstimatorSpec. The first new metric I am trying to log is from my loss function, which is composed of two components: a loss for an age prediction (tf.float32) and a loss for a class prediction (one-hot/multiclass), which I add together to determine a total loss (my model is predicting both a class and an age). The total loss is output just fine during training and shows up on tensorboard, but I would like to track the individual age and the class prediction loss components as well.
I think a solution that is supposed to work is to add a eval_metric_ops argument to the EstimatorSpec as described here (Custom eval_metric_ops in Estimator in Tensorflow). I have not been able to make this approach work, however. I defined a custom metric function that looks like this:
def age_loss_function(labels, ages_pred, ages_true):
per_sample_age_loss = get_age_loss_per_sample(ages_pred, ages_true) ### works fine
#### The error happens on this line:
mean_abs_age_diff, age_loss_update_fn = tf.metrics.Mean(per_sample_age_loss)
######
return mean_abs_age_diff, age_loss_update_fn
eval_metric_ops = {"age_loss": age_loss_function} #### Want to use this in EstimatorSpec
The instructions seem to say that I need both the error metric and the update function which should both be returned from the tf.metrics command as in examples like the one I linked. But this command fails for me with the error message:
tensorflow.python.framework.errors_impl.OperatorNotAllowedInGraphError: using a `tf.Tensor` as a Python `bool` is not allowed in Graph execution. Use Eager execution or decorate this function with #tf.function.
I am probably just misusing the APIs. If someone can guide me on the proper usage I would really appreciate it. Thanks!
It looks like the problem was from a version change. I had updated to tensorflow 2.0 while the instructions I was following were from 1.X. Using tf.compat.v1.metrics.mean() instead gets past this problem.
I'm following the RNN text-generation tutorial with eager execution pretty much line for line. I've trained the model with my own data set and have saved a low loss checkpoint. I'm able to load the weights and generate text but I want to export/save the model so that I can learn how to deploy one using flask. However I can't figure out how. The version I'm using is '1.14.0-rc1'.
The tutorial: https://www.tensorflow.org/tutorials/sequences/text_generation
I have been able to save the model as an HDF5 file but I cannot load it. I've also disabled eager execution but that causes problems with running the code later on. I have tried the following and a few more snippets but those led to nothing as well:
new_model = keras.models.load_model("/content/gdrive/My Drive/ColabNotebooks/ckpt4/my_model.h5")
How ever I get
RuntimeError: tf.placeholder() is not compatible with eager execution.
Lastly I found this in another post and tried it as well but was met with another error:
tf.saved_model.save(model, "/content/gdrive/My Drive/Colab Notebooks/ckpt4/my_model.h5")
error:
AssertionError: Tried to export a function which references untracked object Tensor("StatefulPartitionedCall/args_2:0", shape=(), dtype=resource).TensorFlow objects (e.g. tf.Variable) captured by functions must be tracked by assigning them to an attribute of a tracked object or assigned to an attribute of the main object directly.
I have a (trained) tf.keras model which I would like to convert to a quantized model and retrain with tensorflow's fake quant strategy (using python as frontend).
Could I somehow apply tf.contrib.quantize.create_training_graph directly to the keras model (graph) then retrain?
Seems like there's some problem with the fact that the session is already created when taking the graph from K.get_session().graph.
For example, the following approach:
import tensorflow.contrib.lite as tflite
keras_graph = tf.keras.backend.get_session().graph
from tensorflow.contrib.quantize import create_training_graph
create_training_graph(input_graph=keras_graph,
quant_delay=int(0*(len(X_train) / batch_size)))
...
model.compile(...)
model.fit_generator(...)
results with the message:
"Operation '{name:'act_softmax/sub' id:2294 op device:{} def:{{{node act_softmax/sub}} = Sub[T=DT_FLOAT](conv_preds/act_quant/FakeQuantWithMinMaxVars:0, act_softmax/Max)}}' was changed by updating input tensor after it was run by a session. This mutation will have no effect, and will trigger an error in the future. Either don't modify nodes after running them or create a new session."
And true enough, the error:
tensorflow.python.framework.errors_impl.FailedPreconditionError: Attempting to use uninitialized value conv_preds/act_quant/conv_preds/act_quant/max/biased
(i.e. create_training_graph needs the graph before the session was created? is it possible to get the graph from a keras model before the session was instantiated?)
Alternatively, if this doesn't work, could I convert the (h5) model to a checkpoint, then somehow load the model from this checkpoint to a tensorflow graph and continue working with pure tensorflow?
Would appreciate any help or pointers.
Thank you!