taomanwai / tensorboardcolab Goto Github PK

Every time i rerun the setup, it removes all my saved summaries. While sometimes after a long run model, the ngrok connect breaks, so i have to rerun the tensorboardcolab, and hours of works are gone.

Of course i can always run ngrok http 6006 instead after the setup, but somehow i forget every time.

Please support eager execution, which will be the default mode in TensorFlow v2

Hi all,

I am currently attempting to use Google Colab to graph some machine learning algorithms. Currently, whenever I create a model in the TensorBoard Scalar section I get two folders being the 'Training' and 'Validation' folders showing the data I need. However, when I run another model, the new data gets appended to those files as opposed to a new folder.

Is there a way to make unique named models that could be compared as opposed to appended to the same graph?

Directory path is hardcoded as 'training', right?
training_log_dir = os.path.join(log_dir, 'training')

Hi, when I add this callback to a simple Keras script in Colab I get the error:

FailedPreconditionError: Error while reading resource variable conv_dw_8_1/depthwise_kernel from Container: localhost. This could mean that the variable was uninitialized. Not found: Resource localhost/conv_dw_8_1/depthwise_kernel/N10tensorflow3VarE does not exist.
[[Node: conv_dw_8_1/depthwise/ReadVariableOp = ReadVariableOpdtype=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:GPU:0"]]
[[Node: metrics_1/acc/Mean/_99 = _Recvclient_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device_incarnation=1, tensor_name="edge_6754_metrics_1/acc/Mean", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:CPU:0"]]

Any idea why?

/usr/local/lib/python3.6/dist-packages/tensorboardcolab/callbacks.py in set_model(self, model)
30 self.val_writer = tf.contrib.summary.create_file_writer(self.val_log_dir)
31 else:
---> 32 self.val_writer = tf.summary.FileWriter(self.val_log_dir)
33
34 super(TensorBoardColabCallback, self).set_model(model)

AttributeError: module 'tensorboard.summary._tf.summary' has no attribute 'FileWriter'

The callbacks parameter for Keras is not taking inputs for TensorBoardColabCallback. @taomanwai
https://colab.research.google.com/drive/19_QqEfi8WeNkbhRpxObukE8IggmMFJWY#scrollTo=UwtUJEjI51en

Please have a look or suggest a temporary solution.

Hello!
It seems that the latest 1.x version of Keras+Tensorflow requires an on_train_batch_begin function definition that is missing...

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-14-f2f738e2066b> in <module>()
      2 with session.as_default(), graph.as_default() :
      3   model.set_weights(weights)
----> 4   result = model.fit(X_train, y_train, batch_size=32, epochs=100, verbose=0, shuffle=False, validation_data=(X_test, y_test), callbacks=[PrintDots(),TensorBoardColabCallback(tbc)])
      5 end_time = time.perf_counter()
      6 print( "time = " + str(end_time - start_time) + "s" )

2 frames
/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/callbacks.py in _call_batch_hook(self, mode, hook, batch, logs)
    194     t_before_callbacks = time.time()
    195     for callback in self.callbacks:
--> 196       batch_hook = getattr(callback, hook_name)
    197       batch_hook(batch, logs)
    198     self._delta_ts[hook_name].append(time.time() - t_before_callbacks)

AttributeError: 'TensorBoardColabCallback' object has no attribute 'on_train_batch_begin'

These are the versions of Tensorflow, Keras and tensorboardcolab that I'm using, respectively, which already come pre-installed in Google Colab:

1.13.1
2.2.4-tf
Requirement already up-to-date: tensorboardcolab in /usr/local/lib/python3.6/dist-packages (0.0.22)

Any workaround for this issue?
Thanks!

ImportError Traceback (most recent call last)
/usr/local/lib/python3.6/dist-packages/keras/callbacks.py in init(self, log_dir, histogram_freq, batch_size, write_graph, write_grads, write_images, embeddings_freq, embeddings_layer_names, embeddings_metadata, embeddings_data, update_freq)
744 import tensorflow as tf
--> 745 from tensorflow.contrib.tensorboard.plugins import projector
746 except ImportError:

4 frames
/usr/local/lib/python3.6/dist-packages/tensorflow/contrib/init.py in ()
30 from tensorflow.contrib import cloud
---> 31 from tensorflow.contrib import cluster_resolver
32 from tensorflow.contrib import coder

/usr/local/lib/python3.6/dist-packages/tensorflow/contrib/cluster_resolver/init.py in ()
25 from tensorflow.python.distribute.cluster_resolver.cluster_resolver import UnionClusterResolver
---> 26 from tensorflow.python.distribute.cluster_resolver.gce_cluster_resolver import GceClusterResolver
27 from tensorflow.python.distribute.cluster_resolver.kubernetes_cluster_resolver import KubernetesClusterResolver

ImportError: cannot import name 'GceClusterResolver'

During handling of the above exception, another exception occurred:

ImportError Traceback (most recent call last)
in ()
17 epochs=5,
18 validation_data=(x_test, y_test),
---> 19 callbacks=[TensorBoardColabCallback(tbc)])

/usr/local/lib/python3.6/dist-packages/tensorboardcolab/callbacks.py in init(self, tbc, write_graph, **kwargs)
20
21 training_log_dir = os.path.join(log_dir, 'training')
---> 22 super(TensorBoardColabCallback, self).init(training_log_dir, **kwargs)
23
24 # Log the validation metrics to a separate subdirectory

/usr/local/lib/python3.6/dist-packages/keras/callbacks.py in init(self, log_dir, histogram_freq, batch_size, write_graph, write_grads, write_images, embeddings_freq, embeddings_layer_names, embeddings_metadata, embeddings_data, update_freq)
745 from tensorflow.contrib.tensorboard.plugins import projector
746 except ImportError:
--> 747 raise ImportError('You need the TensorFlow module installed to '
748 'use TensorBoard.')
749

ImportError: You need the TensorFlow module installed to use TensorBoard.

I use pytorch with the following implementation:

from tensorboardcolab import TensorBoardColab
tb = TensorBoardColab()

and this call:

tb.save_value('Validation Accuracy', 'valid_acc',i, validation_acc)

Traceback gives me this error:

AttributeError                            Traceback (most recent call last)

<ipython-input-23-1106bec90267> in <module>()
      6 optimizer = optim.Adam(model.parameters(), lr=0.001, weight_decay=0.0001)
      7 training_epochs = 150
----> 8 metrics = train(model, train_loader, test_loader, loss, optimizer, training_epochs)
      9 
     10 summary(model, (3, 32,32))

1 frames

/usr/local/lib/python3.6/dist-packages/tensorboardcolab/core.py in save_value(self, graph_name, line_name, epoch, value)
     99                 tf.contrib.summary.scalar(graph_name, value)
    100         else:
--> 101             summary = tf.Summary()
    102             summary_value = summary.value.add()
    103             summary_value.simple_value = value

AttributeError: module 'tensorflow' has no attribute 'Summary'

Something seems to be wrong with setup.
I attempted to run the tensorboardcolab tool in Colab as suggested in README,

!pip install -U tensorboardcolab
from tensorboardcolab import *
tbc=TensorBoardColab()

and got

(...)
Wait for 5 seconds...
Setup not passed, retry again (24)
(...)

Hi,
I have run my code and the checkpoints and event files already created in the collab content drive. How can I run this specific event file in tensorborad in Collab?

Thank you