Groups
Conversations
All groups and messages
Send feedback to Google
Help
Sign in
Groups
TPU Users
Conversations
About
TPU Users
1–9 of 9
Community discussion and support for TPU users
Mark all as read
Report abusive group
0 selected
Rohan Mahajan
2
5/3/21
tpu distributed strategy
Hi Russell, Thanks for your response. I am not a google employee(only google cloud user) so can't
unread,
tpu distributed strategy
Hi Russell, Thanks for your response. I am not a google employee(only google cloud user) so can't
5/3/21
Faisal
11/27/20
Matrix.Vector, Vector.Vector multiply, then max of a vector
Firstly I apologize if this is a very basic question, but I am very new to TPU/GPU, so asking this. I
unread,
Matrix.Vector, Vector.Vector multiply, then max of a vector
Firstly I apologize if this is a very basic question, but I am very new to TPU/GPU, so asking this. I
11/27/20
Gil Motta
,
Russell Power
3
11/16/20
TPU vs GPU Loss is bigger on TPU
Hi Russel, Thanks for the suggestion but I have no idea where to look for bfloat16 for the final
unread,
TPU vs GPU Loss is bigger on TPU
Hi Russel, Thanks for the suggestion but I have no idea where to look for bfloat16 for the final
11/16/20
Santosh Gupta
7/28/20
Data fetch bottleneck at inference, but not during training for TPU
This is what my inference setup looks like autotune = tf.data.experimental.AUTOTUNE with strategy.
unread,
Data fetch bottleneck at inference, but not during training for TPU
This is what my inference setup looks like autotune = tf.data.experimental.AUTOTUNE with strategy.
7/28/20
G
7/19/20
How can I see if the ASIC of the Coral Dev Board is working?
Is there a command that shows me the load of the TPU, in order to know if it does accelerate my
unread,
How can I see if the ASIC of the Coral Dev Board is working?
Is there a command that shows me the load of the TPU, in order to know if it does accelerate my
7/19/20
Santosh Gupta
7/1/20
Having trouble getting models to run on TPU “NotImplementedError: TPUStrategy.run(fn, …) does not support pure eager execution…”
I am attempting to make Keras models that use Bert is a component of the overall model architecture,
unread,
Having trouble getting models to run on TPU “NotImplementedError: TPUStrategy.run(fn, …) does not support pure eager execution…”
I am attempting to make Keras models that use Bert is a component of the overall model architecture,
7/1/20
dimple a shajahan ed16d009
,
Paige Bailey
2
6/27/20
Chamfer Loss in tensorflow
You can file a feature request to have Chamfer Loss implemented in TF Addons ( +addons@tensorflow.org
unread,
Chamfer Loss in tensorflow
You can file a feature request to have Chamfer Loss implemented in TF Addons ( +addons@tensorflow.org
6/27/20
Peter Peter
9/10/19
TPU from PyCharm
Hello I have a big matrix with data (about 80GB) and I can't copy it to google drive. Is possible
unread,
TPU from PyCharm
Hello I have a big matrix with data (about 80GB) and I can't copy it to google drive. Is possible
9/10/19
nwo...@ebay.com
, …
Jonathan Hseu
9
8/13/18
TPU pre-trained model with bfloat 16 transformation to traditional tensorflow floating poing
Hey Kyle, Additional things that I mentioned during the call, but it'll be easier to explain them
unread,
TPU pre-trained model with bfloat 16 transformation to traditional tensorflow floating poing
Hey Kyle, Additional things that I mentioned during the call, but it'll be easier to explain them
8/13/18
