tensorflow / tensorflow/probability
Use conv_utils instead of tf_utils for tfp conv layers
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
Is there a reason for convolutional layers in tfp.layers.Convolution* to use tf_utils instead of conv_utils which is used in tf.keras.layers.Conv*? For my specific case, I need 'causal' padding which is a specific padding type for conv nets that's supported by conv_utils but not tf_utils.
Instead of
from tensorflow.python.keras.utils import tf_utils
use
from tensorflow.python.keras.utils import conv_utils
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start at the tfp.layers.Convolution* entry points and compare their use of tensorflow.python.keras.utils.tf_utils with the corresponding tf.keras.layers.Conv* behavior. Verify that using conv_utils enables the requested causal padding for these convolutional layers and check the affected layer behavior before considering the issue done.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- tensorflow
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100