tensorflow / tensorflow/probability
API error in jax substrate?
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
Hello,
If I install tensorflow probability and try to follow the instructions to create a realNVP flow, I get an import error. (This is replicated across several different installs and operating systems, including linux, mac, and google colab, and python 3.9 and 3.10. It happens even in a fresh conda environment where all that's been installed is tensorflow-probability)
The error can be reproduced as follows:
from tensorflow_probability.substrates import jax as tfp
tfb = tfp.bijectors
tfb.real_nvp_default_template(hidden_layers=[512, 512])
This is the smallest reproducible error of trying to run the example from here: https://www.tensorflow.org/probability/api_docs/python/tfp/bijectors/RealNVP
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 by running the minimal Python reproduction in the issue and compare it with the linked TensorFlow Probability RealNVP example. Investigate the JAX substrate path used by tensorflow_probability.substrates.jax and capture the missing import or traceback. Done means the example runs successfully in the reported environments without the import error.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, tensorflow
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100