Registry / ai-ml / tf-nightly-cpu

tf-nightly-cpu

library2.22.0.dev20260530pypypiunverified

Daily built CPU-only preview of TensorFlow, an open source machine learning framework. Version 2.22.0.dev20260530. Released nightly; contains unverified changes and may be unstable.

ai-ml

pip install tf-nightly-cpu

homepage ↗

INSTALL

IMPORT

SIG · TF-NIGHTLY-CPU

tf-nightly-cpu

ai-mlpythonv2.22.0.dev20260530

harness data pending

Install & Compatibility

Where this runs

No compatibility data collected yet for this library.

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

TensorFlow

✓ import tensorflow as tf

✗ import tensorflow_cpu as tf

The package name is `tf-nightly-cpu` but imports still use `tensorflow`.

Minimal TF nightly CPU quickstart: creates and fits a simple Keras model.

import tensorflow as tf import os # Build a simple model model = tf.keras.Sequential([ tf.keras.layers.Dense(10, activation='relu', input_shape=(4,)), tf.keras.layers.Dense(1) ]) model.compile(optimizer='adam', loss='mse') # Create dummy data import numpy as np x = np.random.random((100, 4)) y = np.random.random((100, 1)) model.fit(x, y, epochs=1, verbose=0) print('TensorFlow version:', tf.__version__) print('Quickstart succeeded.')

Debug

Known issues

breakingNightly builds may break existing code due to API changes or removed features. Pin to a specific nightly if reproducibility is needed.

fix

Use `pip install tf-nightly-cpu==<exact-version>` or switch to stable `tensorflow` for production.

affects: all

deprecatedtf.compat.v1 symbols are gradually being removed in nightly. Relying on v1 compatibility may fail without warning.

fix

Migrate to TF2 idioms (e.g., use `tf.function` instead of `tf.Session`).

affects: >=2.16

gotchacuDNN/cuBLAS version mismatches can cause silent failures or crashes even on CPU-only builds if CUDA libraries are accidentally installed.

fix

Ensure no CUDA/CUDNN libraries are in your path. Use `pip show tf-nightly-cpu` to verify no GPU dependencies.

affects: all

gotchaAutomatic tensor type promotion may change behavior between nightly versions (e.g., int32 vs int64).

fix

Explicitly cast tensors with `tf.cast` and avoid relying on implicit promotion.

affects: >=2.18

Upgrade

Version history

2.22.0.dev20260530latest on PyPI · released May 30, 2026

Audit

Dependencies

No dependency data recorded yet.

Agent activity

20 hits · last 30 days

node

OpenAI (training)

Resources

homepagehttps://www.tensorflow.org/ ↗