Registry / ai-ml / tf-models-official

tf-models-official

library2.20.0pypypi✓ verified 86d ago

TensorFlow Official Models (tf-models-official) is a collection of reference implementations for state-of-the-art TensorFlow models, provided by the TensorFlow team. Current version is 2.20.0, supporting TensorFlow 2.x. It provides a modular library of NLP, vision, and recommendation models with reusable building blocks. Release cadence aligns with TensorFlow releases (usually quarterly).

ai-ml

pip install tf-models-official

github ↗homepage ↗

INSTALL

IMPORT

SIG · TF-MODELS-OFFICIAL

tf-models-official

ai-mlpythonv2.20.0

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.

nlp

✓ from official import nlp

✗ from official.nlp import bert

Common mistake: import from official.nlp directly. Use 'official' as top-level and then access nlp module.

vision

✓ from official import vision

✗ import tensorflow_models as tfm

Importing tensorflow_models is deprecated; use 'from official import vision' instead.

Quick example: import ResNet50 and train on dummy data.

import os os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3' import tensorflow as tf from official import vision train_ds = tf.data.Dataset.from_tensor_slices((tf.random.normal([100, 224, 224, 3]), tf.random.uniform([100], maxval=1000, dtype=tf.int64))) model = vision.resnet50() model.compile(optimizer='adam', loss='sparse_categorical_crossentropy') model.fit(train_ds.batch(32), epochs=1)

Debug

Known issues

breakingVersion 2.20.0 requires TensorFlow 2.20.0 or higher. Using an older TensorFlow version will cause import errors.

fix

Update TensorFlow to match: pip install tensorflow>=2.20

affects: >=2.20.0

deprecatedThe 'tensorflow_models' top-level module is deprecated. Use 'official' instead.

fix

Change imports from 'tensorflow_models' to 'official'.

affects: >=2.0.0

gotchaImports from `official.nlp.bert` may not work directly; you need to import the `nlp` module first (e.g., `from official import nlp`).

fix

Use `from official import nlp` then `nlp.bert`.

affects: all

Errors

Common errors & fixes

ModuleNotFoundError: No module named 'official'

Package not installed or environment issue.

fix

Run `pip install tf-models-official` to install the package.

ImportError: cannot import name 'resnet50' from 'official.vision'

Using incorrect import path or the model is not available in that version.

fix

Check available models: `from official import vision; print(dir(vision))` or use `vision.resnet50` (exact name).

AttributeError: module 'official' has no attribute 'nlp'

Trying to import `official.nlp` directly before importing `official`.

fix

Use `from official import nlp` instead of `import official.nlp`.

Upgrade

Version history

2.20.0latest on PyPI · released Feb 11, 2026

Audit

Dependencies

tensorflowrequiredCore dependency; version should match tf-models-official major.minor.

tensorflow-datasetsoptionalUsed for dataset loading in examples.

Agent activity

17 hits · last 30 days

node

Amazon

OpenAI (training)

Resources

githubhttps://github.com/tensorflow/models ↗

homepagehttps://github.com/tensorflow/models ↗