← all repositories
Amin-Tgz/awesome-tensorflow-2

A curated index for the TF2 migration hangover

A hand-maintained directory of TensorFlow 2 tutorials, repos, and videos because Google's own docs weren't enough.

522 stars Learning
awesome-tensorflow-2
Velocity · 7d
+0.2
★ / day
Trend
steady
star history

What it does This repo is a classic “awesome list” — a manually curated collection of TensorFlow 2.x resources sorted into tutorials, sample projects (GANs, RL, NLP, object detection), videos, and blog posts. It also catalogs a few tools and books. Think of it as a community-edited table of contents for a framework that overhauled its entire API.

The interesting bit The list was assembled during the TF 1.x → 2.0 transition, and it shows: the author explicitly frames it as the TF2 successor to jtoy’s older awesome-tensorflow list. There’s something almost archival about capturing that migration moment — the scramble to re-implement YOLO, BERT, and Mario-playing RL agents in eager execution.

Key highlights

  • Covers niche domains often missing from official docs: CartoonGAN, risk parity portfolios, handwritten text recognition, pointnet++ layers
  • Includes non-English resources (Chinese tutorials) and mobile deployment guides (Android/iOS/Browser)
  • Tracks conference content from TensorFlow World 2019, DevSummit 2019, and Google I/O 2019
  • Links to several Packt/Manning book repositories, which is either handy or a warning depending on your view of tech publishing
  • Still actively tagged with TensorFlow 2.3-era topics, though the README’s “TensorFlow 2.3 is now available!” banner suggests maintenance has slowed

Caveats

  • No clear criteria for inclusion; star ratings appear as tiny inline images (3/5, 4/5, 5/5) with unexplained methodology
  • Many links are to external repos with their own maintenance status — link rot is inevitable
  • The “advantages of TF2” section is just two bullet points paraphrasing Google’s marketing copy from 2019

Verdict Worth a bookmark if you’re maintaining legacy TF2 code or exploring specific model implementations (the YOLO and GAN sections are particularly dense). Skip it if you want interactive tutorials or up-to-date guidance — PyTorch won the narrative, and this list mostly froze in time.

heatdrop uses Google Analytics to see which pages get read — nothing else. Your call. How we handle data.