# Hugging Face — Faster TensorFlow models in Hugging Face Transformers

- Company: Hugging Face (huggingface.co)
- Announced: 2021-01-26
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://huggingface.co/blog/tf-serving
- Record: https://forck.live/items/2270-faster-tensorflow-models-in-hugging-face-transformers
- Subject: Platform
- Models affected: BERT, RoBERTa, ELECTRA, MPNet

The Hugging Face team improved TensorFlow models in Transformers (BERT, RoBERTa, ELECTRA, MPNet) for faster computation and added features for TensorFlow Serving deployment.

## Evidence

Verbatim from https://huggingface.co/blog/tf-serving:

> the Hugging Face team has been working hard on improving Transformers’ TensorFlow models to make them more robust and faster. The recent improvements are mainly focused on two aspects: Computational performance: BERT, RoBERTa, ELECTRA and MPNet have been improved in order to have a much faster computation time. This gain of computational performance is noticeable for all the computational aspects: graph/eager mode, TF Serving and for CPU/GPU/TPU devices. TensorFlow Serving: each of these TensorFlow model can be deployed with TensorFlow Serving to benefit of this gain of computational performance for inference.

---

Record: https://forck.live/items/2270-faster-tensorflow-models-in-hugging-face-transformers
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md