Hugging Face Transformers: v4.6.1: Patch release
Hugging Face Transformers v4.6.1. Source highlight: Fix regression in models for sequence classification used for regression tasks #11785
256 tracked entries. Older history is imported progressively; see source coverage.
Hugging Face Transformers v4.6.1. Source highlight: Fix regression in models for sequence classification used for regression tasks #11785
Hugging Face Transformers v4.6.0. Source highlight: v4.6.0: ViT, DeiT, CLIP, LUKE, BigBirdPegasus, MegatronBERT
Hugging Face Transformers v4.5.1. Source highlight: Fix `pipeline` when used with private models (#11123)
Hugging Face Transformers v4.5.0. Source highlight: v4.5.0: BigBird, GPT Neo, Examples, Flax support
Hugging Face Transformers v4.4.2. Source highlight: Add support for detecting intel-tensorflow version
Hugging Face Transformers v4.4.0. Source highlight: v4.4.0: S2T, M2M100, I-BERT, mBART-50, DeBERTa-v2, XLSR-Wav2Vec2
Hugging Face Transformers v4.3.3. Source highlight: This patch fixes an issue with the conversion for ConvBERT models: https://github.com/huggingface/transformers/pull/10314.
Hugging Face Transformers v4.3.2. Source highlight: This patch release fixes the RAG model (#10094) and the detection of whether faiss is available (#10103)
Hugging Face Transformers v4.3.1. Source highlight: This patch release modifies the API of the `Wav2Vec2` model: the `Wav2Vec2ForCTC` was added as a replacement of `Wav2Vec2ForMaskedLM`. `Wav2Vec2ForMaskedLM` is kept for backwards compatibility but is deprecated.
Hugging Face Transformers v4.3.0. Source highlight: Wav2Vec2 from facebook (@patrickvonplaten)
Hugging Face Transformers v4.2.2. Source highlight: This patch contains two fixes:
Hugging Face Transformers v4.2.1. Source highlight: This patch contains three fixes:
Hugging Face Transformers v4.2.0. Source highlight: v4.2.0: LED from AllenAI, encoder-decoder templates, fast imports
Hugging Face Transformers v4.1.1. Source highlight: v4.1.1: TAPAS, MPNet, model parallelization, Sharded DDP, conda, multi-part downloads.
Hugging Face Transformers v4.0.1. Source highlight: This patch releases introduces:
Hugging Face Transformers v4.0.0. Source highlight: Transformers v4.0.0-rc-1: Fast tokenizers, model outputs, file reorganization
Hugging Face Transformers v3.5.1 is published, but the available notes are too limited to assess upgrade impact. Review the official source before upgrading.
Hugging Face Transformers v3.5.0. Source highlight: Model versioning, TensorFlow encoder-decoder models, new scripts, refactor of the `generate` method
Hugging Face Transformers v3.4.0. Source highlight: ProphetNet, Blenderbot, SqueezeBERT, DeBERTa
Hugging Face Transformers v3.3.1. Source highlight: Fixes errors due to the name conflicts between the datasets library and local folder or modules named datasets.
Hugging Face Transformers v3.3.0. Source highlight: The RAG model is a retrieval-augmented generation model that can be leveraged for question-answering tasks using `RagTokenForGeneration` or `RagSequenceForGeneration` as proposed in Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks by Patrick Lewis, Ethan Perez, Aleksandara Piktus, Fa
Hugging Face Transformers v3.2.0. Source highlight: Bert Seq2Seq models, FSMT, Funnel Transformer, LXMERT
Hugging Face Transformers v3.1.0. Source highlight: Pegasus, mBART, DPR, self-documented outputs and new pipelines
Hugging Face Transformers v3.0.2 is published, but the available notes are too limited to assess upgrade impact. Review the official source before upgrading.
Hugging Face Transformers v3.0.1. Source highlight: Better backward-compatibility for tokenizers following v3.0.0 refactoring
Hugging Face Transformers v3.0.0. Source highlight: New tokenizer API, TensorFlow improvements, enhanced documentation & tutorials
Hugging Face Transformers v2.11.0. Source highlight: Longformer (@ibeltagy)
Hugging Face Transformers v2.10.0. Source highlight: Reformer (@patrickvonplaten)
Hugging Face Transformers v2.9.1. Source highlight: A new model architecture, `MarianMTModel` with 1,008+ pretrained weights is available for machine translation in PyTorch.
Hugging Face Transformers v2.9.0. Source highlight: Trainer & TFTrainer (@julien-c)
Hugging Face Transformers v2.8.0. Source highlight: ELECTRA Model (@LysandreJik)
Hugging Face Transformers v2.7.0. Source highlight: T5 Model (@patrickvonplaten, @thomwolf )
Hugging Face Transformers v2.6.0. Source highlight: New Model: BART (added by @sshleifer)
Hugging Face Transformers v2.5.1. Source highlight: AutoTokenizer has been put back to False by default so as to not have a breaking change between 2.4.x and 2.5.x
Hugging Face Transformers v2.5.0. Source highlight: Rust tokenizers (@mfuntowicz, @n1t0 )
Hugging Face Transformers v2.4.1 is published, but the available notes are too limited to assess upgrade impact. Review the official source before upgrading.
Hugging Face Transformers v2.4.0. Source highlight: FlauBERT, MMBT, UmBERTo
Hugging Face Transformers v2.3.0. Source highlight: New class `Pipeline` (beta): easily run and use models on down-stream NLP tasks
Hugging Face Transformers v2.2.2 is published, but the available notes are too limited to assess upgrade impact. Review the official source before upgrading.
Hugging Face Transformers v2.2.1. Source highlight: This patch fixes a bug related to the input shape in several models in TensorFlow.
Hugging Face Transformers v2.2.0. Source highlight: New model architectures: ALBERT, CamemBERT, GPT2-XL, DistilRoberta
Hugging Face Transformers v2.1.1. Source highlight: New model architectures: CTRL, DistilGPT-2
Hugging Face Transformers v2.1.0 is published, but the available notes are too limited to assess upgrade impact. Review the official source before upgrading.
Hugging Face Transformers v2.0.0. Source highlight: Name change: welcome 🤗 Transformers
Hugging Face Transformers 1.2.0. Source highlight: New model architecture: DistilBERT
Hugging Face Transformers 1.1.0. Source highlight: **RoBERTa** (from Facebook), a Robustly Optimized BERT Pretraining Approach by Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du et al.
Hugging Face Transformers v1.0.0. Source highlight: Name change: welcome PyTorch-Transformers 👾
Hugging Face Transformers v0.6.2. Source highlight: Better serialization for all models and tokenizers (BERT, GPT, GPT-2 and Transformer-XL) with best practices for saving/loading in readme and examples.
Hugging Face Transformers v0.6.1 is published, but the available notes are too limited to assess upgrade impact. Review the official source before upgrading.
Hugging Face Transformers v0.6.0 is published, but the available notes are too limited to assess upgrade impact. Review the official source before upgrading.