v4.10.0 is a minor release. Key changes include:

• Add LayoutLMv2 + LayoutXLM #12604 (@NielsRogge);
• Add BEiT #12994 (@NielsRogge).

v4.9.2 is a patch release. Key changes include:

• Tpu tie weights #13030 (@sgugger);
• ONNX fixes & examples: #13048, #13049, #13028, #13014, #12911, (@mfuntowicz, @michaelbenayoun, @LysandreJik).

v4.9.1 is a patch release. Key changes include:

• v4.9.1: Patch release;
• Fix barrier for SM distributed #12853 (@sgugger).

v4.9.0 is a minor release. Key changes include:

• Validating ONNX Model output "last_hidden_state":;
• Validating ONNX Model output "pooler_output":.

v4.8.2 is a patch release. Key changes include:

• Rename detr targets to labels #12280 (@NielsRogge);
• fix ids_to_tokens naming error in tokenizer of deberta v2 #12412 (@hjptriplebee).

v4.8.1 is a patch release. Key changes include:

• Fix default for TensorBoard folder;
• Ray Tune install #12338.

v4.8.0 is a minor release. Key changes include:

• push_to_hub_model_id to control the repo name;
• push_to_hub_organization to specify an organization.

v4.7.0 is a minor release. Key changes include:

• Add DETR #11653 (@NielsRogge);
• Improve DETR #12147 (@NielsRogge).

v4.6.1 is a patch release: fix regression in models for sequence classification used for regression tasks #11785 fix checkpoint deletion when load_bert_model_at_end = true #11748 fix eva…

v4.6.0 is a minor release. Key changes include:

• Add DeiT (PyTorch) #11056 (@NielsRogge);
• CLIP #11445 (@patil-suraj).

v4.5.1 is a patch release. Key changes include:

• Fix pipeline when used with private models;
• Fix loading an architecture in an other.

v4.5.0 is a minor release. Key changes include:

• BigBird #10183 (@vasudevgupta7);
• [BigBird] Fix big bird gpu test #10967 (@patrickvonplaten).

v4.4.2 is a patch release. Key changes include:

• Add support for detecting intel-tensorflow version;
• Fix distributed evaluation on SageMaker with distributed evaluation.

v4.4.0 is a minor release. Key changes include:

• Speech2TextTransformer #10175 (@patil-suraj);
• Add m2m100 #10236 (@patil-suraj).

v4.3.3 is a patch release: this patch fixes an issue with the conversion for convbert models: https://github.com/huggingface/transformers/pull/10314.

v4.3.2 is a patch release: this patch release fixes the rag model (#10094) and the detection of whether faiss is available (#10103)

v4.3.1 is a patch release: deprecate wav2vec2formaskedlm and add wav2vec2forctc #10089 (@patrickvonplaten)

v4.3.0 is a minor release. Key changes include:

• https://colab.research.google.com/drive/18Ms6WjyjpsL-73Y2Vpagh9WdJvVEIoQL?usp=sharing;
• Wav2Vec2 #9659 (@patrickvonplaten).

v4.3.0.rc1 is a unknown release. Key changes include:

• https://colab.research.google.com/drive/18Ms6WjyjpsL-73Y2Vpagh9WdJvVEIoQL?usp=sharing;
• Wav2Vec2 #9659 (@patrickvonplaten).

v4.2.2 is a patch release. Key changes include:

• [TF Led] Fix wrong decoder attention mask behavior #9601 (@patrickvonplaten );
• Fix imports in conversion scripts #9674 (@sgugger).

v4.2.1 is a patch release. Key changes include:

• Fix Trainer with a parallel model #9578 (@sgugger);
• Compliancy with tf-nightly #9570 (@jplu).

v4.2.0 is a minor release. Key changes include:

• Evaluation: https://colab.research.google.com/drive/12INTTR6n64TzS4RrXZxMSXfrOd9Xzamo?usp=sharing;
• Finetuning: https://colab.research.google.com/drive/12LjJazBl7Gam0XBPy_y0CTOJZeZ34c2v?usp=sharing.