[Docs] Add missing language options and fix broken links (#28852)

* Add missing entries to the language selector * Add links to the Colab and AWS Studio notebooks for ONNX * Use anchor links in CONTRIBUTING.md * Fix broken hyperlinks due to spaces * Fix links to OpenAI research articles * Remove confusing footnote symbols from author names, as they are also considered invalid markup
2024-02-06 21:01:01 +01:00
parent 40658be461
commit 1c31b7aa3b
38 changed files with 202 additions and 188 deletions
--- a/docs/source/de/add_new_model.md
+++ b/docs/source/de/add_new_model.md
@@ -556,7 +556,7 @@ demselben Framework wie *brand_new_bert* geschrieben wurde. Normalerweise reicht
 es für Ihren Anwendungsfall leicht anzupassen. Zögern Sie nicht, das Hugging Face Team zu bitten, Sie auf ein ähnliches, bereits vorhandenes
 Konvertierungsskript für Ihr Modell zu finden.

- Wenn Sie ein Modell von TensorFlow nach PyTorch portieren, ist ein guter Ausgangspunkt das Konvertierungsskript von BERT [hier] (https://github.com/huggingface/transformers/blob/7acfa95afb8194f8f9c1f4d2c6028224dbed35a2/src/transformers/models/bert/modeling_bert.py#L91)
+- Wenn Sie ein Modell von TensorFlow nach PyTorch portieren, ist ein guter Ausgangspunkt das Konvertierungsskript von BERT [hier](https://github.com/huggingface/transformers/blob/7acfa95afb8194f8f9c1f4d2c6028224dbed35a2/src/transformers/models/bert/modeling_bert.py#L91)
 - Wenn Sie ein Modell von PyTorch nach PyTorch portieren, ist ein guter Ausgangspunkt das Konvertierungsskript von BART [hier](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bart/convert_bart_original_pytorch_checkpoint_to_pytorch.py)

 Im Folgenden werden wir kurz erklären, wie PyTorch-Modelle Ebenengewichte speichern und Ebenennamen definieren. In PyTorch wird der
--- a/docs/source/de/index.md
+++ b/docs/source/de/index.md
@@ -100,10 +100,10 @@ Die Bibliothek enthält derzeit JAX-, PyTorch- und TensorFlow-Implementierungen,
 1. **[FNet](model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GLPN](model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
+1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
 1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
 1. **[GPT NeoX](model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
-1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
+1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) by Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
 1. **[GPTSAN-japanese](model_doc/gptsan-japanese)** released in the repository [tanreinama/GPTSAN](https://github.com/tanreinama/GPTSAN/blob/main/report/model.md) by Toshiyuki Sakamoto(tanreinama).
 1. **[GroupViT](model_doc/groupvit)** (from UCSD, NVIDIA) released with the paper [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
--- a/docs/source/de/installation.md
+++ b/docs/source/de/installation.md
@@ -94,7 +94,7 @@ Installieren wir 🤗 Transformers aus dem Quellcode mit dem folgenden Befehl:
 pip install git+https://github.com/huggingface/transformers
 ```

-Dieser Befehl installiert die aktuelle `main` Version und nicht die neueste `stable` Version. Die `main`-Version ist nützlich, um mit den neuesten Entwicklungen Schritt zu halten. Zum Beispiel, wenn ein Fehler seit der letzten offiziellen Version behoben wurde, aber eine neue Version noch nicht veröffentlicht wurde. Das bedeutet jedoch, dass die "Hauptversion" nicht immer stabil ist. Wir bemühen uns, die Hauptversion einsatzbereit zu halten, und die meisten Probleme werden normalerweise innerhalb weniger Stunden oder eines Tages behoben. Wenn Sie auf ein Problem stoßen, öffnen Sie bitte ein [Issue] (https://github.com/huggingface/transformers/issues), damit wir es noch schneller beheben können!
+Dieser Befehl installiert die aktuelle `main` Version und nicht die neueste `stable` Version. Die `main`-Version ist nützlich, um mit den neuesten Entwicklungen Schritt zu halten. Zum Beispiel, wenn ein Fehler seit der letzten offiziellen Version behoben wurde, aber eine neue Version noch nicht veröffentlicht wurde. Das bedeutet jedoch, dass die "Hauptversion" nicht immer stabil ist. Wir bemühen uns, die Hauptversion einsatzbereit zu halten, und die meisten Probleme werden normalerweise innerhalb weniger Stunden oder eines Tages behoben. Wenn Sie auf ein Problem stoßen, öffnen Sie bitte ein [Issue](https://github.com/huggingface/transformers/issues), damit wir es noch schneller beheben können!

 Überprüfen wir, ob 🤗 Transformers richtig installiert wurde, indem Sie den folgenden Befehl ausführen:

@@ -245,6 +245,6 @@ Sobald Ihre Datei heruntergeladen und lokal zwischengespeichert ist, geben Sie d

 <Tip>

-Weitere Informationen zum Herunterladen von Dateien, die auf dem Hub gespeichert sind, finden Sie im Abschnitt [Wie man Dateien vom Hub herunterlädt] (https://huggingface.co/docs/hub/how-to-downstream).
+Weitere Informationen zum Herunterladen von Dateien, die auf dem Hub gespeichert sind, finden Sie im Abschnitt [Wie man Dateien vom Hub herunterlädt](https://huggingface.co/docs/hub/how-to-downstream).

 </Tip>
--- a/docs/source/de/llm_tutorial.md
+++ b/docs/source/de/llm_tutorial.md
@@ -149,7 +149,7 @@ Wenn in der Datei [`~generation.GenerationConfig`] nichts angegeben ist, gibt `g

 ### Falscher Generierungsmodus

-Standardmäßig und sofern nicht in der Datei [`~generation.GenerationConfig`] angegeben, wählt `generate` bei jeder Iteration das wahrscheinlichste Token aus (gierige Dekodierung). Je nach Aufgabe kann dies unerwünscht sein; kreative Aufgaben wie Chatbots oder das Schreiben eines Aufsatzes profitieren vom Sampling. Andererseits profitieren Aufgaben, bei denen es auf die Eingabe ankommt, wie z.B. Audiotranskription oder Übersetzung, von der gierigen Dekodierung. Aktivieren Sie das Sampling mit `do_sample=True`. Mehr zu diesem Thema erfahren Sie in diesem [Blogbeitrag] (https://huggingface.co/blog/how-to-generate).
+Standardmäßig und sofern nicht in der Datei [`~generation.GenerationConfig`] angegeben, wählt `generate` bei jeder Iteration das wahrscheinlichste Token aus (gierige Dekodierung). Je nach Aufgabe kann dies unerwünscht sein; kreative Aufgaben wie Chatbots oder das Schreiben eines Aufsatzes profitieren vom Sampling. Andererseits profitieren Aufgaben, bei denen es auf die Eingabe ankommt, wie z.B. Audiotranskription oder Übersetzung, von der gierigen Dekodierung. Aktivieren Sie das Sampling mit `do_sample=True`. Mehr zu diesem Thema erfahren Sie in diesem [Blogbeitrag](https://huggingface.co/blog/how-to-generate).

 ```py
 >>> # Set seed or reproducibility -- you don't need this unless you want full reproducibility
--- a/docs/source/de/pipeline_tutorial.md
+++ b/docs/source/de/pipeline_tutorial.md
@@ -71,7 +71,7 @@ Alle zusätzlichen Parameter für Ihre Aufgabe können auch in die [`pipeline`]

 ### Wählen Sie ein Modell und einen Tokenizer

-Die [`pipeline`] akzeptiert jedes Modell aus dem [Hub] (https://huggingface.co/models). Auf dem Hub gibt es Tags, mit denen Sie nach einem Modell filtern können, das Sie für Ihre Aufgabe verwenden möchten. Sobald Sie ein passendes Modell ausgewählt haben, laden Sie es mit der entsprechenden `AutoModelFor` und [`AutoTokenizer`] Klasse. Laden Sie zum Beispiel die Klasse [`AutoModelForCausalLM`] für eine kausale Sprachmodellierungsaufgabe:
+Die [`pipeline`] akzeptiert jedes Modell aus dem [Hub](https://huggingface.co/models). Auf dem Hub gibt es Tags, mit denen Sie nach einem Modell filtern können, das Sie für Ihre Aufgabe verwenden möchten. Sobald Sie ein passendes Modell ausgewählt haben, laden Sie es mit der entsprechenden `AutoModelFor` und [`AutoTokenizer`] Klasse. Laden Sie zum Beispiel die Klasse [`AutoModelForCausalLM`] für eine kausale Sprachmodellierungsaufgabe:

 ```py
 >>> from transformers import AutoTokenizer, AutoModelForCausalLM
--- a/docs/source/de/preprocessing.md
+++ b/docs/source/de/preprocessing.md
@@ -344,7 +344,7 @@ Laden wir den [food101](https://huggingface.co/datasets/food101) Datensatz für
 >>> dataset = load_dataset("food101", split="train[:100]")
 ```

-Als Nächstes sehen Sie sich das Bild mit dem Merkmal 🤗 Datensätze [Bild] (https://huggingface.co/docs/datasets/package_reference/main_classes?highlight=image#datasets.Image) an:
+Als Nächstes sehen Sie sich das Bild mit dem Merkmal 🤗 Datensätze [Bild](https://huggingface.co/docs/datasets/package_reference/main_classes?highlight=image#datasets.Image) an:

 ```py
 >>> dataset[0]["image"]
--- a/docs/source/de/quicktour.md
+++ b/docs/source/de/quicktour.md
@@ -89,7 +89,7 @@ Importieren sie die [`pipeline`] und spezifizieren sie die Aufgabe, welche sie l
 >>> classifier = pipeline("sentiment-analysis")
 ```

-Die Pipeline lädt ein standardmäßiges [vortrainiertes Modell] (https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english) und einen Tokenizer für die Stimmungs-Analyse herunter und speichert sie. Jetzt können Sie den "Klassifikator" auf Ihren Zieltext anwenden:
+Die Pipeline lädt ein standardmäßiges [vortrainiertes Modell](https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english) und einen Tokenizer für die Stimmungs-Analyse herunter und speichert sie. Jetzt können Sie den "Klassifikator" auf Ihren Zieltext anwenden:

 ```py
 >>> classifier("We are very happy to show you the 🤗 Transformers library.")
@@ -148,7 +148,7 @@ Bei einem größeren Datensatz mit vielen Eingaben (wie bei Sprache oder Bildver

 ### Ein anderes Modell und einen anderen Tokenizer in der Pipeline verwenden

-Die [`pipeline`] kann jedes Modell aus dem [Model Hub] (https://huggingface.co/models) verwenden, wodurch es einfach ist, die [`pipeline`] für andere Anwendungsfälle anzupassen. Wenn Sie beispielsweise ein Modell wünschen, das französischen Text verarbeiten kann, verwenden Sie die Tags im Model Hub, um nach einem geeigneten Modell zu filtern. Das oberste gefilterte Ergebnis liefert ein mehrsprachiges [BERT-Modell](https://huggingface.co/nlptown/bert-base-multilingual-uncased-sentiment), das auf die Stimmungsanalyse abgestimmt ist. Großartig, verwenden wir dieses Modell!
+Die [`pipeline`] kann jedes Modell aus dem [Model Hub](https://huggingface.co/models) verwenden, wodurch es einfach ist, die [`pipeline`] für andere Anwendungsfälle anzupassen. Wenn Sie beispielsweise ein Modell wünschen, das französischen Text verarbeiten kann, verwenden Sie die Tags im Model Hub, um nach einem geeigneten Modell zu filtern. Das oberste gefilterte Ergebnis liefert ein mehrsprachiges [BERT-Modell](https://huggingface.co/nlptown/bert-base-multilingual-uncased-sentiment), das auf die Stimmungsanalyse abgestimmt ist. Großartig, verwenden wir dieses Modell!

 ```py
 >>> model_name = "nlptown/bert-base-multilingual-uncased-sentiment"
--- a/docs/source/de/run_scripts.md
+++ b/docs/source/de/run_scripts.md
@@ -22,7 +22,7 @@ Sie werden auch Skripte finden, die wir in unseren [Forschungsprojekten](https:/

 Es wird nicht erwartet, dass die Beispielskripte bei jedem Problem sofort funktionieren. Möglicherweise müssen Sie das Skript an das Problem anpassen, das Sie zu lösen versuchen. Um Ihnen dabei zu helfen, legen die meisten Skripte vollständig offen, wie die Daten vorverarbeitet werden, so dass Sie sie nach Bedarf für Ihren Anwendungsfall bearbeiten können.

-Für jede Funktion, die Sie in einem Beispielskript implementieren möchten, diskutieren Sie bitte im [Forum] (https://discuss.huggingface.co/) oder in einem [issue] (https://github.com/huggingface/transformers/issues), bevor Sie einen Pull Request einreichen. Wir freuen uns zwar über Fehlerkorrekturen, aber es ist unwahrscheinlich, dass wir einen Pull Request zusammenführen, der mehr Funktionalität auf Kosten der Lesbarkeit hinzufügt.
+Für jede Funktion, die Sie in einem Beispielskript implementieren möchten, diskutieren Sie bitte im [Forum](https://discuss.huggingface.co/) oder in einem [issue](https://github.com/huggingface/transformers/issues), bevor Sie einen Pull Request einreichen. Wir freuen uns zwar über Fehlerkorrekturen, aber es ist unwahrscheinlich, dass wir einen Pull Request zusammenführen, der mehr Funktionalität auf Kosten der Lesbarkeit hinzufügt.

 Diese Anleitung zeigt Ihnen, wie Sie ein Beispiel für ein Trainingsskript zur Zusammenfassung in [PyTorch](https://github.com/huggingface/transformers/tree/main/examples/pytorch/summarization) und [TensorFlow](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/summarization) ausführen können. Sofern nicht anders angegeben, sollten alle Beispiele mit beiden Frameworks funktionieren.

--- a/docs/source/de/testing.md
+++ b/docs/source/de/testing.md
@@ -379,7 +379,7 @@ pytest --random-order-bucket=none

 Standardmäßig ist `--random-order-bucket=module` impliziert, wodurch die Dateien auf den Modulebenen gemischt werden. Es kann auch
 auf den Ebenen `class`, `package`, `global` und `none` mischen. Die vollständigen Details entnehmen Sie bitte der
-[Dokumentation] (https://github.com/jbasko/pytest-random-order).
+[Dokumentation](https://github.com/jbasko/pytest-random-order).

 Eine weitere Alternative zur Randomisierung ist: [`pytest-random`](https://github.com/pytest-dev/pytest-randomly). Dieses
 Modul hat eine sehr ähnliche Funktionalität/Schnittstelle, aber es hat nicht die Eimermodi, die in
--- a/docs/source/es/community.md
+++ b/docs/source/es/community.md
@@ -10,7 +10,7 @@ Esta página agrupa los recursos de 🤗 Transformers desarrollados por la comun

 | Recurso     |      Descripción     |      Autor      |
 |:----------|:-------------|------:|
-| [Hugging Face Transformers Glossary Flashcards](https://www.darigovresearch.com/huggingface-transformers-glossary-flashcards) | Un conjunto de flashcards basadas en el [Glosario de documentos de Transformers] (glosario) que se ha puesto en un formato que se puede aprender/revisar fácilmente usando [Anki] (https://apps.ankiweb.net/) una fuente abierta, aplicación de multiplataforma diseñada específicamente para la retención de conocimientos a largo plazo. Ve este [Introductory video on how to use the flashcards](https://www.youtube.com/watch?v=Dji_h7PILrw). | [Darigov Research](https://www.darigovresearch.com/) |
+| [Hugging Face Transformers Glossary Flashcards](https://www.darigovresearch.com/huggingface-transformers-glossary-flashcards) | Un conjunto de flashcards basadas en el [Glosario de documentos de Transformers] (glosario) que se ha puesto en un formato que se puede aprender/revisar fácilmente usando [Anki](https://apps.ankiweb.net/) una fuente abierta, aplicación de multiplataforma diseñada específicamente para la retención de conocimientos a largo plazo. Ve este [Introductory video on how to use the flashcards](https://www.youtube.com/watch?v=Dji_h7PILrw). | [Darigov Research](https://www.darigovresearch.com/) |

 ## Los cuadernos de la comunidad:

--- a/docs/source/es/index.md
+++ b/docs/source/es/index.md
@@ -90,8 +90,8 @@ La biblioteca actualmente contiene implementaciones de JAX, PyTorch y TensorFlow
 1. **[FNet](model_doc/fnet)** (de Google Research) publicado con el paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) por James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](model_doc/funnel)** (de CMU/Google Brain) publicado con el paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) por Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GLPN](model_doc/glpn)** (de KAIST) publicado con el paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) por Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (de OpenAI) publicado con el paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) por Alec Radford, Karthik Narasimhan, Tim Salimans y Ilya Sutskever.
-1. **[GPT-2](model_doc/gpt2)** (de OpenAI) publicado con el paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) por Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** y Ilya Sutskever**.
+1. **[GPT](model_doc/openai-gpt)** (de OpenAI) publicado con el paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) por Alec Radford, Karthik Narasimhan, Tim Salimans y Ilya Sutskever.
+1. **[GPT-2](model_doc/gpt2)** (de OpenAI) publicado con el paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) por Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei y Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (de EleutherAI) publicado con el repositorio [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) por Ben Wang y Aran Komatsuzaki.
 1. **[GPT Neo](model_doc/gpt_neo)** (de EleutherAI) publicado en el paper [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) por Sid Black, Stella Biderman, Leo Gao, Phil Wang y Connor Leahy.
 1. **[GPTSAN-japanese](model_doc/gptsan-japanese)** released with [GPTSAN](https://github.com/tanreinama/GPTSAN) by Toshiyuki Sakamoto (tanreinama).
--- a/docs/source/es/model_sharing.md
+++ b/docs/source/es/model_sharing.md
@@ -220,4 +220,4 @@ Para asegurarnos que los usuarios entiendan las capacidades de tu modelo, sus li
 * Elaborando y subiendo manualmente el archivo`README.md`.
 * Dando click en el botón **Edit model card** dentro del repositorio.

-Toma un momento para ver la [tarjeta de modelo](https://huggingface.co/distilbert-base-uncased) de DistilBert para que tengas un buen ejemplo del tipo de información que debería incluir. Consulta [la documentación](https://huggingface.co/docs/hub/models-cards) para más detalles acerca de otras opciones que puedes controlar dentro del archivo `README.md` como la huella de carbono del modelo o ejemplos de widgets. Consulta la documentación [aquí] (https://huggingface.co/docs/hub/models-cards).
+Toma un momento para ver la [tarjeta de modelo](https://huggingface.co/distilbert-base-uncased) de DistilBert para que tengas un buen ejemplo del tipo de información que debería incluir. Consulta [la documentación](https://huggingface.co/docs/hub/models-cards) para más detalles acerca de otras opciones que puedes controlar dentro del archivo `README.md` como la huella de carbono del modelo o ejemplos de widgets. Consulta la documentación [aquí](https://huggingface.co/docs/hub/models-cards).
--- a/docs/source/fr/index.md
+++ b/docs/source/fr/index.md
@@ -116,11 +116,11 @@ La documentation est organisée en 5 parties:
 1. **[Funnel Transformer](model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GIT](model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
+1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
 1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
 1. **[GPT NeoX](model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
 1. **[GPT NeoX Japanese](model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
-1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
+1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) by Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
 1. **[GPT-Sw3](model_doc/gpt-sw3)** (from AI-Sweden) released with the paper [Lessons Learned from GPT-SW3: Building the First Large-Scale Generative Language Model for Swedish](http://www.lrec-conf.org/proceedings/lrec2022/pdf/2022.lrec-1.376.pdf) by Ariel Ekgren, Amaru Cuba Gyllensten, Evangelia Gogoulou, Alice Heiman, Severine Verlinden, Joey Öhman, Fredrik Carlsson, Magnus Sahlgren.
 1. **[Graphormer](model_doc/graphormer)** (from Microsoft) released with the paper [Do Transformers Really Perform Bad for Graph Representation?](https://arxiv.org/abs/2106.05234) by Chengxuan Ying, Tianle Cai, Shengjie Luo, Shuxin Zheng, Guolin Ke, Di He, Yanming Shen, Tie-Yan Liu.
--- a/docs/source/hi/pipeline_tutorial.md
+++ b/docs/source/hi/pipeline_tutorial.md
@@ -16,7 +16,7 @@ rendered properly in your Markdown viewer.

 # अनुमान के लिए पाइपलाइन

-[`pipeline`] किसी भी भाषा, कंप्यूटर दृष्टि, भाषण और मल्टीमॉडल कार्यों पर अनुमान लगाने के लिए [Hub] (https://huggingface.co/models) से किसी भी मॉडल का उपयोग करना आसान बनाता है। भले ही आपके पास किसी विशिष्ट तौर-तरीके का अनुभव न हो या आप मॉडलों के पीछे अंतर्निहित कोड से परिचित न हों, फिर भी आप [`pipeline`] के अनुमान के लिए उनका उपयोग कर सकते हैं! यह ट्यूटोरियल आपको ये सिखाएगा:
+[`pipeline`] किसी भी भाषा, कंप्यूटर दृष्टि, भाषण और मल्टीमॉडल कार्यों पर अनुमान लगाने के लिए [Hub](https://huggingface.co/models) से किसी भी मॉडल का उपयोग करना आसान बनाता है। भले ही आपके पास किसी विशिष्ट तौर-तरीके का अनुभव न हो या आप मॉडलों के पीछे अंतर्निहित कोड से परिचित न हों, फिर भी आप [`pipeline`] के अनुमान के लिए उनका उपयोग कर सकते हैं! यह ट्यूटोरियल आपको ये सिखाएगा:

 * अनुमान के लिए [`pipeline`] का उपयोग करें।
 * एक विशिष्ट टोकननाइज़र या मॉडल का उपयोग करें।
@@ -67,7 +67,7 @@ Wav2Vec2.
 {'text': ' I have a dream that one day this nation will rise up and live out the true meaning of its creed.'}
 ```

-अब यह परिणाम अधिक सटीक दिखता है! Wav2Vec2 बनाम व्हिस्पर पर गहन तुलना के लिए, [ऑडियो ट्रांसफॉर्मर्स कोर्स] (https://huggingface.co/learn/audio-course/chapter5/asr_models) देखें।
+अब यह परिणाम अधिक सटीक दिखता है! Wav2Vec2 बनाम व्हिस्पर पर गहन तुलना के लिए, [ऑडियो ट्रांसफॉर्मर्स कोर्स](https://huggingface.co/learn/audio-course/chapter5/asr_models) देखें।
 हम वास्तव में आपको विभिन्न भाषाओं में मॉडल, आपके क्षेत्र में विशेषीकृत मॉडल और बहुत कुछ के लिए हब की जांच करने के लिए प्रोत्साहित करते हैं।
 आप हब पर सीधे अपने ब्राउज़र से मॉडल परिणामों की जांच और तुलना कर सकते हैं कि यह फिट बैठता है या नहीं
 अन्य मामलों की तुलना में कोने के मामलों को बेहतर ढंग से संभालता है।
@@ -114,7 +114,7 @@ transcriber = pipeline(model="openai/whisper-large-v2", device=0)
 ```

 यदि मॉडल एकल GPU के लिए बहुत बड़ा है और आप PyTorch का उपयोग कर रहे हैं, तो आप `device_map="auto"` को स्वचालित रूप से सेट कर सकते हैं
-निर्धारित करें कि मॉडल वज़न को कैसे लोड और संग्रहीत किया जाए। `device_map` तर्क का उपयोग करने के लिए 🤗 [Accelerate] (https://huggingface.co/docs/accelerate) की आवश्यकता होती है
+निर्धारित करें कि मॉडल वज़न को कैसे लोड और संग्रहीत किया जाए। `device_map` तर्क का उपयोग करने के लिए 🤗 [Accelerate](https://huggingface.co/docs/accelerate) की आवश्यकता होती है
 पैकेट:

 ```bash
@@ -131,7 +131,7 @@ transcriber = pipeline(model="openai/whisper-large-v2", device_map="auto")

 ### बैच का आकार

-डिफ़ॉल्ट रूप से, पाइपलाइनें [यहां] (https://huggingface.co/docs/transformers/main_classes/pipelines#pipeline-batching) विस्तार से बताए गए कारणों के लिए बैच अनुमान नहीं लगाएंगी। इसका कारण यह है कि बैचिंग आवश्यक रूप से तेज़ नहीं है, और वास्तव में कुछ मामलों में काफी धीमी हो सकती है।
+डिफ़ॉल्ट रूप से, पाइपलाइनें [यहां](https://huggingface.co/docs/transformers/main_classes/pipelines#pipeline-batching) विस्तार से बताए गए कारणों के लिए बैच अनुमान नहीं लगाएंगी। इसका कारण यह है कि बैचिंग आवश्यक रूप से तेज़ नहीं है, और वास्तव में कुछ मामलों में काफी धीमी हो सकती है।

 लेकिन अगर यह आपके उपयोग के मामले में काम करता है, तो आप इसका उपयोग कर सकते हैं:

--- a/docs/source/it/index.md
+++ b/docs/source/it/index.md
@@ -97,8 +97,8 @@ La libreria attualmente contiene implementazioni in JAX, PyTorch e TensorFlow, p
 1. **[FNet](model_doc/fnet)** (da Google Research) rilasciato con il paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) da James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](model_doc/funnel)** (da CMU/Google Brain) rilasciato con il paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) da Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GLPN](model_doc/glpn)** (da KAIST) rilasciato con il paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) da Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (da OpenAI) rilasciato con il paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) da Alec Radford, Karthik Narasimhan, Tim Salimans e Ilya Sutskever.
-1. **[GPT-2](model_doc/gpt2)** (da OpenAI) rilasciato con il paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) da Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** e Ilya Sutskever**.
+1. **[GPT](model_doc/openai-gpt)** (da OpenAI) rilasciato con il paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) da Alec Radford, Karthik Narasimhan, Tim Salimans e Ilya Sutskever.
+1. **[GPT-2](model_doc/gpt2)** (da OpenAI) rilasciato con il paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) da Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei e Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (da EleutherAI) rilasciato nel repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) da Ben Wang e Aran Komatsuzaki.
 1. **[GPT Neo](model_doc/gpt_neo)** (da EleutherAI) rilasciato nel repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) da Sid Black, Stella Biderman, Leo Gao, Phil Wang e Connor Leahy.
 1. **[GPT NeoX](model_doc/gpt_neox)** (da EleutherAI) rilasciato con il paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) da Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
--- a/docs/source/ja/index.md
+++ b/docs/source/ja/index.md
@@ -112,11 +112,11 @@ rendered properly in your Markdown viewer.
 1. **[Funnel Transformer](https://huggingface.co/docs/transformers/model_doc/funnel)** (CMU/Google Brain から) Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le から公開された研究論文: [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236)
 1. **[GIT](https://huggingface.co/docs/transformers/main/model_doc/git)** (Microsoft Research から) Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang. から公開された研究論文 [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100)
 1. **[GLPN](https://huggingface.co/docs/transformers/model_doc/glpn)** (KAIST から) Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim から公開された研究論文: [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436)
-1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (OpenAI から) Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever から公開された研究論文: [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/)
+1. **[GPT](https://huggingface.co/docs/transformers/model_doc/openai-gpt)** (OpenAI から) Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever から公開された研究論文: [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/)
 1. **[GPT Neo](https://huggingface.co/docs/transformers/model_doc/gpt_neo)** (EleutherAI から) Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy から公開されたレポジトリー : [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo)
 1. **[GPT NeoX](https://huggingface.co/docs/transformers/model_doc/gpt_neox)** (EleutherAI から) Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach から公開された研究論文: [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745)
 1. **[GPT NeoX Japanese](https://huggingface.co/docs/transformers/model_doc/gpt_neox_japanese)** (ABEJA から) Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori からリリース.
-1. **[GPT-2](https://huggingface.co/docs/transformers/model_doc/gpt2)** (OpenAI から) Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever** から公開された研究論文: [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/)
+1. **[GPT-2](https://huggingface.co/docs/transformers/model_doc/gpt2)** (OpenAI から) Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever から公開された研究論文: [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/)
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (EleutherAI から) Ben Wang and Aran Komatsuzaki から公開されたレポジトリー [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) 
 1. **[GPT-Sw3](https://huggingface.co/docs/transformers/main/model_doc/gpt-sw3)** (AI-Sweden から) Ariel Ekgren, Amaru Cuba Gyllensten, Evangelia Gogoulou, Alice Heiman, Severine Verlinden, Joey Öhman, Fredrik Carlsson, Magnus Sahlgren から公開された研究論文: [Lessons Learned from GPT-SW3: Building the First Large-Scale Generative Language Model for Swedish](http://www.lrec-conf.org/proceedings/lrec2022/pdf/2022.lrec-1.376.pdf) 
 1. **[GroupViT](https://huggingface.co/docs/transformers/model_doc/groupvit)** (UCSD, NVIDIA から) Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang から公開された研究論文: [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094)
--- a/docs/source/ja/main_classes/deepspeed.md
+++ b/docs/source/ja/main_classes/deepspeed.md
@@ -481,7 +481,7 @@ deepspeed examples/pytorch/translation/run_translation.py ...
 設定ファイルで使用できる DeepSpeed 設定オプションの完全なガイドについては、次を参照してください。
 [次のドキュメント](https://www.deepspeed.ai/docs/config-json/) にアクセスしてください。

-さまざまな実際のニーズに対応する数十の DeepSpeed 構成例を [DeepSpeedExamples] (https://github.com/microsoft/DeepSpeedExamples)で見つけることができます。
+さまざまな実際のニーズに対応する数十の DeepSpeed 構成例を [DeepSpeedExamples](https://github.com/microsoft/DeepSpeedExamples)で見つけることができます。
 リポジトリ:

 ```bash
--- a/docs/source/ja/tasks/knowledge_distillation_for_image_classification.md
+++ b/docs/source/ja/tasks/knowledge_distillation_for_image_classification.md
@@ -19,7 +19,7 @@ rendered properly in your Markdown viewer.

 知識の蒸留は、より大規模で複雑なモデル (教師) からより小規模で単純なモデル (生徒) に知識を伝達するために使用される手法です。あるモデルから別のモデルに知識を抽出するには、特定のタスク (この場合は画像分類) でトレーニングされた事前トレーニング済み教師モデルを取得し、画像分類でトレーニングされる生徒モデルをランダムに初期化します。次に、学生モデルをトレーニングして、その出力と教師の出力の差を最小限に抑え、動作を模倣します。これは [Distilling the Knowledge in a Neural Network by Hinton et al](https://arxiv.org/abs/1503.02531) で最初に導入されました。このガイドでは、タスク固有の知識の蒸留を行います。これには [Beans データセット](https://huggingface.co/datasets/beans) を使用します。

-このガイドでは、[微調整された ViT モデル](https://huggingface.co/merve/vit-mobilenet-beans-224) (教師モデル) を抽出して [MobileNet](https://huggingface. co/google/mobilenet_v2_1.4_224) (学生モデル) 🤗 Transformers の [Trainer API](https://huggingface.co/docs/transformers/en/main_classes/trainer#trainer) を使用します。
+このガイドでは、[微調整された ViT モデル](https://huggingface.co/merve/vit-mobilenet-beans-224) (教師モデル) を抽出して [MobileNet](https://huggingface.co/google/mobilenet_v2_1.4_224) (学生モデル) 🤗 Transformers の [Trainer API](https://huggingface.co/docs/transformers/en/main_classes/trainer#trainer) を使用します。

 蒸留とプロセスの評価に必要なライブラリをインストールしましょう。

@@ -185,4 +185,4 @@ trainer.train()
 trainer.evaluate(processed_datasets["test"])
 ```

-テスト セットでは、モデルの精度は 72% に達します。蒸留効率の健全性チェックを行うために、同じハイパーパラメータを使用して Bean データセットで MobileNet を最初からトレーニングし、テスト セットで 63% の精度を観察しました。読者の皆様には、さまざまな事前トレーニング済み教師モデル、学生アーキテクチャ、蒸留パラメータを試していただき、その結果を報告していただくようお勧めします。抽出されたモデルのトレーニング ログとチェックポイントは [このリポジトリ](https://huggingface.co/merve/vit-mobilenet-beans-224) にあり、最初からトレーニングされた MobileNetV2 はこの [リポジトリ]( https://huggingface.co/merve/resnet-mobilenet-beans-5)。
+テスト セットでは、モデルの精度は 72% に達します。蒸留効率の健全性チェックを行うために、同じハイパーパラメータを使用して Bean データセットで MobileNet を最初からトレーニングし、テスト セットで 63% の精度を観察しました。読者の皆様には、さまざまな事前トレーニング済み教師モデル、学生アーキテクチャ、蒸留パラメータを試していただき、その結果を報告していただくようお勧めします。抽出されたモデルのトレーニング ログとチェックポイントは [このリポジトリ](https://huggingface.co/merve/vit-mobilenet-beans-224) にあり、最初からトレーニングされた MobileNetV2 はこの [リポジトリ](https://huggingface.co/merve/resnet-mobilenet-beans-5)。
--- a/docs/source/ko/index.md
+++ b/docs/source/ko/index.md
@@ -104,11 +104,11 @@ rendered properly in your Markdown viewer.
 1. **[FNet](model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GLPN](model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
+1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
 1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
 1. **[GPT NeoX](model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
 1. **[GPT NeoX Japanese](model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
-1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
+1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) by Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
 1. **[GPTSAN-japanese](model_doc/gptsan-japanese)** released in the repository [tanreinama/GPTSAN](https://github.com/tanreinama/GPTSAN/blob/main/report/model.md) by Toshiyuki Sakamoto(tanreinama).
 1. **[GroupViT](model_doc/groupvit)** (from UCSD, NVIDIA) released with the paper [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.
--- a/docs/source/ms/index.md
+++ b/docs/source/ms/index.md
@@ -125,11 +125,11 @@ Dokumentasi disusun kepada lima bahagian:
 1. **[Funnel Transformer](model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GIT](model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
+1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
 1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
 1. **[GPT NeoX](model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
 1. **[GPT NeoX Japanese](model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
-1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
+1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) by Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
 1. **[GPT-Sw3](model_doc/gpt-sw3)** (from AI-Sweden) released with the paper [Lessons Learned from GPT-SW3: Building the First Large-Scale Generative Language Model for Swedish](http://www.lrec-conf.org/proceedings/lrec2022/pdf/2022.lrec-1.376.pdf) by Ariel Ekgren, Amaru Cuba Gyllensten, Evangelia Gogoulou, Alice Heiman, Severine Verlinden, Joey Öhman, Fredrik Carlsson, Magnus Sahlgren.
 1. **[GPTBigCode](model_doc/gpt_bigcode)** (from BigCode) released with the paper [SantaCoder: don't reach for the stars!](https://arxiv.org/abs/2301.03988) by Loubna Ben Allal, Raymond Li, Denis Kocetkov, Chenghao Mou, Christopher Akiki, Carlos Munoz Ferrandis, Niklas Muennighoff, Mayank Mishra, Alex Gu, Manan Dey, Logesh Kumar Umapathi, Carolyn Jane Anderson, Yangtian Zi, Joel Lamy Poirier, Hailey Schoelkopf, Sergey Troshin, Dmitry Abulkhanov, Manuel Romero, Michael Lappert, Francesco De Toni, Bernardo García del Río, Qian Liu, Shamik Bose, Urvashi Bhattacharyya, Terry Yue Zhuo, Ian Yu, Paulo Villegas, Marco Zocca, Sourab Mangrulkar, David Lansky, Huu Nguyen, Danish Contractor, Luis Villa, Jia Li, Dzmitry Bahdanau, Yacine Jernite, Sean Hughes, Daniel Fried, Arjun Guha, Harm de Vries, Leandro von Werra.
--- a/docs/source/pt/index.md
+++ b/docs/source/pt/index.md
@@ -103,8 +103,8 @@ Atualmente a biblioteca contém implementações do PyTorch, TensorFlow e JAX, p
 1. **[FNet](model_doc/fnet)** (from Google Research) released with the paper [FNet: Mixing Tokens with Fourier Transforms](https://arxiv.org/abs/2105.03824) by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein, Santiago Ontanon.
 1. **[Funnel Transformer](model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GLPN](model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
-1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
+1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
+1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) by Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
 1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
 1. **[GPTSAN-japanese](model_doc/gptsan-japanese)** released in the repository [tanreinama/GPTSAN](https://github.com/tanreinama/GPTSAN/blob/main/report/model.md) by Toshiyuki Sakamoto(tanreinama).
--- a/docs/source/pt/quicktour.md
+++ b/docs/source/pt/quicktour.md
@@ -331,7 +331,7 @@ Todos os modelos de 🤗 Transformers (PyTorch ou TensorFlow) geram tensores *an

 </Tip>

-Os modelos são um standard [`torch.nn.Module`](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) ou um [`tf.keras.Model`](https: //www.tensorflow.org/api_docs/python/tf/keras/Model) para que você possa usá-los em seu loop de treinamento habitual. No entanto, para facilitar as coisas, 🤗 Transformers fornece uma classe [`Trainer`] para PyTorch que adiciona funcionalidade para treinamento distribuído, precisão mista e muito mais. Para o TensorFlow, você pode usar o método `fit` de [Keras](https://keras.io/). Consulte o [tutorial de treinamento](./training) para obter mais detalhes.
+Os modelos são um standard [`torch.nn.Module`](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) ou um [`tf.keras.Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model) para que você possa usá-los em seu loop de treinamento habitual. No entanto, para facilitar as coisas, 🤗 Transformers fornece uma classe [`Trainer`] para PyTorch que adiciona funcionalidade para treinamento distribuído, precisão mista e muito mais. Para o TensorFlow, você pode usar o método `fit` de [Keras](https://keras.io/). Consulte o [tutorial de treinamento](./training) para obter mais detalhes.

 <Tip>

--- a/docs/source/te/quicktour.md
+++ b/docs/source/te/quicktour.md
@@ -511,7 +511,7 @@ tensor([[0.0021, 0.0018, 0.0115, 0.2121, 0.7725],

 ## TensorFlowతో శిక్షణ పొందండి

-అన్ని మోడల్‌లు ప్రామాణికమైన [`tf.keras.Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model) కాబట్టి వాటిని [Keras]తో TensorFlowలో శిక్షణ పొందవచ్చు(https: //keras.io/) API. 🤗 ట్రాన్స్‌ఫార్మర్‌లు మీ డేటాసెట్‌ని సులభంగా `tf.data.Dataset`గా లోడ్ చేయడానికి [`~TFPreTrainedModel.prepare_tf_dataset`] పద్ధతిని అందజేస్తుంది కాబట్టి మీరు వెంటనే Keras' [`compile`](https://keras.io /api/models/model_training_apis/#compile-method) మరియు [`fit`](https://keras.io/api/models/model_training_apis/#fit-method) పద్ధతులు.
+అన్ని మోడల్‌లు ప్రామాణికమైన [`tf.keras.Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model) కాబట్టి వాటిని [Keras]తో TensorFlowలో శిక్షణ పొందవచ్చు(https: //keras.io/) API. 🤗 ట్రాన్స్‌ఫార్మర్‌లు మీ డేటాసెట్‌ని సులభంగా `tf.data.Dataset`గా లోడ్ చేయడానికి [`~TFPreTrainedModel.prepare_tf_dataset`] పద్ధతిని అందజేస్తుంది కాబట్టి మీరు వెంటనే Keras' [`compile`](https://keras.io/api/models/model_training_apis/#compile-method) మరియు [`fit`](https://keras.io/api/models/model_training_apis/#fit-method) పద్ధతులు.

 1. మీరు [`TFPreTrainedModel`] లేదా [`tf.keras.Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model)తో ప్రారంభిస్తారు:
   ```py
--- a/docs/source/zh/index.md
+++ b/docs/source/zh/index.md
@@ -111,11 +111,11 @@ rendered properly in your Markdown viewer.
 1. **[Funnel Transformer](model_doc/funnel)** (from CMU/Google Brain) released with the paper [Funnel-Transformer: Filtering out Sequential Redundancy for Efficient Language Processing](https://arxiv.org/abs/2006.03236) by Zihang Dai, Guokun Lai, Yiming Yang, Quoc V. Le.
 1. **[GIT](model_doc/git)** (from Microsoft Research) released with the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Jianfeng Wang, Zhengyuan Yang, Xiaowei Hu, Linjie Li, Kevin Lin, Zhe Gan, Zicheng Liu, Ce Liu, Lijuan Wang.
 1. **[GLPN](model_doc/glpn)** (from KAIST) released with the paper [Global-Local Path Networks for Monocular Depth Estimation with Vertical CutDepth](https://arxiv.org/abs/2201.07436) by Doyeon Kim, Woonghyun Ga, Pyungwhan Ahn, Donggyu Joo, Sehwan Chun, Junmo Kim.
-1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://blog.openai.com/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
+1. **[GPT](model_doc/openai-gpt)** (from OpenAI) released with the paper [Improving Language Understanding by Generative Pre-Training](https://openai.com/research/language-unsupervised/) by Alec Radford, Karthik Narasimhan, Tim Salimans and Ilya Sutskever.
 1. **[GPT Neo](model_doc/gpt_neo)** (from EleutherAI) released in the repository [EleutherAI/gpt-neo](https://github.com/EleutherAI/gpt-neo) by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
 1. **[GPT NeoX](model_doc/gpt_neox)** (from EleutherAI) released with the paper [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) by Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, Samuel Weinbach
 1. **[GPT NeoX Japanese](model_doc/gpt_neox_japanese)** (from ABEJA) released by Shinya Otani, Takayoshi Makabe, Anuj Arora, and Kyo Hattori.
-1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://blog.openai.com/better-language-models/) by Alec Radford*, Jeffrey Wu*, Rewon Child, David Luan, Dario Amodei** and Ilya Sutskever**.
+1. **[GPT-2](model_doc/gpt2)** (from OpenAI) released with the paper [Language Models are Unsupervised Multitask Learners](https://openai.com/research/better-language-models/) by Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei and Ilya Sutskever.
 1. **[GPT-J](model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
 1. **[GPT-Sw3](model_doc/gpt-sw3)** (from AI-Sweden) released with the paper [Lessons Learned from GPT-SW3: Building the First Large-Scale Generative Language Model for Swedish](http://www.lrec-conf.org/proceedings/lrec2022/pdf/2022.lrec-1.376.pdf) by Ariel Ekgren, Amaru Cuba Gyllensten, Evangelia Gogoulou, Alice Heiman, Severine Verlinden, Joey Öhman, Fredrik Carlsson, Magnus Sahlgren.
 1. **[GroupViT](model_doc/groupvit)** (from UCSD, NVIDIA) released with the paper [GroupViT: Semantic Segmentation Emerges from Text Supervision](https://arxiv.org/abs/2202.11094) by Jiarui Xu, Shalini De Mello, Sifei Liu, Wonmin Byeon, Thomas Breuel, Jan Kautz, Xiaolong Wang.