[Time series] Add patchtst (#27581)

* add distribution head to forecasting

* formatting

* Add generate function for forecasting

* Add generate function to prediction task

* formatting

* use argsort

* add past_observed_mask ordering

* fix arguments

* docs

* add back test_model_outputs_equivalence test

* formatting

* cleanup

* formatting

* use ACT2CLS

* formatting

* fix add_start_docstrings decorator

* add distribution head and generate function to regression task

add distribution head and generate function to regression task. Also made add PatchTSTForForecastingOutput,  PatchTSTForRegressionOutput.

* add distribution head and generate function to regression task

add distribution head and generate function to regression task. Also made add PatchTSTForForecastingOutput,  PatchTSTForRegressionOutput.

* fix typos

* add forecast_masking

* fixed tests

* use set_seed

* fix doc test

* formatting

* Update docs/source/en/model_doc/patchtst.md
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* better var names

* rename PatchTSTTranspose

* fix argument names and docs string

* remove compute_num_patches and unused class

* remove assert

* renamed to PatchTSTMasking

* use num_labels for classification

* use num_labels

* use default num_labels from super class

* move model_type after docstring

* renamed PatchTSTForMaskPretraining

* bs -> batch_size

* more review fixes

* use hidden_state

* rename encoder layer and block class

* remove commented seed_number

* edit docstring

* Add docstring

* formatting

* use past_observed_mask

* doc suggestion

* make fix-copies

* use Args:

* add docstring

* add docstring

* change some variable names and add PatchTST before some class names

* formatting

* fix argument types

* fix tests

* change x variable to patch_input

* format

* formatting

* fix-copies

* Update tests/models/patchtst/test_modeling_patchtst.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* move loss to forward

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* formatting

* fix a bug when pre_norm is set to True

* output_hidden_states is set to False as default

* set pre_norm=True as default

* format docstring

* format

* output_hidden_states is None by default

* add missing docs

* better var names

* docstring: remove default to False in output_hidden_states

* change labels name to target_values in regression task

* format

* fix tests

* change to forecast_mask_ratios and random_mask_ratio

* change mask names

* change future_values to target_values param in the prediction class

* remove nn.Sequential and make PatchTSTBatchNorm class

* black

* fix argument name for prediction

* add output_attentions option

* add output_attentions to PatchTSTEncoder

* formatting

* Add attention output option to all classes

* Remove PatchTSTEncoderBlock

* create PatchTSTEmbedding class

* use config in PatchTSTPatchify

* Use config in PatchTSTMasking class

* add channel_attn_weights

* Add PatchTSTScaler class

* add output_attentions arg to test function

* format

* Update doc with image patchtst.md

* fix-copies

* rename Forecast <-> Prediction

* change name of a few parameters to match with PatchTSMixer.

* Remove *ForForecasting class to match with other time series models.

* make style

* Remove PatchTSTForForecasting in the test

* remove PatchTSTForForecastingOutput class

* change test_forecast_head to test_prediction_head

* style

* fix docs

* fix tests

* change num_labels to num_targets

* Remove PatchTSTTranspose

* remove arguments in PatchTSTMeanScaler

* remove arguments in PatchTSTStdScaler

* add config as an argument to all the scaler classes

* reformat

* Add norm_eps for batchnorm and layernorm

* reformat.

* reformat

* edit docstring

* update docstring

* change variable name pooling to pooling_type

* fix output_hidden_states as tuple

* fix bug when calling PatchTSTBatchNorm

* change stride to patch_stride

* create PatchTSTPositionalEncoding class and restructure the PatchTSTEncoder

* formatting

* initialize scalers with configs

* edit output_hidden_states

* style

* fix forecast_mask_patches doc string

* doc improvements

* move summary to the start

* typo

* fix docstring

* turn off masking when using prediction, regression, classification

* return scaled output

* adjust output when using distribution head

* remove _num_patches function in the config

* get config.num_patches from patchifier init

* add output_attentions docstring, remove tuple in output_hidden_states

* change SamplePatchTSTPredictionOutput and SamplePatchTSTRegressionOutput to SamplePatchTSTOutput

* remove print("model_class: ", model_class)

* change encoder_attention_heads to num_attention_heads

* change norm to norm_layer

* change encoder_layers to num_hidden_layers

* change shared_embedding to share_embedding, shared_projection to share_projection

* add output_attentions

* more robust check of norm_type

* change dropout_path to path_dropout

* edit docstring

* remove positional_encoding function and add _init_pe in PatchTSTPositionalEncoding

* edit shape of cls_token and initialize it

* add a check on the num_input_channels.

* edit head_dim in the Prediction class to allow the use of cls_token

* remove some positional_encoding_type options, remove learn_pe arg, initalize pe

* change Exception to ValueError

* format

* norm_type is "batchnorm"

* make style

* change cls_token shape

* Change forecast_mask_patches to num_mask_patches. Remove forecast_mask_ratios.

* Bring PatchTSTClassificationHead on top of PatchTSTForClassification

* change encoder_ffn_dim to ffn_dim and edit the docstring.

* update variable names to match with the config

* add generation tests

* change num_mask_patches to num_forecast_mask_patches

* Add examples explaining the use of these models

* make style

* Revert "Revert "[time series] Add PatchTST (#25927)" (#27486)"

This reverts commit 78f6ed6c

.

* make style

* fix default std scaler's minimum_scale

* fix docstring

* close code blocks

* Update docs/source/en/model_doc/patchtst.md
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update tests/models/patchtst/test_modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/configuration_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* Update src/transformers/models/patchtst/modeling_patchtst.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

* fix tests

* add add_start_docstrings

* move examples to the forward's docstrings

* update prepare_batch

* update test

* fix test_prediction_head

* fix generation test

* use seed to create generator

* add output_hidden_states and config.num_patches

* add loc and scale args in PatchTSTForPredictionOutput

* edit outputs if if not return_dict

* use self.share_embedding to check instead checking type.

* remove seed

* make style

* seed is an optional int

* fix test

* generator device

* Fix assertTrue test

* swap order of items in outputs when return_dict=False.

* add mask_type and random_mask_ratio to unittest

* Update modeling_patchtst.py

* add add_start_docstrings for regression model

* make style

* update model path

* Edit the ValueError comment in forecast_masking

* update examples

* make style

* fix commented code

* update examples: remove config from from_pretrained call

* Edit example outputs

* Set default target_values to None

* remove config setting in regression example

* Update configuration_patchtst.py

* Update configuration_patchtst.py

* remove config from examples

* change default d_model and ffn_dim

* norm_eps default

* set has_attentions to Trye and define self.seq_length = self.num_patche

* update docstring

* change variable mask_input to do_mask_input

* fix blank space.

* change logger.debug to logger.warning.

* remove unused PATCHTST_INPUTS_DOCSTRING

* remove all_generative_model_classes

* set test_missing_keys=True

* remove undefined params in the docstring.

---------
Co-authored-by: nnguyen <nnguyen@us.ibm.com>
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Nam Nguyen <namctin@gmail.com>
Co-authored-by: Wesley Gifford <79663411+wgifford@users.noreply.github.com>
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>

README.md

@@ -440,6 +440,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (from Google AI) released with the paper [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) by Matthias Minderer, Alexey Gritsenko, Neil Houlsby.
+1. **[PatchTST](https://huggingface.co/docs/transformers/main/model_doc/patchtst)** (from IBM) released with the paper [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, and Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.

README_es.md

@@ -415,6 +415,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (from Google AI) released with the paper [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) by Matthias Minderer, Alexey Gritsenko, Neil Houlsby.
+1. **[PatchTST](https://huggingface.co/docs/transformers/main/model_doc/patchtst)** (from IBM) released with the paper [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, and Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.

README_hd.md

@@ -389,6 +389,7 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (Google AI से) साथ में कागज [विज़न ट्रांसफॉर्मर्स के साथ सिंपल ओपन-वोकैबुलरी ऑब्जेक्ट डिटेक्शन](https:/ /arxiv.org/abs/2205.06230) मैथियास मिंडरर, एलेक्सी ग्रिट्सेंको, ऑस्टिन स्टोन, मैक्सिम न्यूमैन, डिर्क वीसेनबोर्न, एलेक्सी डोसोवित्स्की, अरविंद महेंद्रन, अनुराग अर्नब, मुस्तफा देहघानी, ज़ुओरन शेन, जिओ वांग, ज़ियाओहुआ झाई, थॉमस किफ़, और नील हॉल्सबी द्वारा पोस्ट किया गया।
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (Google AI से) Matthias Minderer, Alexey Gritsenko, Neil Houlsby. द्वाराअनुसंधान पत्र [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) के साथ जारी किया गया
+1. **[PatchTST](https://huggingface.co/docs/transformers/main/model_doc/patchtst)** (IBM से) Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. द्वाराअनुसंधान पत्र [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) के साथ जारी किया गया
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (Google की ओर से) साथ में दिया गया पेपर [लंबे इनपुट सारांश के लिए ट्रांसफ़ॉर्मरों को बेहतर तरीके से एक्सटेंड करना](https://arxiv .org/abs/2208.04347) जेसन फांग, याओ झाओ, पीटर जे लियू द्वारा।
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (दीपमाइंड से) साथ में पेपर [पर्सीवर आईओ: संरचित इनपुट और आउटपुट के लिए एक सामान्य वास्तुकला] (https://arxiv.org/abs/2107.14795) एंड्रयू जेगल, सेबेस्टियन बोरग्यूड, जीन-बैप्टिस्ट अलायराक, कार्ल डोर्श, कैटलिन इओनेस्कु, डेविड द्वारा डिंग, स्कंद कोप्पुला, डैनियल ज़ोरान, एंड्रयू ब्रॉक, इवान शेलहैमर, ओलिवियर हेनाफ, मैथ्यू एम। बोट्विनिक, एंड्रयू ज़िसरमैन, ओरिओल विनियल्स, जोआओ कैरेरा द्वारा पोस्ट किया गया।

README_ja.md

@@ -449,6 +449,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (Meta AI から) Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al から公開された研究論文: [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068)
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (Google AI から) Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby から公開された研究論文: [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230)
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (Google AI から) Matthias Minderer, Alexey Gritsenko, Neil Houlsby. から公開された研究論文 [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683)
+1. **[PatchTST](https://huggingface.co/docs/transformers/main/model_doc/patchtst)** (IBM から) Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam. から公開された研究論文 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf)
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (Google から) Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu から公開された研究論文: [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777)
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (Google から) Jason Phang, Yao Zhao, and Peter J. Liu から公開された研究論文: [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347)
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (Deepmind から) Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira から公開された研究論文: [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795)

README_ko.md

@@ -364,6 +364,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (Meta AI 에서) Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al 의 [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) 논문과 함께 발표했습니다.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (Google AI 에서) Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby 의 [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) 논문과 함께 발표했습니다.
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (Google AI 에서 제공)은 Matthias Minderer, Alexey Gritsenko, Neil Houlsby.의 [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683)논문과 함께 발표했습니다.
+1. **[PatchTST](https://huggingface.co/docs/transformers/main/model_doc/patchtst)** (IBM 에서 제공)은 Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.의 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf)논문과 함께 발표했습니다.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (Google 에서) Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu 의 [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) 논문과 함께 발표했습니다.
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (Google 에서) Jason Phang, Yao Zhao, Peter J. Liu 의 [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) 논문과 함께 발표했습니다.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (Deepmind 에서) Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira 의 [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) 논문과 함께 발표했습니다.

README_zh-hans.md

@@ -388,6 +388,7 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (来自 Meta AI) 伴随论文 [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) 由 Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al 发布。
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (来自 Google AI) 伴随论文 [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) 由 Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby 发布。
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (来自 Google AI) 伴随论文 [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) 由 Matthias Minderer, Alexey Gritsenko, Neil Houlsby 发布。
+1. **[PatchTST](https://huggingface.co/docs/transformers/main/model_doc/patchtst)** (来自 IBM) 伴随论文 [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) 由 Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam 发布。
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (来自 Google) 伴随论文 [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) 由 Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu 发布。
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (来自 Google) 伴随论文 [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) 由 Jason Phang, Yao Zhao, Peter J. Liu 发布。
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (来自 Deepmind) 伴随论文 [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) 由 Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira 发布。

README_zh-hant.md

@@ -400,6 +400,7 @@ conda install -c huggingface transformers
 1. **[OPT](https://huggingface.co/docs/transformers/master/model_doc/opt)** (from Meta AI) released with the paper [OPT: Open Pre-trained Transformer Language Models](https://arxiv.org/abs/2205.01068) by Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen et al.
 1. **[OWL-ViT](https://huggingface.co/docs/transformers/model_doc/owlvit)** (from Google AI) released with the paper [Simple Open-Vocabulary Object Detection with Vision Transformers](https://arxiv.org/abs/2205.06230) by Matthias Minderer, Alexey Gritsenko, Austin Stone, Maxim Neumann, Dirk Weissenborn, Alexey Dosovitskiy, Aravindh Mahendran, Anurag Arnab, Mostafa Dehghani, Zhuoran Shen, Xiao Wang, Xiaohua Zhai, Thomas Kipf, and Neil Houlsby.
 1. **[OWLv2](https://huggingface.co/docs/transformers/model_doc/owlv2)** (from Google AI) released with the paper [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683) by Matthias Minderer, Alexey Gritsenko, Neil Houlsby.
+1. **[PatchTST](https://huggingface.co/docs/transformers/main/model_doc/patchtst)** (from IBM) released with the paper [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/pdf/2211.14730.pdf) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, Jayant Kalagnanam.
 1. **[Pegasus](https://huggingface.co/docs/transformers/model_doc/pegasus)** (from Google) released with the paper [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://arxiv.org/abs/1912.08777) by Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu.
 1. **[PEGASUS-X](https://huggingface.co/docs/transformers/model_doc/pegasus_x)** (from Google) released with the paper [Investigating Efficiently Extending Transformers for Long Input Summarization](https://arxiv.org/abs/2208.04347) by Jason Phang, Yao Zhao, Peter J. Liu.
 1. **[Perceiver IO](https://huggingface.co/docs/transformers/model_doc/perceiver)** (from Deepmind) released with the paper [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://arxiv.org/abs/2107.14795) by Andrew Jaegle, Sebastian Borgeaud, Jean-Baptiste Alayrac, Carl Doersch, Catalin Ionescu, David Ding, Skanda Koppula, Daniel Zoran, Andrew Brock, Evan Shelhamer, Olivier Hénaff, Matthew M. Botvinick, Andrew Zisserman, Oriol Vinyals, João Carreira.

docs/source/en/_toctree.yml

@@ -755,6 +755,8 @@
         title: Autoformer
       - local: model_doc/informer
         title: Informer
+      - local: model_doc/patchtst
+        title: PatchTST
       - local: model_doc/time_series_transformer
         title: Time Series Transformer
       title: Time series models

docs/source/en/index.md

@@ -214,6 +214,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                           [OPT](model_doc/opt)                           |       ✅        |         ✅         |      ✅      |
 |                       [OWL-ViT](model_doc/owlvit)                        |       ✅        |         ❌         |      ❌      |
 |                         [OWLv2](model_doc/owlv2)                         |       ✅        |         ❌         |      ❌      |
+|                      [PatchTST](model_doc/patchtst)                      |       ✅        |         ❌         |      ❌      |
 |                       [Pegasus](model_doc/pegasus)                       |       ✅        |         ✅         |      ✅      |
 |                     [PEGASUS-X](model_doc/pegasus_x)                     |       ✅        |         ❌         |      ❌      |
 |                     [Perceiver](model_doc/perceiver)                     |       ✅        |         ❌         |      ❌      |

docs/source/en/model_doc/patchtst.md

+<!--Copyright 2023 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+
+⚠️ Note that this file is in Markdown but contain specific syntax for our doc-builder (similar to MDX) that may not be
+rendered properly in your Markdown viewer.
+
+-->
+
+# PatchTST
+
+## Overview
+
+The PatchTST model was proposed in [A Time Series is Worth 64 Words: Long-term Forecasting with Transformers](https://arxiv.org/abs/2211.14730) by Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong and Jayant Kalagnanam.
+
+At a high level the model vectorizes time series into patches of a given size and encodes the resulting sequence of vectors via a Transformer that then outputs the prediction length forecast via an appropriate head. The model is illustrated in the following figure:
+
+![model](https://github.com/namctin/transformers/assets/8100/150af169-29de-419a-8d98-eb78251c21fa)
+
+The abstract from the paper is the following:
+
+*We propose an efficient design of Transformer-based models for multivariate time series forecasting and self-supervised representation learning. It is based on two key components: (i) segmentation of time series into subseries-level patches which are served as input tokens to Transformer; (ii) channel-independence where each channel contains a single univariate time series that shares the same embedding and Transformer weights across all the series. Patching design naturally has three-fold benefit: local semantic information is retained in the embedding; computation and memory usage of the attention maps are quadratically reduced given the same look-back window; and the model can attend longer history. Our channel-independent patch time series Transformer (PatchTST) can improve the long-term forecasting accuracy significantly when compared with that of SOTA Transformer-based models. We also apply our model to self-supervised pre-training tasks and attain excellent fine-tuning performance, which outperforms supervised training on large datasets. Transferring of masked pre-trained representation on one dataset to others also produces SOTA forecasting accuracy.*
+
+This model was contributed by [namctin](https://huggingface.co/namctin), [gsinthong](https://huggingface.co/gsinthong), [diepi](https://huggingface.co/diepi), [vijaye12](https://huggingface.co/vijaye12), [wmgifford](https://huggingface.co/wmgifford), and [kashif](https://huggingface.co/kashif). The original code can be found [here](https://github.com/yuqinie98/PatchTST).
+
+## Usage tips
+
+The model can also be used for time series classification and time series regression. See the respective [`PatchTSTForClassification`] and [`PatchTSTForRegression`] classes.
+
+
+## PatchTSTConfig
+
+[[autodoc]] PatchTSTConfig
+
+## PatchTSTModel
+
+[[autodoc]] PatchTSTModel
+    - forward
+
+## PatchTSTForPrediction
+
+[[autodoc]] PatchTSTForPrediction
+    - forward
+
+## PatchTSTForClassification
+
+[[autodoc]] PatchTSTForClassification
+    - forward
+
+## PatchTSTForPretraining
+
+[[autodoc]] PatchTSTForPretraining
+    - forward
+
+## PatchTSTForRegression
+
+[[autodoc]] PatchTSTForRegression
+    - forward

src/transformers/__init__.py

@@ -499,6 +499,7 @@ _import_structure = {
         "OwlViTTextConfig",
         "OwlViTVisionConfig",
     ],
+    "models.patchtst": ["PATCHTST_PRETRAINED_CONFIG_ARCHIVE_MAP", "PatchTSTConfig"],
     "models.pegasus": ["PEGASUS_PRETRAINED_CONFIG_ARCHIVE_MAP", "PegasusConfig", "PegasusTokenizer"],
     "models.pegasus_x": ["PEGASUS_X_PRETRAINED_CONFIG_ARCHIVE_MAP", "PegasusXConfig"],
     "models.perceiver": ["PERCEIVER_PRETRAINED_CONFIG_ARCHIVE_MAP", "PerceiverConfig", "PerceiverTokenizer"],
@@ -1178,6 +1179,8 @@ else:
             "MODEL_FOR_TEXT_ENCODING_MAPPING",
             "MODEL_FOR_TEXT_TO_SPECTROGRAM_MAPPING",
             "MODEL_FOR_TEXT_TO_WAVEFORM_MAPPING",
+            "MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING",
+            "MODEL_FOR_TIME_SERIES_REGRESSION_MAPPING",
             "MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING",
             "MODEL_FOR_UNIVERSAL_SEGMENTATION_MAPPING",
             "MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING",
@@ -2508,6 +2511,17 @@ else:
             "OwlViTVisionModel",
         ]
     )
+    _import_structure["models.patchtst"].extend(
+        [
+            "PATCHTST_PRETRAINED_MODEL_ARCHIVE_LIST",
+            "PatchTSTForClassification",
+            "PatchTSTForPrediction",
+            "PatchTSTForPretraining",
+            "PatchTSTForRegression",
+            "PatchTSTModel",
+            "PatchTSTPreTrainedModel",
+        ]
+    )
     _import_structure["models.pegasus"].extend(
         ["PegasusForCausalLM", "PegasusForConditionalGeneration", "PegasusModel", "PegasusPreTrainedModel"]
     )
@@ -4729,6 +4743,7 @@ if TYPE_CHECKING:
         OwlViTTextConfig,
         OwlViTVisionConfig,
     )
+    from .models.patchtst import PATCHTST_PRETRAINED_CONFIG_ARCHIVE_MAP, PatchTSTConfig
     from .models.pegasus import PEGASUS_PRETRAINED_CONFIG_ARCHIVE_MAP, PegasusConfig, PegasusTokenizer
     from .models.pegasus_x import PEGASUS_X_PRETRAINED_CONFIG_ARCHIVE_MAP, PegasusXConfig
     from .models.perceiver import PERCEIVER_PRETRAINED_CONFIG_ARCHIVE_MAP, PerceiverConfig, PerceiverTokenizer
@@ -5340,6 +5355,8 @@ if TYPE_CHECKING:
             MODEL_FOR_TEXT_ENCODING_MAPPING,
             MODEL_FOR_TEXT_TO_SPECTROGRAM_MAPPING,
             MODEL_FOR_TEXT_TO_WAVEFORM_MAPPING,
+            MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING,
+            MODEL_FOR_TIME_SERIES_REGRESSION_MAPPING,
             MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING,
             MODEL_FOR_UNIVERSAL_SEGMENTATION_MAPPING,
             MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING,
@@ -6434,6 +6451,15 @@ if TYPE_CHECKING:
             OwlViTTextModel,
             OwlViTVisionModel,
         )
+        from .models.patchtst import (
+            PATCHTST_PRETRAINED_MODEL_ARCHIVE_LIST,
+            PatchTSTForClassification,
+            PatchTSTForPrediction,
+            PatchTSTForPretraining,
+            PatchTSTForRegression,
+            PatchTSTModel,
+            PatchTSTPreTrainedModel,
+        )
         from .models.pegasus import (
             PegasusForCausalLM,
             PegasusForConditionalGeneration,

src/transformers/models/__init__.py

@@ -158,6 +158,7 @@ from . import (
     opt,
     owlv2,
     owlvit,
+    patchtst,
     pegasus,
     pegasus_x,
     perceiver,

src/transformers/models/auto/__init__.py

@@ -77,6 +77,8 @@ else:
         "MODEL_WITH_LM_HEAD_MAPPING",
         "MODEL_FOR_ZERO_SHOT_IMAGE_CLASSIFICATION_MAPPING",
         "MODEL_FOR_ZERO_SHOT_OBJECT_DETECTION_MAPPING",
+        "MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING",
+        "MODEL_FOR_TIME_SERIES_REGRESSION_MAPPING",
         "AutoModel",
         "AutoBackbone",
         "AutoModelForAudioClassification",
@@ -250,6 +252,8 @@ if TYPE_CHECKING:
             MODEL_FOR_TEXT_ENCODING_MAPPING,
             MODEL_FOR_TEXT_TO_SPECTROGRAM_MAPPING,
             MODEL_FOR_TEXT_TO_WAVEFORM_MAPPING,
+            MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING,
+            MODEL_FOR_TIME_SERIES_REGRESSION_MAPPING,
             MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING,
             MODEL_FOR_UNIVERSAL_SEGMENTATION_MAPPING,
             MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING,

src/transformers/models/auto/configuration_auto.py

@@ -164,6 +164,7 @@ CONFIG_MAPPING_NAMES = OrderedDict(
         ("opt", "OPTConfig"),
         ("owlv2", "Owlv2Config"),
         ("owlvit", "OwlViTConfig"),
+        ("patchtst", "PatchTSTConfig"),
         ("pegasus", "PegasusConfig"),
         ("pegasus_x", "PegasusXConfig"),
         ("perceiver", "PerceiverConfig"),
@@ -378,6 +379,7 @@ CONFIG_ARCHIVE_MAP_MAPPING_NAMES = OrderedDict(
         ("opt", "OPT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("owlv2", "OWLV2_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("owlvit", "OWLVIT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
+        ("patchtst", "PATCHTST_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("pegasus", "PEGASUS_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("pegasus_x", "PEGASUS_X_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("perceiver", "PERCEIVER_PRETRAINED_CONFIG_ARCHIVE_MAP"),
@@ -612,6 +614,7 @@ MODEL_NAMES_MAPPING = OrderedDict(
         ("opt", "OPT"),
         ("owlv2", "OWLv2"),
         ("owlvit", "OWL-ViT"),
+        ("patchtst", "PatchTST"),
         ("pegasus", "Pegasus"),
         ("pegasus_x", "PEGASUS-X"),
         ("perceiver", "Perceiver"),

src/transformers/models/auto/modeling_auto.py

@@ -157,6 +157,7 @@ MODEL_MAPPING_NAMES = OrderedDict(
         ("opt", "OPTModel"),
         ("owlv2", "Owlv2Model"),
         ("owlvit", "OwlViTModel"),
+        ("patchtst", "PatchTSTModel"),
         ("pegasus", "PegasusModel"),
         ("pegasus_x", "PegasusXModel"),
         ("perceiver", "PerceiverModel"),
@@ -1133,6 +1134,18 @@ MODEL_FOR_TEXT_ENCODING_MAPPING_NAMES = OrderedDict(
     ]
 )
 
+MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING_NAMES = OrderedDict(
+    [
+        ("patchtst", "PatchTSTForClassification"),
+    ]
+)
+
+MODEL_FOR_TIME_SERIES_REGRESSION_MAPPING_NAMES = OrderedDict(
+    [
+        ("patchtst", "PatchTSTForRegression"),
+    ]
+)
+
 MODEL_FOR_IMAGE_TO_IMAGE_MAPPING_NAMES = OrderedDict(
     [
         ("swin2sr", "Swin2SRForImageSuperResolution"),
@@ -1224,6 +1237,14 @@ MODEL_FOR_MASK_GENERATION_MAPPING = _LazyAutoMapping(CONFIG_MAPPING_NAMES, MODEL
 
 MODEL_FOR_TEXT_ENCODING_MAPPING = _LazyAutoMapping(CONFIG_MAPPING_NAMES, MODEL_FOR_TEXT_ENCODING_MAPPING_NAMES)
 
+MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING = _LazyAutoMapping(
+    CONFIG_MAPPING_NAMES, MODEL_FOR_TIME_SERIES_CLASSIFICATION_MAPPING_NAMES
+)
+
+MODEL_FOR_TIME_SERIES_REGRESSION_MAPPING = _LazyAutoMapping(
+    CONFIG_MAPPING_NAMES, MODEL_FOR_TIME_SERIES_REGRESSION_MAPPING_NAMES
+)
+
 MODEL_FOR_IMAGE_TO_IMAGE_MAPPING = _LazyAutoMapping(CONFIG_MAPPING_NAMES, MODEL_FOR_IMAGE_TO_IMAGE_MAPPING_NAMES)
 
 

src/transformers/models/autoformer/modeling_autoformer.py

@@ -208,71 +208,70 @@ class AutoformerFeatureEmbedder(nn.Module):
         )
 
 
-# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesStdScaler with TimeSeries->Autoformer
+# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesStdScaler with TimeSeriesTransformer->Autoformer,TimeSeries->Autoformer
 class AutoformerStdScaler(nn.Module):
     """
-    Standardize features by calculating the mean and scaling along some given dimension `dim`, and then normalizes it
-    by subtracting from the mean and dividing by the standard deviation.
-
-    Args:
-        dim (`int`):
-            Dimension along which to calculate the mean and standard deviation.
-        keepdim (`bool`, *optional*, defaults to `False`):
-            Controls whether to retain dimension `dim` (of length 1) in the scale tensor, or suppress it.
-        minimum_scale (`float`, *optional*, defaults to 1e-5):
-            Default scale that is used for elements that are constantly zero along dimension `dim`.
+    Standardize features by calculating the mean and scaling along the first dimension, and then normalizes it by
+    subtracting from the mean and dividing by the standard deviation.
     """
 
-    def __init__(self, dim: int, keepdim: bool = False, minimum_scale: float = 1e-5):
+    def __init__(self, config: AutoformerConfig):
         super().__init__()
-        if not dim > 0:
-            raise ValueError("Cannot compute scale along dim = 0 (batch dimension), please provide dim > 0")
-        self.dim = dim
-        self.keepdim = keepdim
-        self.minimum_scale = minimum_scale
+        self.dim = config.scaling_dim if hasattr(config, "scaling_dim") else 1
+        self.keepdim = config.keepdim if hasattr(config, "keepdim") else True
+        self.minimum_scale = config.minimum_scale if hasattr(config, "minimum_scale") else 1e-5
 
-    @torch.no_grad()
-    def forward(self, data: torch.Tensor, weights: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
-        denominator = weights.sum(self.dim, keepdim=self.keepdim)
+    def forward(
+        self, data: torch.Tensor, observed_indicator: torch.Tensor
+    ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+        """
+        Parameters:
+            data (`torch.Tensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                input for Batch norm calculation
+            observed_indicator (`torch.BoolTensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                Calculating the scale on the observed indicator.
+        Returns:
+            tuple of `torch.Tensor` of shapes
+                (`(batch_size, sequence_length, num_input_channels)`,`(batch_size, 1, num_input_channels)`,
+                `(batch_size, 1, num_input_channels)`)
+        """
+        denominator = observed_indicator.sum(self.dim, keepdim=self.keepdim)
         denominator = denominator.clamp_min(1.0)
-        loc = (data * weights).sum(self.dim, keepdim=self.keepdim) / denominator
+        loc = (data * observed_indicator).sum(self.dim, keepdim=self.keepdim) / denominator
 
-        variance = (((data - loc) * weights) ** 2).sum(self.dim, keepdim=self.keepdim) / denominator
+        variance = (((data - loc) * observed_indicator) ** 2).sum(self.dim, keepdim=self.keepdim) / denominator
         scale = torch.sqrt(variance + self.minimum_scale)
         return (data - loc) / scale, loc, scale
 
 
-# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesMeanScaler with TimeSeries->Autoformer
+# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesMeanScaler with TimeSeriesTransformer->Autoformer,TimeSeries->Autoformer
 class AutoformerMeanScaler(nn.Module):
     """
-    Computes a scaling factor as the weighted average absolute value along dimension `dim`, and scales the data
+    Computes a scaling factor as the weighted average absolute value along the first dimension, and scales the data
     accordingly.
-
-    Args:
-        dim (`int`):
-            Dimension along which to compute the scale.
-        keepdim (`bool`, *optional*, defaults to `False`):
-            Controls whether to retain dimension `dim` (of length 1) in the scale tensor, or suppress it.
-        default_scale (`float`, *optional*, defaults to `None`):
-            Default scale that is used for elements that are constantly zero. If `None`, we use the scale of the batch.
-        minimum_scale (`float`, *optional*, defaults to 1e-10):
-            Default minimum possible scale that is used for any item.
     """
 
-    def __init__(
-        self, dim: int = -1, keepdim: bool = True, default_scale: Optional[float] = None, minimum_scale: float = 1e-10
-    ):
+    def __init__(self, config: AutoformerConfig):
         super().__init__()
-        self.dim = dim
-        self.keepdim = keepdim
-        self.minimum_scale = minimum_scale
-        self.default_scale = default_scale
+        self.dim = config.scaling_dim if hasattr(config, "scaling_dim") else 1
+        self.keepdim = config.keepdim if hasattr(config, "keepdim") else True
+        self.minimum_scale = config.minimum_scale if hasattr(config, "minimum_scale") else 1e-10
+        self.default_scale = config.default_scale if hasattr(config, "default_scale") else None
 
-    @torch.no_grad()
     def forward(
         self, data: torch.Tensor, observed_indicator: torch.Tensor
     ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
-        # shape: (N, [C], T=1)
+        """
+        Parameters:
+            data (`torch.Tensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                input for Batch norm calculation
+            observed_indicator (`torch.BoolTensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                Calculating the scale on the observed indicator.
+        Returns:
+            tuple of `torch.Tensor` of shapes
+                (`(batch_size, sequence_length, num_input_channels)`,`(batch_size, 1, num_input_channels)`,
+                `(batch_size, 1, num_input_channels)`)
+        """
         ts_sum = (data * observed_indicator).abs().sum(self.dim, keepdim=True)
         num_observed = observed_indicator.sum(self.dim, keepdim=True)
 
@@ -300,26 +299,29 @@ class AutoformerMeanScaler(nn.Module):
         return scaled_data, torch.zeros_like(scale), scale
 
 
-# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesNOPScaler with TimeSeries->Autoformer
+# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesNOPScaler with TimeSeriesTransformer->Autoformer,TimeSeries->Autoformer
 class AutoformerNOPScaler(nn.Module):
     """
-    Assigns a scaling factor equal to 1 along dimension `dim`, and therefore applies no scaling to the input data.
-
-    Args:
-        dim (`int`):
-            Dimension along which to compute the scale.
-        keepdim (`bool`, *optional*, defaults to `False`):
-            Controls whether to retain dimension `dim` (of length 1) in the scale tensor, or suppress it.
+    Assigns a scaling factor equal to 1 along the first dimension, and therefore applies no scaling to the input data.
     """
 
-    def __init__(self, dim: int, keepdim: bool = False):
+    def __init__(self, config: AutoformerConfig):
         super().__init__()
-        self.dim = dim
-        self.keepdim = keepdim
+        self.dim = config.scaling_dim if hasattr(config, "scaling_dim") else 1
+        self.keepdim = config.keepdim if hasattr(config, "keepdim") else True
 
     def forward(
-        self, data: torch.Tensor, observed_indicator: torch.Tensor
+        self, data: torch.Tensor, observed_indicator: torch.Tensor = None
     ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+        """
+        Parameters:
+            data (`torch.Tensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                input for Batch norm calculation
+        Returns:
+            tuple of `torch.Tensor` of shapes
+                (`(batch_size, sequence_length, num_input_channels)`,`(batch_size, 1, num_input_channels)`,
+                `(batch_size, 1, num_input_channels)`)
+        """
         scale = torch.ones_like(data, requires_grad=False).mean(dim=self.dim, keepdim=self.keepdim)
         loc = torch.zeros_like(data, requires_grad=False).mean(dim=self.dim, keepdim=self.keepdim)
         return data, loc, scale
@@ -1433,11 +1435,11 @@ class AutoformerModel(AutoformerPreTrainedModel):
         super().__init__(config)
 
         if config.scaling == "mean" or config.scaling is True:
-            self.scaler = AutoformerMeanScaler(dim=1, keepdim=True)
+            self.scaler = AutoformerMeanScaler(config)
         elif config.scaling == "std":
-            self.scaler = AutoformerStdScaler(dim=1, keepdim=True)
+            self.scaler = AutoformerStdScaler(config)
         else:
-            self.scaler = AutoformerNOPScaler(dim=1, keepdim=True)
+            self.scaler = AutoformerNOPScaler(config)
 
         if config.num_static_categorical_features > 0:
             self.embedder = AutoformerFeatureEmbedder(

src/transformers/models/informer/modeling_informer.py

@@ -81,71 +81,70 @@ class InformerFeatureEmbedder(nn.Module):
         )
 
 
-# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesStdScaler with TimeSeries->Informer
+# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesStdScaler with TimeSeriesTransformer->Informer,TimeSeries->Informer
 class InformerStdScaler(nn.Module):
     """
-    Standardize features by calculating the mean and scaling along some given dimension `dim`, and then normalizes it
-    by subtracting from the mean and dividing by the standard deviation.
-
-    Args:
-        dim (`int`):
-            Dimension along which to calculate the mean and standard deviation.
-        keepdim (`bool`, *optional*, defaults to `False`):
-            Controls whether to retain dimension `dim` (of length 1) in the scale tensor, or suppress it.
-        minimum_scale (`float`, *optional*, defaults to 1e-5):
-            Default scale that is used for elements that are constantly zero along dimension `dim`.
+    Standardize features by calculating the mean and scaling along the first dimension, and then normalizes it by
+    subtracting from the mean and dividing by the standard deviation.
     """
 
-    def __init__(self, dim: int, keepdim: bool = False, minimum_scale: float = 1e-5):
+    def __init__(self, config: InformerConfig):
         super().__init__()
-        if not dim > 0:
-            raise ValueError("Cannot compute scale along dim = 0 (batch dimension), please provide dim > 0")
-        self.dim = dim
-        self.keepdim = keepdim
-        self.minimum_scale = minimum_scale
+        self.dim = config.scaling_dim if hasattr(config, "scaling_dim") else 1
+        self.keepdim = config.keepdim if hasattr(config, "keepdim") else True
+        self.minimum_scale = config.minimum_scale if hasattr(config, "minimum_scale") else 1e-5
 
-    @torch.no_grad()
-    def forward(self, data: torch.Tensor, weights: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
-        denominator = weights.sum(self.dim, keepdim=self.keepdim)
+    def forward(
+        self, data: torch.Tensor, observed_indicator: torch.Tensor
+    ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+        """
+        Parameters:
+            data (`torch.Tensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                input for Batch norm calculation
+            observed_indicator (`torch.BoolTensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                Calculating the scale on the observed indicator.
+        Returns:
+            tuple of `torch.Tensor` of shapes
+                (`(batch_size, sequence_length, num_input_channels)`,`(batch_size, 1, num_input_channels)`,
+                `(batch_size, 1, num_input_channels)`)
+        """
+        denominator = observed_indicator.sum(self.dim, keepdim=self.keepdim)
         denominator = denominator.clamp_min(1.0)
-        loc = (data * weights).sum(self.dim, keepdim=self.keepdim) / denominator
+        loc = (data * observed_indicator).sum(self.dim, keepdim=self.keepdim) / denominator
 
-        variance = (((data - loc) * weights) ** 2).sum(self.dim, keepdim=self.keepdim) / denominator
+        variance = (((data - loc) * observed_indicator) ** 2).sum(self.dim, keepdim=self.keepdim) / denominator
         scale = torch.sqrt(variance + self.minimum_scale)
         return (data - loc) / scale, loc, scale
 
 
-# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesMeanScaler with TimeSeries->Informer
+# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesMeanScaler with TimeSeriesTransformer->Informer,TimeSeries->Informer
 class InformerMeanScaler(nn.Module):
     """
-    Computes a scaling factor as the weighted average absolute value along dimension `dim`, and scales the data
+    Computes a scaling factor as the weighted average absolute value along the first dimension, and scales the data
     accordingly.
-
-    Args:
-        dim (`int`):
-            Dimension along which to compute the scale.
-        keepdim (`bool`, *optional*, defaults to `False`):
-            Controls whether to retain dimension `dim` (of length 1) in the scale tensor, or suppress it.
-        default_scale (`float`, *optional*, defaults to `None`):
-            Default scale that is used for elements that are constantly zero. If `None`, we use the scale of the batch.
-        minimum_scale (`float`, *optional*, defaults to 1e-10):
-            Default minimum possible scale that is used for any item.
     """
 
-    def __init__(
-        self, dim: int = -1, keepdim: bool = True, default_scale: Optional[float] = None, minimum_scale: float = 1e-10
-    ):
+    def __init__(self, config: InformerConfig):
         super().__init__()
-        self.dim = dim
-        self.keepdim = keepdim
-        self.minimum_scale = minimum_scale
-        self.default_scale = default_scale
+        self.dim = config.scaling_dim if hasattr(config, "scaling_dim") else 1
+        self.keepdim = config.keepdim if hasattr(config, "keepdim") else True
+        self.minimum_scale = config.minimum_scale if hasattr(config, "minimum_scale") else 1e-10
+        self.default_scale = config.default_scale if hasattr(config, "default_scale") else None
 
-    @torch.no_grad()
     def forward(
         self, data: torch.Tensor, observed_indicator: torch.Tensor
     ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
-        # shape: (N, [C], T=1)
+        """
+        Parameters:
+            data (`torch.Tensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                input for Batch norm calculation
+            observed_indicator (`torch.BoolTensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                Calculating the scale on the observed indicator.
+        Returns:
+            tuple of `torch.Tensor` of shapes
+                (`(batch_size, sequence_length, num_input_channels)`,`(batch_size, 1, num_input_channels)`,
+                `(batch_size, 1, num_input_channels)`)
+        """
         ts_sum = (data * observed_indicator).abs().sum(self.dim, keepdim=True)
         num_observed = observed_indicator.sum(self.dim, keepdim=True)
 
@@ -173,26 +172,29 @@ class InformerMeanScaler(nn.Module):
         return scaled_data, torch.zeros_like(scale), scale
 
 
-# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesNOPScaler with TimeSeries->Informer
+# Copied from transformers.models.time_series_transformer.modeling_time_series_transformer.TimeSeriesNOPScaler with TimeSeriesTransformer->Informer,TimeSeries->Informer
 class InformerNOPScaler(nn.Module):
     """
-    Assigns a scaling factor equal to 1 along dimension `dim`, and therefore applies no scaling to the input data.
-
-    Args:
-        dim (`int`):
-            Dimension along which to compute the scale.
-        keepdim (`bool`, *optional*, defaults to `False`):
-            Controls whether to retain dimension `dim` (of length 1) in the scale tensor, or suppress it.
+    Assigns a scaling factor equal to 1 along the first dimension, and therefore applies no scaling to the input data.
     """
 
-    def __init__(self, dim: int, keepdim: bool = False):
+    def __init__(self, config: InformerConfig):
         super().__init__()
-        self.dim = dim
-        self.keepdim = keepdim
+        self.dim = config.scaling_dim if hasattr(config, "scaling_dim") else 1
+        self.keepdim = config.keepdim if hasattr(config, "keepdim") else True
 
     def forward(
-        self, data: torch.Tensor, observed_indicator: torch.Tensor
+        self, data: torch.Tensor, observed_indicator: torch.Tensor = None
     ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+        """
+        Parameters:
+            data (`torch.Tensor` of shape `(batch_size, sequence_length, num_input_channels)`):
+                input for Batch norm calculation
+        Returns:
+            tuple of `torch.Tensor` of shapes
+                (`(batch_size, sequence_length, num_input_channels)`,`(batch_size, 1, num_input_channels)`,
+                `(batch_size, 1, num_input_channels)`)
+        """
         scale = torch.ones_like(data, requires_grad=False).mean(dim=self.dim, keepdim=self.keepdim)
         loc = torch.zeros_like(data, requires_grad=False).mean(dim=self.dim, keepdim=self.keepdim)
         return data, loc, scale
@@ -1447,11 +1449,11 @@ class InformerModel(InformerPreTrainedModel):
         super().__init__(config)
 
         if config.scaling == "mean" or config.scaling is True:
-            self.scaler = InformerMeanScaler(dim=1, keepdim=True)
+            self.scaler = InformerMeanScaler(config)
         elif config.scaling == "std":
-            self.scaler = InformerStdScaler(dim=1, keepdim=True)
+            self.scaler = InformerStdScaler(config)
         else:
-            self.scaler = InformerNOPScaler(dim=1, keepdim=True)
+            self.scaler = InformerNOPScaler(config)
 
         if config.num_static_categorical_features > 0:
             self.embedder = InformerFeatureEmbedder(

src/transformers/models/patchtst/__init__.py

+# Copyright 2023 The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from typing import TYPE_CHECKING
+
+# rely on isort to merge the imports
+from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available
+
+
+_import_structure = {
+    "configuration_patchtst": [
+        "PATCHTST_PRETRAINED_CONFIG_ARCHIVE_MAP",
+        "PatchTSTConfig",
+    ],
+}
+
+try:
+    if not is_torch_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["modeling_patchtst"] = [
+        "PATCHTST_PRETRAINED_MODEL_ARCHIVE_LIST",
+        "PatchTSTModel",
+        "PatchTSTPreTrainedModel",
+        "PatchTSTForPrediction",
+        "PatchTSTForPretraining",
+        "PatchTSTForRegression",
+        "PatchTSTForClassification",
+    ]
+
+
+if TYPE_CHECKING:
+    from .configuration_patchtst import PATCHTST_PRETRAINED_CONFIG_ARCHIVE_MAP, PatchTSTConfig
+
+    try:
+        if not is_torch_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .modeling_patchtst import (
+            PATCHTST_PRETRAINED_MODEL_ARCHIVE_LIST,
+            PatchTSTForClassification,
+            PatchTSTForPrediction,
+            PatchTSTForPretraining,
+            PatchTSTForRegression,
+            PatchTSTModel,
+            PatchTSTPreTrainedModel,
+        )
+
+else:
+    import sys
+
+    sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)

src/transformers/models/patchtst/configuration_patchtst.py

+# coding=utf-8
+# Copyright 2023 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""PatchTST model configuration"""
+
+from typing import List, Optional, Union
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+PATCHTST_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "ibm/patchtst-base": "https://huggingface.co/ibm/patchtst-base/resolve/main/config.json",
+    # See all PatchTST models at https://huggingface.co/ibm/models?filter=patchtst
+}
+
+
+class PatchTSTConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of an [`PatchTSTModel`]. It is used to instantiate an
+    PatchTST model according to the specified arguments, defining the model architecture.
+    [ibm/patchtst](https://huggingface.co/ibm/patchtst) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        num_input_channels (`int`, *optional*, defaults to 1):
+            The size of the target variable which by default is 1 for univariate targets. Would be > 1 in case of
+            multivariate targets.
+        context_length (`int`, *optional*, defaults to 32):
+            The context length of the input sequence.
+        distribution_output (`str`, *optional*, defaults to `"student_t"`):
+            The distribution emission head for the model when loss is "nll". Could be either "student_t", "normal" or
+            "negative_binomial".
+        loss (`str`, *optional*, defaults to `"mse"`):
+            The loss function for the model corresponding to the `distribution_output` head. For parametric
+            distributions it is the negative log likelihood ("nll") and for point estimates it is the mean squared
+            error "mse".
+        patch_length (`int`, *optional*, defaults to 1):
+            Define the patch length of the patchification process.
+        patch_stride (`int`, *optional*, defaults to 1):
+            Define the stride of the patchification process.
+        num_hidden_layers (`int`, *optional*, defaults to 3):
+            Number of hidden layers.
+        d_model (`int`, *optional*, defaults to 128):
+            Dimensionality of the transformer layers.
+        num_attention_heads (`int`, *optional*, defaults to 4):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        share_embedding (`bool`, *optional*, defaults to `True`):
+            Sharing the input embedding across all channels.
+        channel_attention (`bool`, *optional*, defaults to `False`):
+            Activate channel attention block in the Transformer to allow channels to attend each other.
+        ffn_dim (`int`, *optional*, defaults to 512):
+            Dimension of the "intermediate" (often named feed-forward) layer in the Transformer encoder.
+        norm_type (`str` , *optional*, defaults to `"batchnorm"`):
+            Normalization at each Transformer layer. Can be `"batchnorm"` or `"layernorm"`.
+        norm_eps (`float`, *optional*, defaults to 1e-05):
+            A value added to the denominator for numerical stability of normalization.
+        attention_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout probability for the attention probabilities.
+        dropout (`float`, *optional*, defaults to 0.0):
+            The dropout probability for all fully connected layers in the Transformer.
+        positional_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout probability in the positional embedding layer.
+        path_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout path in the residual block.
+        ff_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout probability used between the two layers of the feed-forward networks.
+        bias (`bool`, *optional*, defaults to `True`):
+            Whether to add bias in the feed-forward networks.
+        activation_function (`str`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (string) in the Transformer.`"gelu"` and `"relu"` are supported.
+        pre_norm (`bool`, *optional*, defaults to `True`):
+            Normalization is applied before self-attention if pre_norm is set to `True`. Otherwise, normalization is
+            applied after residual block.
+        positional_encoding_type (`str`, *optional*, defaults to `"sincos"`):
+            Positional encodings. Options `"random"` and `"sincos"` are supported.
+        use_cls_token (`bool`, *optional*, defaults to `False`):
+            Whether cls token is used.
+        init_std (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated normal weight initialization distribution.
+        share_projection (`bool`, *optional*, defaults to `True`):
+            Sharing the projection layer across different channels in the forecast head.
+        scaling (`Union`, *optional*, defaults to `"std"`):
+            Whether to scale the input targets via "mean" scaler, "std" scaler or no scaler if `None`. If `True`, the
+            scaler is set to "mean".
+        do_mask_input (`bool`, *optional*):
+            Apply masking during the pretraining.
+        mask_type (`str`, *optional*, defaults to `"random"`):
+            Masking type. Only `"random"` and `"forecast"` are currently supported.
+        random_mask_ratio (`float`, *optional*, defaults to 0.5):
+            Masking ratio applied to mask the input data during random pretraining.
+        num_forecast_mask_patches (`int` or `list`, *optional*, defaults to `[2]`):
+            Number of patches to be masked at the end of each batch sample. If it is an integer,
+            all the samples in the batch will have the same number of masked patches. If it is a list,
+            samples in the batch will be randomly masked by numbers defined in the list. This argument is only used
+            for forecast pretraining.
+        channel_consistent_masking (`bool`, *optional*, defaults to `False`):
+            If channel consistent masking is True, all the channels will have the same masking pattern.
+        unmasked_channel_indices (`list`, *optional*):
+            Indices of channels that are not masked during pretraining. Values in the list are number between 1 and
+            `num_input_channels`
+        mask_value (`int`, *optional*, defaults to 0):
+            Values in the masked patches will be filled by `mask_value`.
+        pooling_type (`str`, *optional*, defaults to `"mean"`):
+            Pooling of the embedding. `"mean"`, `"max"` and `None` are supported.
+        head_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout probability for head.
+        prediction_length (`int`, *optional*, defaults to 24):
+            The prediction horizon that the model will output.
+        num_targets (`int`, *optional*, defaults to 1):
+            Number of targets for regression and classification tasks. For classification, it is the number of
+            classes.
+        output_range (`list`, *optional*):
+            Output range for regression task. The range of output values can be set to enforce the model to produce
+            values within a range.
+        num_parallel_samples (`int`, *optional*, defaults to 100):
+            The number of samples is generated in parallel for probabilistic prediction.
+
+
+    ```python
+    >>> from transformers import PatchTSTConfig, PatchTSTModel
+
+    >>> # Initializing an PatchTST configuration with 12 time steps for prediction
+    >>> configuration = PatchTSTConfig(prediction_length=12)
+
+    >>> # Randomly initializing a model (with random weights) from the configuration
+    >>> model = PatchTSTModel(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "patchtst"
+    attribute_map = {
+        "hidden_size": "d_model",
+        "num_attention_heads": "num_attention_heads",
+        "num_hidden_layers": "num_hidden_layers",
+    }
+
+    def __init__(
+        self,
+        # time series specific configuration
+        num_input_channels: int = 1,
+        context_length: int = 32,
+        distribution_output: str = "student_t",
+        loss: str = "mse",
+        # PatchTST arguments
+        patch_length: int = 1,
+        patch_stride: int = 1,
+        # Transformer architecture configuration
+        num_hidden_layers: int = 3,
+        d_model: int = 128,
+        num_attention_heads: int = 4,
+        share_embedding: bool = True,
+        channel_attention: bool = False,
+        ffn_dim: int = 512,
+        norm_type: str = "batchnorm",
+        norm_eps: float = 1e-05,
+        attention_dropout: float = 0.0,
+        dropout: float = 0.0,
+        positional_dropout: float = 0.0,
+        path_dropout: float = 0.0,
+        ff_dropout: float = 0.0,
+        bias: bool = True,
+        activation_function: str = "gelu",
+        pre_norm: bool = True,
+        positional_encoding_type: str = "sincos",
+        use_cls_token: bool = False,
+        init_std: float = 0.02,
+        share_projection: bool = True,
+        scaling: Optional[Union[str, bool]] = "std",
+        # mask pretraining
+        do_mask_input: Optional[bool] = None,
+        mask_type: str = "random",
+        random_mask_ratio: float = 0.5,
+        num_forecast_mask_patches: Optional[Union[List[int], int]] = [2],
+        channel_consistent_masking: Optional[bool] = False,
+        unmasked_channel_indices: Optional[List[int]] = None,
+        mask_value: int = 0,
+        # head
+        pooling_type: str = "mean",
+        head_dropout: float = 0.0,
+        prediction_length: int = 24,
+        num_targets: int = 1,
+        output_range: Optional[List] = None,
+        # distribution head
+        num_parallel_samples: int = 100,
+        **kwargs,
+    ):
+        # time series specific configuration
+        self.context_length = context_length
+        self.num_input_channels = num_input_channels  # n_vars
+        self.loss = loss
+        self.distribution_output = distribution_output
+        self.num_parallel_samples = num_parallel_samples
+
+        # Transformer architecture configuration
+        self.d_model = d_model
+        self.num_attention_heads = num_attention_heads
+        self.ffn_dim = ffn_dim
+        self.num_hidden_layers = num_hidden_layers
+        self.dropout = dropout
+        self.attention_dropout = attention_dropout
+        self.share_embedding = share_embedding
+        self.channel_attention = channel_attention
+        self.norm_type = norm_type
+        self.norm_eps = norm_eps
+        self.positional_dropout = positional_dropout
+        self.path_dropout = path_dropout
+        self.ff_dropout = ff_dropout
+        self.bias = bias
+        self.activation_function = activation_function
+        self.pre_norm = pre_norm
+        self.positional_encoding_type = positional_encoding_type
+        self.use_cls_token = use_cls_token
+        self.init_std = init_std
+        self.scaling = scaling
+
+        # PatchTST parameters
+        self.patch_length = patch_length
+        self.patch_stride = patch_stride
+
+        # Mask pretraining
+        self.do_mask_input = do_mask_input
+        self.mask_type = mask_type
+        self.random_mask_ratio = random_mask_ratio  # for random masking
+        self.num_forecast_mask_patches = num_forecast_mask_patches  # for forecast masking
+        self.channel_consistent_masking = channel_consistent_masking
+        self.unmasked_channel_indices = unmasked_channel_indices
+        self.mask_value = mask_value
+
+        # general head params
+        self.pooling_type = pooling_type
+        self.head_dropout = head_dropout
+
+        # For prediction head
+        self.share_projection = share_projection
+        self.prediction_length = prediction_length
+
+        # For prediction and regression head
+        self.num_parallel_samples = num_parallel_samples
+
+        # Regression
+        self.num_targets = num_targets
+        self.output_range = output_range
+
+        super().__init__(**kwargs)