BLOOM (#17474)

* adding template

* update model

* model update

* update conf for debug model

* update conversion

* update conversion script

* update conversion script

* fix missing keys check

* add tests to test the tokenizer in the local machine

* Change variable name

* add tests on xnli dataset

* add more description

* add descriptions + clearer code

* clearer code

* adding new tests + skipping few tests because of env problems

* change comment

* add dtype on the configuration

* add test embeddings

* add hardcoded test

* fix dtype issue

* adding torch.float16 to config

* adding more metrics (min, max, mean)

* add sum

* now the test passes with almost equal

* add files for conversion - test passes on cpu  gpu

* add final changes

* cleaning code

* add new args in the docstring

* fix one liner function

* remove macros

* remove forward attention

* clean up init funtion

* add comments on the issue

* rm scale mask softmax

* do make style

* fix dtype in init

* fixing for loop on att probs

* fix style with black

* fix style + doc error

* fix and debug CI errors (docs + style)

* some updates

- change new operations
- finally add scaled softmax
- added new args in the config

* make use cache working

* add changes

- save sharded models
- final changes on the modeling script

* add changes

- comment on alibi
- add TODO on seq length

* test commit

- added a text to test the commit
Co-authored-by: thomasw21 <24695242+thomasw21@users.noreply.github.com>

* final changes

- attention mask change
- generation works on BS176b
Co-authored-by: thomasw21 <24695242+thomasw21@users.noreply.github.com>

* changes - model + conversion

* move to correct dir

* put ,

* fex fixes

* fix tokenizer autodoc

* fix minor CI issues

* fix minor CI issues

* fix minor CI issues

* fix style issue

* fix minor import issues

* fix few issues

* remove def main on the test

* add require torch

* replace decorator with 'with'

* fix style

* change to bloom

* add quick fix tokenizer

* fix tokenizer file

* fix tokenizer

- merge tests
- small fixes

* fix import issue

* add bloom to readme

* fix consistency

* Update docs/source/en/model_doc/bloom.mdx
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Apply suggestions from code review

fix comment issues on file headers
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* fix doc issue

* small fix - modeling test

* some changes

- refactor some code
- taking into account reviews
- more tests should pass
- removed pruning tests

* remove useless division

* more tests should pass

* more tests should pass

* more tests should pass

* let's try this one

-add alibi offset
- remove all permutes to make the grad operations work
- finger crossed

* refactor

- refactor code
- style changes
- add new threshold for test

* major changes

- change BLOOM to Bloom
- add quick doc on bloom.mdx
- move embeddings test on modeling test

* modify readme

* small fixes

* small fix

- better threshold for a test

* remove old test file from fetcher

* fix small typo

* major change

- change BloomLMHead to BloomForCausalLM

* remove onnx config

* major changes

- refactor the code
- remove asserts
- change tol for test

* make style

* small change

* adding a slow test + commenting old ones for now

* make style

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* make style

* fix duplicates

* cleaning comments on config

* clean a bit conversion file

* refacor a bit modeling file

* refactor tokenizer file

* fix tokenization test issue

* fix tokenization issue #2

* fix tokenization issue second try

* fix test issue

* make style + add suggestions

* change test fetcher

* try this one

- slow tests should pass
- finger crossed

* possible final changes

* make style

* try fix padding side issue

* fix side

* fix padding issue

* fix ko-readme

* fix config auto

* cleaning modeling file

* keep bloom in caps in ko

* update config docs

* remove pretraining_pp

* remove model parallel

* update config

- add correct config files

* fix duplicates

* fix fetcher

* fix refactor issue

- remove divide function

* try to remove alibi

* small fixes

- fix alibi
- remove seq length
- refactor a bit the code

* put correct values

- fix bos and eos token ids

* fix attention mask loop
Co-authored-by: thomasw21 <24695242+thomasw21@users.noreply.github.com>

* small fixes:

- remove skip bias add

* small fixes

- fix typo in readme
- fix typos in config

* small changes

- remove a test
- add reconstruction test
- change config

* small changes

- change Scaled Softmax to BloomScaledSoftmax

* small fixes

- fix alibi dtype

* major changes

- removing explicit dtype when loading modules
- fixing test args (torch_dtype=auto)
- add dosctring

* fix readmes

* major changes

- now bloom supports alibi shifting
- refactor a bit the code
- better test tolerance now

* refactor a bit

* refactor a bit

* put correct name on test

* change docstring

* small changes

- fix docstring modeling
- fix test tolerance

* fix small nit

- take dtype from tensors in the conversion script

* minor fix

- fix mdx issue

* minor fix

- change config docstring

* forward contrib credits from PR14084

* Apply suggestions from code review
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>

* apply modifications
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>

* resolve softmax upcast

* Apply suggestions from code review
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>

* Update src/transformers/models/bloom/modeling_bloom.py
Co-authored-by: Niklas Muennighoff <n.muennighoff@gmail.com>

* final changes modeling
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>

* Merge commit 'd156898f

'

* merge commit

* Apply suggestions from code review
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>

* apply suggestions

Apply suggestions from Stas comments
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>

* Fix gradient checkpointing
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>

* add slow but exact

* add accelerate compatibility
Co-authored-by: Nicolas Patry <Narsil@users.noreply.github.com>

* forward contrib credits
Co-authored-by: thomasw21 <thomasw21@users.noreply.github.com>
Co-authored-by: sgugger <sgugger@users.noreply.github.com>
Co-authored-by: patrickvonplaten <patrickvonplaten@users.noreply.github.com>
Co-authored-by: Niklas Muennighoff <n.muennighoff@gmail.com>
Co-authored-by: LysandreJik <LysandreJik@users.noreply.github.com>

* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* fix torch device on tests

* make style

* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* fix nits

Co-authored-by: patrickvonplaten<patrickvonplaten@users.noreply.github.com>

* remove final nits

* fix doc

- add more details on the doc
- add links to checkpoints

* Update src/transformers/__init__.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Update src/transformers/models/bloom/modeling_bloom.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* apply suggestions
Co-authored-by: sgugger <sgugger@users.noreply.github.com>

* put test torchscript to false

* Update src/transformers/models/bloom/modeling_bloom.py
Co-authored-by: justheuristic <justheuristic@gmail.com>

* fix alibi

- create alibi only once

* add small doc

* make quality

* replace torch.nn

* remove token type emb

* fix fused op + output bias

* add fused op

- now can control fused operation from config

* remove fused op

* make quality

* small changes

- remove unsed args on config
- removed bias gelu file
- make the model torchscriptable
- add torchscript slow tests

* Update src/transformers/models/bloom/modeling_bloom.py

* fix slow

* make style

* add accelerate support

* add bloom to deepspeed tests

* minor changes

* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* minor change

* slow tests pass

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Update docs/source/en/model_doc/bloom.mdx
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* minor changes:

- change docstring
- add link to paper
Co-authored-by: Thomwolf <thomwolf@gmail.com>
Co-authored-by: Thomas Wolf <thomas@huggingface.co>
Co-authored-by: thomasw21 <24695242+thomasw21@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: sIncerass <sheng.s@berkeley.edu>
Co-authored-by: Stas Bekman <stas00@users.noreply.github.com>
Co-authored-by: Niklas Muennighoff <n.muennighoff@gmail.com>
Co-authored-by: Nicolas Patry <Narsil@users.noreply.github.com>
Co-authored-by: thomasw21 <thomasw21@users.noreply.github.com>
Co-authored-by: sgugger <sgugger@users.noreply.github.com>
Co-authored-by: patrickvonplaten <patrickvonplaten@users.noreply.github.com>
Co-authored-by: LysandreJik <LysandreJik@users.noreply.github.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: justheuristic <justheuristic@gmail.com>
Co-authored-by: Stas Bekman <stas@stason.org>

README.md

@@ -240,6 +240,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
+1. **[BLOOM](https://huggingface.co/docs/transformers/main/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/).
 1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (from Alexa) released with the paper [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) by Adrian de Wynter and Daniel J. Perry.
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (from Inria/Facebook/Sorbonne) released with the paper [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) by Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot.

README_ko.md

@@ -221,6 +221,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
+1. **[BLOOM](https://huggingface.co/docs/transformers/main/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/).
 1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (from Alexa) released with the paper [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) by Adrian de Wynter and Daniel J. Perry.
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (from Inria/Facebook/Sorbonne) released with the paper [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) by Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot.

README_zh-hans.md

@@ -245,6 +245,7 @@ conda install -c huggingface transformers
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (来自 Google Research) 伴随论文 [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) 由 Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed 发布。
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (来自 Facebook) 伴随论文 [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) 由 Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston 发布。
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (来自 Facebook) 伴随论文 [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) 由 Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston 发布。
+1. **[BLOOM](https://huggingface.co/docs/transformers/main/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/). 
 1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (来自 Alexa) 伴随论文 [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) 由 Adrian de Wynter and Daniel J. Perry 发布。
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (来自 Google Research) 伴随论文 [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) 由 Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel 发布。
 1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (来自 Inria/Facebook/Sorbonne) 伴随论文 [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) 由 Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot 发布。

README_zh-hant.md

@@ -257,6 +257,7 @@ conda install -c huggingface transformers
 1. **[BigBird-RoBERTa](https://huggingface.co/docs/transformers/model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](https://huggingface.co/docs/transformers/model_doc/blenderbot)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
 1. **[BlenderbotSmall](https://huggingface.co/docs/transformers/model_doc/blenderbot-small)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
+1. **[BLOOM](https://huggingface.co/docs/transformers/main/model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/). 
 1. **[BORT](https://huggingface.co/docs/transformers/model_doc/bort)** (from Alexa) released with the paper [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) by Adrian de Wynter and Daniel J. Perry.
 1. **[ByT5](https://huggingface.co/docs/transformers/model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert)** (from Inria/Facebook/Sorbonne) released with the paper [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) by Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot.

docs/source/en/_toctree.yml

@@ -176,6 +176,8 @@
       title: Blenderbot
     - local: model_doc/blenderbot-small
       title: Blenderbot Small
+    - local: model_doc/bloom
+      title: BLOOM
     - local: model_doc/bort
       title: BORT
     - local: model_doc/byt5

docs/source/en/index.mdx

@@ -63,6 +63,7 @@ The library currently contains JAX, PyTorch and TensorFlow implementations, pret
 1. **[BigBird-RoBERTa](model_doc/big_bird)** (from Google Research) released with the paper [Big Bird: Transformers for Longer Sequences](https://arxiv.org/abs/2007.14062) by Manzil Zaheer, Guru Guruganesh, Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed.
 1. **[Blenderbot](model_doc/blenderbot)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
 1. **[BlenderbotSmall](model_doc/blenderbot-small)** (from Facebook) released with the paper [Recipes for building an open-domain chatbot](https://arxiv.org/abs/2004.13637) by Stephen Roller, Emily Dinan, Naman Goyal, Da Ju, Mary Williamson, Yinhan Liu, Jing Xu, Myle Ott, Kurt Shuster, Eric M. Smith, Y-Lan Boureau, Jason Weston.
+1. **[BLOOM](model_doc/bloom)** (from BigScience workshop) released by the [BigSicence Workshop](https://bigscience.huggingface.co/).
 1. **[BORT](model_doc/bort)** (from Alexa) released with the paper [Optimal Subarchitecture Extraction For BERT](https://arxiv.org/abs/2010.10499) by Adrian de Wynter and Daniel J. Perry.
 1. **[ByT5](model_doc/byt5)** (from Google Research) released with the paper [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel.
 1. **[CamemBERT](model_doc/camembert)** (from Inria/Facebook/Sorbonne) released with the paper [CamemBERT: a Tasty French Language Model](https://arxiv.org/abs/1911.03894) by Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot.
@@ -193,6 +194,7 @@ Flax), PyTorch, and/or TensorFlow.
 |       BigBird-Pegasus       |       ❌       |       ❌       |       ✅        |         ❌         |      ❌      |
 |         Blenderbot          |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
 |       BlenderbotSmall       |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |
+|            BLOOM            |       ❌       |       ✅       |       ✅        |         ❌         |      ❌      |
 |          CamemBERT          |       ✅       |       ✅       |       ✅        |         ✅         |      ❌      |
 |           CANINE            |       ✅       |       ❌       |       ✅        |         ❌         |      ❌      |
 |            CLIP             |       ✅       |       ✅       |       ✅        |         ✅         |      ✅      |

docs/source/en/model_doc/bloom.mdx

+<!--Copyright 2022 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.
+-->
+
+# BLOOM
+
+## Overview
+
+The BLOOM model has been proposed with its various versions through the [BigScience Workshop](https://bigscience.huggingface.co/). BigScience is inspired by other open science initiatives where researchers have pooled their time and resources to collectively achieve a higher impact.
+The architecture of BLOOM is essentially similar to GPT3 (auto-regressive model for next token prediction), but has been trained on different 46 languages including code.
+Several smaller versions of the models have been trained on the same dataset. BLOOM is available in the following versions:
+
+- [bloom-350m](https://huggingface.co/bigscience/bloom-350m)
+- [bloom-760m](https://huggingface.co/bigscience/bloom-760m)
+- [bloom-1b3](https://huggingface.co/bigscience/bloom-1b3)
+- [bloom-2b5](https://huggingface.co/bigscience/bloom-2b5)
+- [bloom-6b3](https://huggingface.co/bigscience/bloom-6b3)
+- [bloom](https://huggingface.co/bigscience/bloom) (175B parameters)
+
+
+## BloomConfig
+
+[[autodoc]] BloomConfig
+    - all
+
+## BloomModel
+
+[[autodoc]] BloomModel
+    - forward
+
+## BloomTokenizerFast
+
+[[autodoc]] BloomTokenizerFast
+    - all
+
+## BloomForCausalLM
+
+[[autodoc]] BloomForCausalLM
+    - forward
\ No newline at end of file

src/transformers/__init__.py

@@ -156,6 +156,7 @@ _import_structure = {
         "BlenderbotSmallConfig",
         "BlenderbotSmallTokenizer",
     ],
+    "models.bloom": ["BLOOM_PRETRAINED_CONFIG_ARCHIVE_MAP", "BloomConfig"],
     "models.bort": [],
     "models.byt5": ["ByT5Tokenizer"],
     "models.camembert": ["CAMEMBERT_PRETRAINED_CONFIG_ARCHIVE_MAP", "CamembertConfig"],
@@ -497,6 +498,7 @@ else:
     _import_structure["models.big_bird"].append("BigBirdTokenizerFast")
     _import_structure["models.blenderbot"].append("BlenderbotTokenizerFast")
     _import_structure["models.blenderbot_small"].append("BlenderbotSmallTokenizerFast")
+    _import_structure["models.bloom"].append("BloomTokenizerFast")
     _import_structure["models.camembert"].append("CamembertTokenizerFast")
     _import_structure["models.clip"].append("CLIPTokenizerFast")
     _import_structure["models.convbert"].append("ConvBertTokenizerFast")
@@ -858,6 +860,14 @@ else:
             "BigBirdPegasusPreTrainedModel",
         ]
     )
+    _import_structure["models.bloom"].extend(
+        [
+            "BLOOM_PRETRAINED_MODEL_ARCHIVE_LIST",
+            "BloomForCausalLM",
+            "BloomModel",
+            "BloomPreTrainedModel",
+        ]
+    )
     _import_structure["models.blenderbot"].extend(
         [
             "BLENDERBOT_PRETRAINED_MODEL_ARCHIVE_LIST",
@@ -2755,6 +2765,7 @@ if TYPE_CHECKING:
         BlenderbotSmallConfig,
         BlenderbotSmallTokenizer,
     )
+    from .models.bloom import BLOOM_PRETRAINED_CONFIG_ARCHIVE_MAP, BloomConfig
     from .models.byt5 import ByT5Tokenizer
     from .models.camembert import CAMEMBERT_PRETRAINED_CONFIG_ARCHIVE_MAP, CamembertConfig
     from .models.canine import CANINE_PRETRAINED_CONFIG_ARCHIVE_MAP, CanineConfig, CanineTokenizer
@@ -3064,6 +3075,7 @@ if TYPE_CHECKING:
         from .models.big_bird import BigBirdTokenizerFast
         from .models.blenderbot import BlenderbotTokenizerFast
         from .models.blenderbot_small import BlenderbotSmallTokenizerFast
+        from .models.bloom import BloomTokenizerFast
         from .models.camembert import CamembertTokenizerFast
         from .models.clip import CLIPTokenizerFast
         from .models.convbert import ConvBertTokenizerFast
@@ -3382,6 +3394,12 @@ if TYPE_CHECKING:
             BlenderbotSmallModel,
             BlenderbotSmallPreTrainedModel,
         )
+        from .models.bloom import (
+            BLOOM_PRETRAINED_MODEL_ARCHIVE_LIST,
+            BloomForCausalLM,
+            BloomModel,
+            BloomPreTrainedModel,
+        )
         from .models.camembert import (
             CAMEMBERT_PRETRAINED_MODEL_ARCHIVE_LIST,
             CamembertForCausalLM,

src/transformers/models/__init__.py

@@ -31,6 +31,7 @@ from . import (
     bigbird_pegasus,
     blenderbot,
     blenderbot_small,
+    bloom,
     bort,
     byt5,
     camembert,

src/transformers/models/auto/configuration_auto.py

@@ -38,6 +38,7 @@ CONFIG_MAPPING_NAMES = OrderedDict(
         ("bigbird_pegasus", "BigBirdPegasusConfig"),
         ("blenderbot", "BlenderbotConfig"),
         ("blenderbot-small", "BlenderbotSmallConfig"),
+        ("bloom", "BloomConfig"),
         ("camembert", "CamembertConfig"),
         ("canine", "CanineConfig"),
         ("clip", "CLIPConfig"),
@@ -51,7 +52,6 @@ CONFIG_MAPPING_NAMES = OrderedDict(
         ("deberta", "DebertaConfig"),
         ("deberta-v2", "DebertaV2Config"),
         ("decision_transformer", "DecisionTransformerConfig"),
-        ("decision_transformer", "DecisionTransformerConfig"),
         ("deit", "DeiTConfig"),
         ("detr", "DetrConfig"),
         ("distilbert", "DistilBertConfig"),
@@ -155,6 +155,7 @@ CONFIG_ARCHIVE_MAP_MAPPING_NAMES = OrderedDict(
         ("bigbird_pegasus", "BIGBIRD_PEGASUS_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("blenderbot", "BLENDERBOT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("blenderbot-small", "BLENDERBOT_SMALL_PRETRAINED_CONFIG_ARCHIVE_MAP"),
+        ("bloom", "BLOOM_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("camembert", "CAMEMBERT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("canine", "CANINE_PRETRAINED_CONFIG_ARCHIVE_MAP"),
         ("clip", "CLIP_PRETRAINED_CONFIG_ARCHIVE_MAP"),
@@ -262,6 +263,7 @@ MODEL_NAMES_MAPPING = OrderedDict(
         ("bigbird_pegasus", "BigBird-Pegasus"),
         ("blenderbot", "Blenderbot"),
         ("blenderbot-small", "BlenderbotSmall"),
+        ("bloom", "BLOOM"),
         ("bort", "BORT"),
         ("byt5", "ByT5"),
         ("camembert", "CamemBERT"),
@@ -362,7 +364,6 @@ MODEL_NAMES_MAPPING = OrderedDict(
         ("van", "VAN"),
         ("vilt", "ViLT"),
         ("vision-encoder-decoder", "Vision Encoder decoder"),
-        ("vision-encoder-decoder", "Vision Encoder decoder"),
         ("vision-text-dual-encoder", "VisionTextDualEncoder"),
         ("visual_bert", "VisualBERT"),
         ("vit", "ViT"),

src/transformers/models/auto/modeling_auto.py

@@ -37,6 +37,7 @@ MODEL_MAPPING_NAMES = OrderedDict(
         ("bigbird_pegasus", "BigBirdPegasusModel"),
         ("blenderbot", "BlenderbotModel"),
         ("blenderbot-small", "BlenderbotSmallModel"),
+        ("bloom", "BloomModel"),
         ("camembert", "CamembertModel"),
         ("canine", "CanineModel"),
         ("clip", "CLIPModel"),
@@ -50,7 +51,6 @@ MODEL_MAPPING_NAMES = OrderedDict(
         ("deberta", "DebertaModel"),
         ("deberta-v2", "DebertaV2Model"),
         ("decision_transformer", "DecisionTransformerModel"),
-        ("decision_transformer", "DecisionTransformerModel"),
         ("decision_transformer_gpt2", "DecisionTransformerGPT2Model"),
         ("deit", "DeiTModel"),
         ("detr", "DetrModel"),
@@ -144,6 +144,7 @@ MODEL_FOR_PRETRAINING_MAPPING_NAMES = OrderedDict(
         ("bart", "BartForConditionalGeneration"),
         ("bert", "BertForPreTraining"),
         ("big_bird", "BigBirdForPreTraining"),
+        ("bloom", "BloomForCausalLM"),
         ("camembert", "CamembertForMaskedLM"),
         ("ctrl", "CTRLLMHeadModel"),
         ("data2vec-text", "Data2VecTextForMaskedLM"),
@@ -194,6 +195,7 @@ MODEL_WITH_LM_HEAD_MAPPING_NAMES = OrderedDict(
         ("big_bird", "BigBirdForMaskedLM"),
         ("bigbird_pegasus", "BigBirdPegasusForConditionalGeneration"),
         ("blenderbot-small", "BlenderbotSmallForConditionalGeneration"),
+        ("bloom", "BloomForCausalLM"),
         ("camembert", "CamembertForMaskedLM"),
         ("convbert", "ConvBertForMaskedLM"),
         ("ctrl", "CTRLLMHeadModel"),
@@ -252,6 +254,7 @@ MODEL_FOR_CAUSAL_LM_MAPPING_NAMES = OrderedDict(
         ("bigbird_pegasus", "BigBirdPegasusForCausalLM"),
         ("blenderbot", "BlenderbotForCausalLM"),
         ("blenderbot-small", "BlenderbotSmallForCausalLM"),
+        ("bloom", "BloomForCausalLM"),
         ("camembert", "CamembertForCausalLM"),
         ("ctrl", "CTRLLMHeadModel"),
         ("data2vec-text", "Data2VecTextForCausalLM"),

src/transformers/models/auto/tokenization_auto.py

@@ -76,6 +76,7 @@ else:
             ("bigbird_pegasus", ("PegasusTokenizer", "PegasusTokenizerFast" if is_tokenizers_available() else None)),
             ("blenderbot", ("BlenderbotTokenizer", "BlenderbotTokenizerFast")),
             ("blenderbot-small", ("BlenderbotSmallTokenizer", None)),
+            ("bloom", (None, "BloomTokenizerFast" if is_tokenizers_available() else None)),
             ("byt5", ("ByT5Tokenizer", None)),
             (
                 "camembert",

src/transformers/models/bloom/__init__.py

+# flake8: noqa
+# There's no way to ignore "F401 '...' imported but unused" warnings in this
+# module, but to preserve other warnings. So, don't check this module at all.
+
+# Copyright 2022 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
+
+from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_tokenizers_available, is_torch_available
+
+
+_import_structure = {
+    "configuration_bloom": [
+        "BLOOM_PRETRAINED_CONFIG_ARCHIVE_MAP",
+        "BloomConfig",
+    ],
+}
+try:
+    if not is_tokenizers_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["tokenization_bloom_fast"] = ["BloomTokenizerFast"]
+
+try:
+    if not is_torch_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["modeling_bloom"] = [
+        "BLOOM_PRETRAINED_MODEL_ARCHIVE_LIST",
+        "BloomForCausalLM",
+        "BloomModel",
+        "BloomPreTrainedModel",
+    ]
+
+if TYPE_CHECKING:
+    from .configuration_bloom import BLOOM_PRETRAINED_CONFIG_ARCHIVE_MAP, BloomConfig
+
+    try:
+        if not is_tokenizers_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .tokenization_bloom_fast import BloomTokenizerFast
+
+    try:
+        if not is_torch_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .modeling_bloom import (
+            BLOOM_PRETRAINED_MODEL_ARCHIVE_LIST,
+            BloomForCausalLM,
+            BloomModel,
+            BloomPreTrainedModel,
+        )
+
+else:
+    import sys
+
+    sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)

src/transformers/models/bloom/configuration_bloom.py

+# coding=utf-8
+# Copyright 2022 the Big Science Workshop and 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.
+""" Bloom configuration"""
+from ...configuration_utils import PretrainedConfig
+from ...utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+BLOOM_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "bigscience/bloom": "https://huggingface.co/bigscience/bloom/resolve/main/config.json",
+    "bigscience/bloom-350m": "https://huggingface.co/bigscience/bloom-350m/blob/main/config.json",
+    "bigscience/bloom-760m": "https://huggingface.co/bigscience/bloom-760m/blob/main/config.json",
+    "bigscience/bloom-1b3": "https://huggingface.co/bigscience/bloom-1b3/blob/main/config.json",
+    "bigscience/bloom-2b5": "https://huggingface.co/bigscience/bloom-2b5/blob/main/config.json",
+    "bigscience/bloom-6b3": "https://huggingface.co/bigscience/bloom-6b3/blob/main/config.json",
+}
+
+
+class BloomConfig(PretrainedConfig):
+    """
+    This is the configuration class to store the configuration of a [`BloomModel`]. It is used to instantiate a Bloom
+    model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
+    defaults will yield a similar configuration to the Bloom architecture
+    [bigscience/bloom](https://huggingface.co/bigscience/bloom).
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 50257):
+            Vocabulary size of the Bloom model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`BloomModel`].
+        hidden_size (`int`, *optional*, defaults to 768):
+            Dimensionality of the embeddings and hidden states.
+        n_layer (`int`, *optional*, defaults to 12):
+            Number of hidden layers in the Transformer encoder.
+        n_head (`int`, *optional*, defaults to 12):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        attn_pdrop (`float`, *optional*, defaults to 0.1):
+            The dropout ratio for the attention.
+        layer_norm_epsilon (`float`, *optional*, defaults to 1e-5):
+            The epsilon to use in the layer normalization layers.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        apply_residual_connection_post_layernorm (`bool`, *optional*, defaults to `False`):
+            If enabled, use the layer norm of the hidden states as the residual in the transformer blocks
+        skip_bias_add (`bool`, *optional*, defaults to `True`):
+            If set to `True`, it will skip bias add for each linear layer in the transformer blocks
+        skip_bias_add_qkv (`bool`, *optional*, defaults to `False`):
+            If set to `True`, it will skip bias add for the first linear layer in the transformer blocks
+        attention_softmax_in_fp32 (`bool`, *optional*, defaults to `True`):
+            If set to `True` and the `dtype` is set to `float16` it will scale the input of the Softmax function to
+            `fp32`
+        hidden_dropout (`float`, *optional*, defaults to 0.1):
+            Dropout rate of the dropout function on the bias dropout.
+        attention_dropout (`float`, *optional*, defaults to 0.1):
+            Dropout rate applied to the attention probs
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models).
+        dtype (`str`, *optional*, defaults to `"bfloat16"`):
+            Precision that has been used for the model's training in Megatron. Please load the model in the correct
+            precision by doing `model = BloomModel.from_pretrained(model_name, torch_dtype="auto")`.`
+        pretraining_tp (`int`, *optional*, defaults to `1`):
+            Experimental feature. Tensor parallelism rank used during pretraining with Megatron. Please refer to [this
+            document](https://huggingface.co/docs/transformers/parallelism) to understand more about it. This value is
+            necessary to ensure exact reproducibility of the pretraining results. Please refer to [this
+            issue](https://github.com/pytorch/pytorch/issues/76232). Note also that this is enabled only when
+            `slow_but_exact=True`.
+        slow_but_exact (`bool`, *optional*, defaults to `False`):
+            Experimental feature. Whether to use slow but exact implementation of the attention mechanism. While
+            merging the TP rank tensors, due to slicing operations the results may be slightly different between the
+            model trained on Megatron and our model. Please refer to [this
+            issue](https://github.com/pytorch/pytorch/issues/76232). A solution to obtain more accurate results is to
+            enable this feature. Enabling this will hurt the computational time of the inference. Will be probably
+            resolved in the future once the main model has been fine-tuned with TP_rank=1.
+
+    Example:
+
+    ```python
+    >>> from transformers import BloomModel, BloomConfig
+
+    >>> # Initializing a Bloom configuration
+    >>> configuration = BloomConfig()
+
+    >>> # Initializing a model from the configuration
+    >>> model = BloomModel(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "bloom"
+    keys_to_ignore_at_inference = ["past_key_values"]
+    attribute_map = {
+        "num_hidden_layers": "n_layer",
+        "n_head": "num_attention_heads",
+        "hidden_size": "n_embed",
+        "dtype": "torch_dtype",
+    }
+
+    def __init__(
+        self,
+        vocab_size=250880,
+        hidden_size=64,
+        n_layer=2,
+        n_head=8,
+        masked_softmax_fusion=True,
+        layer_norm_epsilon=1e-5,
+        initializer_range=0.02,
+        use_cache=False,
+        bos_token_id=1,
+        eos_token_id=2,
+        apply_residual_connection_post_layernorm=False,
+        hidden_dropout=0.0,
+        attention_dropout=0.0,
+        attention_softmax_in_fp32=True,
+        pretraining_tp=1,  # TP rank used when training with megatron
+        dtype="bfloat16",
+        slow_but_exact=False,
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.hidden_size = hidden_size
+        self.n_layer = n_layer
+        self.n_head = n_head
+        self.masked_softmax_fusion = masked_softmax_fusion
+        self.layer_norm_epsilon = layer_norm_epsilon
+        self.initializer_range = initializer_range
+        self.use_cache = use_cache
+        self.pretraining_tp = pretraining_tp
+        self.apply_residual_connection_post_layernorm = apply_residual_connection_post_layernorm
+        self.hidden_dropout = hidden_dropout
+        self.attention_dropout = attention_dropout
+        self.attention_softmax_in_fp32 = attention_softmax_in_fp32
+
+        self.bos_token_id = bos_token_id
+        self.eos_token_id = eos_token_id
+        self.dtype = dtype
+        self.slow_but_exact = slow_but_exact
+
+        super().__init__(bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)

src/transformers/models/bloom/convert_bloom_original_checkpoint_to_pytorch.py

+# coding=utf-8
+# Copyright 2022 The HuggingFace Inc. team.
+#
+# 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.
+"""Convert BigScience BLOOM checkpoint."""
+
+
+import argparse
+import json
+import os
+import re
+
+import torch
+
+from transformers import BloomConfig, BloomModel
+from transformers.file_utils import CONFIG_NAME, WEIGHTS_NAME
+from transformers.utils import logging
+
+
+logging.set_verbosity_info()
+
+WEIGHTS_TO_AVERAGE_ENDSWITH = [
+    "word_embeddings_layernorm.weight",
+    "word_embeddings_layernorm.bias",
+    "input_layernorm.weight",
+    "input_layernorm.bias",
+    "post_attention_layernorm.weight",
+    "post_attention_layernorm.bias",
+    "self_attention.dense.bias",
+    "mlp.dense_4h_to_h.bias",
+    "ln_f.weight",
+    "ln_f.bias",
+]
+
+WEIGHTS_WITH_ROW_PARALLELISM_CONTAIN = [
+    "mlp.dense_4h_to_h.weight",
+    "self_attention.dense.weight",
+]
+
+
+def layer_name_mapping(key, file):
+    """Convert Megatron-DeepSpeed TP/PP weights mapping in transformers PP only"""
+    # Handle first and last layers
+    layer_rename_map = {
+        "word_embeddings.weight": "word_embeddings.weight",
+        "word_embeddings.norm.weight": "word_embeddings_layernorm.weight",
+        "word_embeddings.norm.bias": "word_embeddings_layernorm.bias",
+        "weight": "ln_f.weight",
+        "bias": "ln_f.bias",
+    }
+
+    if key in layer_rename_map:
+        return layer_rename_map[key]
+
+    # Handle transformer blocks
+    layer_number = int(re.match(r".*layer_(\d*).*", file)[1])
+    layer_number -= 3
+    return f"h.{layer_number}." + key
+
+
+def get_dtype_size(dtype):
+    if dtype == torch.bool:
+        return 1 / 8
+    bit_search = re.search("[^\d](\d+)$", str(dtype))
+    if bit_search is None:
+        raise ValueError(f"`dtype` is not a valid dtype: {dtype}.")
+    bit_size = int(bit_search.groups()[0])
+    return bit_size // 8
+
+
+def convert_bloom_checkpoint_to_pytorch(
+    bloom_checkpoint_path, bloom_config_file, pytorch_dump_folder_path, shard_model, pretraining_tp
+):
+    # Construct model
+    if bloom_config_file == "":
+        config = BloomConfig()
+    else:
+        config = BloomConfig.from_json_file(bloom_config_file)
+
+    if shard_model:
+        file_names = os.listdir(bloom_checkpoint_path)
+        file_names = list(sorted(filter(lambda s: s.startswith("layer") and "model_00" in s, file_names)))
+
+        index_dict = {"weight_map": {}, "metadata": {}}
+        total_size = 0
+
+        missing_keys = None
+
+        config = BloomConfig()
+
+        for j, file in enumerate(file_names):
+            print("Processing file: {}".format(file))
+            tensors = None
+
+            for i in range(pretraining_tp):
+                # load all TP files
+                f_name = file.replace("model_00", f"model_0{i}")
+                temp = torch.load(os.path.join(bloom_checkpoint_path, f_name), map_location="cpu")
+
+                # Rename keys in the transformers names
+                keys = list(temp.keys())
+                for key in keys:
+                    temp[layer_name_mapping(key, file)] = temp.pop(key)
+
+                if tensors is None:
+                    tensors = temp
+                else:
+                    for key in tensors.keys():
+                        if any(key.endswith(end) for end in WEIGHTS_TO_AVERAGE_ENDSWITH):
+                            # We average (sum and then divide) some weights accross TP ranks (see https://github.com/bigscience-workshop/Megatron-DeepSpeed/blob/olruwase/sync_layer_norms/megatron/training.py#L425)
+                            tensors[key] += temp[key]
+                        else:
+                            # Some weights are RowParallelLinear in Megatron-Deepspeed, others are ColumnParallel
+                            cat_dim = 1 if any(text in key for text in WEIGHTS_WITH_ROW_PARALLELISM_CONTAIN) else 0
+                            # We concatenate these weights accross TP ranks
+                            tensors[key] = torch.cat([tensors[key], temp[key]], dim=cat_dim)
+
+            # Divide by the number of TP the weights we want to average
+            for key in tensors.keys():
+                if any(key.endswith(end) for end in WEIGHTS_TO_AVERAGE_ENDSWITH):
+                    tensors[key] = tensors[key] / pretraining_tp
+            torch.save(
+                tensors,
+                os.path.join(
+                    pytorch_dump_folder_path,
+                    "pytorch_model_{}-of-{}.bin".format(str(j + 1).zfill(5), str(len(file_names)).zfill(5)),
+                ),
+            )
+
+            for key in tensors.keys():
+                value = tensors[key]
+                total_size += value.numel() * get_dtype_size(value.dtype)
+                if key not in index_dict["weight_map"]:
+                    index_dict["weight_map"][key] = "pytorch_model_{}-of-{}.bin".format(
+                        str(j + 1).zfill(5), str(len(file_names)).zfill(5)
+                    )
+
+        config = BloomConfig()
+        pytorch_config_dump_path = pytorch_dump_folder_path + "/" + CONFIG_NAME
+        index_dict["metadata"]["total_size"] = total_size
+        with open(pytorch_config_dump_path, "w", encoding="utf-8") as f:
+            f.write(config.to_json_string())
+        with open(os.path.join(pytorch_dump_folder_path, WEIGHTS_NAME + ".index.json"), "w", encoding="utf-8") as f:
+            json_config = json.dumps(index_dict, indent=2, sort_keys=True) + "\n"
+            f.write(json_config)
+    else:
+        model = BloomModel(config)
+
+        file_names = os.listdir(bloom_checkpoint_path)
+        file_names = list(sorted(filter(lambda s: s.startswith("layer") and "model_00" in s, file_names)))
+
+        missing_keys = None
+        for i, file in enumerate(file_names):
+            tensors = None
+            for i in range(pretraining_tp):
+                # load all TP files
+                f_name = file.replace("model_00", f"model_0{i}")
+                temp = torch.load(os.path.join(bloom_checkpoint_path, f_name), map_location="cpu")
+
+                # Rename keys in the transformers names
+                keys = list(temp.keys())
+                for key in keys:
+                    temp[layer_name_mapping(key, file)] = temp.pop(key)
+
+                if tensors is None:
+                    tensors = temp
+                else:
+                    for key in tensors.keys():
+                        # We average (sum and then divide) some weights accross TP ranks (see https://github.com/bigscience-workshop/Megatron-DeepSpeed/blob/olruwase/sync_layer_norms/megatron/training.py#L425)
+                        if any(key.endswith(end) for end in WEIGHTS_TO_AVERAGE_ENDSWITH):
+                            tensors[key] += temp[key]
+                        else:
+                            # Some weights are RowParallelLinear in Megatron-Deepspeed, others are ColumnParallel
+                            cat_dim = 1 if any(text in key for text in WEIGHTS_WITH_ROW_PARALLELISM_CONTAIN) else 0
+                            # We concatenate these weights accross TP ranks
+                            tensors[key] = torch.cat([tensors[key], temp[key]], dim=cat_dim)
+
+            # Divide by the number of TP the weights we want to average
+            for key in tensors.keys():
+                if any(key.endswith(end) for end in WEIGHTS_TO_AVERAGE_ENDSWITH):
+                    tensors[key] = tensors[key] / pretraining_tp
+
+            other_keys = model.load_state_dict(tensors, strict=False)
+            assert not other_keys.unexpected_keys
+            if missing_keys is None:
+                missing_keys = set(other_keys.missing_keys)
+            else:
+                missing_keys = missing_keys.intersection(set(other_keys.missing_keys))
+
+        assert not missing_keys
+
+        # Save pytorch-model
+        os.makedirs(pytorch_dump_folder_path, exist_ok=True)
+        pytorch_weights_dump_path = pytorch_dump_folder_path + "/" + WEIGHTS_NAME
+        pytorch_config_dump_path = pytorch_dump_folder_path + "/" + CONFIG_NAME
+        print(f"Save PyTorch model to {pytorch_weights_dump_path}")
+        torch.save(model.state_dict(), pytorch_weights_dump_path)
+        print(f"Save configuration file to {pytorch_config_dump_path}")
+        with open(pytorch_config_dump_path, "w", encoding="utf-8") as f:
+            f.write(config.to_json_string())
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    # Required parameters
+    parser.add_argument(
+        "--bloom_checkpoint_path",
+        default=None,
+        type=str,
+        required=True,
+        help="Path to the Megatron-LM checkpoint path.",
+    )
+    parser.add_argument(
+        "--pytorch_dump_folder_path", default=None, type=str, required=True, help="Path to the output PyTorch model."
+    )
+    parser.add_argument(
+        "--bloom_config_file",
+        default="",
+        type=str,
+        help=(
+            "An optional config json file corresponding to the pre-trained model. \n"
+            "This specifies the model architecture."
+        ),
+    )
+    parser.add_argument(
+        "--shard_model",
+        action="store_true",
+        help="An optional setting to shard the output model \nThis enables sharding the converted checkpoint",
+    )
+    parser.add_argument(
+        "--pretraining_tp",
+        default=4,
+        type=int,
+        help="Pretraining TP rank that has been used when training the model in Megatron-LM \n",
+    )
+    args = parser.parse_args()
+    convert_bloom_checkpoint_to_pytorch(
+        args.bloom_checkpoint_path,
+        args.bloom_config_file,
+        args.pytorch_dump_folder_path,
+        args.shard_model,
+        args.pretraining_tp,
+    )

src/transformers/models/bloom/modeling_bloom.py

+# coding=utf-8
+# Copyright 2022 HuggingFace Inc. team and BigScience workshop.
+#
+# 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.
+"""PyTorch BLOOM model."""
+
+import math
+from typing import Tuple
+
+import torch
+import torch.utils.checkpoint
+from torch import nn
+from torch.nn import CrossEntropyLoss, LayerNorm
+
+from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
+from ...modeling_outputs import BaseModelOutputWithPastAndCrossAttentions, CausalLMOutputWithCrossAttentions
+from ...modeling_utils import PreTrainedModel
+from ...utils import logging
+from .configuration_bloom import BloomConfig
+
+
+logger = logging.get_logger(__name__)
+
+_CHECKPOINT_FOR_DOC = "bigscience/Bloom"
+_CONFIG_FOR_DOC = "BloomConfig"
+_TOKENIZER_FOR_DOC = "BloomTokenizer"
+
+BLOOM_PRETRAINED_MODEL_ARCHIVE_LIST = [
+    "bigscience/bigscience-small-testing",
+    "bigscience/bloom-350m",
+    "bigscience/bloom-760m",
+    "bigscience/bloom-1b3",
+    "bigscience/bloom-2b5",
+    "bigscience/bloom-6b3",
+    "bigscience/bloom-176b",
+]
+
+
+def split_tensor_along_last_dim(tensor, num_partitions, contiguous_split_chunks=False):
+    """Split a tensor along its last dimension.
+
+    Args:
+        tensor: ([`torch.tensor`], *required*):
+            input tensor to split
+        num_partitions ([`int`], *required*):
+            number of partitions to split the tensor
+        contiguous_split_chunks ([`bool`], *optional*, default=`False`)::
+            If True, make each chunk contiguous in memory.
+    """
+    # Get the size and dimension.
+    last_dim = tensor.dim() - 1
+    numerator, denominator = tensor.size()[last_dim], num_partitions
+    if not (numerator % denominator == 0):
+        raise ValueError(f"{numerator} is not divisible by {denominator}")
+    last_dim_size = numerator // denominator
+    # Split.
+    tensor_list = torch.split(tensor, last_dim_size, dim=last_dim)
+    # Note: torch.split does not create contiguous tensors by default.
+    if contiguous_split_chunks:
+        return tuple(chunk.contiguous() for chunk in tensor_list)
+
+    return tensor_list
+
+
+def attention_mask_func(attention_scores, attention_mask, causal_mask):
+    if attention_mask.dtype == torch.bool:
+        attention_mask_bool = ~attention_mask
+    else:
+        attention_mask_bool = (1 - attention_mask).bool()
+
+    query_length, key_length, n_heads = attention_scores.size(2), attention_scores.size(3), attention_scores.size(1)
+    padded_causal_mask = (
+        attention_mask_bool[:, None, key_length - query_length : key_length, None]
+        + ~causal_mask[:, :, key_length - query_length : key_length, :key_length]
+    ).bool()
+    padded_causal_mask = padded_causal_mask + attention_mask_bool[:, None, None, :key_length].bool()
+    # Make use of floats
+    return (
+        attention_scores.masked_fill_(padded_causal_mask.expand(-1, n_heads, -1, -1), -10000.0),
+        padded_causal_mask,
+    )
+
+
+def build_alibi_tensor(max_seq_len, n_head, dtype=torch.bfloat16):
+    """
+    Link to paper: https://arxiv.org/abs/2108.12409 Alibi tensor is not causal as the original paper mentions, it
+    relies on a translation invariance of softmax for quick implementation: with l being a tensor, and a fixed value
+    `softmax(l+a) = softmax(l)`. Based on
+    https://github.com/ofirpress/attention_with_linear_biases/blob/a35aaca144e0eb6b789dfcb46784c4b8e31b7983/fairseq/models/transformer.py#L742
+
+    Args:
+    Returns tensor shaped (n_head, 1, max_seq_len)
+        max_seq_len: (`int`, *required*):
+            max sequence length
+        n_head: (`int`, *required*):
+            number of heads
+        dtype: (`torch.dtype`, *optional*, default=`torch.bfloat16`):
+            dtype of the output tensor
+    """
+
+    def get_slopes(n):
+        def get_slopes_power_of_2(n):
+            start = 2 ** (-(2 ** -(math.log2(n) - 3)))
+            ratio = start
+            return [start * ratio**i for i in range(n)]
+
+        if math.log2(n).is_integer():
+            return get_slopes_power_of_2(n)
+        else:
+            closest_power_of_2 = 2 ** math.floor(math.log2(n))
+            return (
+                get_slopes_power_of_2(closest_power_of_2)
+                + get_slopes(2 * closest_power_of_2)[0::2][: n - closest_power_of_2]
+            )
+
+    slopes = torch.Tensor(get_slopes(n_head)).unsqueeze(1).unsqueeze(1)
+    arange_tensor = torch.arange(max_seq_len).unsqueeze(0).unsqueeze(0)
+    alibi = slopes * arange_tensor.expand(n_head, -1, -1)
+
+    alibi = alibi.to(dtype)
+
+    return alibi
+
+
+def pre_process_alibi_for_pad(alibi, attention_mask, num_heads):
+    """
+    Args:
+    Pre-process the alibi tensor for padding.
+        alibi: ([`torch.tensor`], *required*):
+            alibi tensor to pre-process
+        attention_mask: ([`torch.tensor`], *required*):
+            attention mask to pre-process"""
+
+    # Sanity check if we are not inferring less tokens than the total sequence length
+    # This usually happens when the inference is done with past_key_values
+    # In this case we re-create the alibi tensor with the correct sequence length
+    if attention_mask.shape[-1] != alibi.shape[-1]:
+        alibi = build_alibi_tensor(attention_mask.shape[-1], num_heads, alibi.dtype).repeat(
+            attention_mask.shape[0], 1, 1
+        )
+    # Get the indexes of the padding tokens
+    index_x0, index_y0 = torch.where(attention_mask == 0.0)
+    index_x1, index_y1 = torch.where(attention_mask == 1.0)
+
+    # Clone the embeddings  - we can detach because the embeddings are not learned
+    # Get a refence tensor
+    slice_reference_alibi = build_alibi_tensor(alibi.shape[-1], num_heads, alibi.dtype)
+
+    # Loop over the batch where the padding is and replace the alibi tensor by the reference tensor
+    # Only where you do not have padding. Replace padding tokens by zeros
+    # This operation can be seen as a shifting operation.
+    for i, index in enumerate(torch.unique(index_x0)):
+        slice_to_modify = torch.zeros_like(slice_reference_alibi)
+        index_shift = index_y1[index_x1 == index]
+        shift_value = len(index_shift)
+        slice_to_modify[:, :, index_shift] = slice_reference_alibi[:, :, :shift_value]
+        alibi[index * num_heads : (index + 1) * num_heads] = slice_to_modify
+    return alibi
+
+
+def dropout_add(x, residual, prob, training):
+    """
+    Dropout add function
+
+    Args:
+        x (`torch.tensor`, *required*):
+            input tensor
+        residual (`torch.tensor`, *rquired*):
+            esidual tensor
+        prob (`float`, *required*):
+            dropout probability
+        training (`bool`, *required*):
+            training mode
+    """
+    out = nn.functional.dropout(x, p=prob, training=training)
+    out = residual + out
+    return out
+
+
+def bloom_gelu_forward(x):
+    """
+    Custom bias GELU function. Adapted from Megatron-DeepSpeed code. Here we use a simple implementation (inference) to
+    make the model jitable.
+
+    Args:
+        x (`torch.tensor`, *required*):
+            input hidden states
+    """
+    return x * 0.5 * (1.0 + torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x)))
+
+
+def bloom_gelu_back(g, x):
+    """
+    gradient of tanh approximation of gelu gradient of actual gelu is: 0.5 * (1. + torch.erf(x * 0.70710678)) +
+    0.3989423 * x * torch.exp(-0.5 * x * x)
+
+    Args:
+        g (`torch.tensor`, *required*):
+            gradient output tensor
+        x (`torch.tensor`, *required*):
+            input tensor
+    """
+    x = x[0]  # x is a tuple of 1 element, needs to unpack it first
+    tanh_out = torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x))
+    # sqrt(2/pi) * 3 * 0.044715 -> 0.1070322243
+    ff = 0.5 * x * ((1 - tanh_out * tanh_out) * (0.79788456 + 0.1070322243 * x * x)) + 0.5 * (1 + tanh_out)
+    return ff * g
+
+
+class GeLUFunction(torch.autograd.Function):
+    @staticmethod
+    def forward(ctx, input):
+        ctx.save_for_backward(input)
+        return bloom_gelu_forward(input)
+
+    @staticmethod
+    def backward(ctx, grad_output):
+        input = ctx.saved_tensors
+        tmp = bloom_gelu_back(grad_output, input)
+        return tmp
+
+
+class BloomGelu(nn.Module):
+    """
+    BloomBiasGelu wrapper function that make use of the simple function on inference mode to make the model
+    torchscriptable and use the autograd function in training mode to get the accurate results of the gradients Partly
+    copied from Megatron-DeepSpeed code and adapted for our needs
+
+    See here why autograd functions are not torchscriptable: https://github.com/pytorch/pytorch/issues/22329
+
+    """
+
+    def __init__(self):
+        super().__init__()
+
+    def forward(self, x):
+        if self.training:
+            return GeLUFunction.apply(x)
+        else:
+            return bloom_gelu_forward(x)
+
+
+class BloomScaledSoftmax(nn.Module):
+    """
+    fused operation: scaling + mask + softmax
+
+    Args:
+        input_in_fp16 (`bool`, *required*):
+            flag to indicate if input in fp16 data format.
+        input_in_bf16 (`bool`, *required*):
+            flag to indicate if input in bf16 data format.
+        scaled_masked_softmax_fusion (`bool`, *required*):
+            flag to indicate user want to use softmax fusion
+        mask_func (`function`, *required*):
+            mask function to be applied.
+        softmax_in_fp32 (`bool`, *required*):
+            if true, softmax in performed at fp32 precision.
+        scale (`float`, *required*):
+            scaling factor used in input tensor scaling.
+    """
+
+    def __init__(self, scaled_masked_softmax_fusion, mask_func, softmax_in_fp32, scale):
+        super().__init__()
+        self.scaled_masked_softmax_fusion = scaled_masked_softmax_fusion
+        self.mask_func = mask_func
+        self.softmax_in_fp32 = softmax_in_fp32
+        self.scale = scale
+
+        if not (self.scale is None or softmax_in_fp32):
+            raise ValueError("softmax should be in fp32 when scaled")
+
+    def forward(self, input, mask, max_positions):
+        input_dtype = input.dtype
+        input_in_16bit = input_dtype in [torch.float16, torch.bfloat16]
+        softmax_dtype = torch.float32 if self.softmax_in_fp32 else input_dtype
+
+        if self.scale is not None:
+            input = input * self.scale
+
+        if mask is not None:
+            mask = mask.to(input.device)
+            causal_mask = (
+                torch.tril(torch.ones((max_positions, max_positions), dtype=torch.bool))
+                .view(1, 1, max_positions, max_positions)
+                .to(input.device)
+            )
+            mask_output, padded_causal_mask = self.mask_func(input, mask, causal_mask)
+            probs = nn.functional.softmax(mask_output, dim=-1, dtype=softmax_dtype) * (~padded_causal_mask)
+        else:
+            probs = nn.functional.softmax(input, dim=-1, dtype=softmax_dtype)
+
+        if input_in_16bit and self.softmax_in_fp32:
+            probs = probs.to(dtype=input_dtype)
+
+        return probs
+
+
+class BloomAttention(nn.Module):
+    def __init__(self, config, layer_number=None):
+        super().__init__()
+
+        self.pretraining_tp = config.pretraining_tp
+        self.slow_but_exact = config.slow_but_exact
+
+        self.hidden_size = config.hidden_size
+        self.num_heads = config.n_head
+        self.head_dim = self.hidden_size // self.num_heads
+        self.split_size = self.hidden_size
+        self.attention_softmax_in_fp32 = config.attention_softmax_in_fp32
+        self.masked_softmax_fusion = config.masked_softmax_fusion
+        self.hidden_dropout = config.hidden_dropout
+
+        if self.head_dim * self.num_heads != self.hidden_size:
+            raise ValueError(
+                f"`hidden_size` must be divisible by num_heads (got `hidden_size`: {self.hidden_size} and `num_heads`:"
+                f" {self.num_heads})."
+            )
+
+        # Layer-wise attention scaling
+        self.layer_number = max(1, layer_number)
+        self.norm_factor = math.sqrt(self.head_dim) * self.layer_number
+
+        # Scaled Softmax
+        self.scale_mask_softmax = BloomScaledSoftmax(
+            self.masked_softmax_fusion,
+            attention_mask_func,
+            self.attention_softmax_in_fp32,
+            self.layer_number,
+        )
+
+        self.query_key_value = nn.Linear(self.hidden_size, 3 * self.hidden_size, bias=True)
+        self.dense = nn.Linear(self.hidden_size, self.hidden_size)
+        self.attention_dropout = nn.Dropout(config.attention_dropout)
+
+    def forward(
+        self,
+        hidden_states,
+        residual,
+        layer_past=None,
+        attention_mask=None,
+        alibi=None,
+        head_mask=None,
+        use_cache=False,
+        output_attentions=False,
+    ):
+        # hidden_states: [batch_size, seq_length, hidden_size]
+        # repeat alibi tensor with the batch size
+        alibi = alibi.repeat(hidden_states.shape[0], 1, 1).to(hidden_states.device)
+
+        # apply preprocessing if the input is padded
+        if attention_mask is not None and 0 in attention_mask:
+            alibi = pre_process_alibi_for_pad(alibi, attention_mask, self.num_heads)
+
+        mixed_x_layer = self.query_key_value(hidden_states)
+
+        # [batch_size, seq_length, 3 x hidden_size] --> [batch_size, seq_length, num_heads, 3 x head_dim]
+        new_tensor_shape = mixed_x_layer.size()[:-1] + (self.num_heads, 3 * self.head_dim)
+        mixed_x_layer = mixed_x_layer.view(*new_tensor_shape)
+
+        # [batch_size, seq_length, num_heads, 3 x head_dim] --> 3  [batch_size, seq_length, num_heads, head_dim]
+        (query_layer, key_layer, value_layer) = split_tensor_along_last_dim(mixed_x_layer, 3)
+
+        if layer_past is not None:
+            past_key, past_value = layer_past
+            key_layer = torch.cat((past_key.type_as(key_layer), key_layer), dim=1)
+            value_layer = torch.cat((past_value.type_as(value_layer), value_layer), dim=1)
+
+        if use_cache is True:
+            present = (key_layer, value_layer)
+        else:
+            present = None
+
+        # [batch_size, head_dim, q_length, k_length]
+        output_size = (query_layer.size(0), query_layer.size(2), query_layer.size(1), key_layer.size(1))
+
+        # [batch_size, q_length, num_heads, head_dim] -> [q_length, batch_size * num_heads, head_dim]
+        query_layer = query_layer.transpose(1, 0).reshape(output_size[2], output_size[0] * output_size[1], -1)
+
+        # [batch_size, k_length, num_heads, head_dim] -> [k_length, batch_size * num_heads, head_dim]
+        key_layer = key_layer.transpose(1, 0).reshape(output_size[3], output_size[0] * output_size[1], -1)
+
+        # slice alibi tensor until the query length
+        sliced_alibi = alibi[: output_size[0] * output_size[1], :, : output_size[3]]
+
+        # Raw attention scores. [batch_size * num_heads, q_length, k_length]
+        beta = 1.0 / self.layer_number
+
+        matmul_result = torch.baddbmm(
+            sliced_alibi,
+            query_layer.transpose(1, 0),
+            key_layer.transpose(1, 0).transpose(1, 2),
+            beta=beta,
+            alpha=(1.0 / self.norm_factor),
+        )
+
+        # change view to [batch_size, num_heads, q_length, k_length]
+        attention_scores = matmul_result.view(*output_size)
+
+        # attention scores and attention mask [b, np, sq, sk]
+        max_positions = max(attention_scores.shape[-1], attention_scores.shape[-2])
+        attention_probs = self.scale_mask_softmax(attention_scores, attention_mask, max_positions).to(
+            value_layer.dtype
+        )
+        attention_probs = self.attention_dropout(attention_probs)
+
+        if head_mask is not None:
+            attention_probs = attention_probs * head_mask
+
+        # context layer shape: [batch_size, num_heads, q_length, head_dim]
+        output_size = (value_layer.size(0), value_layer.size(2), query_layer.size(0), value_layer.size(3))
+
+        # change view [k_length, batch_size x num_heads, head_dim]
+        value_layer = value_layer.transpose(1, 0).reshape(value_layer.size(1), output_size[0] * output_size[1], -1)
+
+        # change view [batch_size x num_heads, q_length, k_length]
+        attention_probs_reshaped = attention_probs.view(output_size[0] * output_size[1], output_size[2], -1)
+
+        # matmul: [batch_size * num_heads, q_length, head_dim]
+        context_layer = torch.bmm(attention_probs_reshaped, value_layer.transpose(0, 1))
+
+        # change view [batch_size, num_heads, q_length, head_dim]
+        context_layer = context_layer.view(*output_size)
+
+        # [batchs_size, num_heads, q_length, head_dim] --> [q_length, batch_size, num_heads, head_dim]
+        context_layer = context_layer.permute(2, 0, 1, 3).contiguous()
+
+        # [q_length, batch_size, num_heads, head_dim] --> [q_length, batch_size, hidden_size]
+        new_context_layer_shape = context_layer.size()[:-2] + (self.hidden_size,)
+
+        context_layer = context_layer.view(*new_context_layer_shape)
+
+        # Output. [q_length, batch_size, hidden_size]
+
+        # aggregate results across tp ranks. See here: https://github.com/pytorch/pytorch/issues/76232
+        if self.pretraining_tp > 1 and self.slow_but_exact:
+            slices = context_layer.shape[-1] / self.pretraining_tp
+            output_tensor = torch.zeros_like(context_layer)
+            for i in range(self.pretraining_tp):
+                output_tensor = output_tensor + nn.functional.linear(
+                    context_layer[:, :, int(i * slices) : int((i + 1) * slices)],
+                    self.dense.weight[:, int(i * slices) : int((i + 1) * slices)],
+                )
+        else:
+            output_tensor = self.dense(context_layer)
+
+        output = output_tensor.transpose(1, 0)
+
+        output = dropout_add(output, residual, self.hidden_dropout, self.training)
+
+        outputs = (output, present)
+        if output_attentions:
+            outputs += (attention_probs,)
+
+        return outputs
+
+
+class BloomMLP(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        hidden_size = config.hidden_size
+
+        self.pretraining_tp = config.pretraining_tp
+        self.slow_but_exact = config.slow_but_exact
+        self.dense_h_to_4h = nn.Linear(hidden_size, 4 * hidden_size)
+        self.dense_4h_to_h = nn.Linear(4 * hidden_size, hidden_size)
+        self.hidden_dropout = config.hidden_dropout
+        self.gelu_impl = BloomGelu()
+
+    def forward(self, hidden_states, residual):
+        hidden_states = self.gelu_impl(self.dense_h_to_4h(hidden_states))
+
+        if self.pretraining_tp > 1 and self.slow_but_exact:
+            intermediate_output = torch.zeros_like(residual)
+            slices = self.dense_4h_to_h.weight.shape[-1] / self.pretraining_tp
+            for i in range(self.pretraining_tp):
+                intermediate_output = intermediate_output + nn.functional.linear(
+                    hidden_states[:, :, int(i * slices) : int((i + 1) * slices)],
+                    self.dense_4h_to_h.weight[:, int(i * slices) : int((i + 1) * slices)],
+                )
+        else:
+            intermediate_output = self.dense_4h_to_h(hidden_states)
+
+        output = dropout_add(intermediate_output, residual, self.hidden_dropout, self.training)
+
+        return output
+
+
+class BloomBlock(nn.Module):
+    def __init__(self, config, layer_number=None):
+        super().__init__()
+        hidden_size = config.hidden_size
+
+        self.input_layernorm = LayerNorm(hidden_size, eps=config.layer_norm_epsilon)
+        self.n_head = config.n_head
+        self.self_attention = BloomAttention(config, layer_number=layer_number)
+        self.post_attention_layernorm = LayerNorm(hidden_size, eps=config.layer_norm_epsilon)
+
+        self.mlp = BloomMLP(config)
+
+        self.apply_residual_connection_post_layernorm = config.apply_residual_connection_post_layernorm
+        self.hidden_dropout = config.hidden_dropout
+
+    def forward(
+        self,
+        hidden_states,
+        layer_past=None,
+        attention_mask=None,
+        head_mask=None,
+        use_cache=False,
+        output_attentions=False,
+        alibi=None,
+    ):
+        # hidden_states: [batch_size, seq_length, hidden_size]
+
+        # Layer norm at the beginning of the transformer layer.
+        layernorm_output = self.input_layernorm(hidden_states)
+
+        # Layer norm post the self attention.
+        if self.apply_residual_connection_post_layernorm:
+            residual = layernorm_output
+        else:
+            residual = hidden_states
+
+        # Self attention.
+        attn_outputs = self.self_attention(
+            layernorm_output,
+            residual,
+            layer_past=layer_past,
+            attention_mask=attention_mask,
+            alibi=alibi,
+            head_mask=head_mask,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+        )
+
+        attention_output = attn_outputs[0]
+
+        outputs = attn_outputs[1:]
+
+        layernorm_output = self.post_attention_layernorm(attention_output)
+
+        # Get residual
+        if self.apply_residual_connection_post_layernorm:
+            residual = layernorm_output
+        else:
+            residual = attention_output
+
+        # MLP.
+        output = self.mlp(layernorm_output, residual)
+
+        if use_cache:
+            outputs = (output,) + outputs
+        else:
+            outputs = (output,) + outputs[1:]
+
+        return outputs  # hidden_states, present, attentions
+
+
+class BloomPreTrainedModel(PreTrainedModel):
+    _keys_to_ignore_on_load_missing = [r"h.*.self_attention.scale_mask_softmax.causal_mask", r"lm_head.weight"]
+    """
+    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
+    models.
+    """
+
+    config_class = BloomConfig
+    base_model_prefix = "transformer"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["BloomBlock"]
+
+    def __init__(self, *inputs, **kwargs):
+        super().__init__(*inputs, **kwargs)
+
+    def _init_weights(self, module):
+        """Initialize the weights."""
+        if isinstance(module, (nn.Linear)):
+            # Slightly different from the TF version which uses truncated_normal for initialization
+            # cf https://github.com/pytorch/pytorch/pull/5617
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+        elif isinstance(module, LayerNorm):
+            module.bias.data.zero_()
+            module.weight.data.fill_(1.0)
+
+    def _set_gradient_checkpointing(self, module, value=False):
+        if isinstance(module, BloomModel):
+            module.gradient_checkpointing = value
+
+
+BLOOM_START_DOCSTRING = r"""
+
+    This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
+    library implements for all its model (such as downloading or saving, resizing the input embeddings etc.)
+
+    This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
+    Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
+    and behavior.
+
+    Parameters:
+        config ([`BloomConfig`]): Model configuration class with all the parameters of the model.
+            Initializing with a config file does not load the weights associated with the model, only the
+            configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights.
+"""
+
+BLOOM_INPUTS_DOCSTRING = r"""
+    Args:
+        input_ids (`torch.LongTensor` of shape `(batch_size, input_ids_length)`):
+            `input_ids_length` = `sequence_length` if `past_key_values` is `None` else
+            `past_key_values[0][0].shape[-2]` (`sequence_length` of input past key value states). Indices of input
+            sequence tokens in the vocabulary.
+
+            If `past_key_values` is used, only `input_ids` that do not have their past calculated should be passed as
+            `input_ids`.
+
+            Indices can be obtained using [`BloomTokenizer`]. See [`PreTrainedTokenizer.encode`] and
+            [`PreTrainedTokenizer.__call__`] for details.
+
+            [What are input IDs?](../glossary#input-ids)
+        past_key_values (`Tuple[Tuple[torch.Tensor]]` of length `config.n_layers`):
+            Contains precomputed hidden-states (key and values in the attention blocks) as computed by the model (see
+            `past_key_values` output below). Can be used to speed up sequential decoding. The `input_ids` which have
+            their past given to this model should not be passed as `input_ids` as they have already been computed.
+        attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+
+            [What are attention masks?](../glossary#attention-mask)
+        position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
+            config.max_position_embeddings - 1]`.
+
+            [What are position IDs?](../glossary#position-ids)
+        head_mask (`torch.FloatTensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*):
+            Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`:
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+
+        inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
+            Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
+            is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
+            model's internal embedding lookup matrix.
+
+            If `past_key_values` is used, optionally only the last `inputs_embeds` have to be input (see
+            `past_key_values`).
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+        output_attentions (`bool`, *optional*):
+            Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned
+            tensors for more detail.
+        output_hidden_states (`bool`, *optional*):
+            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
+            more detail.
+        return_dict (`bool`, *optional*):
+            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
+"""
+
+
+@add_start_docstrings(
+    "The bare Bloom Model transformer outputting raw hidden-states without any specific head on top.",
+    BLOOM_START_DOCSTRING,
+)
+class BloomModel(BloomPreTrainedModel):
+    def __init__(self, config):
+        super().__init__(config)
+
+        self.embed_dim = config.hidden_size
+        self.n_head = config.n_head
+
+        # Embedding + LN Embedding
+        self.word_embeddings = nn.Embedding(config.vocab_size, self.embed_dim)
+
+        self.word_embeddings_layernorm = LayerNorm(self.embed_dim, eps=config.layer_norm_epsilon)
+
+        # Transformer blocks
+        self.h = nn.ModuleList([BloomBlock(config, layer_number=i) for i in range(config.num_hidden_layers)])
+
+        # Final Layer Norm
+        self.ln_f = LayerNorm(self.embed_dim, eps=config.layer_norm_epsilon)
+
+        self.gradient_checkpointing = False
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.word_embeddings
+
+    def set_input_embeddings(self, new_embeddings):
+        self.word_embeddings = new_embeddings
+
+    @add_start_docstrings_to_model_forward(BLOOM_INPUTS_DOCSTRING)
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=BaseModelOutputWithPastAndCrossAttentions,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    def forward(
+        self,
+        input_ids=None,
+        past_key_values=None,
+        attention_mask=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        use_cache=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            input_shape = input_ids.size()
+            input_ids = input_ids.view(-1, input_shape[-1])
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        if past_key_values is None:
+            past_key_values = tuple([None] * len(self.h))
+
+        # Prepare head mask if needed
+        # 1.0 in head_mask indicate we keep the head
+        # attention_probs has shape bsz x n_head x N x N
+        # head_mask has shape n_layer x batch x n_head x N x N
+        head_mask = self.get_head_mask(head_mask, self.config.n_layer)
+
+        if inputs_embeds is None:
+            inputs_embeds = self.word_embeddings(input_ids)
+
+        hidden_states = self.word_embeddings_layernorm(inputs_embeds)
+
+        output_shape = input_shape + (hidden_states.size(-1),)
+
+        presents = () if use_cache else None
+        all_self_attentions = () if output_attentions else None
+        all_hidden_states = () if output_hidden_states else None
+
+        # Compute alibi tensor: check build_alibi_tensor documentation
+        current_sequence_length = hidden_states.shape[1]
+        if past_key_values[0] is not None:
+            current_sequence_length += past_key_values[0][0].shape[1]
+        alibi = build_alibi_tensor(current_sequence_length, self.n_head, hidden_states.dtype)
+
+        for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)):
+
+            if output_hidden_states:
+                all_hidden_states = all_hidden_states + (hidden_states,)
+
+            if self.gradient_checkpointing and self.training:
+
+                if use_cache:
+                    logger.warning(
+                        "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
+                    )
+                    use_cache = False
+
+                def create_custom_forward(module):
+                    def custom_forward(*inputs):
+                        # None for past_key_value
+                        return module(*inputs, use_cache, output_attentions, alibi)
+
+                    return custom_forward
+
+                outputs = torch.utils.checkpoint.checkpoint(
+                    create_custom_forward(block),
+                    hidden_states,
+                    None,
+                    attention_mask,
+                    head_mask[i],
+                )
+            else:
+                outputs = block(
+                    hidden_states,
+                    layer_past=layer_past,
+                    attention_mask=attention_mask,
+                    head_mask=head_mask[i],
+                    use_cache=use_cache,
+                    output_attentions=output_attentions,
+                    alibi=alibi,
+                )
+
+            hidden_states = outputs[0]
+            if use_cache is True:
+                presents = presents + (outputs[1],)
+
+            if output_attentions:
+                all_self_attentions = all_self_attentions + (outputs[2 if use_cache else 1],)
+
+        # Add last hidden state
+        hidden_states = self.ln_f(hidden_states)
+
+        if output_hidden_states:
+            all_hidden_states = all_hidden_states + (hidden_states,)
+
+        hidden_states = hidden_states.view(output_shape)
+
+        if not return_dict:
+            return tuple(v for v in [hidden_states, presents, all_hidden_states, all_self_attentions] if v is not None)
+
+        return BaseModelOutputWithPastAndCrossAttentions(
+            last_hidden_state=hidden_states,
+            past_key_values=presents,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attentions,
+        )
+
+
+@add_start_docstrings(
+    """
+    The Bloom Model transformer with a language modeling head on top (linear layer with weights tied to the input
+    embeddings).
+    """,
+    BLOOM_START_DOCSTRING,
+)
+class BloomForCausalLM(BloomPreTrainedModel):
+    _keys_to_ignore_on_load_missing = [r"h.*.self_attention.scale_mask_softmax.causal_mask", r"lm_head.weight"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.transformer = BloomModel(config)
+        self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def prepare_inputs_for_generation(self, input_ids, past=None, **kwargs):
+        # only last token for inputs_ids if past is defined in kwargs
+        if past:
+            input_ids = input_ids[:, -1].unsqueeze(-1)
+
+        attention_mask = kwargs.get("attention_mask", None)
+        position_ids = kwargs.get("position_ids", None)
+
+        if attention_mask is not None and position_ids is None:
+            # create position_ids on the fly for batch generation
+            position_ids = attention_mask.long().cumsum(-1) - 1
+            position_ids.masked_fill_(attention_mask == 0, 1)
+            if past:
+                position_ids = position_ids[:, -1].unsqueeze(-1)
+        else:
+            position_ids = None
+        return {
+            "input_ids": input_ids,
+            "past_key_values": past,
+            "use_cache": kwargs.get("use_cache"),
+            "position_ids": position_ids,
+            "attention_mask": attention_mask,
+        }
+
+    @add_start_docstrings_to_model_forward(BLOOM_INPUTS_DOCSTRING)
+    @add_code_sample_docstrings(
+        processor_class=_TOKENIZER_FOR_DOC,
+        checkpoint=_CHECKPOINT_FOR_DOC,
+        output_type=CausalLMOutputWithCrossAttentions,
+        config_class=_CONFIG_FOR_DOC,
+    )
+    def forward(
+        self,
+        input_ids=None,
+        past_key_values=None,
+        attention_mask=None,
+        position_ids=None,
+        head_mask=None,
+        inputs_embeds=None,
+        labels=None,
+        use_cache=None,
+        output_attentions=None,
+        output_hidden_states=None,
+        return_dict=None,
+    ):
+        r"""
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for language modeling. Note that the labels **are shifted** inside the model, i.e. you can set
+            `labels = input_ids` Indices are selected in `[-100, 0, ..., config.vocab_size]` All labels set to `-100`
+            are ignored (masked), the loss is only computed for labels in `[0, ..., config.vocab_size]`
+        """
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        transformer_outputs = self.transformer(
+            input_ids,
+            past_key_values=past_key_values,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        hidden_states = transformer_outputs[0]
+
+        lm_logits = self.lm_head(hidden_states)
+
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            shift_logits = lm_logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))
+
+        if not return_dict:
+            output = (lm_logits,) + transformer_outputs[1:]
+            return ((loss,) + output) if loss is not None else output
+
+        return CausalLMOutputWithCrossAttentions(
+            loss=loss,
+            logits=lm_logits,
+            past_key_values=transformer_outputs.past_key_values,
+            hidden_states=transformer_outputs.hidden_states,
+            attentions=transformer_outputs.attentions,
+        )
+
+    @staticmethod
+    def _reorder_cache(past: Tuple[Tuple[torch.Tensor]], beam_idx: torch.Tensor) -> Tuple[Tuple[torch.Tensor]]:
+        """
+        This function is used to re-order the `past_key_values` cache if [`~PreTrainedModel.beam_search`] or
+        [`~PreTrainedModel.beam_sample`] is called. This is required to match `past_key_values` with the correct
+        beam_idx at every generation step.
+        """
+        return tuple(
+            tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past)
+            for layer_past in past
+        )

src/transformers/models/bloom/tokenization_bloom_fast.py

+# coding=utf-8
+# Copyright 2022 The HuggingFace Inc. team.
+#
+# 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.
+"""Tokenization classes for Bloom."""
+
+
+import json
+from typing import TYPE_CHECKING, List, Optional, Tuple
+
+from tokenizers import pre_tokenizers
+
+from ...tokenization_utils_base import BatchEncoding
+from ...tokenization_utils_fast import PreTrainedTokenizerFast
+from ...utils import logging
+
+
+if TYPE_CHECKING:
+    from transformers.pipelines.conversational import Conversation
+
+
+logger = logging.get_logger(__name__)
+
+VOCAB_FILES_NAMES = {"tokenizer_file": "tokenizer.json"}
+
+PRETRAINED_VOCAB_FILES_MAP = {
+    "tokenizer_file": {
+        "bigscience/tokenizer": "https://huggingface.co/bigscience/tokenizer/blob/main/tokenizer.json",
+        "bigscience/bloom-350m": "https://huggingface.co/bigscience/bloom-350m/blob/main/tokenizer.json",
+        "bigscience/bloom-760m": "https://huggingface.co/bigscience/bloom-760m/blob/main/tokenizer.json",
+        "bigscience/bloom-1b3": "https://huggingface.co/bigscience/bloom-1b3/blob/main/tokenizer.json",
+        "bigscience/bloom-2b5": "https://huggingface.co/bigscience/bloom-2b5/blob/main/tokenizer.json",
+        "bigscience/bloom-6b3": "https://huggingface.co/bigscience/bloom-2b5/blob/main/tokenizer.json",
+        "bigscience/bloom": "https://huggingface.co/bigscience/bloom/blob/main/tokenizer.json",
+    },
+}
+
+PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
+    "bigscience/tokenizer": 1024,
+    "bigscience/bloom-350m": 1024,
+    "bigscience/bloom-760m": 1024,
+    "bigscience/bloom-1b3": 1024,
+    "bigscience/bloom-2b5": 1024,
+    "bigscience/bloom-6b3": 1024,
+    "bigscience/bloom": 1024,
+}
+
+
+class BloomTokenizerFast(PreTrainedTokenizerFast):
+    """
+    Construct a "fast" Bloom tokenizer (backed by HuggingFace's *tokenizers* library). Based on byte-level
+    Byte-Pair-Encoding.
+
+    This tokenizer has been trained to treat spaces like parts of the tokens (a bit like sentencepiece) so a word will
+    be encoded differently whether it is at the beginning of the sentence (without space) or not:
+
+    ```
+    >>> from transformers import BloomTokenizerFast
+    >>> tokenizer = BloomTokenizerFast.from_pretrained("bigscience/bloom")
+    >>> tokenizer("Hello world")['input_ids']
+    [15496, 995]
+    >>> tokenizer(" Hello world")['input_ids']
+    [18435, 995]
+    ```
+
+    You can get around that behavior by passing `add_prefix_space=True` when instantiating this tokenizer, but since
+    the model was not pretrained this way, it might yield a decrease in performance.
+
+    <Tip>
+
+    When used with `is_split_into_words=True`, this tokenizer needs to be instantiated with `add_prefix_space=True`.
+
+    </Tip>
+
+    This tokenizer inherits from [`PreTrainedTokenizerFast`] which contains most of the main methods. Users should
+    refer to this superclass for more information regarding those methods.
+
+    Args:
+        vocab_file (`str`):
+            Path to the vocabulary file.
+        merges_file (`str`):
+            Path to the merges file.
+        errors (`str`, *optional*, defaults to `"replace"`):
+            Paradigm to follow when decoding bytes to UTF-8. See
+            [bytes.decode](https://docs.python.org/3/library/stdtypes.html#bytes.decode) for more information.
+        unk_token (`str`, *optional*, defaults to `<|endoftext|>`):
+            The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this
+            token instead.
+        bos_token (`str`, *optional*, defaults to `<|endoftext|>`):
+            The beginning of sequence token.
+        eos_token (`str`, *optional*, defaults to `<|endoftext|>`):
+            The end of sequence token.
+        add_prefix_space (`bool`, *optional*, defaults to `False`):
+            Whether or not to add an initial space to the input. This allows to treat the leading word just as any
+            other word. (Bloom tokenizer detect beginning of words by the preceding space).
+        trim_offsets (`bool`, *optional*, defaults to `True`):
+            Whether or not the post-processing step should trim offsets to avoid including whitespaces.
+    """
+
+    vocab_files_names = VOCAB_FILES_NAMES
+    pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
+    max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
+    model_input_names = ["input_ids", "attention_mask"]
+    slow_tokenizer_class = None
+
+    def __init__(
+        self,
+        vocab_file=None,
+        merges_file=None,
+        tokenizer_file=None,
+        unk_token="<unk>",
+        bos_token="<s>",
+        eos_token="</s>",
+        pad_token="<pad>",
+        add_prefix_space=False,
+        **kwargs
+    ):
+        super().__init__(
+            vocab_file,
+            merges_file,
+            tokenizer_file=tokenizer_file,
+            unk_token=unk_token,
+            bos_token=bos_token,
+            eos_token=eos_token,
+            pad_token=pad_token,
+            add_prefix_space=add_prefix_space,
+            **kwargs,
+        )
+
+        pre_tok_state = json.loads(self.backend_tokenizer.pre_tokenizer.__getstate__())
+        if pre_tok_state.get("add_prefix_space", add_prefix_space) != add_prefix_space:
+            pre_tok_class = getattr(pre_tokenizers, pre_tok_state.pop("type"))
+            pre_tok_state["add_prefix_space"] = add_prefix_space
+            self.backend_tokenizer.pre_tokenizer = pre_tok_class(**pre_tok_state)
+
+        self.add_prefix_space = add_prefix_space
+
+    def _batch_encode_plus(self, *args, **kwargs) -> BatchEncoding:
+        is_split_into_words = kwargs.get("is_split_into_words", False)
+        if not (self.add_prefix_space or not is_split_into_words):
+            raise Exception(
+                f"You need to instantiate {self.__class__.__name__} with add_prefix_space=True to use it with"
+                " pretokenized inputs."
+            )
+
+        return super()._batch_encode_plus(*args, **kwargs)
+
+    def _encode_plus(self, *args, **kwargs) -> BatchEncoding:
+        is_split_into_words = kwargs.get("is_split_into_words", False)
+
+        if not (self.add_prefix_space or not is_split_into_words):
+            raise Exception(
+                f"You need to instantiate {self.__class__.__name__} with add_prefix_space=True to use it with"
+                " pretokenized inputs."
+            )
+
+        return super()._encode_plus(*args, **kwargs)
+
+    def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
+        files = self._tokenizer.model.save(save_directory, name=filename_prefix)
+        return tuple(files)
+
+    def _build_conversation_input_ids(self, conversation: "Conversation") -> List[int]:
+        """This corresponds to DialoGPT variants of models."""
+        input_ids = []
+        for is_user, text in conversation.iter_texts():
+            input_ids.extend(self.encode(text, add_special_tokens=False) + [self.eos_token_id])
+
+        if len(input_ids) > self.model_max_length:
+            input_ids = input_ids[-self.model_max_length :]
+        return input_ids

src/transformers/utils/dummy_pt_objects.py

@@ -966,6 +966,30 @@ class BlenderbotSmallPreTrainedModel(metaclass=DummyObject):
         requires_backends(self, ["torch"])
 
 
+BLOOM_PRETRAINED_MODEL_ARCHIVE_LIST = None
+
+
+class BloomForCausalLM(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class BloomModel(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
+class BloomPreTrainedModel(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+
 CAMEMBERT_PRETRAINED_MODEL_ARCHIVE_LIST = None
 
 

src/transformers/utils/dummy_tokenizers_objects.py

@@ -52,6 +52,13 @@ class BlenderbotSmallTokenizerFast(metaclass=DummyObject):
         requires_backends(self, ["tokenizers"])
 
 
+class BloomTokenizerFast(metaclass=DummyObject):
+    _backends = ["tokenizers"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["tokenizers"])
+
+
 class CamembertTokenizerFast(metaclass=DummyObject):
     _backends = ["tokenizers"]
 

tests/deepspeed/test_model_zoo.py

@@ -58,6 +58,7 @@ BERT_TINY = "hf-internal-testing/tiny-bert"
 BIGBIRD_PEGASUS_TINY = "hf-internal-testing/tiny-random-bigbird_pegasus"
 BIG_BIRD_TINY = "hf-internal-testing/tiny-random-big_bird"
 BLENDERBOT_TINY = "hf-internal-testing/tiny-random-blenderbot"
+BLOOM_TINY = "bigscience/bigscience-small-testing"
 DEBERTA_TINY = "hf-internal-testing/tiny-random-deberta"
 DEBERTA_V2_TINY = "hf-internal-testing/tiny-random-deberta-v2"
 DISTILBERT_TINY = "sshleifer/tiny-distilbert-base-cased"
@@ -183,6 +184,7 @@ def make_task_cmds():
             "big_bird",
             "bigbird_pegasus",
             "blenderbot",
+            "bloom",
             "gpt2",
             "gpt_neo",
             "gptj",