SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

GPT Neo (#10848)

Browse Source 
* lets begin

* boom boom

* fix out proj in attn

* fix attention

* fix local attention

* add tokenizer

* fix imports

* autotokenizer

* fix checkpoint name

* cleanup

* more clean-up

* more cleanup

* output attentions

* fix attn mask creation

* fix imports

* config doc

* add tests

* add slow tests

* quality

* add conversion script

* copyright

* typo

* another bites the dust

* fix attention tests

* doc

* add embed init in convert function

* fix copies

* remove tokenizer

* enable caching

* address review comments

* improve config and create attn layer list internally

* more consistent naming

* init hf config from mesh-tf config json file

* remove neo tokenizer from doc

* handle attention_mask in local attn layer

* attn_layers => attention_layers

* add tokenizer_class in config

* fix docstring

* raise if len of attention_layers is not same as num_layers

* remove tokenizer_class from config

* more consistent naming

* fix doc

* fix checkpoint names

* fp16 compat

* Apply suggestions from code review

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
This commit is contained in:
Suraj Patil
2021-03-30 19:12:30 +05:30
committed by GitHub
parent a04eb8d369
commit 860264379f
14 changed files with 1953 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files

57
docs/source/index.rst
View File

@@ -151,79 +151,81 @@ and conversion utilities for the following models:
22. :doc:`GPT-2 <model_doc/gpt2>` (from OpenAI) released with the paper `Language Models are Unsupervised Multitask
Learners <https://blog.openai.com/better-language-models/>`__ by Alec Radford*, Jeffrey Wu*, Rewon Child, David
Luan, Dario Amodei** and Ilya Sutskever**.
23. :doc:`I-BERT <model_doc/ibert>` (from Berkeley) released with the paper `I-BERT: Integer-only BERT Quantization
23. :doc:`GPT Neo <model_doc/gpt_neo>` (from EleutherAI) released in the repository `EleutherAI/gpt-neo
<https://github.com/EleutherAI/gpt-neo>`__ by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy.
24. :doc:`I-BERT <model_doc/ibert>` (from Berkeley) released with the paper `I-BERT: Integer-only BERT Quantization
<https://arxiv.org/abs/2101.01321>`__ by Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney, Kurt Keutzer
24. :doc:`LayoutLM <model_doc/layoutlm>` (from Microsoft Research Asia) released with the paper `LayoutLM: Pre-training
25. :doc:`LayoutLM <model_doc/layoutlm>` (from Microsoft Research Asia) released with the paper `LayoutLM: Pre-training
of Text and Layout for Document Image Understanding <https://arxiv.org/abs/1912.13318>`__ by Yiheng Xu, Minghao Li,
Lei Cui, Shaohan Huang, Furu Wei, Ming Zhou.
25. :doc:`LED <model_doc/led>` (from AllenAI) released with the paper `Longformer: The Long-Document Transformer
26. :doc:`LED <model_doc/led>` (from AllenAI) released with the paper `Longformer: The Long-Document Transformer
<https://arxiv.org/abs/2004.05150>`__ by Iz Beltagy, Matthew E. Peters, Arman Cohan.
26. :doc:`Longformer <model_doc/longformer>` (from AllenAI) released with the paper `Longformer: The Long-Document
27. :doc:`Longformer <model_doc/longformer>` (from AllenAI) released with the paper `Longformer: The Long-Document
Transformer <https://arxiv.org/abs/2004.05150>`__ by Iz Beltagy, Matthew E. Peters, Arman Cohan.
27. :doc:`LXMERT <model_doc/lxmert>` (from UNC Chapel Hill) released with the paper `LXMERT: Learning Cross-Modality
28. :doc:`LXMERT <model_doc/lxmert>` (from UNC Chapel Hill) released with the paper `LXMERT: Learning Cross-Modality
Encoder Representations from Transformers for Open-Domain Question Answering <https://arxiv.org/abs/1908.07490>`__
by Hao Tan and Mohit Bansal.
28. :doc:`M2M100 <model_doc/m2m_100>` (from Facebook) released with the paper `Beyond English-Centric Multilingual
29. :doc:`M2M100 <model_doc/m2m_100>` (from Facebook) released with the paper `Beyond English-Centric Multilingual
Machine Translation <https://arxiv.org/abs/2010.11125>`__ by by Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi
Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman
Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, Armand Joulin.
29. :doc:`MarianMT <model_doc/marian>` Machine translation models trained using `OPUS <http://opus.nlpl.eu/>`__ data by
30. :doc:`MarianMT <model_doc/marian>` Machine translation models trained using `OPUS <http://opus.nlpl.eu/>`__ data by
Jörg Tiedemann. The `Marian Framework <https://marian-nmt.github.io/>`__ is being developed by the Microsoft
Translator Team.
30. :doc:`MBart <model_doc/mbart>` (from Facebook) released with the paper `Multilingual Denoising Pre-training for
31. :doc:`MBart <model_doc/mbart>` (from Facebook) released with the paper `Multilingual Denoising Pre-training for
Neural Machine Translation <https://arxiv.org/abs/2001.08210>`__ by Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li,
Sergey Edunov, Marjan Ghazvininejad, Mike Lewis, Luke Zettlemoyer.
31. :doc:`MBart-50 <model_doc/mbart>` (from Facebook) released with the paper `Multilingual Translation with Extensible
32. :doc:`MBart-50 <model_doc/mbart>` (from Facebook) released with the paper `Multilingual Translation with Extensible
Multilingual Pretraining and Finetuning <https://arxiv.org/abs/2008.00401>`__ by Yuqing Tang, Chau Tran, Xian Li,
Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, Angela Fan.
32. :doc:`MPNet <model_doc/mpnet>` (from Microsoft Research) released with the paper `MPNet: Masked and Permuted
33. :doc:`MPNet <model_doc/mpnet>` (from Microsoft Research) released with the paper `MPNet: Masked and Permuted
Pre-training for Language Understanding <https://arxiv.org/abs/2004.09297>`__ by Kaitao Song, Xu Tan, Tao Qin,
Jianfeng Lu, Tie-Yan Liu.
33. :doc:`MT5 <model_doc/mt5>` (from Google AI) released with the paper `mT5: A massively multilingual pre-trained
34. :doc:`MT5 <model_doc/mt5>` (from Google AI) released with the paper `mT5: A massively multilingual pre-trained
text-to-text transformer <https://arxiv.org/abs/2010.11934>`__ by Linting Xue, Noah Constant, Adam Roberts, Mihir
Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel.
34. :doc:`Pegasus <model_doc/pegasus>` (from Google) released with the paper `PEGASUS: Pre-training with Extracted
35. :doc:`Pegasus <model_doc/pegasus>` (from Google) released with the paper `PEGASUS: Pre-training with Extracted
Gap-sentences for Abstractive Summarization <https://arxiv.org/abs/1912.08777>`__> by Jingqing Zhang, Yao Zhao,
Mohammad Saleh and Peter J. Liu.
35. :doc:`ProphetNet <model_doc/prophetnet>` (from Microsoft Research) released with the paper `ProphetNet: Predicting
36. :doc:`ProphetNet <model_doc/prophetnet>` (from Microsoft Research) released with the paper `ProphetNet: Predicting
Future N-gram for Sequence-to-Sequence Pre-training <https://arxiv.org/abs/2001.04063>`__ by Yu Yan, Weizhen Qi,
Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
36. :doc:`Reformer <model_doc/reformer>` (from Google Research) released with the paper `Reformer: The Efficient
37. :doc:`Reformer <model_doc/reformer>` (from Google Research) released with the paper `Reformer: The Efficient
Transformer <https://arxiv.org/abs/2001.04451>`__ by Nikita Kitaev, Łukasz Kaiser, Anselm Levskaya.
37. :doc:`RoBERTa <model_doc/roberta>` (from Facebook), released together with the paper a `Robustly Optimized BERT
38. :doc:`RoBERTa <model_doc/roberta>` (from Facebook), released together with the paper a `Robustly Optimized BERT
Pretraining Approach <https://arxiv.org/abs/1907.11692>`__ by Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar
Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, Veselin Stoyanov.
38. :doc:`SpeechToTextTransformer <model_doc/speech_to_text>` (from Facebook), released together with the paper
39. :doc:`SpeechToTextTransformer <model_doc/speech_to_text>` (from Facebook), released together with the paper
`fairseq S2T: Fast Speech-to-Text Modeling with fairseq <https://arxiv.org/abs/2010.05171>`__ by Changhan Wang, Yun
Tang, Xutai Ma, Anne Wu, Dmytro Okhonko, Juan Pino.
39. :doc:`SqueezeBert <model_doc/squeezebert>` released with the paper `SqueezeBERT: What can computer vision teach NLP
40. :doc:`SqueezeBert <model_doc/squeezebert>` released with the paper `SqueezeBERT: What can computer vision teach NLP
about efficient neural networks? <https://arxiv.org/abs/2006.11316>`__ by Forrest N. Iandola, Albert E. Shaw, Ravi
Krishna, and Kurt W. Keutzer.
40. :doc:`T5 <model_doc/t5>` (from Google AI) released with the paper `Exploring the Limits of Transfer Learning with a
41. :doc:`T5 <model_doc/t5>` (from Google AI) released with the paper `Exploring the Limits of Transfer Learning with a
Unified Text-to-Text Transformer <https://arxiv.org/abs/1910.10683>`__ by Colin Raffel and Noam Shazeer and Adam
Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu.
41. :doc:`TAPAS <model_doc/tapas>` (from Google AI) released with the paper `TAPAS: Weakly Supervised Table Parsing via
42. :doc:`TAPAS <model_doc/tapas>` (from Google AI) released with the paper `TAPAS: Weakly Supervised Table Parsing via
Pre-training <https://arxiv.org/abs/2004.02349>`__ by Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Müller,
Francesco Piccinno and Julian Martin Eisenschlos.
42. :doc:`Transformer-XL <model_doc/transformerxl>` (from Google/CMU) released with the paper `Transformer-XL:
43. :doc:`Transformer-XL <model_doc/transformerxl>` (from Google/CMU) released with the paper `Transformer-XL:
Attentive Language Models Beyond a Fixed-Length Context <https://arxiv.org/abs/1901.02860>`__ by Zihang Dai*,
Zhilin Yang*, Yiming Yang, Jaime Carbonell, Quoc V. Le, Ruslan Salakhutdinov.
43. :doc:`Wav2Vec2 <model_doc/wav2vec2>` (from Facebook AI) released with the paper `wav2vec 2.0: A Framework for
44. :doc:`Wav2Vec2 <model_doc/wav2vec2>` (from Facebook AI) released with the paper `wav2vec 2.0: A Framework for
Self-Supervised Learning of Speech Representations <https://arxiv.org/abs/2006.11477>`__ by Alexei Baevski, Henry
Zhou, Abdelrahman Mohamed, Michael Auli.
44. :doc:`XLM <model_doc/xlm>` (from Facebook) released together with the paper `Cross-lingual Language Model
45. :doc:`XLM <model_doc/xlm>` (from Facebook) released together with the paper `Cross-lingual Language Model
Pretraining <https://arxiv.org/abs/1901.07291>`__ by Guillaume Lample and Alexis Conneau.
45. :doc:`XLM-ProphetNet <model_doc/xlmprophetnet>` (from Microsoft Research) released with the paper `ProphetNet:
46. :doc:`XLM-ProphetNet <model_doc/xlmprophetnet>` (from Microsoft Research) released with the paper `ProphetNet:
Predicting Future N-gram for Sequence-to-Sequence Pre-training <https://arxiv.org/abs/2001.04063>`__ by Yu Yan,
Weizhen Qi, Yeyun Gong, Dayiheng Liu, Nan Duan, Jiusheng Chen, Ruofei Zhang and Ming Zhou.
46. :doc:`XLM-RoBERTa <model_doc/xlmroberta>` (from Facebook AI), released together with the paper `Unsupervised
47. :doc:`XLM-RoBERTa <model_doc/xlmroberta>` (from Facebook AI), released together with the paper `Unsupervised
Cross-lingual Representation Learning at Scale <https://arxiv.org/abs/1911.02116>`__ by Alexis Conneau*, Kartikay
Khandelwal*, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke
Zettlemoyer and Veselin Stoyanov.
47. :doc:`XLNet <model_doc/xlnet>` (from Google/CMU) released with the paper `​XLNet: Generalized Autoregressive
48. :doc:`XLNet <model_doc/xlnet>` (from Google/CMU) released with the paper `​XLNet: Generalized Autoregressive
Pretraining for Language Understanding <https://arxiv.org/abs/1906.08237>`__ by Zhilin Yang*, Zihang Dai*, Yiming
Yang, Jaime Carbonell, Ruslan Salakhutdinov, Quoc V. Le.
48. :doc:`XLSR-Wav2Vec2 <model_doc/xlsr_wav2vec2>` (from Facebook AI) released with the paper `Unsupervised
49. :doc:`XLSR-Wav2Vec2 <model_doc/xlsr_wav2vec2>` (from Facebook AI) released with the paper `Unsupervised
Cross-Lingual Representation Learning For Speech Recognition <https://arxiv.org/abs/2006.13979>`__ by Alexis
Conneau, Alexei Baevski, Ronan Collobert, Abdelrahman Mohamed, Michael Auli.
@@ -280,6 +282,8 @@ TensorFlow and/or Flax.
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| Funnel Transformer | ✅ | ✅ | ✅ | ✅ | ❌ |
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| GPT Neo | ❌ | ❌ | ✅ | ❌ | ❌ |
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| I-BERT | ❌ | ❌ | ✅ | ❌ | ❌ |
+-----------------------------+----------------+----------------+-----------------+--------------------+--------------+
| LED | ✅ | ✅ | ✅ | ✅ | ❌ |
@@ -443,6 +447,7 @@ TensorFlow and/or Flax.
model_doc/mt5
model_doc/gpt
model_doc/gpt2
model_doc/gpt_neo
model_doc/pegasus
model_doc/phobert
model_doc/prophetnet

65
docs/source/model_doc/gpt_neo.rst Normal file
View File

@@ -0,0 +1,65 @@
..
Copyright 2021 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.
GPT Neo
-----------------------------------------------------------------------------------------------------------------------
Overview
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The GPTNeo model was released in the `EleutherAI/gpt-neo <https://github.com/EleutherAI/gpt-neo>`__ repository by Sid
Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy. It is a GPT2 like causal language model trained on the
`Pile <https://pile.eleuther.ai/>`__ dataset.
The architecture is similar to GPT2 except that GPT Neo uses local attention in every other layer with a window size of
256 tokens.
Generation
_______________________________________________________________________________________________________________________
The :obj:`generate()` method can be used to generate text using GPT Neo model.
.. code-block::
>>> from transformers import GPTNeoForCausalLM, GPT2Tokenizer
>>> model = GPTNeoForCausalLM.from_pretrained("EleutherAI/gpt_neo_xl")
>>> tokenizer = GPT2Tokenizer.from_pretrained("EleutherAI/gpt_neo_xl")
>>> prompt = "In a shocking finding, scientists discovered a herd of unicorns living in a remote, " \
... "previously unexplored valley, in the Andes Mountains. Even more surprising to the " \
... "researchers was the fact that the unicorns spoke perfect English."
>>> input_ids = tokenizer(unicorns, return_tensors="pt").input_ids
>>> gen_tokens = model.generate(ids, do_sample=True, temperature=0.9, max_length=100,)
>>> gen_text = tokenizer.batch_decode(gen_tokens)[0]
GPTNeoConfig
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: transformers.GPTNeoConfig
:members:
GPTNeoModel
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: transformers.GPTNeoModel
:members: forward
GPTNeoForCausalLM
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. autoclass:: transformers.GPTNeoForCausalLM
:members: forward

6
docs/source/pretrained_models.rst
View File

@@ -139,6 +139,12 @@ For the full list, refer to `https://huggingface.co/models <https://huggingface.
| | ``gpt2-xl`` | | 48-layer, 1600-hidden, 25-heads, 1558M parameters. |
| | | | OpenAI's XL-sized GPT-2 English model |
+--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
| GPTNeo | ``EleutherAI/gpt_neo_xl`` | | 24-layer, 2048-hidden, 16-heads, 1.3B parameters. |
| | | | EleutherAI's GPT-3 like language model. |
| +------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
| | ``EleutherAI/gpt_neo_2.7B`` | | 32-layer, 2560-hidden, 20-heads, 2.7B parameters. |
| | | | EleutherAI's GPT-3 like language model. |
+--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
| Transformer-XL | ``transfo-xl-wt103`` | | 18-layer, 1024-hidden, 16-heads, 257M parameters. |
| | | | English model trained on wikitext-103 |
+--------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+