Text Generation
Transformers
PyTorch
bloom
text-generation-inference
Inference Endpoints
bigscience-bot commited on
Commit
a233cdc
1 Parent(s): b8cb7fd

Add README

Browse files
Files changed (1) hide show
  1. README.md +591 -0
README.md ADDED
@@ -0,0 +1,591 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: bigscience-bloom-rail-1.0
3
+ language:
4
+ - ak
5
+ - ar
6
+ - as
7
+ - bm
8
+ - bn
9
+ - ca
10
+ - code
11
+ - en
12
+ - es
13
+ - eu
14
+ - fon
15
+ - fr
16
+ - gu
17
+ - hi
18
+ - id
19
+ - ig
20
+ - ki
21
+ - kn
22
+ - lg
23
+ - ln
24
+ - ml
25
+ - mr
26
+ - ne
27
+ - nso
28
+ - ny
29
+ - or
30
+ - pa
31
+ - pt
32
+ - rn
33
+ - rw
34
+ - sn
35
+ - st
36
+ - sw
37
+ - ta
38
+ - te
39
+ - tn
40
+ - ts
41
+ - tum
42
+ - tw
43
+ - ur
44
+ - vi
45
+ - wo
46
+ - xh
47
+ - yo
48
+ - zh
49
+ - zhs
50
+ - zht
51
+ - zu
52
+ ---
53
+
54
+ # <span style="color:red"><b>WARNING:</b> The checkpoints on this repo are not fully trained model. Evaluations of intermediary checkpoints and the final model will be added when conducted (see below).</span>
55
+
56
+ # <p>BLOOM LM<br/> _BigScience Large Open-science Open-access Multilingual Language Model_ <br/>Model Card</p>
57
+ <img src="https://assets.website-files.com/6139f3cdcbbff3a68486761d/613cd8997b270da063e230c5_Tekengebied%201-p-500.png" alt="BigScience Logo" width="200"/>
58
+
59
+
60
+ Version 1.3 / 11.July.2022 - Available intermediary checkpoints - global steps:
61
+
62
+ + `1000`, `10000`, `50000`, `100000`, `150000`, `200000`, `250000`, `300000`, `350000`
63
+
64
+ You can check the available checkpoints by clicking on the branches section of the repo
65
+
66
+ # How to load a specific version
67
+
68
+ We use `git tags` to load a model in a specific version (eg. `global_step1000`):
69
+
70
+ ```python
71
+ from transformers import AutoModelForCausalLM
72
+
73
+ model = AutoModelForCausalLM.from_pretrained(
74
+ "bigscience/bloom-6b3-intermediate",
75
+ revision="global_step1000",
76
+ torch_dtype="auto",
77
+ )
78
+ ```
79
+
80
+
81
+ # Table of Contents
82
+ 1. [Model Details](#model-details)
83
+ 2. [Uses](#uses)
84
+ 3. [Training Data](#training-data)
85
+ 4. [Risks and Limitations](#risks-and-limitations)
86
+ 5. [Evaluation](#evaluation)
87
+ 6. [Recommendations](#recommendations)
88
+ 7. [Glossary and Calculations](#glossary-and-calculations)
89
+ 8. [More Information](#more-information)
90
+ 9. [Model Card Authors](#model-card-authors)
91
+
92
+ ---
93
+
94
+ # Model Details
95
+
96
+ BLOOM is a type of language model, which is a probability distribution over sequences of words. Specifically, BLOOM is a Large Language Model (LLM), meaning that it is trained on vast amounts of text data using industrial-scale computational resources. As such, the model is able to capture the statistical tendencies of words, phrases, sentences, and larger spans of text that it is exposed to in the training data.
97
+
98
+ ## Basics
99
+ *This section provides information about the model type, version, license, funders, release date, developers, and contact information.*
100
+ *It is useful for anyone who wants to reference the model.*
101
+
102
+ <details>
103
+ <summary>Click to expand</summary>
104
+
105
+ **Developed by:** BigScience ([website](https://bigscience.huggingface.co))
106
+
107
+ *All collaborators are either volunteers or have an agreement with their employer. (Further breakdown of participants forthcoming.)*
108
+
109
+ **Model Type:** Transformer-based Language Model
110
+
111
+ **Version:** 1.0.0
112
+
113
+ **Languages:** Multiple; see [training data](#training-data)
114
+
115
+ **License:** RAIL License v1.0 ([link](https://huggingface.co/spaces/bigscience/license))
116
+
117
+ **Release Date Estimate:** Monday, 11.July.2022
118
+
119
+ **Send Questions to:** bigscience-contact@googlegroups.com
120
+
121
+ **Cite as:** BigScience, _BigScience Language Open-science Open-access Multilingual (BLOOM) Language Model_. International, May 2021-May 2022
122
+
123
+ **Funded by:**
124
+
125
+ * The French government.
126
+
127
+ * Hugging Face ([website](https://huggingface.co)).
128
+
129
+ * Organizations of contributors. *(Further breakdown of organizations forthcoming.)*
130
+
131
+ </details>
132
+
133
+ ## Technical Specifications
134
+ *This section includes details about the model objective and architecture, and the compute infrastructure.*
135
+ *It is useful for people interested in model development.*
136
+
137
+ <details>
138
+ <summary>Click to expand</summary>
139
+
140
+ Please see [the BLOOM training README](https://github.com/bigscience-workshop/bigscience/tree/master/train/tr11-176B-ml#readme) for full details on replicating training.
141
+
142
+ ### Model Architecture and Objective
143
+
144
+ * Modified from Megatron-LM GPT2 (see [paper](https://arxiv.org/abs/1909.08053), [BLOOM Megatron code](https://github.com/bigscience-workshop/Megatron-DeepSpeed)):
145
+
146
+ * Decoder-only architecture
147
+
148
+ * Layer normalization applied to word embeddings layer (`StableEmbedding`; see [code](https://github.com/facebookresearch/bitsandbytes), [paper](https://arxiv.org/pdf/2110.02861.pdf))
149
+
150
+ * ALiBI positional encodings (see [paper](https://arxiv.org/pdf/2108.12409.pdf)), with GeLU activation functions
151
+
152
+ * 176 billion parameters:
153
+
154
+ * 70 layers, 112 attention heads
155
+
156
+ * Hidden layers are 14336-dimensional
157
+
158
+ * Sequence length of 2048 tokens used (see [BLOOM tokenizer](https://huggingface.co/bigscience/tokenizer), [tokenizer description](#tokenization))
159
+
160
+ **Objective Function:** Cross Entropy with mean reduction (see [API documentation](https://pytorch.org/docs/stable/generated/torch.nn.CrossEntropyLoss.html#torch.nn.CrossEntropyLoss)).
161
+
162
+ ### Compute infrastructure
163
+ Jean Zay Public Supercomputer, provided by the French government (see [announcement](https://www.enseignementsup-recherche.gouv.fr/fr/signature-du-marche-d-acquisition-de-l-un-des-supercalculateurs-les-plus-puissants-d-europe-46733)).
164
+
165
+ #### Hardware
166
+
167
+ * 384 A100 80GB GPUs (48 nodes)
168
+
169
+ * Additional 32 A100 80GB GPUs (4 nodes) in reserve
170
+
171
+ * 8 GPUs per node Using NVLink 4 inter-gpu connects, 4 OmniPath links
172
+
173
+ * CPU: AMD
174
+
175
+ * CPU memory: 512GB per node
176
+
177
+ * GPU memory: 640GB per node
178
+
179
+ * Inter-node connect: Omni-Path Architecture (OPA)
180
+
181
+ * NCCL-communications network: a fully dedicated subnet
182
+
183
+ * Disc IO network: shared network with other types of nodes
184
+
185
+ #### Software
186
+
187
+ * Megatron-DeepSpeed ([Github link](https://github.com/bigscience-workshop/Megatron-DeepSpeed))
188
+
189
+ * DeepSpeed ([Github link](https://github.com/microsoft/DeepSpeed))
190
+
191
+ * PyTorch (pytorch-1.11 w/ CUDA-11.5; see [Github link](https://github.com/pytorch/pytorch))
192
+
193
+ * apex ([Github link](https://github.com/NVIDIA/apex))
194
+
195
+ </details>
196
+
197
+ ---
198
+
199
+ # Training
200
+ *This section provides information about the training data, the speed and size of training elements, and the environmental impact of training.*
201
+ *It is useful for people who want to learn more about the model inputs and training footprint.*
202
+
203
+ <details>
204
+ <summary>Click to expand</summary>
205
+
206
+ ## Training Data
207
+ *This section provides a high-level overview of the training data. It is relevant for anyone who wants to know the basics of what the model is learning.*
208
+
209
+ Details for each dataset are provided in individual [Data Cards](https://huggingface.co/spaces/bigscience/BigScienceCorpus).
210
+
211
+ Training data includes:
212
+
213
+ - 45 natural languages
214
+
215
+ - 12 programming languages
216
+
217
+ - In 1.5TB of pre-processed text, converted into 350B unique tokens (see [the tokenizer section](#tokenization) for more.)
218
+
219
+ ### Languages
220
+
221
+ The pie chart shows the distribution of languages in training data.
222
+
223
+ ![pie chart showing the distribution of languages in training data](https://github.com/bigscience-workshop/model_card/blob/main/assets/data/pie_chart.svg?raw=true)
224
+
225
+
226
+ The following tables shows the further distribution of Niger-Congo & Indic languages and programming languages in the training data.
227
+
228
+ Distribution of Niger Congo and Indic languages.
229
+
230
+ | Niger Congo | Percentage | | Indic | Percentage |
231
+ |----------------|------------ |------ |-----------|------------|
232
+ | Chi Tumbuka | 0.00002 | | Assamese | 0.01 |
233
+ | Kikuyu | 0.00004 | | Odia | 0.04 |
234
+ | Bambara | 0.00004 | | Gujarati | 0.04 |
235
+ | Akan | 0.00007 | | Marathi | 0.05 |
236
+ | Xitsonga | 0.00007 | | Punjabi | 0.05 |
237
+ | Sesotho | 0.00007 | | Kannada | 0.06 |
238
+ | Chi Chewa | 0.0001 | | Nepali | 0.07 |
239
+ | Setswana | 0.0002 | | Telugu | 0.09 |
240
+ | Northern Sotho | 0.0002 | | Malayalam | 0.10 |
241
+ | Fon | 0.0002 | | Urdu | 0.10 |
242
+ | Kirundi | 0.0003 | | Tamil | 0.20 |
243
+ | Wolof | 0.0004 | | Bengali | 0.50 |
244
+ | Kuganda | 0.0004 | | Hindi | 0.70 |
245
+ | Chi Shona | 0.001 |
246
+ | Isi Zulu | 0.001 |
247
+ | Igbo | 0.001 |
248
+ | Xhosa | 0.001 |
249
+ | Kinyarwanda | 0.003 |
250
+ | Yoruba | 0.006 |
251
+ | Swahili | 0.02 |
252
+
253
+ Distribution of programming languages.
254
+
255
+ | Extension | Language | Number of files |
256
+ |----------------|------------|-----------------|
257
+ | java | Java | 5,407,724 |
258
+ | php | PHP | 4,942,186 |
259
+ | cpp | C++ | 2,503,930 |
260
+ | py | Python | 2,435,072 |
261
+ | js | JavaScript | 1,905,518 |
262
+ | cs | C# | 1,577,347 |
263
+ | rb | Ruby | 6,78,413 |
264
+ | cc | C++ | 443,054 |
265
+ | hpp | C++ | 391,048 |
266
+ | lua | Lua | 352,317 |
267
+ | go | GO | 227,763 |
268
+ | ts | TypeScript | 195,254 |
269
+ | C | C | 134,537 |
270
+ | scala | Scala | 92,052 |
271
+ | hh | C++ | 67,161 |
272
+ | H | C++ | 55,899 |
273
+ | tsx | TypeScript | 33,107 |
274
+ | rs | Rust | 29,693 |
275
+ | phpt | PHP | 9,702 |
276
+ | c++ | C++ | 1,342 |
277
+ | h++ | C++ | 791 |
278
+ | php3 | PHP | 540 |
279
+ | phps | PHP | 270 |
280
+ | php5 | PHP | 166 |
281
+ | php4 | PHP | 29 |
282
+
283
+ ### Preprocessing
284
+
285
+ **Tokenization:** The BLOOM tokenizer ([link](https://huggingface.co/bigscience/tokenizer)), a learned subword tokenizer trained using:
286
+
287
+ - A byte-level Byte Pair Encoding (BPE) algorithm
288
+
289
+ - A simple pre-tokenization rule, no normalization
290
+
291
+ - A vocabulary size of 250,680
292
+
293
+ It was trained on a subset of a preliminary version of the corpus using alpha-weighting per language.
294
+
295
+ ## Speeds, Sizes, Times
296
+
297
+ Training logs: [Tensorboard link](https://huggingface.co/tensorboard/bigscience/tr11-176B-ml-logs/)
298
+
299
+ - Dates:
300
+
301
+ - Started 11th March, 2022 11:42am PST
302
+
303
+ - Estimated end: 5th July, 2022
304
+
305
+ - Checkpoint size:
306
+
307
+ - Bf16 weights: 329GB
308
+
309
+ - Full checkpoint with optimizer states: 2.3TB
310
+
311
+ - Training throughput: About 150 TFLOP per GPU per second
312
+
313
+ - Number of epochs: 1
314
+
315
+ - Estimated cost of training: Equivalent of $2-5M in cloud computing (including preliminary experiments)
316
+
317
+ - Server training location: Île-de-France, France
318
+
319
+
320
+ ## Environmental Impact
321
+
322
+ The training supercomputer, Jean Zay ([website](http://www.idris.fr/eng/jean-zay/jean-zay-presentation-eng.html)), uses mostly nuclear energy. The heat generated by it is reused for heating campus housing.
323
+
324
+ **Estimated carbon emissions:** *(Forthcoming.)*
325
+
326
+ **Estimated electricity usage:** *(Forthcoming.)*
327
+
328
+ </details>
329
+
330
+ ---
331
+
332
+ # Uses
333
+
334
+ *This section addresses questions around how the model is intended to be used, discusses the foreseeable users of the model (including those affected by the model), and describes uses that are considered out of scope or misuse of the model.*
335
+ *It is useful for anyone considering using the model or who is affected by the model.*
336
+
337
+ <details>
338
+ <summary>Click to expand</summary>
339
+
340
+ ## Intended Use
341
+
342
+ This model is being created in order to enable public research on large language models (LLMs). LLMs are intended to be used for language generation or as a pretrained base model that can be further fine-tuned for specific tasks. Use cases below are not exhaustive.
343
+
344
+ ### Direct Use
345
+
346
+ - Text generation
347
+
348
+ - Exploring characteristics of language generated by a language model
349
+
350
+ - Examples: Cloze tests, counterfactuals, generations with reframings
351
+
352
+ ### Downstream Use
353
+
354
+ - Tasks that leverage language models include: Information Extraction, Question Answering, Summarization
355
+
356
+ ### Misuse and Out-of-scope Use
357
+ *This section addresses what users ought not do with the model.*
358
+
359
+ See the [BLOOM License](https://huggingface.co/spaces/bigscience/license), Attachment A, for detailed usage restrictions. The below list is non-exhaustive, but lists some easily foreseeable problematic use cases.
360
+
361
+ #### Out-of-scope Uses
362
+
363
+ Using the model in [high-stakes](#high-stakes) settings is out of scope for this model. The model is not designed for [critical decisions](#critical-decisions) nor uses with any material consequences on an individual's livelihood or wellbeing. The model outputs content that appears factual but is not correct.
364
+
365
+ Out-of-scope Uses Include:
366
+
367
+ - Usage in biomedical domains, political and legal domains, or finance domains
368
+
369
+ - Usage for evaluating or scoring individuals, such as for employment, education, or credit
370
+
371
+ - Applying the model for critical automatic decisions, generating factual content, creating reliable summaries, or generating predictions that must be correct
372
+
373
+ #### Misuse
374
+
375
+ Intentionally using the model for harm, violating [human rights](#human-rights), or other kinds of malicious activities, is a misuse of this model. This includes:
376
+
377
+ - Spam generation
378
+
379
+ - Disinformation and influence operations
380
+
381
+ - Disparagement and defamation
382
+
383
+ - Harassment and abuse
384
+
385
+ - [Deception](#deception)
386
+
387
+ - Unconsented impersonation and imitation
388
+
389
+ - Unconsented surveillance
390
+
391
+ - Generating content without attribution to the model, as specified in the [RAIL License, Use Restrictions](https://huggingface.co/spaces/bigscience/license)
392
+
393
+ ## Intended Users
394
+
395
+ ### Direct Users
396
+
397
+ - General Public
398
+
399
+ - Researchers
400
+
401
+ - Students
402
+
403
+ - Educators
404
+
405
+ - Engineers/developers
406
+
407
+ - Non-commercial entities
408
+
409
+ - Community advocates, including human and civil rights groups
410
+
411
+ ### Indirect Users
412
+
413
+ - Users of derivatives created by Direct Users, such as those using software with an [intended use](#intended-use)
414
+
415
+ - Users of [Derivatives of the Model, as described in the License](https://huggingface.co/spaces/bigscience/license)
416
+
417
+ ### Others Affected (Parties Prenantes)
418
+
419
+ - People and groups referred to by the LLM
420
+
421
+ - People and groups exposed to outputs of, or decisions based on, the LLM
422
+
423
+ - People and groups whose original work is included in the LLM
424
+
425
+ </details>
426
+
427
+ ---
428
+
429
+ # Risks and Limitations
430
+ *This section identifies foreseeable harms and misunderstandings.*
431
+
432
+ <details>
433
+ <summary>Click to expand</summary>
434
+
435
+ Model may:
436
+
437
+ - Overrepresent some viewpoints and underrepresent others
438
+
439
+ - Contain stereotypes
440
+
441
+ - Contain [personal information](#personal-data-and-information)
442
+
443
+ - Generate:
444
+
445
+ - Hateful, abusive, or violent language
446
+
447
+ - Discriminatory or prejudicial language
448
+
449
+ - Content that may not be appropriate for all settings, including sexual content
450
+
451
+ - Make errors, including producing incorrect information as if it were factual
452
+
453
+ - Generate irrelevant or repetitive outputs
454
+
455
+ </details>
456
+
457
+ ---
458
+
459
+ # Evaluation
460
+ *This section describes the evaluation protocols and provides the results.*
461
+
462
+
463
+ <details>
464
+ <summary>Click to expand</summary>
465
+
466
+ ## Metrics
467
+ *This section describes the different ways performance is calculated and why.*
468
+
469
+
470
+ Includes:
471
+
472
+ | Metric | Why chosen |
473
+ |--------------------|--------------------------------------------------------------------|
474
+ | [Perplexity](#perplexity) | Standard metric for quantifying model improvements during training |
475
+ | Cross Entropy [Loss](#loss) | Standard objective for language models. |
476
+
477
+ And multiple different metrics for specific tasks. _(More evaluation metrics forthcoming upon completion of evaluation protocol.)_
478
+
479
+ ## Factors
480
+ *This section lists some different aspects of what BLOOM models. Its focus is on those aspects that are likely to give rise to high variance in model behavior.*
481
+
482
+ - Language, such as English or Yoruba
483
+
484
+ - Domain, such as newswire or stories
485
+
486
+ - Demographic characteristics, such as gender or nationality
487
+
488
+ ## Results
489
+ *Results are based on the [Factors](#factors) and [Metrics](#metrics).*
490
+
491
+ **Train-time Evaluation:**
492
+
493
+ As of 25.May.2022, 15:00 PST:
494
+
495
+ - Training Loss: 2.0
496
+
497
+ - Validation Loss: 2.2
498
+
499
+ - Perplexity: 8.9
500
+
501
+ (More evaluation scores forthcoming.)
502
+
503
+ </details>
504
+
505
+ ---
506
+
507
+ # Recommendations
508
+
509
+ *This section provides information on warnings and potential mitigations.*
510
+
511
+ <details>
512
+ <summary>Click to expand</summary>
513
+
514
+ - Indirect users should be made aware when the content they're working with is created by the LLM.
515
+
516
+ - Users should be aware of [Risks and Limitations](#risks-and-limitations), and include an appropriate age disclaimer or blocking interface as necessary.
517
+
518
+ - Models trained or finetuned downstream of BLOOM LM should include an updated Model Card.
519
+
520
+ - Users of the model should provide mechanisms for those affected to provide feedback, such as an email address for comments.
521
+
522
+ </details>
523
+
524
+ ---
525
+
526
+ # Glossary and Calculations
527
+
528
+ *This section defines common terms and how metrics are calculated.*
529
+ <details>
530
+ <summary>Click to expand</summary>
531
+
532
+ - <a name="loss">**Loss:**</a> A calculation of the difference between what the model has learned and what the data shows ("groundtruth"). The lower the loss, the better. The training process aims to minimize the loss.
533
+
534
+ - <a name="perplexity">**Perplexity:**</a> This is based on what the model estimates the probability of new data is. The lower the perplexity, the better. If the model is 100% correct at predicting the next token it will see, then the perplexity is 1. Mathematically this is calculated using entropy.
535
+
536
+ - <a name="high-stakes">**High-stakes settings:**</a> Such as those identified as "high-risk AI systems" and "unacceptable risk AI systems" in the European Union's proposed [Artificial Intelligence (AI) Act](https://artificialintelligenceact.eu/annexes/).
537
+
538
+ - <a name="critical-decisions">**Critical decisions:**</a> Such as those defined in [the United States' proposed Algorithmic Accountability Act](https://www.congress.gov/117/bills/s3572/BILLS-117s3572is.pdf).
539
+
540
+ - <a name="human-rights">**Human rights:**</a> Includes those rights defined in the [Universal Declaration of Human Rights](https://www.un.org/sites/un2.un.org/files/2021/03/udhr.pdf).
541
+
542
+ - <a name="personal-data-and-information">**Personal Data and Personal Information:**</a> Personal data and information is defined in multiple data protection regulations, such as "[personal data](https://gdpr-info.eu/issues/personal-data/)" in the [European Union's General Data Protection Regulation](https://gdpr-info.eu); and "personal information" in the Republic of South Africa's [Protection of Personal Information Act](https://www.gov.za/sites/default/files/gcis_document/201409/3706726-11act4of2013popi.pdf), The People's Republic of China's [Personal information protection law](http://en.npc.gov.cn.cdurl.cn/2021-12/29/c_694559.htm).
543
+
544
+ - <a name="sensitive-characteristics">**Sensitive characteristics:**</a> This includes specifically protected categories in human rights (see [UHDR, Article 2](https://www.un.org/sites/un2.un.org/files/2021/03/udhr.pdf)) and personal information regulation (see GDPR, [Article 9; Protection of Personal Information Act, Chapter 1](https://www.gov.za/sites/default/files/gcis_document/201409/3706726-11act4of2013popi.pdf))
545
+
546
+ - <a name="deception">**Deception:**</a> Doing something to intentionally mislead individuals to believe something that is false, such as by creating deadbots or chatbots on social media posing as real people, or generating text documents without making consumers aware that the text is machine generated.
547
+
548
+ </details>
549
+
550
+ ---
551
+
552
+ # More Information
553
+ *This section provides links to writing on dataset creation, technical specifications, lessons learned, and initial results.*
554
+
555
+ <details>
556
+ <summary>Click to expand</summary>
557
+
558
+ ## Dataset Creation
559
+
560
+ Blog post detailing the design choices during the dataset creation: https://bigscience.huggingface.co/blog/building-a-tb-scale-multilingual-dataset-for-language-modeling
561
+
562
+ ## Technical Specifications
563
+
564
+ Blog post summarizing how the architecture, size, shape, and pre-training duration where selected: https://bigscience.huggingface.co/blog/what-language-model-to-train-if-you-have-two-million-gpu-hours
565
+
566
+ More details on the architecture/optimizer: https://github.com/bigscience-workshop/bigscience/tree/master/train/tr11-176B-ml
567
+
568
+ Blog post on the hardware/engineering side: https://bigscience.huggingface.co/blog/which-hardware-to-train-a-176b-parameters-model
569
+
570
+ Details on the distributed setup used for the training: https://github.com/bigscience-workshop/bigscience/tree/master/train/tr11-176B-ml
571
+
572
+ Tensorboard updated during the training: https://huggingface.co/bigscience/tr11-176B-ml-logs/tensorboard#scalars&tagFilter=loss
573
+
574
+ ## Lessons
575
+
576
+ Insights on how to approach training, negative results: https://github.com/bigscience-workshop/bigscience/blob/master/train/lessons-learned.md
577
+
578
+ Details on the obstacles overcome during the preparation on the engineering side (instabilities, optimization of training throughput, so many technical tricks and questions): https://github.com/bigscience-workshop/bigscience/blob/master/train/tr11-176B-ml/chronicles.md
579
+
580
+ ## Initial Results
581
+
582
+ Initial prompting experiments using interim checkpoints: https://huggingface.co/spaces/bigscience/bloom-book
583
+
584
+ </details>
585
+
586
+ ---
587
+
588
+ # Model Card Authors
589
+ *Ordered roughly chronologically and by amount of time spent.*
590
+
591
+ Margaret Mitchell, Giada Pistilli, Yacine Jernite, Ezinwanne Ozoani, Marissa Gerchick, Nazneen Rajani, Sasha Luccioni, Irene Solaiman, Maraim Masoud, Somaieh Nikpoor, Carlos Muñoz Ferrandis, Stas Bekman, Christopher Akiki, Danish Contractor, David Lansky, Angelina McMillan-Major, Tristan Thrush, Suzana Ilić, Gérard Dupont, Shayne Longpre, Manan Dey, Stella Biderman, Douwe Kiela, Emi Baylor, Teven Le Scao, Aaron Gokaslan, Julien Launay, Niklas Muennighoff