NLP Lineage

20 October 2022 · Scale and data

Flan-T5 / Flan-PaLM

Scaling Instruction-Finetuned Language Models

H. Chung, L. Hou, S. Longpre, et al. (35 authors)

Date
20 October 2022
What the date marks
arXiv submission
Kind
Training method
Organisation
Google
Access
Apache 2.0
Threads
In the guided tour
Chapter 6. Teaching models what people want
Learning phases
2. How LLMs are trained

Notable: no later milestone in the data is documented as descending from it. How levels are decided

Read the primary source (arxiv.org)

Full summary in progress. Until then, this page shows the milestone's record from the research data.

What it changed

Before it

Instruction tuning had been shown on dozens of tasks at moderate scale.

The problem

Did it keep working with more tasks, bigger models and reasoning data mixed in?

What it introduced

Scale to 1,800 tasks and add chain-of-thought examples to the instruction mixture.

Why it matters

It made instruction tuning a scalable recipe, and the released Flan-T5 models became a research workhorse.

What modern models inherit

Instruction mixtures that include reasoning traces.

What came next

The Flan collection fed open instruction datasets; the recipe is standard post-training practice.

Before and after

The milestones it descends from, and the milestones that descend from it. Work developed alongside it is not counted. Weak links are left out unless you include them.

Before: where it came from

Flan-T5 / Flan-PaLM has 35 documented ancestors across 7 chapters, reaching back to Recurrent networks in 1990. Showing its direct parents and the 15 turning points among them.

Flan-T5 / Flan-PaLM has 35 documented ancestors across 7 chapters, reaching back to Recurrent networks in 1990. Showing all of them.

Flan-T5 / Flan-PaLM has 37 recorded ancestors, including indirect ones, across 7 chapters, reaching back to Recurrent networks in 1990. Showing its direct parents and the 15 turning points among them.

Flan-T5 / Flan-PaLM has 37 recorded ancestors, including indirect ones, across 7 chapters, reaching back to Recurrent networks in 1990. Showing all of them.

  1. 2022Flan-T5 / Flan-PaLMThis milestone.
  2. 2022PaLMFlan-T5 / Flan-PaLM built directly on thisChapter 5. Scale becomes a science
  3. 2022Chain of thought Turning pointFlan-T5 / Flan-PaLM built directly on thisChapter 12. Thinking longer
  4. 2021FLANFlan-T5 / Flan-PaLM built directly on thisChapter 6. Teaching models what people want
  5. 2020GShard Turning pointPaLM was influenced by thisChapter 5. Scale becomes a science
  6. 2020GPT-3 Turning pointFLAN was influenced by this
    Chain of thought was influenced by this    Chapter 5. Scale becomes a science
  7. 2020Scaling laws Turning pointGPT-3 was influenced by thisChapter 5. Scale becomes a science
  8. 2019T5 Turning pointFlan-T5 / Flan-PaLM built directly on thisChapter 4. Pretrain once, use everywhere
  9. 2019C4 Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  10. 2019Sparse Transformer Turning pointGPT-3 built directly on thisChapter 4. Pretrain once, use everywhere
  11. 2019GPT-2 Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  12. 2019WebTextGPT-2 was influenced by thisChapter 4. Pretrain once, use everywhere
  13. 2018BERT Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  14. 2018SentencePieceT5 built directly on thisChapter 4. Pretrain once, use everywhere
  15. 2018GPT-1BERT was influenced by this
    GPT-2 built directly on this    Chapter 4. Pretrain once, use everywhere
  16. 2018ELMoBERT was influenced by thisChapter 4. Pretrain once, use everywhere
  17. 2018ULMFiTBERT was influenced by thisChapter 4. Pretrain once, use everywhere
  18. 2017CoVeELMo was influenced by thisChapter 3. Attention, then only attention
  19. 2017Transformer Turning pointT5 built directly on thisChapter 3. Attention, then only attention
  20. 2017Sparse MoE Turning pointGShard built directly on thisChapter 3. Attention, then only attention
  21. 2016Layer normTransformer built directly on this
    GPT-2 built directly on this    Chapter 3. Attention, then only attention
  22. 2015Residual connectionsTransformer was influenced by thisChapter 3. Attention, then only attention
  23. 2015LM pretraining Turning pointBERT was influenced by thisChapter 2. Words become vectors
  24. 2015BPE tokenizationGPT-2 built directly on thisChapter 2. Words become vectors
  25. 2015Luong attentionTransformer was influenced by thisChapter 3. Attention, then only attention
  26. 2014AdamTransformer was influenced by thisChapter 2. Words become vectors
  27. 2014GloVeELMo built directly on thisChapter 2. Words become vectors
  28. 2014Seq2seq Turning pointTransformer built directly on thisChapter 2. Words become vectors
  29. 2014Attention Turning pointTransformer built directly on thisChapter 3. Attention, then only attention
  30. 2014GRU / encoder-decoderAttention built directly on thisChapter 2. Words become vectors
  31. 2013word2vecELMo was influenced by thisChapter 2. Words become vectors
  32. 2011NLP from scratchELMo was influenced by thisChapter 1. Language as probability
  33. 2003Neural LMNLP from scratch was influenced by this
    word2vec was influenced by this    Chapter 1. Language as probability
  34. 1997LSTM Turning pointSeq2seq built directly on this
    Sparse MoE built directly on this
    ELMo built directly on this    Chapter 1. Language as probability
  35. 1991Mixture of expertsSparse MoE built directly on thisChapter 1. Language as probability
  36. 1990Recurrent networksLSTM was influenced by thisChapter 1. Language as probability
  1. 2022Flan-T5 / Flan-PaLMThis milestone.
  2. 2022PaLMFlan-T5 / Flan-PaLM built directly on thisChapter 5. Scale becomes a science
  3. 2022Chain of thought Turning pointFlan-T5 / Flan-PaLM built directly on thisChapter 12. Thinking longer
  4. 2021FLANFlan-T5 / Flan-PaLM built directly on thisChapter 6. Teaching models what people want
  5. 2020GShard Turning pointPaLM was influenced by thisChapter 5. Scale becomes a science
  6. 2020GPT-3 Turning pointFLAN was influenced by this
    Chain of thought was influenced by this    Chapter 5. Scale becomes a science
  7. 2020Scaling laws Turning pointGPT-3 was influenced by thisChapter 5. Scale becomes a science
  8. 2019T5 Turning pointFlan-T5 / Flan-PaLM built directly on thisChapter 4. Pretrain once, use everywhere
  9. 2019C4 Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  10. 2019Sparse Transformer Turning pointGPT-3 built directly on thisChapter 4. Pretrain once, use everywhere
  11. 2019GPT-2 Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  12. 2019WebTextGPT-2 was influenced by this
    C4 was influenced by this (indirect influence)    Chapter 4. Pretrain once, use everywhere
  13. 2019Transformer-XLSparse Transformer was influenced by this (indirect influence)Chapter 4. Pretrain once, use everywhere
  14. 2018BERT Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  15. 2018SentencePieceT5 built directly on thisChapter 4. Pretrain once, use everywhere
  16. 2018GPT-1BERT was influenced by this
    GPT-2 built directly on this    Chapter 4. Pretrain once, use everywhere
  17. 2018ELMoBERT was influenced by thisChapter 4. Pretrain once, use everywhere
  18. 2018ULMFiTBERT was influenced by thisChapter 4. Pretrain once, use everywhere
  19. 2017CoVeELMo was influenced by thisChapter 3. Attention, then only attention
  20. 2017Transformer Turning pointT5 built directly on thisChapter 3. Attention, then only attention
  21. 2017ConvS2STransformer was influenced by this (indirect influence)Chapter 3. Attention, then only attention
  22. 2017Sparse MoE Turning pointGShard built directly on thisChapter 3. Attention, then only attention
  23. 2016Layer normTransformer built directly on this
    GPT-2 built directly on this    Chapter 3. Attention, then only attention
  24. 2015Residual connectionsTransformer was influenced by thisChapter 3. Attention, then only attention
  25. 2015LM pretraining Turning pointBERT was influenced by thisChapter 2. Words become vectors
  26. 2015BPE tokenizationGPT-2 built directly on thisChapter 2. Words become vectors
  27. 2015Luong attentionTransformer was influenced by thisChapter 3. Attention, then only attention
  28. 2014AdamTransformer was influenced by thisChapter 2. Words become vectors
  29. 2014GloVeELMo built directly on thisChapter 2. Words become vectors
  30. 2014Seq2seq Turning pointTransformer built directly on thisChapter 2. Words become vectors
  31. 2014Attention Turning pointTransformer built directly on thisChapter 3. Attention, then only attention
  32. 2014GRU / encoder-decoderAttention built directly on thisChapter 2. Words become vectors
  33. 2013word2vecELMo was influenced by thisChapter 2. Words become vectors
  34. 2011NLP from scratchELMo was influenced by thisChapter 1. Language as probability
  35. 2003Neural LMNLP from scratch was influenced by this
    word2vec was influenced by this    Chapter 1. Language as probability
  36. 1997LSTM Turning pointSeq2seq built directly on this
    Sparse MoE built directly on this
    ELMo built directly on this    Chapter 1. Language as probability
  37. 1991Mixture of expertsSparse MoE built directly on thisChapter 1. Language as probability
  38. 1990Recurrent networksLSTM was influenced by thisChapter 1. Language as probability

After: what it led to

No documented descendants are recorded for Flan-T5 / Flan-PaLM without weak links.

No documented descendants are recorded for Flan-T5 / Flan-PaLM without weak links.

No recorded descendants are recorded for Flan-T5 / Flan-PaLM.

No recorded descendants are recorded for Flan-T5 / Flan-PaLM.

Direct links and their sources

Where it came from

Built directly on

What it led to

None recorded in the sources used.

Terms