Long-range sequence Compressive Transformers
No estimate
No hardware requirements for this model
The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.
On record
Full specification
Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.
Origin
Who built this model, where, and when it was published.
- Organisation
- DeepMind
- Organisation type
- Industry
- Country
- United Kingdom of Great Britain and Northern Ireland
- Published
- 13 November 2019
- Authors
- Jack W. Rae, Anna Potapenko, Siddhant M. Jayakumar, Timothy P. Lillicrap
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Language
- Task
- Language modeling
Size
How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.
- Training data
- 103,000,000 tokens
- Epochs
- 328.32
- Batch size
- 65,536
"for word-based LM we used 16, 000 warmup steps with 500, 000 decay steps" "a sequence window size all equal to 512" " a total batch size of 128" 516000*512*128/103000000 = 328.32 epochs
512*128
Training compute
The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.
- Training compute
- 1 × 10²⁰ FLOP
- How it was established
- Hardware
123000000000000 FLOP / second/ GPU * 64 GPUs * 12 hours * 3600 sec/hour * 0.3 [assumed utilization] = 1.0202112e+20 FLOP
The training run
What it physically took to train: which chips, how many, for how long, and what that drew from the wall.
- Training hardware
- Google TPU v3
- Chips used
- 64
- Wall-clock time
- 12 hours
- Power draw
- 59.0 kW
"The model converged in a little over 12 hours."
Availability
Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.
- Weights
- Closed — provider access only
- Model access
- Unreleased
- Training code
- Unreleased
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Confident
- Citations
- 845
- Benchmark data
- Compressive Transformers for Long-Range Sequence Modelling
Sources
Where this record came from and when it was last checked.
- Reference
- Compressive Transformers for Long-Range Sequence Modelling
- Last updated
- 25 May 2026
What the numbers mean
Background
Long-range sequence Compressive Transformers was published by DeepMind, in the country recorded as United Kingdom of Great Britain and Northern Ireland, during November 2019. It comes out of an organisation categorised as industry.
It works in the domain of Language, and is recorded as performing the task of language modeling.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
How it was trained
The training run consumed about 1 × 10²⁰ FLOP, on hardware recorded as Google TPU v3. That figure measures what producing the model cost, and has no bearing on how fast it answers.
It was trained on a corpus of about 103,000,000 tokens of text.
Answers
Long-range sequence Compressive Transformers — common questions
Long-range sequence Compressive Transformers— when was it released?
It was published in November 2019. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
Long-range sequence Compressive Transformers— what is it used for?
It works in the domain of Language, and is recorded as handling the task of language modeling. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.
Long-range sequence Compressive Transformers— how much compute was used to train it?
Training consumed around 1 × 10²⁰ FLOP, on hardware recorded as Google TPU v3. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.
Long-range sequence Compressive Transformers— what GPU do I need to run it?
None. This is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.
Long-range sequence Compressive Transformers— is it open source?
No. Its weights have not been published, so it exists only as a service controlled by its owner.
Long-range sequence Compressive Transformers— how many parameters does it have?
No parameter count has been published for it, which is why no memory or speed figure appears on this page.
Long-range sequence Compressive Transformers— who created it?
It was published by DeepMind, based in United Kingdom of Great Britain and Northern Ireland, an organisation categorised as industry.
The other direction
Looking at it from the other side?
This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.