GPT-2 (AMPS)
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
- University of California (UC) Berkeley
- Organisation type
- Academia
- Country
- United States of America
- Published
- 8 November 2021
- Authors
- Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, Jacob Steinhardt
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Mathematics, Language
- Task
- Language modeling/generation, Quantitative reasoning
- Base model
- GPT-2 (1.5B)
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.
- Parameters
- 1.5M
- Training data
- tokens
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.
- How it was established
- Hardware
- Fine-tuning compute
- 6.5 × 10¹⁹ FLOP
8 GPUs * 24 hours * 3600 sec / hour * 311.84 * 10^12 FLOP / sec * 0.3 [assumed utilization] = 64663142400000000000 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
- NVIDIA A100
- Chips used
- 8
- Wall-clock time
- 24 hours
- Power draw
- 6.5 kW
Models are trained with 8 A100 GPUs, each requiring less than a day. Unless otherwise specified, for GPT-2 we use the default HuggingFace (Wolf et al., 2020) generation parameters, except that we use beam search.
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Speculative
Sources
Where this record came from and when it was last checked.
- Reference
- Measuring Mathematical Problem Solving With the MATH Dataset
- Last updated
- 11 February 2026
What the numbers mean
About this model
GPT-2 (AMPS) was published by University of California (UC) Berkeley, in the country recorded as United States of America, during November 2021. The publishing organisation is categorised as academia.
It works in the domain of Mathematics, Language, and is recorded as performing the task of language modeling/generation, Quantitative reasoning.
It builds on GPT-2 (1.5B). That is the usual way a specialised model is produced.
Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.
Answers
GPT-2 (AMPS) — common questions
GPT-2 (AMPS)— how many parameters does it have?
It has a parameter count of 1.5M. That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.
GPT-2 (AMPS)— who created it?
It was published by University of California (UC) Berkeley, based in United States of America, an organisation categorised as academia.
GPT-2 (AMPS)— when was it released?
It was published in November 2021. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
GPT-2 (AMPS)— what is it used for?
It works in the domain of Mathematics, Language, and is recorded as handling the task of language modeling/generation, Quantitative reasoning. These are the areas it was designed around; they describe intent rather than a hard boundary.
GPT-2 (AMPS)— 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.
GPT-2 (AMPS)— is it open source?
The licensing was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.
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.