NMST+GPT-2
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
- New York University (NYU)
- Organisation type
- Academia
- Country
- United States of America
- Published
- 3 October 2022
- Authors
- Eugene Choi, Cheolhyoung Lee, Kyunghyun Cho
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Language
- Task
- Language modeling
- Base model
- GPT-2 (124M)
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
- 124M
- Training data
- 103,000,000 tokens
- Epochs
- 4.97
124M (the same as GPT-2)
500, 000 steps "For computational efficiency, we bucket the dataset into sequences of similar lengths, and each batch contains a maximum of 1,024 total tokens." 500000*1024 = 512 000 000 updates 512 000 000 / 103 000 000 = 4.97087378641 epochs
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
- 1.2 × 10²⁰ FLOP
- Fine-tuning compute
- 3.8 × 10¹⁷ FLOP
assuming that GPT-2 (124M) compute is similar to GPT-2 (117M) compute, which is 120000000000000000000 FLOP = 1.2 × 10^20 FLOP 120000000000000000000 FLOP + 380928000000000000 FLOP = 1.204×10^20 FLOP
6 FLOP / token / parameter * 512 000 000 tokens * 124 000 000 parameters = 3.80928 × 10^17 FLOP
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
- Open source
BSD 3-Clause License. Looks like just training code https://github.com/nyu-dl/non-monotonic-self-terminating-lm
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Confident
- Benchmark data
- NMST+GPT-2
Sources
Where this record came from and when it was last checked.
- Reference
- A Non-monotonic Self-terminating Language Model
- Last updated
- 28 November 2025
What the numbers mean
Where it came from
NMST+GPT-2 was published by New York University (NYU), in United States of America, in October 2022. academia is the category the publisher falls under.
It works in Language, and is recorded as doing language modeling.
Its starting point was GPT-2 (124M) — most models at this scale are adapted from an existing base rather than built from nothing.
Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.
How it was trained
Training it took roughly 1.2 × 10²⁰ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.
Around 103,000,000 tokens went into training it.
Answers
NMST+GPT-2 — common questions
Who created NMST+GPT-2?
NMST+GPT-2 was published by New York University (NYU), based in United States of America, categorised as academia.
When was NMST+GPT-2 released?
NMST+GPT-2 was published in October 2022. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
What is NMST+GPT-2 used for?
NMST+GPT-2 works in Language, and is recorded as handling language modeling. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.
How much compute was used to train NMST+GPT-2?
Around 1.2 × 10²⁰ FLOP. 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.
What GPU do I need to run NMST+GPT-2?
None. NMST+GPT-2 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.
Is NMST+GPT-2 open source?
No. NMST+GPT-2 has not had its weights published, so it exists only as a service controlled by its owner.
How many parameters does NMST+GPT-2 have?
NMST+GPT-2 has 124M parameters. 124M (the same as GPT-2). 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.
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.