GPT2+CoreLM+Fine-Tuning

Closed weights Aristotle University of Thessaloniki 132M parameters November 2021

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
Aristotle University of Thessaloniki
Organisation type
Academia
Country
Greece
Published
4 November 2021
Authors
Nikolaos Stylianou, Ioannis Vlahavas

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Language
Task
Language modeling/generation, Named entity recognition (NER)
Base model
GPT-2 (124M)
Numerical format
FP16

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
132M

"In all our experiments we use the GPT2-small configuration with 124M parameters, with 12 layers and 12 attention heads each for our base model." "Our model has 132M parameters, a 6% increase, after the addition of the Entity-Gating layer and the entity representations."

Training data
tokens

batch size of 128

Epochs
10

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

GPT-2 124M compute ~ GPT-2 117M compute = 1.1699999999999999 × 10^20 FLOP 1.17 * 10^20 FLOP + 2.57472e+17 FLOP = 1.1725747e+20 FLOP

How it was established
Hardware
Fine-tuning compute
2.6 × 10¹⁷ FLOP

29800000000000 FLOP / s [Titan V, fp16 - reported] * 8 hours * 3600 sec / hour * 1 GPU * 0.3 [assumed utilization] = 2.57472e+17 FLOP (it is probably the upper bound since the 8 hours time seems to account for all models fine-tuned in the paper)

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 Titan V
Chips used
1
Wall-clock time
8 hours

"In this setup, fine-tuning takes approximately 8 hours"

Power draw
277 W

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
Likely
Citations
2
Benchmark data
GPT2+CoreLM+Fine-Tuning

Sources

Where this record came from and when it was last checked.

Reference
CoreLM: Coreference-aware Language Model Fine-Tuning
Last updated
28 November 2025

What the numbers mean

Background

GPT2+CoreLM+Fine-Tuning was published by Aristotle University of Thessaloniki, in the country recorded as Greece, during November 2021. The category the publisher falls under is academia.

It works in the domain of Language, and is recorded as performing the task of language modeling/generation, Named entity recognition (NER).

Rather than being trained from scratch, it is derived from GPT-2 (124M). Most models at this scale are adapted from an existing base rather than built from nothing.

This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.

Training and provenance

The training run consumed about 1.2 × 10²⁰ FLOP, on hardware recorded as NVIDIA Titan V. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Answers

GPT2+CoreLM+Fine-Tuning — common questions

01

GPT2+CoreLM+Fine-Tuning— how much compute was used to train it?

Training consumed around 1.2 × 10²⁰ FLOP, on hardware recorded as NVIDIA Titan V. 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.

02

GPT2+CoreLM+Fine-Tuning— 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.

03

GPT2+CoreLM+Fine-Tuning— is it open source?

No. Its weights have not been published, so it exists only as a service controlled by its owner.

04

GPT2+CoreLM+Fine-Tuning— how many parameters does it have?

It has a parameter count of 132M. "In all our experiments we use the GPT2-small configuration with 124M parameters, with 12 layers and 12 attention heads each for our base model." "Our model has 132M parameters, a 6% increase, after the addition of the Entity-Gating layer and the entity representations.". 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.

05

GPT2+CoreLM+Fine-Tuning— who created it?

It was published by Aristotle University of Thessaloniki, based in Greece, an organisation categorised as academia.

06

GPT2+CoreLM+Fine-Tuning— 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.

07

GPT2+CoreLM+Fine-Tuning— what is it used for?

It works in the domain of Language, and is recorded as handling the task of language modeling/generation, Named entity recognition (NER). Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Source

Original publication

Record last updated 28 November 2025

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.