whatmodel.sh / GPT-Neo 1.3B

GPT-Neo 1.3B

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

GPT-Neo 1.3B is an open-weights language model from EleutherAI, released March 2021. Its published size is 1.3B parameters and a 2k-token context window. Licence: mit.

13.6±2.0#704 of 719 LLMsfrom 11 benchmark results across 4 sources

$ how it compares

The LLMs ranked either side of GPT-Neo 1.3B, out of 828 with a score, then more from EleutherAI.

#modelscorewhereRAMreleased
810BLOOM 1.7B BigScience14±1local~3.1GB2022-07
811Flan-T5 Small Google14±3local~1.8GB2022-10
812GPT-Neo 1.3B EleutherAI14±2local~7.1GB2021-03
813GPT-2 XL (1.5B) OpenAI11±1local~2.8GB2019-11
814BLOOM 1.1B BigScience11±1local~2.6GB2022-07
more from EleutherAI
756GPT-NeoX-20B EleutherAI28±1local~14GB2022-02
777Pythia 12B EleutherAI25±1local~9.5GB2023-02
780GPT-J 6B EleutherAI25±1local~14GB2021-06

see all LLMs · find the ones that fit your machine

$ what it's good at

math█···█······█ #3 of 3 * · 8.7% of its score · 1 test

Trails GPT-2 XL (1.5B) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
MathQA lab24.1% ✎█···█··█···█ beats 1/3hard · avg 46%▇▇▇ 8.7%too hard for it
other tests10 not in a task · 90% of its score
testresultvs the others in its rundifficultycountswhy
HellaSwag lab48.9% ✎▄·█·█··▄▄··▄ beats 1/7very easy · avg 92%▇▇▇▇▇▇▇▇▇▇ 26%at its level
LAMBADA (OpenAI version), last-word accuracy lab57.2% ✎█·▄··█··█··▄ beats 2/7very easy · avg 90%▇▇▇▇▇▇ 15%blunt · at its level
WinoGrande lab55.0% ✎▄·▄·▄▄▄··█·▄ beats 1/7easy · avg 89%▇▇▇▇▇▇ 15%typical
Natural Questions (closed book)4.0% ✎▁·▁·▁▃▁▂·▅▂█ beats 4/28medium · avg 58%▇▇▇▇▇ 13%too hard for it
PIQA lab71.1% ✎█··▃··▃▃·▃·▃ beats 2/7easy · avg 89%▇▇▇▇ 10%blunt · at its level
TriviaQA14.2% ✎▁··▁█▃▃▁▄▅▄█ beats 9/28easy · avg 79%▇ 3.0%noisy
PubMedQA lab54.4% ✎▆·▄▆·▄▂▄▄▂█▂ beats 2/20medium · avg 75%▇ 2.8%blunt · at its level
TruthfulQA MC239.6% ✎▅▃█▆▁▂▅···▁▁ beats 8/31medium · avg 64%▇ 2.6%blunt · at its level
2 more tests
testresultvs the others in its rundifficultycountswhy
Open PT-LLM Leaderboard average (9 Portuguese tasks)26.7% ✎▃▁▁▁▁▁▁▁▂█▅▁ beats 22/276medium · avg 73%▇ 2.1%too hard for it · noisy
HaluEval QA hallucination detection (accuracy)46.3% ✎█▁▁·▁▁·▁▁▁▁▁ beats 15/27medium · avg 70%▇ 0.4%sharp · too hard for it · noisy

$ where to run it

Other ways to run it

  1. Run it as published · needs ~5.31 GB · from huggingface.co/EleutherAI (official)
    uv run --with torch --with 'transformers<5' python -c "from transformers import pipeline; print(pipeline('text-generation', model='EleutherAI/gpt-neo-1.3B')('Once upon a time', max_new_tokens=40)[0]['generated_text'])"

Official weights and documentation: huggingface.co/EleutherAI ↗

$ how the score was worked out · EleutherAI GPT-J-6B model car… 66%, Hallucinations Leaderboard re… 19%, EleutherAI GPT-Neo 2.7B model… 12% and 1 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
EleutherAI GPT-J-6B model card — EleutherAI lm-evaluation-harness runs · lab-reported4▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 66%Apache-2.0
Hallucinations Leaderboard results (HF dataset hallucinations-leaderboard/results)4▇▇▇▇▇▇ 19%Apache-2.0
EleutherAI GPT-Neo 2.7B model card — EleutherAI evaluation-harness runs · lab-reported2▇▇▇ 12%MIT
Open Portuguese LLM Leaderboard (archived)1▇ 2%Apache-2.0
prior · pull toward the middle▇ 2%
weighted average= 13.6

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json