PolyPythias - a EleutherAI Collection

EleutherAI 's Collections

Pythia Scaling Suite

Quirky Models and Datasets

Quirky Math (archive)

Pile-T5 T5x Checkpoints

Sparse Autoencoders

PolyPythias

updated Apr 9, 2025

EleutherAI/polypythias-evals

Preview • Updated Sep 11, 2024 • 1.76k
EleutherAI/pile-preshuffled-seeds

Updated Feb 12 • 153 • 1

Note Training data information for each seed.
EleutherAI/pythia-14m-deduped

Text Generation • 39.2M • Updated Feb 12 • 9.91k • 30
EleutherAI/pythia-14m-seed1

Text Generation • Updated Jan 11 • 1.54k
EleutherAI/pythia-14m-seed2

Text Generation • Updated Jan 11 • 1.36k
EleutherAI/pythia-14m-seed3

Text Generation • Updated Jan 11 • 1.11k
EleutherAI/pythia-14m-seed4

Text Generation • Updated Jan 11 • 1.03k
EleutherAI/pythia-14m-seed5

Text Generation • Updated Jan 11 • 1.27k
EleutherAI/pythia-14m-seed6

Text Generation • Updated Jan 11 • 1.42k
EleutherAI/pythia-14m-seed7

Text Generation • Updated Jan 11 • 1.29k
EleutherAI/pythia-14m-seed8

Text Generation • Updated Jan 11 • 1.27k
EleutherAI/pythia-14m-seed9

Text Generation • Updated Jan 11 • 1.25k
EleutherAI/pythia-31m-deduped

Text Generation • 55.7M • Updated Feb 12 • 1.97k • 5
EleutherAI/pythia-31m-seed1

Text Generation • Updated Jan 11 • 1.19k
EleutherAI/pythia-31m-seed2

Text Generation • Updated Jan 11 • 1.25k
EleutherAI/pythia-31m-seed3

Text Generation • Updated Jan 11 • 1.19k
EleutherAI/pythia-31m-seed4

Text Generation • Updated Jan 11 • 1.16k
EleutherAI/pythia-31m-seed5

Text Generation • Updated Jan 11 • 1.19k
EleutherAI/pythia-31m-seed6

Text Generation • Updated Jan 11 • 1.18k
EleutherAI/pythia-31m-seed7

Text Generation • Updated Jan 11 • 1.34k
EleutherAI/pythia-31m-seed8

Text Generation • Updated Jan 11 • 1.23k
EleutherAI/pythia-31m-seed9

Text Generation • Updated Jan 11 • 1.19k
EleutherAI/pythia-70m

95.6M • Updated Nov 21, 2023 • 577k • 85
EleutherAI/pythia-70m-seed1

Text Generation • Updated Jan 11 • 1.48k
EleutherAI/pythia-70m-seed2

Text Generation • Updated Jan 11 • 1.2k
EleutherAI/pythia-70m-seed3

Text Generation • Updated Jan 11 • 1.04k
EleutherAI/pythia-70m-seed4

Text Generation • Updated Jan 11 • 1.01k
EleutherAI/pythia-70m-seed5

Text Generation • Updated Jan 11 • 975
EleutherAI/pythia-70m-seed6

Text Generation • Updated Jan 11 • 953
EleutherAI/pythia-70m-seed7

Text Generation • Updated Jan 11 • 925
EleutherAI/pythia-70m-seed8

Text Generation • Updated Jan 11 • 1.04k
EleutherAI/pythia-70m-seed9

Text Generation • Updated Jan 11 • 941
EleutherAI/pythia-160m

Text Generation • 0.2B • Updated Jul 9, 2023 • 2.85M • 44
EleutherAI/pythia-160m-seed1

Text Generation • 0.2B • Updated Jan 11 • 2.6k
EleutherAI/pythia-160m-seed2

Text Generation • 0.2B • Updated Jan 11 • 1.94k
EleutherAI/pythia-160m-seed3

Text Generation • 0.2B • Updated Jan 11 • 1.72k
EleutherAI/pythia-160m-seed4

Text Generation • Updated Jan 11 • 1.35k • 1
EleutherAI/pythia-160m-seed5

Text Generation • Updated Jan 11 • 1.01k
EleutherAI/pythia-160m-seed6

Text Generation • Updated Jan 11 • 1.01k
EleutherAI/pythia-160m-seed7

Text Generation • Updated Jan 11 • 998
EleutherAI/pythia-160m-seed8

Text Generation • Updated Jan 11 • 1.01k
EleutherAI/pythia-160m-seed9

Text Generation • Updated Jan 11 • 1.06k
EleutherAI/pythia-410m

Text Generation • 0.5B • Updated Jul 9, 2023 • 329k • 39
EleutherAI/pythia-410m-seed1

Text Generation • Updated Jan 11 • 1.55k
EleutherAI/pythia-410m-seed2

Text Generation • Updated Jan 11 • 1.78k
EleutherAI/pythia-410m-seed3

Text Generation • Updated Jan 11 • 1.69k
EleutherAI/pythia-410m-seed4

Text Generation • Updated Jan 11 • 1.23k
EleutherAI/pythia-410m-seed5

Text Generation • Updated Jan 11 • 1.16k
EleutherAI/pythia-410m-seed6

Text Generation • Updated Jan 11 • 1.13k • 1
EleutherAI/pythia-410m-seed7

Text Generation • Updated Jan 11 • 1.63k
EleutherAI/pythia-410m-seed8

Text Generation • Updated Jan 11 • 1.48k
EleutherAI/pythia-410m-seed9

Text Generation • Updated Jan 11 • 1.5k
EleutherAI/pythia-160m-data-seed1

Text Generation • Updated Jan 11 • 64.5k

Note Version where the data order and weight initialization seeds are decoupled. Here, only the data seed is different from pythia-160m ("seed 0").
EleutherAI/pythia-160m-data-seed2

Text Generation • Updated Jan 11 • 51.9k

Note Version where the data order and weight initialization seeds are decoupled. Here, only the data seed is different from pythia-160m ("seed 0").
EleutherAI/pythia-160m-data-seed3

Text Generation • Updated Jan 11 • 51.1k

Note Version where the data order and weight initialization seeds are decoupled. Here, only the data seed is different from pythia-160m ("seed 0").
EleutherAI/pythia-160m-weight-seed1

Text Generation • Updated Jan 11 • 55.3k

Note Version where the data order and weight initialization seeds are decoupled. Here, only the weight initialization seed is different from pythia-160m ("seed 0").
EleutherAI/pythia-160m-weight-seed2

Text Generation • Updated Jan 11 • 51.5k

Note Version where the data order and weight initialization seeds are decoupled. Here, only the weight initialization seed is different from pythia-160m ("seed 0").
EleutherAI/pythia-160m-weight-seed3

Text Generation • Updated Jan 11 • 50.7k

Note Version where the data order and weight initialization seeds are decoupled. Here, only the weight initialization seed is different from pythia-160m ("seed 0").
PolyPythias: Stability and Outliers across Fifty Language Model Pre-Training Runs

Paper • 2503.09543 • Published Mar 12, 2025