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 • 196
EleutherAI/pile-preshuffled-seeds

Updated Feb 12 • 326 • 1

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

Text Generation • 39.2M • Updated Feb 12 • 10k • 30
EleutherAI/pythia-14m-seed1

Text Generation • Updated Jan 11 • 2.13k
EleutherAI/pythia-14m-seed2

Text Generation • Updated Jan 11 • 1.56k
EleutherAI/pythia-14m-seed3

Text Generation • Updated Jan 11 • 1.65k
EleutherAI/pythia-14m-seed4

Text Generation • Updated Jan 11 • 1.37k
EleutherAI/pythia-14m-seed5

Text Generation • Updated Jan 11 • 1.7k
EleutherAI/pythia-14m-seed6

Text Generation • Updated Jan 11 • 1.45k
EleutherAI/pythia-14m-seed7

Text Generation • Updated Jan 11 • 1.7k
EleutherAI/pythia-14m-seed8

Text Generation • Updated Jan 11 • 1.45k
EleutherAI/pythia-14m-seed9

Text Generation • Updated Jan 11 • 1.68k
EleutherAI/pythia-31m-deduped

Text Generation • 55.7M • Updated Feb 12 • 2.08k • 5
EleutherAI/pythia-31m-seed1

Text Generation • Updated Jan 11 • 1.57k
EleutherAI/pythia-31m-seed2

Text Generation • Updated Jan 11 • 1.47k
EleutherAI/pythia-31m-seed3

Text Generation • Updated Jan 11 • 1.41k
EleutherAI/pythia-31m-seed4

Text Generation • Updated Jan 11 • 1.42k
EleutherAI/pythia-31m-seed5

Text Generation • Updated Jan 11 • 1.42k
EleutherAI/pythia-31m-seed6

Text Generation • Updated Jan 11 • 1.41k
EleutherAI/pythia-31m-seed7

Text Generation • Updated Jan 11 • 1.43k
EleutherAI/pythia-31m-seed8

Text Generation • Updated Jan 11 • 1.41k
EleutherAI/pythia-31m-seed9

Text Generation • Updated Jan 11 • 1.57k
EleutherAI/pythia-70m

95.6M • Updated Nov 21, 2023 • 520k • 85
EleutherAI/pythia-70m-seed1

Text Generation • Updated Jan 11 • 8.81k
EleutherAI/pythia-70m-seed2

Text Generation • Updated Jan 11 • 8.39k
EleutherAI/pythia-70m-seed3

Text Generation • Updated Jan 11 • 1.63k
EleutherAI/pythia-70m-seed4

Text Generation • Updated Jan 11 • 1.56k
EleutherAI/pythia-70m-seed5

Text Generation • Updated Jan 11 • 1.58k
EleutherAI/pythia-70m-seed6

Text Generation • Updated Jan 11 • 1.78k
EleutherAI/pythia-70m-seed7

Text Generation • Updated Jan 11 • 1.74k
EleutherAI/pythia-70m-seed8

Text Generation • Updated Jan 11 • 1.74k
EleutherAI/pythia-70m-seed9

Text Generation • Updated Jan 11 • 1.6k
EleutherAI/pythia-160m

Text Generation • 0.2B • Updated Jul 9, 2023 • 2.4M • 45
EleutherAI/pythia-160m-seed1

Text Generation • 0.2B • Updated Jan 11 • 25.6k
EleutherAI/pythia-160m-seed2

Text Generation • 0.2B • Updated Jan 11 • 25.3k
EleutherAI/pythia-160m-seed3

Text Generation • 0.2B • Updated Jan 11 • 6.78k
EleutherAI/pythia-160m-seed4

Text Generation • Updated Jan 11 • 3.95k • 1
EleutherAI/pythia-160m-seed5

Text Generation • Updated Jan 11 • 4.26k
EleutherAI/pythia-160m-seed6

Text Generation • Updated Jan 11 • 3.11k
EleutherAI/pythia-160m-seed7

Text Generation • Updated Jan 11 • 2.34k
EleutherAI/pythia-160m-seed8

Text Generation • Updated Jan 11 • 1.61k
EleutherAI/pythia-160m-seed9

Text Generation • Updated Jan 11 • 1.71k
EleutherAI/pythia-410m

Text Generation • 0.5B • Updated Jul 9, 2023 • 259k • 39
EleutherAI/pythia-410m-seed1

Text Generation • Updated Jan 11 • 13.4k
EleutherAI/pythia-410m-seed2

Text Generation • Updated Jan 11 • 13.3k
EleutherAI/pythia-410m-seed3

Text Generation • Updated Jan 11 • 2.2k
EleutherAI/pythia-410m-seed4

Text Generation • Updated Jan 11 • 1.84k
EleutherAI/pythia-410m-seed5

Text Generation • Updated Jan 11 • 1.89k
EleutherAI/pythia-410m-seed6

Text Generation • Updated Jan 11 • 1.73k • 1
EleutherAI/pythia-410m-seed7

Text Generation • Updated Jan 11 • 1.68k
EleutherAI/pythia-410m-seed8

Text Generation • Updated Jan 11 • 1.58k
EleutherAI/pythia-410m-seed9

Text Generation • Updated Jan 11 • 1.65k
EleutherAI/pythia-160m-data-seed1

Text Generation • Updated Jan 11 • 143k

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 • 58k

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 • 54.9k

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-weight-seed1

Text Generation • Updated Jan 11 • 68.8k

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 • 62.1k

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 • 54.4k

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