PolyPythias
- Preview • Updated • 982
EleutherAI/pile-preshuffled-seeds
Updated • 953 • 1Note Training data information for each seed.
-
EleutherAI/pythia-14m-deduped
Text Generation • 39.2M • Updated • 5.42k • 32 -
EleutherAI/pythia-14m-seed1
Text Generation • Updated • 2.52k • 1 -
EleutherAI/pythia-14m-seed2
Text Generation • Updated • 2.17k -
EleutherAI/pythia-14m-seed3
Text Generation • Updated • 1.68k -
EleutherAI/pythia-14m-seed4
Text Generation • Updated • 1.55k -
EleutherAI/pythia-14m-seed5
Text Generation • Updated • 1.72k -
EleutherAI/pythia-14m-seed6
Text Generation • Updated • 1.63k -
EleutherAI/pythia-14m-seed7
Text Generation • Updated • 1.58k -
EleutherAI/pythia-14m-seed8
Text Generation • Updated • 1.45k -
EleutherAI/pythia-14m-seed9
Text Generation • Updated • 1.58k -
EleutherAI/pythia-31m-deduped
Text Generation • 55.7M • Updated • 1.93k • 5 -
EleutherAI/pythia-31m-seed1
Text Generation • Updated • 1.89k -
EleutherAI/pythia-31m-seed2
Text Generation • Updated • 1.68k -
EleutherAI/pythia-31m-seed3
Text Generation • Updated • 1.55k -
EleutherAI/pythia-31m-seed4
Text Generation • Updated • 1.48k -
EleutherAI/pythia-31m-seed5
Text Generation • Updated • 1.51k -
EleutherAI/pythia-31m-seed6
Text Generation • Updated • 1.55k -
EleutherAI/pythia-31m-seed7
Text Generation • Updated • 1.35k -
EleutherAI/pythia-31m-seed8
Text Generation • Updated • 1.37k -
EleutherAI/pythia-31m-seed9
Text Generation • Updated • 1.4k -
EleutherAI/pythia-70m
95.6M • Updated • 230k • 88 -
EleutherAI/pythia-70m-seed1
Text Generation • Updated • 5.41k -
EleutherAI/pythia-70m-seed2
Text Generation • Updated • 4.21k -
EleutherAI/pythia-70m-seed3
Text Generation • Updated • 3.37k -
EleutherAI/pythia-70m-seed4
Text Generation • Updated • 2.35k -
EleutherAI/pythia-70m-seed5
Text Generation • Updated • 2.53k -
EleutherAI/pythia-70m-seed6
Text Generation • Updated • 2.79k -
EleutherAI/pythia-70m-seed7
Text Generation • Updated • 2.11k -
EleutherAI/pythia-70m-seed8
Text Generation • Updated • 1.91k -
EleutherAI/pythia-70m-seed9
Text Generation • Updated • 2.11k -
EleutherAI/pythia-160m
Text Generation • 0.2B • Updated • 3.49M • 66 -
EleutherAI/pythia-160m-seed1
Text Generation • 0.2B • Updated • 30.7k -
EleutherAI/pythia-160m-seed2
Text Generation • 0.2B • Updated • 16.5k -
EleutherAI/pythia-160m-seed3
Text Generation • 0.2B • Updated • 15.8k -
EleutherAI/pythia-160m-seed4
Text Generation • Updated • 3.13k • 1 -
EleutherAI/pythia-160m-seed5
Text Generation • Updated • 2.98k -
EleutherAI/pythia-160m-seed6
Text Generation • Updated • 2.26k -
EleutherAI/pythia-160m-seed7
Text Generation • Updated • 2.38k -
EleutherAI/pythia-160m-seed8
Text Generation • Updated • 2.26k -
EleutherAI/pythia-160m-seed9
Text Generation • Updated • 2.42k -
EleutherAI/pythia-410m
Text Generation • 0.5B • Updated • 92.6k • 41 -
EleutherAI/pythia-410m-seed1
Text Generation • Updated • 4.05k -
EleutherAI/pythia-410m-seed2
Text Generation • Updated • 3.15k -
EleutherAI/pythia-410m-seed3
Text Generation • Updated • 3.1k -
EleutherAI/pythia-410m-seed4
Text Generation • Updated • 2.83k -
EleutherAI/pythia-410m-seed5
Text Generation • Updated • 2.35k -
EleutherAI/pythia-410m-seed6
Text Generation • Updated • 2.17k • 1 -
EleutherAI/pythia-410m-seed7
Text Generation • Updated • 2.16k -
EleutherAI/pythia-410m-seed8
Text Generation • Updated • 1.9k -
EleutherAI/pythia-410m-seed9
Text Generation • Updated • 2.09k
EleutherAI/pythia-160m-data-seed1
Text Generation • Updated • 2.12kNote 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 • 1.9kNote 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 • 1.82kNote 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 • 2.15kNote 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 • 2.02kNote 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 • 1.41kNote 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
EleutherAI/pythia-160m-weight-seed1-corrected
Text Generation • Updated • 443Note Corrected weight-seed 1: independent initialization with the original Pythia data order.
EleutherAI/pythia-160m-weight-seed2-corrected
Text Generation • Updated • 778Note Corrected weight-seed 1: independent initialization with the original Pythia data order.