PolyPythias
-
Preview • Updated • 1.93k
EleutherAI/pile-preshuffled-seeds
Updated • 108 • 1Note Training data information for each seed.
-
EleutherAI/pythia-14m
Text Generation • 39.2M • Updated • 51.4k • 25 -
EleutherAI/pythia-14m-seed1
Updated • 78 -
EleutherAI/pythia-14m-seed2
Updated • 58 -
EleutherAI/pythia-14m-seed3
Updated • 49 -
EleutherAI/pythia-14m-seed4
Updated • 46 -
EleutherAI/pythia-14m-seed5
Updated • 21 -
EleutherAI/pythia-14m-seed6
Updated • 21 -
EleutherAI/pythia-14m-seed7
Updated • 22 -
EleutherAI/pythia-14m-seed8
Updated • 21 -
EleutherAI/pythia-14m-seed9
Updated • 39 -
EleutherAI/pythia-31m
Text Generation • 55.7M • Updated • 6.29k • 5 -
EleutherAI/pythia-31m-seed1
Updated • 8 -
EleutherAI/pythia-31m-seed2
Updated • 6 -
EleutherAI/pythia-31m-seed3
Updated • 7 -
EleutherAI/pythia-31m-seed4
Updated • 7 -
EleutherAI/pythia-31m-seed5
Updated • 1 -
EleutherAI/pythia-31m-seed6
Updated -
EleutherAI/pythia-31m-seed7
Updated -
EleutherAI/pythia-31m-seed8
Updated -
EleutherAI/pythia-31m-seed9
Updated -
EleutherAI/pythia-70m
95.6M • Updated • 141k • 76 -
EleutherAI/pythia-70m-seed1
Updated • 943 -
EleutherAI/pythia-70m-seed2
Updated • 890 -
EleutherAI/pythia-70m-seed3
Updated • 892 -
EleutherAI/pythia-70m-seed4
Updated • 7 -
EleutherAI/pythia-70m-seed5
Updated • 9 -
EleutherAI/pythia-70m-seed6
Updated • 14 -
EleutherAI/pythia-70m-seed7
Updated -
EleutherAI/pythia-70m-seed8
Updated • 1 -
EleutherAI/pythia-70m-seed9
Updated • 4 -
EleutherAI/pythia-160m
Text Generation • 0.2B • Updated • 84.3k • 36 -
EleutherAI/pythia-160m-seed1
Text Generation • 0.2B • Updated • 803 -
EleutherAI/pythia-160m-seed2
Text Generation • 0.2B • Updated • 784 -
EleutherAI/pythia-160m-seed3
Text Generation • 0.2B • Updated • 750 -
EleutherAI/pythia-160m-seed4
Updated • 8 • 1 -
EleutherAI/pythia-160m-seed5
Updated • 2 -
EleutherAI/pythia-160m-seed6
Updated • 2 -
EleutherAI/pythia-160m-seed7
Updated • 2 -
EleutherAI/pythia-160m-seed8
Updated • 2 -
EleutherAI/pythia-160m-seed9
Updated • 2 -
EleutherAI/pythia-410m
Text Generation • 0.5B • Updated • 46.8k • 36 -
EleutherAI/pythia-410m-seed1
Updated • 27 -
EleutherAI/pythia-410m-seed2
Updated • 18 -
EleutherAI/pythia-410m-seed3
Updated • 6 -
EleutherAI/pythia-410m-seed4
Updated • 7 -
EleutherAI/pythia-410m-seed5
Updated • 1 -
EleutherAI/pythia-410m-seed6
Updated • 2 -
EleutherAI/pythia-410m-seed7
Updated • 1 -
EleutherAI/pythia-410m-seed8
Updated • 2 -
EleutherAI/pythia-410m-seed9
Updated • 2
EleutherAI/pythia-160m-data-seed1
Updated • 2Note 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
Updated • 1Note 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
Updated • 2Note 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
Updated • 2Note 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
Updated • 2Note 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
Updated • 2Note 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