PolyPythias
-
Preview • Updated • 96
EleutherAI/pile-preshuffled-seeds
Updated • 296 • 1Note Training data information for each seed.
-
EleutherAI/pythia-14m-deduped
Text Generation • 39.2M • Updated • 56.9k • 28 -
EleutherAI/pythia-14m-seed1
Text Generation • Updated • 119 -
EleutherAI/pythia-14m-seed2
Text Generation • Updated • 217 -
EleutherAI/pythia-14m-seed3
Text Generation • Updated • 113 -
EleutherAI/pythia-14m-seed4
Text Generation • Updated • 107 -
EleutherAI/pythia-14m-seed5
Text Generation • Updated • 143 -
EleutherAI/pythia-14m-seed6
Text Generation • Updated • 133 -
EleutherAI/pythia-14m-seed7
Text Generation • Updated • 143 -
EleutherAI/pythia-14m-seed8
Text Generation • Updated • 138 -
EleutherAI/pythia-14m-seed9
Text Generation • Updated • 135 -
EleutherAI/pythia-31m-deduped
Text Generation • 55.7M • Updated • 2.99k • 5 -
EleutherAI/pythia-31m-seed1
Text Generation • Updated • 111 -
EleutherAI/pythia-31m-seed2
Text Generation • Updated • 112 -
EleutherAI/pythia-31m-seed3
Text Generation • Updated • 103 -
EleutherAI/pythia-31m-seed4
Text Generation • Updated • 98 -
EleutherAI/pythia-31m-seed5
Text Generation • Updated • 106 -
EleutherAI/pythia-31m-seed6
Text Generation • Updated • 110 -
EleutherAI/pythia-31m-seed7
Text Generation • Updated • 108 -
EleutherAI/pythia-31m-seed8
Text Generation • Updated • 104 -
EleutherAI/pythia-31m-seed9
Text Generation • Updated • 106 -
EleutherAI/pythia-70m
95.6M • Updated • 78k • 78 -
EleutherAI/pythia-70m-seed1
Text Generation • Updated • 908 -
EleutherAI/pythia-70m-seed2
Text Generation • Updated • 588 -
EleutherAI/pythia-70m-seed3
Text Generation • Updated • 554 -
EleutherAI/pythia-70m-seed4
Text Generation • Updated • 525 -
EleutherAI/pythia-70m-seed5
Text Generation • Updated • 551 -
EleutherAI/pythia-70m-seed6
Text Generation • Updated • 526 -
EleutherAI/pythia-70m-seed7
Text Generation • Updated • 530 -
EleutherAI/pythia-70m-seed8
Text Generation • Updated • 531 -
EleutherAI/pythia-70m-seed9
Text Generation • Updated • 518 -
EleutherAI/pythia-160m
Text Generation • Updated • 1.22M • 38 -
EleutherAI/pythia-160m-seed1
Text Generation • 0.2B • Updated • 967 -
EleutherAI/pythia-160m-seed2
Text Generation • 0.2B • Updated • 913 -
EleutherAI/pythia-160m-seed3
Text Generation • 0.2B • Updated • 906 -
EleutherAI/pythia-160m-seed4
Text Generation • Updated • 848 • 1 -
EleutherAI/pythia-160m-seed5
Text Generation • Updated • 608 -
EleutherAI/pythia-160m-seed6
Text Generation • Updated • 592 -
EleutherAI/pythia-160m-seed7
Text Generation • Updated • 587 -
EleutherAI/pythia-160m-seed8
Text Generation • Updated • 594 -
EleutherAI/pythia-160m-seed9
Text Generation • Updated • 579 -
EleutherAI/pythia-410m
Text Generation • 0.5B • Updated • 69k • 36 -
EleutherAI/pythia-410m-seed1
Text Generation • Updated • 648 -
EleutherAI/pythia-410m-seed2
Text Generation • Updated • 1.25k -
EleutherAI/pythia-410m-seed3
Text Generation • Updated • 609 -
EleutherAI/pythia-410m-seed4
Text Generation • Updated • 568 -
EleutherAI/pythia-410m-seed5
Text Generation • Updated • 530 -
EleutherAI/pythia-410m-seed6
Text Generation • Updated • 846 • 1 -
EleutherAI/pythia-410m-seed7
Text Generation • Updated • 538 -
EleutherAI/pythia-410m-seed8
Text Generation • Updated • 571 -
EleutherAI/pythia-410m-seed9
Text Generation • Updated • 712
EleutherAI/pythia-160m-data-seed1
Text Generation • Updated • 138Note 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 • 142Note 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 • 144Note 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 • 245Note 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 • 213Note 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 • 203Note 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