PolyPythias
-
Preview • Updated • 275
EleutherAI/pile-preshuffled-seeds
Updated • 192 • 1Note Training data information for each seed.
-
EleutherAI/pythia-14m
Text Generation • 39.2M • Updated • 76.4k • 25 -
EleutherAI/pythia-14m-seed1
Updated • 103 -
EleutherAI/pythia-14m-seed2
Updated • 94 -
EleutherAI/pythia-14m-seed3
Updated • 94 -
EleutherAI/pythia-14m-seed4
Updated • 94 -
EleutherAI/pythia-14m-seed5
Updated • 92 -
EleutherAI/pythia-14m-seed6
Updated • 93 -
EleutherAI/pythia-14m-seed7
Updated • 95 -
EleutherAI/pythia-14m-seed8
Updated • 93 -
EleutherAI/pythia-14m-seed9
Updated • 94 -
EleutherAI/pythia-31m
Text Generation • 55.7M • Updated • 7.29k • 5 -
EleutherAI/pythia-31m-seed1
Updated -
EleutherAI/pythia-31m-seed2
Updated -
EleutherAI/pythia-31m-seed3
Updated -
EleutherAI/pythia-31m-seed4
Updated -
EleutherAI/pythia-31m-seed5
Updated -
EleutherAI/pythia-31m-seed6
Updated • 1 -
EleutherAI/pythia-31m-seed7
Updated -
EleutherAI/pythia-31m-seed8
Updated -
EleutherAI/pythia-31m-seed9
Updated -
EleutherAI/pythia-70m
95.6M • Updated • 64k • 74 -
EleutherAI/pythia-70m-seed1
Updated • 58 -
EleutherAI/pythia-70m-seed2
Updated • 59 -
EleutherAI/pythia-70m-seed3
Updated • 58 -
EleutherAI/pythia-70m-seed4
Updated • 54 -
EleutherAI/pythia-70m-seed5
Updated • 56 -
EleutherAI/pythia-70m-seed6
Updated • 54 -
EleutherAI/pythia-70m-seed7
Updated • 52 -
EleutherAI/pythia-70m-seed8
Updated • 54 -
EleutherAI/pythia-70m-seed9
Updated • 52 -
EleutherAI/pythia-160m
Text Generation • 0.2B • Updated • 68k • 35 -
EleutherAI/pythia-160m-seed1
Text Generation • 0.2B • Updated • 1.8k -
EleutherAI/pythia-160m-seed2
Text Generation • 0.2B • Updated • 1.8k -
EleutherAI/pythia-160m-seed3
Text Generation • 0.2B • Updated • 1.82k -
EleutherAI/pythia-160m-seed4
Updated • 688 • 1 -
EleutherAI/pythia-160m-seed5
Updated • 676 -
EleutherAI/pythia-160m-seed6
Updated • 671 -
EleutherAI/pythia-160m-seed7
Updated • 669 -
EleutherAI/pythia-160m-seed8
Updated • 673 -
EleutherAI/pythia-160m-seed9
Updated • 673 -
EleutherAI/pythia-410m
Text Generation • 0.5B • Updated • 38.3k • 35 -
EleutherAI/pythia-410m-seed1
Updated • 681 -
EleutherAI/pythia-410m-seed2
Updated • 676 -
EleutherAI/pythia-410m-seed3
Updated • 672 -
EleutherAI/pythia-410m-seed4
Updated • 664 -
EleutherAI/pythia-410m-seed5
Updated • 675 -
EleutherAI/pythia-410m-seed6
Updated • 672 -
EleutherAI/pythia-410m-seed7
Updated • 672 -
EleutherAI/pythia-410m-seed8
Updated • 674 -
EleutherAI/pythia-410m-seed9
Updated • 672
EleutherAI/pythia-160m-data-seed1
UpdatedNote 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
UpdatedNote 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
UpdatedNote 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
UpdatedNote 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
UpdatedNote 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
UpdatedNote 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