PolyPythias
- Preview • Updated • 53
EleutherAI/pile-preshuffled-seeds
Updated • 242 • 1Note Training data information for each seed.
-
EleutherAI/pythia-14m-deduped
Text Generation • 39.2M • Updated • 21.6k • 29 -
EleutherAI/pythia-14m-seed1
Text Generation • Updated • 5.05k -
EleutherAI/pythia-14m-seed2
Text Generation • Updated • 5.19k -
EleutherAI/pythia-14m-seed3
Text Generation • Updated • 5.14k -
EleutherAI/pythia-14m-seed4
Text Generation • Updated • 4.08k -
EleutherAI/pythia-14m-seed5
Text Generation • Updated • 4.35k -
EleutherAI/pythia-14m-seed6
Text Generation • Updated • 4.88k -
EleutherAI/pythia-14m-seed7
Text Generation • Updated • 5.12k -
EleutherAI/pythia-14m-seed8
Text Generation • Updated • 4.09k -
EleutherAI/pythia-14m-seed9
Text Generation • Updated • 5.3k -
EleutherAI/pythia-31m-deduped
Text Generation • 55.7M • Updated • 2.89k • 5 -
EleutherAI/pythia-31m-seed1
Text Generation • Updated • 4.05k -
EleutherAI/pythia-31m-seed2
Text Generation • Updated • 3.92k -
EleutherAI/pythia-31m-seed3
Text Generation • Updated • 4.27k -
EleutherAI/pythia-31m-seed4
Text Generation • Updated • 2.73k -
EleutherAI/pythia-31m-seed5
Text Generation • Updated • 3.66k -
EleutherAI/pythia-31m-seed6
Text Generation • Updated • 3.55k -
EleutherAI/pythia-31m-seed7
Text Generation • Updated • 3.87k -
EleutherAI/pythia-31m-seed8
Text Generation • Updated • 2.72k -
EleutherAI/pythia-31m-seed9
Text Generation • Updated • 4.16k -
EleutherAI/pythia-70m
95.6M • Updated • 320k • 79 -
EleutherAI/pythia-70m-seed1
Text Generation • Updated • 3.86k -
EleutherAI/pythia-70m-seed2
Text Generation • Updated • 3.26k -
EleutherAI/pythia-70m-seed3
Text Generation • Updated • 3.41k -
EleutherAI/pythia-70m-seed4
Text Generation • Updated • 2.9k -
EleutherAI/pythia-70m-seed5
Text Generation • Updated • 3.14k -
EleutherAI/pythia-70m-seed6
Text Generation • Updated • 2.87k -
EleutherAI/pythia-70m-seed7
Text Generation • Updated • 3.98k -
EleutherAI/pythia-70m-seed8
Text Generation • Updated • 3.93k -
EleutherAI/pythia-70m-seed9
Text Generation • Updated • 2.87k -
EleutherAI/pythia-160m
Text Generation • Updated • 2.58M • 39 -
EleutherAI/pythia-160m-seed1
Text Generation • 0.2B • Updated • 5.08k -
EleutherAI/pythia-160m-seed2
Text Generation • 0.2B • Updated • 5.29k -
EleutherAI/pythia-160m-seed3
Text Generation • 0.2B • Updated • 5.3k -
EleutherAI/pythia-160m-seed4
Text Generation • Updated • 3.77k • 1 -
EleutherAI/pythia-160m-seed5
Text Generation • Updated • 4.17k -
EleutherAI/pythia-160m-seed6
Text Generation • Updated • 2.85k -
EleutherAI/pythia-160m-seed7
Text Generation • Updated • 3.98k -
EleutherAI/pythia-160m-seed8
Text Generation • Updated • 2.83k -
EleutherAI/pythia-160m-seed9
Text Generation • Updated • 2.98k -
EleutherAI/pythia-410m
Text Generation • 0.5B • Updated • 167k • 36 -
EleutherAI/pythia-410m-seed1
Text Generation • Updated • 3.63k -
EleutherAI/pythia-410m-seed2
Text Generation • Updated • 3.52k -
EleutherAI/pythia-410m-seed3
Text Generation • Updated • 3.6k -
EleutherAI/pythia-410m-seed4
Text Generation • Updated • 2.88k -
EleutherAI/pythia-410m-seed5
Text Generation • Updated • 2.94k -
EleutherAI/pythia-410m-seed6
Text Generation • Updated • 2.81k • 1 -
EleutherAI/pythia-410m-seed7
Text Generation • Updated • 2.84k -
EleutherAI/pythia-410m-seed8
Text Generation • Updated • 2.81k -
EleutherAI/pythia-410m-seed9
Text Generation • Updated • 2.82k
EleutherAI/pythia-160m-data-seed1
Text Generation • Updated • 3.32kNote 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 • 3.25kNote 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 • 3.41kNote 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 • 2kNote 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 • 1.98kNote 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 • 1.98kNote 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