allegrolab/hubble-8b-100b_toks-paraphrased-perturbed-hf
Text Generation
•
8B
•
Updated
•
8
Hubble - Paraphrase
Two models trained on paraphrased YAGO biographies and MMLU test sets to study how paraphrased information is memorized.
-
-
allegrolab/hubble-1b-100b_toks-paraphrased-perturbed-hf
Text Generation • 1B • Updated • 9
allegrolab/dclm-baseline-500b_toks
Updated • 255Note Use revision 'perturbed-100b-paraphrased'
-
allegrolab/hubble-8b-100b_toks-paraphrased-perturbed-neox
Text Generation • Updated -
allegrolab/hubble-1b-100b_toks-paraphrased-perturbed-neox
Text Generation • Updated