Upload Minisun Trained using model.fit on NeelNanda/pile-10k[0-5000],lr 1e-4,cw 128,2 epoch,batch size 8,cosine with restart

Upload Minisun Trained using model.fit on stas/openwebtext-10k[0-5000],lr 1e-4,cw 128,2 epoch,batch size 4,cosine with restart

Upload Minisun Trained using model.fit on lss dataset[400],lr 2e-4, 2 epoch,batch size 2,cosine with restart