seungkukim/dit4dit_gr1_posttrain_attn_align_cosine_routed_3x6_align70k_ws_pretrain100k_8gpu-100k 10B • Updated Jul 14 • 4
seungkukim/dit4dit_gr1_posttrain_attn_align_2pair_early_layers_cosine_align70k_ws_pretrain100k-100k 10B • Updated Jul 8 • 3
seungkukim/dit4dit_gr1_posttrain_attn_align_2pair_early_layers_cosine_align60k_ws_pretrain100k-100k 10B • Updated Jul 8 • 3
seungkukim/dit4dit_gr1_posttrain_attn_align_2pair_early_layers_cosine_align75k_ws_pretrain100k-100k 10B • Updated Jul 8 • 4
seungkukim/dit4dit_gr1_posttrain_attn_align_2pair_cosine_align60k_ws_pretrain100k-100k 10B • Updated Jul 8 • 4
seungkukim/dit4dit_gr1_posttrain_attn_align_2pair_cosine_align60k_ws_pretrain100k-60k 10B • Updated Jul 7 • 2
seungkukim/dit4dit_gr1_posttrain_attn_align_2pair_early_layers_cosine_align60k-100k 10B • Updated Jul 7 • 3