DeepSeek-V4-Flash-Vision-Exp(huggingface.co)
18 pointsby amulyabaral8 hours ago1 comment
  • syntaxing7 hours ago
    I’m honestly surprised this is better benchmark wise than the text only model. I figured the addition of vision would take away from some of the text capabilities.
    • mcbuilder7 hours ago
      One of the early results from multimodal training is that it kinda works like cross training. Training vision helps with text tasks and visa versa.
    • Llamamoe4 hours ago
      I believe that multimodal training increases the robustness of latent representations regardless of which modality is being processed.
      • nateb20224 hours ago
        Tangentially this makes me wonder how large Opus really is. Perhaps Opus is a lot smaller than most of the 1T+ assumptions, just a lot more post-training/ finetuning on a 300-400B sized MoE model.