在新选项卡中打开链接
  1. GitHub - lucidrains/DALLE2-pytorch: Implementation of DALL-E 2, …

    • Implementation of DALL-E 2, OpenAI's updated text-to-image synthesis neural network, in Pytorch.
      Yannic Kilcher summary | AssemblyAI explainer… 展开

    Status

    •A research group has used the code in this repository to train a functional diffusion prior for their CLIP generations. Will share their work once they release their preprint. This, and Katherine's … 展开

    Github
    Pre-Trained Models

    •LAION is training prior models. Checkpoints are available on 🤗huggingface and the training statistics are available on 🐝WANDB.
    •Decoder - In-pr… 展开

    Github
    Appreciation

    This library would not have gotten to this working state without the help of
    •Zion for the distributed training code for the diffusion prior
    •Aidan for the di… 展开

    Github
    Usage

    To train DALLE-2 is a 3 step process, with the training of CLIP being the most important
    To train CLIP, you can either use x-clip package, or join … 展开

    Github
    Training on Preprocessed CLIP Embeddings

    It is likely, when scaling up, that you would first preprocess your images and text into corresponding embeddings before training the prior network. You can do so easily by simply passing i… 展开

    Github
     
  1. DALL·E 2 - OpenAI

  2. 你可能喜欢的搜索

  3. AIGC学习笔记——DALL-E2详解+测试 - CSDN博客

    2024年2月24日 · dall-e 2 是一种基于语言的人工智能图像生成器,可以根据文本提示创建高质量的图像和艺术作品。它使用clip、先验和 unclip 模型来生成图像,其质量取决于文本提示的具体性。

  4. 李沐论文精读系列五:DALL·E2(生成模型串讲,从GAN…

    2022年11月8日 · 相比 dall·e ,dall·e2 可以生成真实和更准确的画像:综合文本描述中给出的概念、属性与风格等三个元素,生成「现实主义」图像与艺术作品!分辨率更是提高了4倍!

  5. GitHub - marinone94/divedeep-DALLE2-pytorch: DIVE DEEP into ...

  6. GitHub - knowrohit/DALLE-open_source: Implementation of DALL …

  7. 【论文精读】DALL·E2 - 知乎 - 知乎专栏

  8. DALL·E 2 解读 | 结合预训练CLIP和扩散模型实现文本-图 …

    DALL·E 2 模型中,作者团队尝试了两种先验模型:自回归式Autoregressive (AR) prior 和扩散模型Diffusion prior [1]。 实验效果上发现两种模型的性能相似,而因为扩散模型效率较高,因此最终选择了扩散模型作为prior模块。

  9. [2204.13807] A very preliminary analysis of DALL-E 2 - arXiv.org

  10. Communication in Human–AI Co-Creation: Perceptual …

    2022年11月8日 · The main novelty of DALL-E 2 seems to be an extra layer of indirection with the prior network, which predicts an image embedding based on the text embedding from CLIP. Specifically, this repository will only build out …