Shanshan Zhong
Shanshan Zhong
记得在 `/python/oneflow/framework/docstr/__init__.py` import doc
https://github.com/haotian-liu/LLaVA/issues/626#issuecomment-1772025961
We have provided an example checkpoint of SUR-adapter [[Google Drive](https://drive.google.com/drive/folders/1UyC9_AqTezmHXmj4dh0A-9RBKKx_JmJZ?usp=share_link)]. Please try it!
感谢你的关注！我们在 [论文](https://arxiv.org/pdf/2305.05189.pdf) 的 **Knowledge from LLM** 这部分内容对于语义特征的处理进行了描述：  我们在 **Knowledge from LLM** 中说明了我们在 token 维度取了均值，以使得 token 对齐。也即您只需要通过 `h.mean(1)` 就可以获取 SUR-adapter 蒸馏过程中使用的语义表征。
您好，我们提供了checkpoint [[Google Drive](https://drive.google.com/drive/folders/1UyC9_AqTezmHXmj4dh0A-9RBKKx_JmJZ?usp=share_link)]，您可以测试一下模型效果
Maybe you'll need VPN to access these goodies in Hugging Face🤣. But fear not! We've got a workaround – you can snag our [dataset](https://modelscope.cn/datasets/shan233/CLoT-Oogiri-GO/summary) and [checkpoint](https://modelscope.cn/models/shan233/CLoT-cn/summary) via ModelScope. If you...
**Creativity is challenging**. The images used in the Oogiri game are general images. In fact, as mentioned in our paper, innovation itself is a tough endeavor. Even top human players...
Our testing was conducted on Linux. Due to a lack of Apple Mac devices, we're unable to address your query. Here is a suggestion. "Out of memory" errors often occur...
Our model is based on `Qwen/Qwen-VL-Chat`, and using the int4 version of `Qwen-VL` directly might compromise the quality of responses. Humor is subjective. You can try generating a few more...
Our testing was conducted on Linux. Due to a lack of Apple Mac devices, we're unable to address your query. If you intend to run the model on an Apple...