eric8607242

Hello! Is there any method to keep the file in the multipart into the memory directly without dumping the file into the tempfile now? Very thanks for any response!

Hi, I create this [repo](https://github.com/eric8607242/ImageCaption_Inference) for generating the caption of an image. In this repo, I incorporate the repo [image-captioning-DLCT](https://github.com/luo3300612/image-captioning-DLCT) and [grid-feats-vqa](https://github.com/facebookresearch/grid-feats-vqa) to make image caption model inference more easily....

Sorry for the unclear README. [VAL_PATH] is the path to the directory of Imagenet validation dataset. If there is any question, please feel free to let me know. Thank you

Sorry for the late reply, and thanks for the Sungwool's response. If there is any question, please feel free to let me know. Thank you. I have pushed a new...

Thanks for the pull request! To make sure the gradient flow and update the architecture generator, I think that the architecture parameter may not be detached in line 222 of...

@feifeibear Hi, Thanks for the quick response! The conflict fix done!

Hi @1SAA , Thanks for the great work and awesome function. I think this is exactly what I want! But the new function is coupled with the `ZeroDDP.state_dict`. Implementing the...

Hi @1SAA , I think this is a great idea. Thanks for your suggestion! I will follow this rule to modify the method as soon as possible.

Hi @1SAA, I rewrite the `state_dict` model with a new bool variable `strict` to specify which version of`state_dict` would be returned. The default value is `True` to return the `state_dict`...

Hi @1SAA , Thanks for review! I modify the argument `GeminiDDP` module to `ZeroDDP` module in `get_static_torch_model` function to be compatible with the lower version (v0.1.9 and v0.1.10).