appliedml42 pfp

appliedml42

@appliedml42

18 Following
61 Followers
appliedml42 pfp
appliedml42
@appliedml42
https://www.youtube.com/watch?v=sBmcprlAPb0
0 reply
0 recast
2 reactions
appliedml42 pfp
appliedml42
@appliedml42
Check out https://youtu.be/QPPFM8NyuaQ
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Reversible Vision Transformers | GPU memory requirement | Vision Transformer | Multiscale Vision Transformers | image classification | object detection | video classification | reduced memory footprint | additional computational burden
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
A memory efficient architecture design for visual recognition which decouples the GPU memory requirement from the depth of the model. We benchmark extensively across both model sizes and tasks and show up to 15.5x reduced memory footprint at roughly identical model complexity, parameters and accuracy.
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Reversible Vision Transformers https://arxiv.org/abs/2302.04869v1 https://i.imgur.com/XJlssKp.png
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Keywords: training data extraction | language models | privacy protection | text generation | text ranking | token-level criteria
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
This paper proposes tricks for improving training data extraction from language models, such as sampling strategies and token-level criteria. Experiments show that these tricks can significantly outperform the baseline in most cases, providing a stronger baseline for future research.
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Bag of Tricks for Training Data Extraction from Language Models abs:https://arxiv.org/abs/2302.04460v1 https://i.imgur.com/MT3TcVP.png
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
https://arxiv.org/abs/2302.01676v1 This paper proposes MERLIN, a novel multimodal deep learning framework designed to predict the NFT selling prices. MERLIN exploits only NFT images and textual descriptions. https://i.imgur.com/1PmQQ3M.png
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
This is an amazing read. Cramming: Training a Language Model on a Single GPU in One Day abs: https://arxiv.org/abs/2212.14034 🔥summary thread from Lucas https://twitter.com/giffmana/status/1608568387583737856?s=61&t=qPnqsJlJqse2GDklDp4hww
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
If anybody is looking to get working knowledge of Stable Diffusion. I would recommend checking out this series. This was super helpful for me when I was learning the basics of SD. https://github.com/huggingface/diffusion-models-class
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
@thethriller https://github.com/showlab/Tune-A-Video
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
@douglass another interesting diffusion based video paper that came out recently.
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Keywords: video diffusion models | text-driven image | video editing | motion editability | image animation | subject-driven video generation
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Text-based video editing approach using diffusion models. Fuses low-res video info with synthesized info to match the text prompt. Method has a mixed objective with full temporal attention and attention masking to improve motion editability. Includes a framework for image animation and subject-driven video generation.
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Dreamix: Video Diffusion Models are General Video Editors https://dreamix-video-editing.github.io/
2 replies
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Yup Karpathy’s capacity to explain something intuitively is insane. I have read many attention explanations but this one is the best.
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
screenshot from my discord arxiv bot project. https://i.imgur.com/2eWzWfD.png
0 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
Language Quantized AutoEncoders: Towards Unsupervised Text-Image Alignment https://arxiv.org/abs/2302.00902v1 https://i.imgur.com/y5DbdyX.png
1 reply
0 recast
0 reaction
appliedml42 pfp
appliedml42
@appliedml42
An amazing explanation of Attention Mechanism by the OG Karpathy. https://youtu.be/kCc8FmEb1nY
1 reply
0 recast
0 reaction