I was plan to participate the neurips workshop. But this was my first time so I was confused how to select the workshop organization,paper idea I was totally confused. I have in-depth knowledge about the ml and deep learning . I was interested in every subject filed so I was wonder how to choose also confused about the workshop what they expect. Help me give me advice, suggestion 🙏 ..
Hi everyone! 👋
I wrote an extensive technical deep-dive into Moonshot AI's Kimi K3, analyzing its architectural innovations, training stability tricks, and benchmark performance.
The blog covers:
I’d love to get your feedback and discuss any specific parts of the architecture!
Blog: https://imrancoder786.github.io/blog-post.html?post=kimi-k3-deep-dive
Hi everyone! 👋
I built and trained the complete Transformer architecture from scratch using pure PyTorch (`torch.nn` primitives) based on the original "Attention Is All You Need" paper.
I trained the model on an English-to-Tamil parallel translation dataset (`gopi30/english-tamil` on Hugging Face) using dual NVIDIA T4 GPUs on Kaggle.
I wrote a detailed mathematical breakdown and step-by-step tutorial covering every equation, tensor shape transformation, and PyTorch block.
Full Blog Post: https://imrancoder786.github.io/blog-post.html?post=transformer-from-scratch
GitHub Repository:
https://github.com/imrancoder786/ML_FROM_SCRATCH/tree/main/Transformer_from_scratch
I’d love to hear your feedback, suggestions, or any questions on the code/math!
I’d love to hear your feedback, suggestions, or any questions on the code/math!
Hey there! I'm a final-year CS undergrad looking for partners to work on ML/DL research problems. I have a solid understanding of the math behind AI and core ML/DL concepts, and I'm good with PyTorch. Hit me up if you want to collaborate!
Hey there! I'm a final-year CS undergrad looking for partners to work on ML/DL research problems. I have a solid understanding of the math behind AI and core ML/DL concepts, and I'm good with PyTorch. Hit me up if you want to collaborate!
(love to work on complex problem)