Have started a youtube series on algorithmic design and thinking for interviews and engineering. https://t.co/wck6TbZxMt #compsci #algorithms #dsa #interviews #faang #Engineering

Run llama.cpp with Intel GPU with modern SYCL Backend. @IntelDevTools #llama.cpp #sycl #compiler #llm #inference #llama #cuda https://t.co/BFf3cDL8CH

Thanks to @usingstdcpp for having me at uc3m Leganes, Madrid. It was great to present about #SYCL runtime to C++ WG members and developers. https://t.co/zh754t1u0K #c++ #intel #sycl #deeplearning

I will be speaking at c++ Russia on Sycl language and it's features along with dpct compiler. Thanks to @cpp_russia for organizing. #sycl #cpp #c++Russia #intel

New era of exa scale computing. Glad to see its usage in prominent national labs across the globe. @argonne @intel

Great effort by @huggingface @GuggerSylvain to help enable for Intel GPUs. @IntelDevTools

Optimizing LLMs and diffusion models have become easier with #intel Cpu and GPU support for popular #huggingface frameworks for #pytorch ecosystem . Speed up training on multi CPU or GPUs with #accelerate to leverage efficient flops and compute performa… https://t.co/QXsI7AEOPd

Distributed deep learning is of paramount importance when it comes to scaling large models. TensorFlow XLA is a complex and efficient backend which helps in enabling better training and optimization of performance for large scale… https://t.co/znSatKWojs

There has been a lot of hype regarding #gpt4 and #chatgpt regarding their ability to perform generative modelling of multimodal spaces. Although it is not inherently #AGI, but provides a strong foundation of merging on-policy acto… https://t.co/vEhBiStmJH

PyTorch is one of the most used distributed deep learning framework .We at Intel Labs are striving to optimize PyTorch for Intel devices - from cpus to gpus. Stay tuned to understand how drastically we can speed up distributed training and inference of la… https://t.co/h05ODY9blx

Thanks @AnalyticsVidhya for having me, enjoyed delivering the session. PS: repo link : https://t.co/JFEpnZoOXN #analyticsvidhya #datahour #diffusion #AI #AIart

Excited to announce that our paper with collaboration from @imperialcollege has been accepted at @lrec2022 . The topic includes unsupervised knowledge discovery using graphs; pre-prints and recordings to be uploaded soon. #NLProc #lrec #graphs #Neuralnetworks

Today marks 5 years since the public release of PyTorch! We didn't expect to come this far, but here we're🙂- 2K Contributors, 90K Projects, 3.9M lines of "import torch" on GitHub. More importantly, we're still receiving lots of love and having a great ride. Here's to the future!

A medium blog detailing the research behind CartoonGAN :

Cartoonizing images is now easier with CartoonGAN : https://t.co/UYePqkvSF8 Based on CVPR 2018 ( https://t.co/o9aZrH0pGE) #GANs #CartoonGAN #Tensorflow

Interested in creating "longer" versions of (any) Pegasus models? LongPegasus is a tf package which is used to incude Longformer self attention on Pegasus models - you can even finetune the "long" model from HF https://t.co/jNl4MrCR2e @huggingface @TensorFlow #pegasus #longformer

Great work from @huggingface AutoNLP team which would allow users to train on their own tasks. Try it out by clicking on "Train" on these pages: https://t.co/XXzW7Mg65r https://t.co/VQUIbojuBr https://t.co/tNkQWPAcgS https://t.co/l508CF47Mw https://t.co/R9sxGcRCir @huggingface

Graph Attention Network Embeddings provide more node level details by neighborhood importance sampling.GraphAttentionNetworks is a package which extracts Multihead GAT embeddings from KGs along with trainable Layer. Repo: https://t.co/XgGvR3fe36 Made with @TensorFlow #DL

SpectralEmbeddings an embedding generator library used for creating Graph Convolution Network and Graph AutoEncoder Embeddings from Knowledge Graphs This allows projection of higher order network dependencies for creating the node embeddings https://t.co/NjUgbkpuqm

My notebook on Quantum TRPO/PPO for memory optimized quantum control in Deep RL is live on #Pennylane (1st one): https://t.co/P0r8jsvtwe Special thanks to @XanaduAI @pennylaneai .