#quantization search results

Bitsandbytes unlocks accessible LLMs with INT8 and 4-bit quantization in PyTorch. Huge for memory-efficient training and inference. A must-know for scaling your LLM projects. #LLMs #Python #Quantization

Training Dynamics Impact Post-Training Quantization Robustness 👥 Albert Catalan-Tatjer, Niccolò Ajroldi & Jonas Geiping #AIResearch #MachineLearning #Quantization #NLP #DeepLearning 🔗 trendtoknow.ai

#Quantization is based on #chord structure and is a chord phenomenon. #physics,#quantum,#Theoreticalphysics,#stringtheory,#Quantumfield,#QuantumPhysics,#CERN,#painting,#music zenodo.org/records/172180…

Training Dynamics Impact Post-Training Quantization Robustness 👥 Albert Catalan-Tatjer, Niccolò Ajroldi & Jonas Geiping #AIResearch #MachineLearning #Quantization #NLP #DeepLearning 🔗 trendtoknow.ai

#COLM2025 #LLM #Quantization #ReasoningModels #EfficientAI 🚀 Thrilled to introduce our recent work at COLM 2025: “Quantization Hurts Reasoning? An Empirical Study on Quantized Reasoning Models”, presented today at Poster Session 3, #74.

Mind-blown by the elegance of FP4 bit-packing. 🤯 Halve your AI model's size by packing two 4-bit floats into a single uint8. No native FP4 type needed. byte = low_fp4 | (high_fp4 << 4) Simple and powerful. #AI #LLM #Quantization #DeepLearning #Optimization

Ever wondered how to make Large Language Models (LLMs) run faster and cheaper — without hurting performance? Let’s talk about #Quantization — the secret sauce behind efficient #LLM deployment 👇

On the Role of Discrete Representation in Sparse Mixture of Experts Giang Do, Kha Pham, Hung Le, Truyen Tran. Action editor: Naigang Wang. openreview.net/forum?id=GTWKm… #quantization #sparse #vqmoe

On the Role of Discrete Representation in Sparse Mixture of Experts Giang Do, Kha Pham, Hung Le, Truyen Tran tmlr.infinite-conf.org/paper_pages/GT… #quantization #sparse #vqmoe

.@Huawei unveils #SINQ, an open-source #quantization tech that slashes #LLM memory by 60-70%, enabling deployment on affordable hardware like consumer #GPUs. Efficiency unlocked. 💡 #AI #OpenSource #ML @VentureBeat @carlfranzen venturebeat.com/ai/huaweis-new…

pytorch-playground - Predefined PyTorch models on popular datasets for learning and benchmarking. #PyTorch #DeepLearning #Quantization

Happy to share our new study on the interaction between #optimizers and #quantization! We show how optimizer choice affects quantized model quality and why outlier-based metrics (like Kurtosis and MMR) often fail to predict performance. Paper: arxiv.org/pdf/2509.23500 [1/5]

SSTQ edges out OpenAI's MXFP4 with semantic-aware precision—10-20% better accuracy on key tasks! Complements xAI's distillation for 3x efficiency. Challenges GPT-OSS-120b, Grok-4; boosts LLaMA-2 13B, Mixtral 8x7B, Falcon 40B. OSS soon—DM for beta! #AI #LLM #Quantization 🔗

🚀 Unveiling SSTQ: Semantic-aware quantization slashing LLM inference costs by 80%! Unified sparsity, precision, & caching via novel math. OSS coming, enterprise beta open! DM for access. #AI #LLM #Quantization 🔗 zetareticula.com

Just implemented the Post Training Quantization on a model and reduced the size by 44% (32bit -> 8bit). Learned tons: PTQ vs QAT, symmetric vs asymmetric quantization, and how a full PTQ pipeline (CLE, AdaRound, bias correction, activation calibration) #Pytorch #Quantization

Happy to share our new study on the interaction between #optimizers and #quantization! We show how optimizer choice affects quantized model quality and why outlier-based metrics (like Kurtosis and MMR) often fail to predict performance. Paper: arxiv.org/pdf/2509.23500 [1/5]

Tchebichef Transforms for #Image #Compression Using Variable #Quantization #computerscience scirp.org/journal/PaperI…

#COLM2025 #LLM #Quantization #ReasoningModels #EfficientAI 🚀 Thrilled to introduce our recent work at COLM 2025: “Quantization Hurts Reasoning? An Empirical Study on Quantized Reasoning Models”, presented today at Poster Session 3, #74.

medium.com/@tspann/how-go… #Milvus #Quantization

What is LLM Quantization and How to Use Them? sabrepc.com/blog/deep-lear… #LLM #Quantization

#Riemann #Complex-Surface Unified-#Quantization 'Replacing' #FeynmanPI ow.ly/Z2Pk3013vs8 #Mathematics #Physics

#Quantization of SpaceTime Based on a '#SpaceTime Interval Operator'! ow.ly/4n27IV #Mathematics #Physics

58% of companies are not optimizing their machine learning models, despite the performance gains techniques like #quantization and #pruning can offer. Why? @mjohnk11 has a theory (hint: it's hard!) and is excited to demo easy model optimization solutions at @odsc next week.

Let's do a simple #quantization ... Practice ... #tensorflow meetup #Zurich

Book review: Nanomaterials, Vol. 2: Quantization and Entropy #DeGruyter #Nanomaterials #Quantization #Entropy Read more here: ow.ly/H5se50CzZgj

HQ-VAE: Hierarchical Discrete Representation Learning with Variational Bayes openreview.net/forum?id=1rowo… #autoencoder #quantization #autoencoding

arxiv.org/abs/2007.06919 fcos/retinanet에 대한 quantization. 늘 그렇듯(?) 배치놈과 씨름을 좀 하고 quantized 모델의 파인튜닝 과정을 개선. 공개되어있는 레포가 흥미로움. (github.com/blueardour/mod…) #quantization #detection

Want to boost your #AI model’s performance? The top techniques, like pruning, #quantization, and hyperparameter tuning, can make a big difference: helping you run models faster and tackle issues like model drift. Know more: bit.ly/3UV4YzW #AImodel #DeepLearning #ARTiBA

Home sweet home. Back to my cozy stuff for this winter vacation. #Quantization #QuantumMechanics #Polarization #Oscilliator

Quantization is widely used in data compression, digital image processing, and signal processing. Learn more: i.mtr.cool/arecnwqhxx #Quantization

الحمد لله، تم نشر بحثتنا بعنوان: Enabling Efficient Training of Convolutional Neural Networks for Histopathology Images #DeepLearning #Quantization #ComputationalPathology link.springer.com/chapter/10.100… وهنا ملخص البحث: youtu.be/vao1KQaktWo

Quantized LLMs have pros and cons. The advantages are smaller models, increased scalability, and faster inference. The main disadvantage is a potential loss of accuracy. #NLP #Quantization

📣Have you heard? Criteo is open-sourcing its automatic KNN indexing library. Get ready to build state-of-the-art indices with no effort! To know more, check our latest @Medium article 👉 tinyurl.com/3vceh9rd #Quantization #Faiss #knnindex #Python

Exploiting Latent Properties to Optimize Neural Codecs openreview.net/forum?id=Sv0FW… #codecs #decoding #quantization

Variation-aware Vision Transformer Quantization openreview.net/forum?id=yVyta… #cnns #cnn #quantization