BitsAndBytes

mentions 2 type Organization feed RSS

// recent coverage 2 mentions

18:30

2026-07-15

hello-fri-end.github.io

machine-learning

Early Breakthroughs in Transformer Quantization

Early breakthroughs in transformer quantization addressed the outlier problem that breaks naive INT8 quantization. Tim Dettmers et al. introduced LLM.int8(), which keeps outlier features in full preci…

07:14

2026-05-20

dev.to

large-language-models

I Thought Fine-Tuning LLMs Needed Expensive GPUs. I Was Wrong.

The author successfully fine-tuned a 1.1 billion parameter TinyLlama model using QLoRA on consumer hardware, training only 0.2% of the model's parameters via low-rank adapter matrices. The project inv…

// co-occurs with top 8 entities

TinyLlama 1 QLoRA 1 PEFT 1 TRL 1 Hugging Face 1 FastAPI 1 LoRA 1 Tim Dettmers 1