Home
Menu
Home
Start here
AI
Kubernetes
Networking
About
|
ES
EN
|
LIGHT
DARK
Slm
2026
QLoRA runbook: from dataset to adapter served in multi-LoRA (operational procedure)
Jun 9, 2026
QLoRA and multi-LoRA at the limit on small models
Jun 9, 2026
Aggressive RAG in small models: trading parameters for retrieval
Jun 9, 2026
Aggressive quantisation (state of the art): from 4-bit to ternary
Jun 9, 2026
Test-time quantization: quantising on the fly with no calibration dataset
Jun 9, 2026
Device-native architectures: fine-grained MoE and pre-attention router
Jun 9, 2026
Self-speculative decoding: the model that gets ahead of itself
Jun 9, 2026
The roofline flips: why optimising small models is a different performance game
Jun 9, 2026