Mitigating Quantization Errors Due to Activation Spikes in Gated Linear Unit-Based Large Language Models
Modern large language models (LLMs) achieve state-of-the-art performance through architectural advancements but require high computational costs for inference. Post-training quantization is a widely adopted approach to reduce these costs by quantizing weights and activations to lower precision, such...
Uloženo v:
| Hlavní autoři: | , , , |
|---|---|
| Médium: | Artigo |
| Jazyk: | Inglês |
| Vydáno: |
MDPI AG
2025-04-01
|
| Edice: | Future Internet |
| Témata: | |
| On-line přístup: | https://www.mdpi.com/1999-5903/17/4/185 |
| Tagy: |
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|