Latent Weight Quantization for Integerized Training of Deep Neural Networks
Existing methods for integerized training speed up deep learning by using low-bitwidth integerized weights, activations, gradients, and optimizer buffers. However, they overlook the issue of full-precision latent weights, which consume excessive memory to accumulate gradient-based updates for optimi...
Ausführliche Beschreibung
Bibliographische Detailangaben
| Veröffentlicht in: | IEEE transactions on pattern analysis and machine intelligence. - 1979. - 47(2025), 4 vom: 15. Apr., Seite 2816-2832
|
|---|
| 1. Verfasser: |
Fei, Wen
(VerfasserIn) |
|---|
| Weitere Verfasser: |
Dai, Wenrui,
Zhang, Liang,
Zhang, Luoming,
Li, Chenglin,
Zou, Junni,
Xiong, Hongkai |
|---|
| Format: | Online-Aufsatz
|
|---|
| Sprache: | English |
|---|
| Veröffentlicht: |
2025
|
|---|
| Zugriff auf das übergeordnete Werk: | IEEE transactions on pattern analysis and machine intelligence
|
|---|
| Schlagworte: | Journal Article |
|---|