Hamster, Ulf A. ; Lee, Ji-Ung ; Geyken, Alexander ; Gurevych, Iryna (2023)
Rediscovering Hashed Random Projections for Efficient Quantization of Contextualized Sentence Embeddings.
doi: 10.48550/arXiv.2304.02481
Report, Bibliographie
Kurzbeschreibung (Abstract)
Training and inference on edge devices often requires an efficient setup due to computational limitations. While pre-computing data representations and caching them on a server can mitigate extensive edge device computation, this leads to two challenges. First, the amount of storage required on the server that scales linearly with the number of instances. Second, the bandwidth required to send extensively large amounts of data to an edge device. To reduce the memory footprint of pre-computed data representations, we propose a simple, yet effective approach that uses randomly initialized hyperplane projections. To further reduce their size by up to 98.96%, we quantize the resulting floating-point representations into binary vectors. Despite the greatly reduced size, we show that the embeddings remain effective for training models across various English and German sentence classification tasks that retain 94%--99% of their floating-point.
| Typ des Eintrags: | Report |
|---|---|
| Erschienen: | 2023 |
| Autor(en): | Hamster, Ulf A. ; Lee, Ji-Ung ; Geyken, Alexander ; Gurevych, Iryna |
| Art des Eintrags: | Bibliographie |
| Titel: | Rediscovering Hashed Random Projections for Efficient Quantization of Contextualized Sentence Embeddings |
| Sprache: | Englisch |
| Publikationsjahr: | 16 Mai 2023 |
| Verlag: | arXiv |
| Reihe: | Computation and Language |
| Kollation: | 14 Seiten |
| DOI: | 10.48550/arXiv.2304.02481 |
| URL / URN: | https://arxiv.org/abs/2304.02481 |
| Kurzbeschreibung (Abstract): | Training and inference on edge devices often requires an efficient setup due to computational limitations. While pre-computing data representations and caching them on a server can mitigate extensive edge device computation, this leads to two challenges. First, the amount of storage required on the server that scales linearly with the number of instances. Second, the bandwidth required to send extensively large amounts of data to an edge device. To reduce the memory footprint of pre-computed data representations, we propose a simple, yet effective approach that uses randomly initialized hyperplane projections. To further reduce their size by up to 98.96%, we quantize the resulting floating-point representations into binary vectors. Despite the greatly reduced size, we show that the embeddings remain effective for training models across various English and German sentence classification tasks that retain 94%--99% of their floating-point. |
| Freie Schlagworte: | UKP_p_EVIDENCE |
| Zusätzliche Informationen: | 2. Version |
| Fachbereich(e)/-gebiet(e): | 20 Fachbereich Informatik 20 Fachbereich Informatik > Ubiquitäre Wissensverarbeitung |
| TU-Projekte: | DFG|GU798/27-1|EVIDENCE: Computer-u |
| Hinterlegungsdatum: | 12 Jun 2023 12:37 |
| Letzte Änderung: | 19 Dez 2024 11:35 |
| PPN: | 510471323 |
| Export: | |
| Suche nach Titel in: | TUfind oder in Google |
 |
Frage zum Eintrag |
Optionen (nur für Redakteure)
 |
Redaktionelle Details anzeigen |
Drucken
Impressum