LLMR: Knowledge Distillation with a Large Language Model-Induced Reward
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909319734231040 |
|---|---|
| author | Li, Dongheng Hao, Yongchang Mou, Lili |
| author_facet | Li, Dongheng Hao, Yongchang Mou, Lili |
| contents | Large language models have become increasingly popular and demonstrated remarkable performance in various natural language processing (NLP) tasks. However, these models are typically computationally expensive and difficult to be deployed in resource-constrained environments. In this paper, we propose LLMR, a novel knowledge distillation (KD) method based on a reward function induced from large language models. We conducted experiments on multiple datasets in the dialogue generation and summarization tasks. Empirical results demonstrate that our LLMR approach consistently outperforms traditional KD methods in different tasks and datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_12500 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LLMR: Knowledge Distillation with a Large Language Model-Induced Reward Li, Dongheng Hao, Yongchang Mou, Lili Computation and Language Artificial Intelligence Large language models have become increasingly popular and demonstrated remarkable performance in various natural language processing (NLP) tasks. However, these models are typically computationally expensive and difficult to be deployed in resource-constrained environments. In this paper, we propose LLMR, a novel knowledge distillation (KD) method based on a reward function induced from large language models. We conducted experiments on multiple datasets in the dialogue generation and summarization tasks. Empirical results demonstrate that our LLMR approach consistently outperforms traditional KD methods in different tasks and datasets. |
| title | LLMR: Knowledge Distillation with a Large Language Model-Induced Reward |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2409.12500 |