LESS: Selecting Influential Data for Targeted Instruction Tuning
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913389365690368 |
|---|---|
| author | Xia, Mengzhou Malladi, Sadhika Gururangan, Suchin Arora, Sanjeev Chen, Danqi |
| author_facet | Xia, Mengzhou Malladi, Sadhika Gururangan, Suchin Arora, Sanjeev Chen, Danqi |
| contents | Instruction tuning has unlocked powerful capabilities in large language models (LLMs), effectively using combined datasets to develop generalpurpose chatbots. However, real-world applications often require a specialized suite of skills (e.g., reasoning). The challenge lies in identifying the most relevant data from these extensive datasets to effectively develop specific capabilities, a setting we frame as targeted instruction tuning. We propose LESS, an optimizer-aware and practically efficient algorithm to effectively estimate data influences and perform Low-rank gradiEnt Similarity Search for instruction data selection. Crucially, LESS adapts existing influence formulations to work with the Adam optimizer and variable-length instruction data. LESS first constructs a highly reusable and transferable gradient datastore with low-dimensional gradient features and then selects examples based on their similarity to few-shot examples embodying a specific capability. Experiments show that training on a LESS-selected 5% of the data can often outperform training on the full dataset across diverse downstream tasks. Furthermore, the selected data is highly transferable: smaller models can be leveraged to select useful data for larger models and models from different families. Our qualitative analysis shows that our method goes beyond surface form cues to identify data that exemplifies the necessary reasoning skills for the intended downstream application. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_04333 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LESS: Selecting Influential Data for Targeted Instruction Tuning Xia, Mengzhou Malladi, Sadhika Gururangan, Suchin Arora, Sanjeev Chen, Danqi Computation and Language Artificial Intelligence Machine Learning Instruction tuning has unlocked powerful capabilities in large language models (LLMs), effectively using combined datasets to develop generalpurpose chatbots. However, real-world applications often require a specialized suite of skills (e.g., reasoning). The challenge lies in identifying the most relevant data from these extensive datasets to effectively develop specific capabilities, a setting we frame as targeted instruction tuning. We propose LESS, an optimizer-aware and practically efficient algorithm to effectively estimate data influences and perform Low-rank gradiEnt Similarity Search for instruction data selection. Crucially, LESS adapts existing influence formulations to work with the Adam optimizer and variable-length instruction data. LESS first constructs a highly reusable and transferable gradient datastore with low-dimensional gradient features and then selects examples based on their similarity to few-shot examples embodying a specific capability. Experiments show that training on a LESS-selected 5% of the data can often outperform training on the full dataset across diverse downstream tasks. Furthermore, the selected data is highly transferable: smaller models can be leveraged to select useful data for larger models and models from different families. Our qualitative analysis shows that our method goes beyond surface form cues to identify data that exemplifies the necessary reasoning skills for the intended downstream application. |
| title | LESS: Selecting Influential Data for Targeted Instruction Tuning |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2402.04333 |