hi, I have around 100k samples of emails which are unlabelled. I want to fine tune a model on this data. What strategies are available to identify the most relevant / different samples to label. ?
Colin
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Advice on too many labels | 0 | 256 | July 10, 2023 | |
| Advice on an email classification problem | 3 | 131 | August 27, 2024 | |
| Email classification, labeling and entity classification/extraction | 2 | 427 | June 6, 2024 | |
| 10_000 samples & 10_000 labels | 0 | 502 | July 31, 2021 | |
| Fine-tuning Zero-shot models | 4 | 6218 | February 7, 2023 |