Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2501.18511 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915299512549376 |
|---|---|
| author | Feuer, Benjamin Hegde, Chinmay |
| author_facet | Feuer, Benjamin Hegde, Chinmay |
| contents | Language model (LLM) post-training, from DPO to distillation, can refine behaviors and unlock new skills, but the open science supporting these post-training techniques is still in its infancy. One limiting factor has been the difficulty of conducting large-scale comparative analyses of synthetic data generating models and LLM judges. To close this gap, we introduce WILDCHAT-50M, the largest public chat dataset to date. We extend the existing WildChat dataset to include responses not only from GPT, but from over 50 different open-weight models, ranging in size from 0.5B to 104B parameters. We conduct an extensive comparative analysis and demonstrate the potential of this dataset by creating RE-WILD, our own public SFT mix, which outperforms the recent Tulu-3 SFT mixture from Allen AI with only 40% as many samples. Our dataset, samples and code are available at https://github.com/penfever/wildchat-50m. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_18511 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | WILDCHAT-50M: A Deep Dive Into the Role of Synthetic Data in Post-Training Feuer, Benjamin Hegde, Chinmay Machine Learning Computation and Language Language model (LLM) post-training, from DPO to distillation, can refine behaviors and unlock new skills, but the open science supporting these post-training techniques is still in its infancy. One limiting factor has been the difficulty of conducting large-scale comparative analyses of synthetic data generating models and LLM judges. To close this gap, we introduce WILDCHAT-50M, the largest public chat dataset to date. We extend the existing WildChat dataset to include responses not only from GPT, but from over 50 different open-weight models, ranging in size from 0.5B to 104B parameters. We conduct an extensive comparative analysis and demonstrate the potential of this dataset by creating RE-WILD, our own public SFT mix, which outperforms the recent Tulu-3 SFT mixture from Allen AI with only 40% as many samples. Our dataset, samples and code are available at https://github.com/penfever/wildchat-50m. |
| title | WILDCHAT-50M: A Deep Dive Into the Role of Synthetic Data in Post-Training |
| topic | Machine Learning Computation and Language |
| url | https://arxiv.org/abs/2501.18511 |