File size: 392 Bytes
debd545
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
---
license: mit
---
# GP-L-Init
This model serves as a initial checkpoint to reproduce results in paper **SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training**.



## Related links

Website: https://tianzhechu.com/SFTvsRL/

Github: https://github.com/LeslieTrue/SFTvsRL

Arxiv: https://arxiv.org/abs/2501.17161v1

HF: https://huggingface.co/papers/2501.17161