Table 5 Cold-Start Hyperparameter Settings
From: Large language models learning to write rhyming Tang poetry A Xunzi Yayun R1 case study
Hyperparameter | Description | Parameter Value |
|---|---|---|
Batch_size | Batch size for training | 2 |
Learning_rate | Learning rate | 1e-4 |
Max_len | Maximum context length | 2048 |
Num_epochs | Number of training epochs | 3 |
gradient_accumulation_steps | Gradient accumulation steps | 4 |
