File size: 554 Bytes
0121069
 
 
 
 
 
 
 
9bf875b
775458a
 
0121069
 
 
9bf875b
0121069
fea0be1
7a433bb
0121069
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
This model is trained with Iterative DPO in OpenRLHF

Datasets and Hyperparameters

```
Reward Model:https://huggingface.co/OpenLLMAI/Llama-3-8b-rm-700k
SFT Model: https://huggingface.co/OpenLLMAI/Llama-3-8b-sft-mixture
Prompt Dataset: https://huggingface.co/datasets/OpenLLMAI/prompt-collection-v0.1

Max Prompt Length: 2048
Max Response Length: 2048
best_of_n: 2 (2 samples for each prompt)
Learning Rate: 5e-7
Beta: 0.1
Scheduler: Cosine with Warmup (0.03) and MinLR (0.1)
Rollout Batch Size: 20000
Training Batch Size: 256
Number of Iterations: 9
```