PyTorch
Safetensors
English
llama
File size: 1,358 Bytes
ad64752
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0063346
ad64752
 
 
 
 
 
 
 
 
 
 
 
 
 
fd3bc36
 
 
 
 
ad64752
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
---
license: apache-2.0
datasets:
- gair-prox/RedPajama-pro
language:
- en
tags:
- llama
---

# RedPJ-ProX-1.7B

<p align="center">
  <img src="prox-teaser.png">
</p>

[ArXiv](http://arxiv.org/abs/2409.17115) | [Models](https://huggingface.co/gair-prox/RedPJ-ProX-1.7B) | [Data](https://huggingface.co/datasets/gair-prox/RedPajama-pro) | [Code](https://github.com/GAIR-NLP/program-every-example)

**RedPJ-ProX-1.7B** is a small language model. It was and trained on the [RedPajama-V2-pro](https://huggingface.co/datasets/gair-prox/RedPajama-pro) for 50B tokens. 

## Evaluations

ProX models are evaluated over 10 language model benchmarks in zero-shot setting.

|                       | ArC-c | ARC-e | CSQA  | HellaS | MMLU  | OBQA  | PiQA  | SIQA  | WinoG | SciQ  | AVG  |
|-----------------------|-------|-------|-------|-----------|-------|-------|-------|-------|-------|-------|------|
| raw | 26.9 | 51.4 | 32.4 | 47.3 | 29.3 | 32.2 | 69.7 | 39.6 | 52.1 | 79.1 | 46.0 |
| ours | 31.1 | 60.7 | 29.8 | 51.0 | 31.7 | 33.2 | 70.9 | 39.2 | 53.3 | 79.1 | 48.0 |

### Citation
```
@article{zhou2024programming,
  title={Programming Every Example: Lifting Pre-training Data Quality like Experts at Scale},
  author={Zhou, Fan and Wang, Zengzhi and Liu, Qian and Li, Junlong and Liu, Pengfei},
  journal={arXiv preprint arXiv:2409.17115},
  year={2024}
}
```