Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,39 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
---
|
4 |
+
|
5 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
6 |
+
|
7 |
+
An unofficial reproduced PRepBN-Llama-350M checkpoints for [SLAB](https://github.com/xinghaochen/SLAB/).
|
8 |
+
|
9 |
+
### Model Sources [optional]
|
10 |
+
|
11 |
+
<!-- Provide the basic links for the model. -->
|
12 |
+
|
13 |
+
- **Repository:** [https://github.com/xinghaochen/SLAB/]
|
14 |
+
- **Paper [optional]:** [https://arxiv.org/abs/2405.11582]
|
15 |
+
|
16 |
+
|
17 |
+
## Evaluation
|
18 |
+
|
19 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
20 |
+
|
21 |
+
https://github.com/xinghaochen/SLAB/tree/main/llama
|
22 |
+
|
23 |
+
```
|
24 |
+
python evaluation.py --ckpt <checkpoint-path>
|
25 |
+
```
|
26 |
+
|
27 |
+
[Results](https://github.com/xinghaochen/SLAB/blob/main/docs/llama.png)
|
28 |
+
|
29 |
+
|
30 |
+
**BibTeX:**
|
31 |
+
|
32 |
+
```
|
33 |
+
@inproceedings{guo2024slab,
|
34 |
+
title={SLAB: Efficient Transformers with Simplified Linear Attention and Progressive Re-parameterized Batch Normalization},
|
35 |
+
author={Guo, Jialong and Chen, Xinghao and Tang, Yehui and Wang, Yunhe},
|
36 |
+
booktitle={International Conference on Machine Learning},
|
37 |
+
year={2024}
|
38 |
+
}
|
39 |
+
```
|