File size: 3,867 Bytes
6e436f4
 
3296965
 
6e436f4
3a6d648
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3296965
b34fe67
fa76e65
3296965
3a6d648
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3296965
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
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
---
license: mit
datasets:
- DAMO-NLP-SG/LongCorpus-2.5B
---
# CLEX: Continuous Length Extrapolation for Large Language Models
This repo stores the checkpoint of CLEX-Mixtral-8x7B-32K.


## Features and Highlights of CLEX
![CLEX_diagram](https://github.com/DAMO-NLP-SG/CLEX/assets/18526640/063ffe34-0116-4759-92bf-e22fc7264cdf)

- **Simple and Clear**: _MINIMAL_ code and architecture changes. Only one up-and-down projection layer introduced, _NO_ recurrent memory caching or sparse attention required.
- **Train Short, Test Long**: _NO_ performance drop on the sequences _4x~8x longer_ than the training ones (see [here](https://github.com/DAMO-NLP-SG/CLEX#language-modelling)). 
- **Continuous Length Extrapolation**: Explicitly modeling the continuous dynamics of context window size during length extrapolation.

If you have any questions, feel free to contact us. (Emails: guanzzh.chen@gmail.com, lixin4ever@gmail.com)

## Model Zoo
<div align="center">

| Model Name | Model Type | Starting Point | Train Data |Train Length | MAX Test Length | HF Repo |
|:-----|:-----|:-----------|:-----------|:-----------|:-----------|:------:|
| CLEX-LLaMA-2-7B-16K | base | LLaMA-2-7B | [Redpajama-Book](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T) | 16K | 64K | [link](https://huggingface.co/DAMO-NLP-SG/CLEX-7B-16K) |
| CLEX-LLaMA-2-7B-Chat-16K | chat | CLEX-7B-16K | [UltraChat](https://github.com/thunlp/UltraChat) | 16K | 64K | [link](https://huggingface.co/DAMO-NLP-SG/CLEX-7B-Chat-16K) |
| CLEX-LLaMA-2-7B-64K | base | LLaMA-2-7B | [Redpajama-Book](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T) | 64k | 256K | [link](https://huggingface.co/DAMO-NLP-SG/CLEX-LLaMA-2-7B-64K) |
| CLEX-Phi-2-32K | base | Phi-2-2.7B | [LongCorpus-2.5B](https://huggingface.co/datasets/DAMO-NLP-SG/LongCorpus-2.5B) | 32k | 128K | [link](https://huggingface.co/DAMO-NLP-SG/CLEX-Phi-2-32K) |
| **CLEX-Mixtral-8x7B-32K** (this checkpoint) | base | Mixtral-8x7B-v0.1 | [LongCorpus-2.5B](https://huggingface.co/datasets/DAMO-NLP-SG/LongCorpus-2.5B) | 32k | >128K | [link](https://huggingface.co/DAMO-NLP-SG/CLEX-Mixtral-8x7B-32K) |
| CLEX-Mixtral-8x7B-Chat-32k | chat | CLEX-Mixtral-8x7B-32K | [Ultrachat 200k](https://huggingface.co/datasets/HuggingFaceH4/ultrachat_200k) | 32k | >128K | [link](https://huggingface.co/DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K) |
</div>


## Usage


```bash
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("DAMO-NLP-SG/CLEX-Mixtral-8x7B-32K", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("DAMO-NLP-SG/CLEX-Mixtral-8x7B-32K", torch_dtype=torch.bfloat16)
inputs = tokenizer("What is CLEX?", return_tensors="pt")
sample = model.generate(**inputs, max_length=128)
print(tokenizer.decode(sample[0]))
```




## Evaluation
### Language Modelling




The CLEX-Phi-2-2.7B and CLEX-Mixtral-8x7B are trained on [LongCorpus-2.5B](https://huggingface.co/datasets/DAMO-NLP-SG/LongCorpus-2.5B), where the eval results on test set are listed below.

|                   | Train Length | Eval.(32k) | Eval.(64k) | Eval.(128k) | Eval.(256k) |
| ----------------- | ------------ | ---------- | ---------- | ----------- | ----------- |
| Mixtral-8x7B      | 32k          | 2.78       | 3.44       | 5.88        | 14.20       |
| CLEX-Mixtral-8x7B | 32k          | 2.56       | 2.53       | 2.57        | 3.78        |





## Citation
If you find our project useful, hope you can star our repo and cite our paper as follows:
```
@article{damonlpsg2023clex,
  author = {Chen, Guanzheng and Li, Xin and Meng, Zaiqiao and Liang, Shangsong and Bing, Lidong},
  title = {CLEX: Continuous Length Extrapolation for Large Language Models},
  year = 2023,
  journal = {arXiv preprint arXiv:2310.16450},
  url = {https://arxiv.org/abs/2310.16450}
}
```