File size: 8,695 Bytes
cabb4bc 69a93c1 09471ed cabb4bc 8721983 69a93c1 d2d47d3 69a93c1 e0f398f 8721983 14da305 8721983 fd7275d 8721983 09471ed |
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 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 |
---
license: mit
datasets:
- DAMO-NLP-SG/LongCorpus-2.5B
model-index:
- name: CLEX-Mixtral-8x7B-Chat-32K
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 66.38
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 86.48
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 70.12
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 56.47
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 82.56
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 50.49
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K
name: Open LLM Leaderboard
---
# CLEX: Continuous Length Extrapolation for Large Language Models
This repo stores the checkpoint of CLEX-Mixtral-8x7B-Chat-32K.
## Features and Highlights of CLEX
- **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 | 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** (this checkpoint) | 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-Chat-32K", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("DAMO-NLP-SG/CLEX-Mixtral-8x7B-Chat-32K", torch_dtype=torch.bfloat16, trust_remote_code=True)
inputs = tokenizer("What is CLEX?", return_tensors="pt")
sample = model.generate(**inputs, max_length=128)
print(tokenizer.decode(sample[0]))
```
## Evaluation
## InfiniteBench
We also evaluate CLEX-Mixtral-8x7B-Chat-32k on [InfiniteBench](https://github.com/OpenBMB/InfiniteBench), which is a 128k-length benchmark covering various tasks. We compare our CLEX-Mixtral-8x7B-Chat-32k with GPT-4, Claude, KimiChat, and vanilla Mixtral-8x7B.
| Task Name | GPT-4 | YaRN-Mistral-7B | Kimi-Chat | Claude 2 | CLEX-Mixtral-8x7B-Chat-32k | Mixtral-8x7B-Instruct-v0.1 |
| ------------------- | ------ | --------------- | --------- | -------- | -------------------------- | -------------------------- |
| Retrieve.PassKey | 100% | 92.71% | 98.14% | 97.80% | 99.72% | 96.78% |
| **Retrieve.Number** | 100% | 56.61% | 95.42% | 98.14% | 76.10% | 76.61% |
| **Retrieve.KV** | 89.00% | < 5% | 53.60% | 65.40% | <5% | <5% |
| En.Sum | 14.73% | 9.09% | 17.93% | 14.45% | 15.48% | 14.3% |
| En.QA | 22.22% | 9.55% | 16.52% | 11.97% | 15.52% | 16.81% |
| En.MC | 67.25% | 27.95% | 72.49% | 62.88% | 58.96% | 56.77% |
| En.Dia | 8.50% | 7.50% | 11.50% | 46.50% | 9% | <5% |
| Code.Debug | 39.59% | < 5% | 18.02% | < 5% | 21.32% | <5% |
| Code.Run | 23.25% | < 5% | < 5% | < 5% | < 5% | <5% |
| Math.Calc | < 5% | < 5% | < 5% | < 5% | < 5% | <5% |
| Math.Find | 60.00% | 17.14% | 12.57% | 32.29% | 28% | 26.57% |
## 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}
}
```
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_DAMO-NLP-SG__CLEX-Mixtral-8x7B-Chat-32K)
| Metric |Value|
|---------------------------------|----:|
|Avg. |68.75|
|AI2 Reasoning Challenge (25-Shot)|66.38|
|HellaSwag (10-Shot) |86.48|
|MMLU (5-Shot) |70.12|
|TruthfulQA (0-shot) |56.47|
|Winogrande (5-shot) |82.56|
|GSM8k (5-shot) |50.49|
|