zhaicunqi's picture
Upload 7 files
96e1ce7 verified
|
raw
history blame
22.6 kB
<div align="center">
<h1>
360Zhinao (360智脑)
</h1>
</div>
<div align="center">
🤖 <a href="https://www.modelscope.cn/profile/qihoo360">ModelScope</a>&nbsp&nbsp | &nbsp&nbsp
🔥 <a href="https://github.com/Qihoo360/360zhinao/blob/main/assets/WeChat.png">GitHub</a>&nbsp&nbsp | &nbsp&nbsp
💬 <a href="https://github.com/Qihoo360/360zhinao/tree/main/assets/WeChat.png">WeChat (微信)</a>&nbsp&nbsp
</div>
<br>
<p align="center">
Feel free to visit 360Zhinao's official website<a href="https://ai.360.com"> https://ai.360.com</a> for more experience.
</p>
<br>
# Models Introduction
🎉🎉🎉We open-source the 360Zhinao model series:
- **360Zhinao-7B-Base**
- **360Zhinao-7B-Chat-4K**
- **360Zhinao-7B-Chat-32K**
- **360Zhinao-7B-Chat-360K**
The characteristics of the 360Zhinao open-source models are:
- **Base Model:** Leveraging a high-quality corpus of 3.4 trillion Tokens which mainly consist of Chinese, English and code, we achieved competitive performance on relevant benchmark evaluations of the same model scale.
- **Chat Model:** Powerful chat capabilities and three different sequence lengths of 4K, 32K and 360K. 360K (about 500k Chinese characters) is the longest sequcence length among open-sourced Chinese models until now.
<br>
# News and Updates
- 2024.04.11 We release **360Zhinao-7B** 1.0 version, include the base model and three chat model with sequence lengths of 4K, 32K adn 360K.
<br>
# Table of contents
- [Download URL](#Download-URL)
- [Model Evaluation](#Model-Evaluation)
- [Quickstart](#Quickstart)
- [Model Inference](#Model-Inference)
- [Model Finetune](#Model-Finetune)
- [License](#License)
<br>
# Download URL
See the following table for this release and download links:
| Size | Model | BF16 | Int4|
|-|-|-|-|
| 7B | 360Zhinao-7B-Base | <a href="https://www.modelscope.cn/models/qihoo360/360Zhinao-7B-Base/summary">🤖</a> <a href="https://huggingface.co/qihoo360/360Zhinao-7B-Base">🤗</a> | |
| 7B | 360Zhinao-7B-Chat-4K | <a href="https://www.modelscope.cn/models/qihoo360/360Zhinao-7B-Chat-4K/summary">🤖</a> <a href="https://huggingface.co/qihoo360/360Zhinao-7B-Chat-4K">🤗</a> | <a href="https://www.modelscope.cn/models/qihoo360/360Zhinao-7B-Chat-4K-Int4/summary">🤖</a> <a href="https://huggingface.co/qihoo360/360Zhinao-7B-Chat-4K-Int4">🤗</a> |
| 7B | 360Zhinao-7B-Chat-32K | <a href="https://www.modelscope.cn/models/qihoo360/360Zhinao-7B-Chat-32K/summary">🤖</a> <a href="https://huggingface.co/qihoo360/360Zhinao-7B-Chat-32K">🤗</a> | <a href="https://www.modelscope.cn/models/qihoo360/360Zhinao-7B-Chat-32K-Int4/summary">🤖</a> <a href="https://huggingface.co/qihoo360/360Zhinao-7B-Chat-32K-Int4">🤗</a> |
| 7B | 360Zhinao-7B-Chat-360K | <a href="https://www.modelscope.cn/models/qihoo360/360Zhinao-7B-Chat-360K/summary">🤖</a> <a href="https://huggingface.co/qihoo360/360Zhinao-7B-Chat-360K">🤗</a> | <a href="https://www.modelscope.cn/models/qihoo360/360Zhinao-7B-Chat-360K-Int4/summary">🤖</a> <a href="https://huggingface.co/qihoo360/360Zhinao-7B-Chat-360K-Int4">🤗</a> |
<br>
# Model Evaluation
## Base Model
We evaluate the performance of our model on the OpenCompass evaluation datasets, including C-Eval, AGIEval, MMLU, CMMLU, HellaSwag, MATH, GSM8K, HumanEval, MBPP, BBH, LAMBADA. The ablity evaluated of model include natural language understanding, knowledge, mathematical computation and reasoning, code generation, logical reasoning, etc.
| <div style="width: 100pt">Model</div> | AVG | CEval | AGIEval | MMLU | CMMLU | HellaSwag | MATH | GSM8K | HumanEval | MBPP | BBH | LAMBADA |
|:----------------------|:---------:|:---------:|:---------:|:---------:|:---------:|:---------:|:---------:|:---------:|:---------:|:---------:|:---------:|:---------:|
| Baichuan2-7B | 41.49 | 56.3 | 34.6 | 54.7 | 57 | 67 | 5.4 | 24.6 | 17.7 | 24 | 41.8 | 73.3 |
| Baichuan-7B | 31.94 | 44.7 | 24.6 | 41.5 | 44.6 | 68.4 | 2.5 | 9.6 | 9.1 | 6.4 | 32.8 | 67.1 |
| ChatGLM3-6B | **58.67** | 67 | 47.4 | 62.8 | 66.5 | 76.5 | 19.2 | 61 | 44.5 | **57.2** | **66.2** | 77.1 |
| DeepSeek-7B | 39.8 | 45 | 24 | 49.3 | 46.8 | 73.4 | 4.2 | 18.3 | 25 | 36.4 | 42.8 | 72.6 |
| InternLM2-7B | 58.01 | 65.7 | 50.2 | 65.5 | 66.2 | 79.6 | 19.9 | **70.6** | 41.5 | 42.4 | 64.4 | 72.1 |
| InternLM-7B | 39.33 | 53.4 | 36.9 | 51 | 51.8 | 70.6 | 6.3 | 31.2 | 13.4 | 14 | 37 | 67 |
| LLaMA-2-7B | 33.27 | 32.5 | 21.8 | 46.8 | 31.8 | 74 | 3.3 | 16.7 | 12.8 | 14.8 | 38.2 | 73.3 |
| LLaMA-7B | 30.35 | 27.3 | 20.6 | 35.6 | 26.8 | 74.3 | 2.9 | 10 | 12.8 | 16.8 | 33.5 | 73.3 |
| Mistral-7B-v0.1 | 47.67 | 47.4 | 32.8 | 64.1 | 44.7 | 78.9 | 11.3 | 47.5 | 27.4 | 38.6 | 56.7 | 75 |
| MPT-7B | 30.06 | 23.5 | 21.3 | 27.5 | 25.9 | 75 | 2.9 | 9.1 | 17.1 | 22.8 | 35.6 | 70 |
| Qwen1.5-7B | 55.12 | 73.57 | **50.8** | 62.15 | 71.84 | 72.62 | **20.36** | 54.36 | **53.05** | 36.8 | 40.01 | 70.74 |
| Qwen-7B | 49.53 | 63.4 | 45.3 | 59.7 | 62.5 | 75 | 13.3 | 54.1 | 27.4 | 31.4 | 45.2 | 67.5 |
| XVERSE-7B | 34.27 | 61.1 | 39 | 58.4 | 60.8 | 73.7 | 2.2 | 11.7 | 4.9 | 10.2 | 31 | 24 |
| Yi-6B | 47.8 | 73 | 44.3 | 64 | **73.5** | 73.1 | 6.3 | 39.9 | 15.2 | 23.6 | 44.9 | 68 |
| **360Zhinao-7B** | 56.15 | **74.11** | 49.49 | **67.44** | 72.38 | **83.05** | 16.38 | 53.83 | 35.98 | 42.4 | 43.95 | **78.59** |
The above results could be viewed or reproduced on [Opencompass](https://rank.opencompass.org.cn/leaderboard-llm).
## Chat Models
We adopted a two-stage approach to train the long context models.
**First stage**: We increased RoPE base and extended the context length to 32K.
- Firstly, we performed Continual Pretraining on approximately 5B tokens with a 32K context window.
- Then during the SFT stage, we fine-tuned the model using long data from various sources, including high-quality human-labeled 32K data.
**Second stage**: We extended the context length to 360K, training with the following data:
- A small amount of high-quality human-labeled super-long data.
- Due to the scarcity of annotated super-long data, we constructed various forms of synthetic data.
- Multi-Doc QA: Similar to [Ziya-Reader](https://arxiv.org/abs/2311.09198), we generated multi-document QA pairs based on 360's database. Multiple QA pairs are constructed for one row of Multi-Doc QA data input, resulting in a multi-turn format and significantly improving the training efficiency.
- Single-Doc QA: Similar to [LLama2 Long](https://arxiv.org/abs/2309.16039), we constructed multi-turn QA data based on different segments within one row of long-text input.
We evaluated our models across various lengths and benchmarks.
- ### Long Context Benchmarks
We evaluated our 32K and 360K models on [LongBench](https://github.com/THUDM/LongBench), a multi-task bilingual benchmark for long contexts. We report results on Chinese tasks that are the most relevant to downstream applications: Single/Multi-Doc QA, Summarization, Few-Shot Learning and Code Completion.
| Model | Avg | 单文档QA | 多文档QA | 摘要 | Few-shot学习 | 代码补全 |
| :------------------------ |:---------:|:--------:|:---------:|:---------:|:------------:|:---------:|
| GPT-3.5-Turbo-16k | 37.84 | 61.2 | 28.7 | 16 | 29.2 | 54.1 |
| ChatGLM2-6B-32k | 37.16 | 51.6 | 37.6 | 16.2 | 27.7 | 52.7 |
| ChatGLM3-6B-32k | 44.62 | **62.3** | 44.8 | 17.8 | 42 | 56.2 |
| InternLM2-Chat-7B | 42.20 | 56.65 | 29.15 | **17.99** | 43.5 | **63.72** |
| Qwen1.5-Chat-7B | 36.75 | 52.85 | 30.08 | 14.28 | 32 | 54.55 |
| Qwen1.5-Chat-14B | 39.80 | 60.39 | 27.99 | 14.77 | 37 | 58.87 |
| 360Zhinao-7B-Chat-32K | **45.18** | 57.18 | **48.06** | 15.03 | **44** | 61.64 |
- ### 360Zhinao-7B-Chat-360K on "NeedleInAHaystack"
[NeedleInAHaystack](https://github.com/gkamradt/LLMTest_NeedleInAHaystack) places one small piece of information in different positions of long text and queries this information as a test of LLM's long-context capabilities.
360Zhinao-7B-Chat-360K could achieve over 98% accuracy on both English and Chinese NeedleInAHaystack tasks.
- English version(same as [NeedleInAHaystack](https://github.com/gkamradt/LLMTest_NeedleInAHaystack))
<p align="center">
<img src="assets/360Zhinao-7B-Chat-360K.en_score.png" width="600" />
<p>
**needle**:The best thing to do in San Francisco is eat a sandwich and sit in Dolores Park on a sunny day.
**query**:What is the best thing to do in San Francisco?
- Chinese version
<p align="center">
<img src="assets/360Zhinao-7B-Chat-360K.zh_score.png" width="600" />
<p>
We constructed the Chinese version following the [SuperCLUE-200K benchmark](https://mp.weixin.qq.com/s/QgoRf2LB-7vc3vTFOHJkpw):
**haystack**:Chinese novels.
**needle**:(in Chinese) 王莽是一名勤奋的店员,他每天凌晨就起床,赶在第一缕阳光照亮大地之前到达店铺,为即将开始的一天做准备。他清扫店铺,整理货架,为顾客提供方便。他对五金的种类和用途了如指掌,无论顾客需要什么,他总能准确地找到。\n然而,他的老板刘秀却总是对他吹毛求疵。刘秀是个挑剔的人,他总能在王莽的工作中找出一些小错误,然后以此为由扣他的工资。他对王莽的工作要求非常严格,甚至有些过分。即使王莽做得再好,刘秀也总能找出一些小问题,让王莽感到非常沮丧。\n王莽虽然对此感到不满,但他并没有放弃。他知道,只有通过自己的努力,才能获得更好的生活。他坚持每天早起,尽管他知道那天可能会再次被刘秀扣工资。他始终保持微笑,尽管他知道刘秀可能会再次对他挑剔。
**query**:(in Chinese) 王莽在谁的手下工作?
<br>
# Quickstart
Simple examples to illustrate how to use 360Zhinao-7B-Base and 360Zhinao-7B-Chat quickly using 🤖 ModelScope and 🤗 Transformers
## Dependency Installation
- python 3.8 and above
- pytorch 2.0 and above
- transformers 4.37.2 and above
- CUDA 11.4 and above are recommended.
```shell
pip install -r requirements.txt
```
We recommend installing Flash-Attention (which currently supports flash attention 2) to increase your performance and reduce your memory footprint. (flash-attention is optional and will work without installation)
>flash-attn >= 2.3.6
```shell
FLASH_ATTENTION_FORCE_BUILD=TRUE pip install flash-attn==2.3.6
```
## 🤗 Transformers
### Demonstration of Base Model Inference
This code demonstrates fast inference with 360Zhinao-7B-Base models using transformers.
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
from transformers.generation import GenerationConfig
MODEL_NAME_OR_PATH = "qihoo360/360Zhinao-7B-Base"
tokenizer = AutoTokenizer.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
MODEL_NAME_OR_PATH,
device_map="auto",
trust_remote_code=True)
generation_config = GenerationConfig.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
inputs = tokenizer('中国二十四节气\n1. 立春\n2. 雨水\n3. 惊蛰\n4. 春分\n5. 清明\n', return_tensors='pt')
inputs = inputs.to(model.device)
pred = model.generate(input_ids=inputs["input_ids"], generation_config=generation_config)
print("outputs:\n", tokenizer.decode(pred.cpu()[0], skip_special_tokens=True))
```
### Demonstration of Chat Model Inference
This code demo uses transformers to quickly use the 360Zhinao-7B-Chat-4K model for inference.
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
from transformers.generation import GenerationConfig
MODEL_NAME_OR_PATH = "qihoo360/360Zhinao-7B-Chat-4K"
tokenizer = AutoTokenizer.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
MODEL_NAME_OR_PATH,
device_map="auto",
trust_remote_code=True)
generation_config = GenerationConfig.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
messages = []
#round-1
messages.append({"role": "user", "content": "介绍一下刘德华"})
response = model.chat(tokenizer=tokenizer, messages=messages, generation_config=generation_config)
messages.append({"role": "assistant", "content": response})
print(messages)
#round-2
messages.append({"role": "user", "content": "他有什么代表作?"})
response = model.chat(tokenizer=tokenizer, messages=messages, generation_config=generation_config)
messages.append({"role": "assistant", "content": response})
print(messages)
```
## 🤖 ModelScope
### Demonstration of Base Model Inference
This code demonstrates using ModelScope to quickly use the 360Zhinao-7B-Base model for inference.
```python
from modelscope import AutoModelForCausalLM, AutoTokenizer
from modelscope import GenerationConfig
MODEL_NAME_OR_PATH = "qihoo360/360Zhinao-7B-Base"
tokenizer = AutoTokenizer.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
MODEL_NAME_OR_PATH,
device_map="auto",
trust_remote_code=True)
generation_config = GenerationConfig.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
inputs = tokenizer('中国二十四节气\n1. 立春\n2. 雨水\n3. 惊蛰\n4. 春分\n5. 清明\n', return_tensors='pt')
inputs = inputs.to(model.device)
pred = model.generate(input_ids=inputs["input_ids"], generation_config=generation_config)
print("outputs:\n", tokenizer.decode(pred.cpu()[0], skip_special_tokens=True))
```
### Demonstration of Chat Model Inference
This code demonstrates using ModelScope to quickly use the 360Zhinao-7B-Chat-4K model for inference.
```python
from modelscope import AutoModelForCausalLM, AutoTokenizer
from modelscope import GenerationConfig
MODEL_NAME_OR_PATH = "qihoo360/360Zhinao-7B-Chat-4K"
tokenizer = AutoTokenizer.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
MODEL_NAME_OR_PATH,
device_map="auto",
trust_remote_code=True)
generation_config = GenerationConfig.from_pretrained(
MODEL_NAME_OR_PATH,
trust_remote_code=True)
messages = []
#round-1
messages.append({"role": "user", "content": "介绍一下刘德华"})
response = model.chat(tokenizer=tokenizer, messages=messages, generation_config=generation_config)
messages.append({"role": "assistant", "content": response})
print(messages)
#round-2
messages.append({"role": "user", "content": "他有什么代表作?"})
response = model.chat(tokenizer=tokenizer, messages=messages, generation_config=generation_config)
messages.append({"role": "assistant", "content": response})
print(messages)
```
## CLI Demo
Use terminal interaction for a fast experience
```shell
python cli_demo.py
```
<p align="center">
<img src="assets/cli_demo.gif" width="600" />
<p>
## Web Demo
You can also use web interaction for a quick experience
```shell
streamlit run web_demo.py
```
<p align="center">
<img src="assets/web_demo.gif" width="600" />
<p>
## API Demo
Start command
```shell
python openai_api.py
```
Request parameter
```shell
curl --location --request POST 'http://localhost:8360/v1/chat/completions' \
--header 'Content-Type: application/json' \
--data-raw '{
"max_new_tokens": 200,
"do_sample": true,
"top_k": 0,
"top_p": 0.8,
"temperature": 1.0,
"repetition_penalty": 1.0,
"messages": [
{
"role": "user",
"content": "你叫什么名字?"
}
]
}'
```
<br>
# Model Inference
## Quantization
We provide quantization schemes based on AutoGPTQ and open source the Int4 quantization models.
## Deployment
### vLLM Installation
If you want to deploy and accelerate inference, we recommend using `vLLM==0.3.3`
If you are using **CUDA 12.1 and PyTorch 2.1**, you can install vLLM directly with the following command.
```shell
pip install vllm==0.3.3
```
Otherwise, please refer to the official vLLM [Installation Instructions](https://docs.vllm.ai/en/latest/getting_started/installation.html)。
>Once the installation is complete, you will need to do the following
1. Copy the vllm/zhinao.py file to the vllm/model_executor/models directory corresponding to your env environment.
2. Copy the vllm/serving_chat.py file to the vllm/entrypoints/openai corresponding to your env environment.
3. Then add a line to vllm/model_executor/models/\_\_init\_\_.py
```shell
"ZhinaoForCausalLM": ("zhinao", "ZhinaoForCausalLM"),
```
### vLLM Service Start
Starting the service
```shell
python -m vllm.entrypoints.openai.api_server \
--served-model-name 360Zhinao-7B-Chat-4K \
--model qihoo360/360Zhinao-7B-Chat-4K \
--trust-remote-code \
--tensor-parallel-size 1 \
--max-model-len 4096 \
--host 0.0.0.0 \
--port 8360
```
Use curl to request the service
```shell
curl http://localhost:8360/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "360Zhinao-7B-Chat-4K",
"max_tokens": 200,
"top_k": -1,
"top_p": 0.8,
"temperature": 1.0,
"presence_penalty": 0.0,
"frequency_penalty": 0.0,
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "你好"}
],
"stop": [
"<eod>",
"<|im_end|>",
"<|im_start|>"
]
}'
```
Use python to request the service
```python
from openai import OpenAI
openai_api_key = "EMPTY"
openai_api_base = "http://localhost:8360/v1"
client = OpenAI(
api_key=openai_api_key,
base_url=openai_api_base,
)
chat_response = client.chat.completions.create(
model="360Zhinao-7B-Chat-4K",
messages=[
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "你好"},
],
stop=[
"<eod>",
"<|im_end|>",
"<|im_start|>"
],
presence_penalty=0.0,
frequency_penalty=0.0
)
print("Chat response:", chat_response)
```
> Notice: If you need to enable repetition penalty, recommended to use *presence_penalty* and *frequency_penalty* parameters.
>
<br>
# Model Finetune
## Training data
Training Data: data/training_data_sample.json. The sample data is 10,000 pieces sampled from [multiturn_chat_0.8M](https://huggingface.co/datasets/BelleGroup/multiturn_chat_0.8M) and format converted.
Data Format:
```json
[
{
"id": 1,
"conversations": [
{
"from": "system",
"value": "You are a helpful assistant."
},
{
"from": "user",
"value": "您好啊"
},
{
"from": "assistant",
"value": "你好!我今天能为您做些什么?有什么问题或需要帮助吗? 我在这里为您提供服务。"
}
]
}
]
```
## Fine-tuning scripts
```shell
set -x
HOSTFILE=hostfile
DS_CONFIG=./finetune/ds_config_zero2.json
# PARAMS
LR=5e-6
EPOCHS=3
MAX_LEN=4096
BATCH_SIZE=4
NUM_NODES=1
NUM_GPUS=8
MASTER_PORT=29500
IS_CONCAT=False # Whether to concatenate to maximum length (MAX_LEN)
DATA_PATH="./data/training_data_sample.json"
MODEL_PATH="qihoo360/360Zhinao-7B-Base"
OUTPUT_DIR="./outputs/"
deepspeed --hostfile ${HOSTFILE} \
--master_port ${MASTER_PORT} \
--num_nodes ${NUM_NODES} \
--num_gpus ${NUM_GPUS} \
finetune.py \
--report_to "tensorboard" \
--data_path ${DATA_PATH} \
--model_name_or_path ${MODEL_PATH} \
--output_dir ${OUTPUT_DIR} \
--model_max_length ${MAX_LEN} \
--num_train_epochs ${EPOCHS} \
--per_device_train_batch_size ${BATCH_SIZE} \
--gradient_accumulation_steps 1 \
--save_strategy steps \
--save_steps 200 \
--learning_rate ${LR} \
--lr_scheduler_type cosine \
--adam_beta1 0.9 \
--adam_beta2 0.95 \
--adam_epsilon 1e-8 \
--max_grad_norm 1.0 \
--weight_decay 0.1 \
--warmup_ratio 0.01 \
--gradient_checkpointing True \
--bf16 True \
--tf32 True \
--deepspeed ${DS_CONFIG} \
--is_concat ${IS_CONCAT} \
--logging_steps 1 \
--log_on_each_node False
```
```shell
bash finetune/ds_finetune.sh
```
- By configuring the **hostfile**, single-machine and multi-machine training can be realized.
- By configuring **ds_config**, realize zero2 and zero3 training
- By configuring the **fp16**、**bf16** realize mixed precision training, bf16 is recommended to be consistent with the pre-trained model.
- By configuring **is_concat**, Whether the training data is concatenated or not is controlled. When the magnitude of the training data is large, the training efficiency can be improved by concatenation.
<br>
# License
The source code of this warehouse follows the open source license Apache 2.0.
The 360 ​Zhinao open source model supports commercial use. If you need to use this model and its derivative models for commercial purposes, please contact us via email (g-zhinao-opensource@360.cn) to apply. For the specific license agreement, please see [《360 Zhinao Open Source Model License》](./360智脑开源模型许可证.txt).